Original Research

Take-Home Points

• Clinical and radiographic outcomes of patients treated with non-spanning external fixation are comparable to those treated with open reduction and internal volar locked plate fixation.
• Non-spanning external fixation can lead to satisfactory outcomes based on the following features: fragment specific fixation, subchondral support, fixed angle strength, limited dissection, distraction/length adjustment, joint distraction avoidance, and ability to perform early rehabilitation.
• Non-spanning external fixation should be considered as a treatment option for complicated unstable comminuted intra-articular distal radius fractures, specifically in the elderly.

In the United States, distal radius fractures (DRFs) are among the most common fractures, comprising about 15% of all extremity fractures.¹ With a DRF, the primary treatment goal is anatomical reduction with restoration of radiographic parameters and stable fixation of the fracture to restore wrist function.

This fracture type has a variety of treatment alternatives, including nonoperative closed reduction and casting of stable fractures, open reduction and internal fixation (ORIF) with dorsal or volar locking plates, and external fixation. Optimal surgical management of unstable DRFs remains controversial.² Closed reduction with percutaneous pinning or external fixation has become less common with a trend toward using volar locking plates for internal fixation.³

External fixation of DRFs traditionally has involved either spanning or simple nonspanning devices. Spanning fixation is particularly useful in open or highly comminuted fractures with an unstable soft-tissue envelope. In the past, nonspanning external fixation typically was reserved for fractures with a noncomminuted extra-articular distal fragment to which several large pins or Kirschner wires (K-wires) could be secured. The Non-Bridging External Fixator (NBX; Nutek Orthopaedics) may be used in cases that traditionally might be treated with locked plating or fragment-specific fixation. Specifically, this device is indicated for comminuted intra-articular DRFs in which bone quality may be less than ideal. The NBX, also suitable in open fractures with a stable soft-tissue envelope, can restore and maintain articular alignment by providing subchondral support and stability with fragment-specific fixation. A key advantage of this type of external fixation is that it involves percutaneous fixation and allows for early postoperative range of motion (ROM).

Numerous studies have found excellent outcomes of treating unstable DRFs with ORIF with volar locking plates.^4-6 However, few studies have compared the clinical and radiographic outcomes of ORIF with those of nonspanning external fixation in the treatment of unstable comminuted intra-articular DRFs. Windolf and colleagues⁷ found that, in cadaveric unstable intra-articular DRFs, nonspanning external fixation with multiplanar K-wires had biomechanical characteristics comparable to those of volar locking plates. Other suitable DRF treatment options have been found: an alternative nonbridging external fixator with multiplanar K-wires (Gradl and colleagues⁸) and the Cross-Pin Fixation system (A.M. Surgical) (Mirza and colleagues⁹).

We conducted a study to compare functional and radiographic outcomes of unstable comminuted intra-articular DRFs treated with a nonspanning external fixation device (NBX) with outcomes achieved with volar locking plates in a historical control group.

Materials and Methods

This retrospective case-control study was approved by our Institutional Review Board and conducted at 2 institutions. Included in the study were 25 consecutive patients (2 institutions) who underwent closed reduction and external fixation (CREF) with NBX as treatment for unstable DRFs (diagnosis based on radiographic parameters or inability to maintain acceptable alignment after closed reduction and casting). Of these 25 patients, 11 were available for clinical follow-up and medical records review; the other 14 were not available for followup but had their charts reviewed for radiographic data and treatment details. Six of the 14 patients declined to participate in the study, and the other 8 were lost to follow-up because of nonstandardized follow-up protocols. Patients were excluded from the study if their final follow-up had not occurred, or if it occurred before 6 months. For their participation in clinical follow-up, patients received nominal time compensation and mileage reimbursement through a grant from the NBX manufacturer.

The 25 patients underwent CREF with NBX between November 2008 and March 2013. Indications for external fixation consideration were intra-articular extension or significant comminution in patients with poor soft tissue or in patients who wanted to avoid invasive surgery or a permanent implant. Of the 11 patients who agreed to participate in the study, 7 were women and 4 were men; mean age was 64 years (range, 15-81 years). Of the 14 patients unable to follow up, 11 were women and 3 were men; mean age was 63 years (range, 26-89 years). At the last available follow-up, each of the 25 patients was doing well, was satisfied with treatment received and function regained, and had a healed DRF. In almost every case, the mechanism of injury was a fall onto an outstretched hand; most fractures were type C per AO (Arbeitsgemeinschaft für Osteosynthesefragen) classification (Table 1).

The surgical technique for this nonspanning external fixator involves closed reduction with longitudinal traction using ligamentotaxis to grossly align the fracture fragments, with small adjustments made throughout the procedure. A dorsally placed radiolucent fixator is used with fluoroscopic guidance to percutaneously affix a subchondral raft of smooth bicortical .062-inch K-wires. The fixator’s abundant pin holes allow for each specific distal fragment to be captured by pins that are a part of the external fixation construct. Furthermore, radially based pins that use a side bar allow for a “weave” of fixation. Radial length is then obtained and maintained by attaching the distal complex to proximal pins in the radial diaphysis. After pins are cut and wrist and digits are taken through full ROM to ensure smooth tracking, fluoroscopy is used to confirm final fracture fixation and alignment (Figure 1).

In ideal scenarios with good fixation, patients can begin gentle ROM exercises within 1 week after surgery. This regimen can progress to more aggressive motion exercises and even light strengthening (Figure 2).

The 11 clinical follow-up patients underwent directed clinical examination, including ROM and strength evaluation, by Dr. Dwyer and Dr. Crosby. Follow-up also included completion of questionnaires and review of radiographs.

During the clinical follow-up, a standard goniometer was used to evaluate active ROM (wrist flexion and extension and wrist radial and ulnar deviation, measured down the long axis of the forearm and the index ray), and forearm pronation and supination were measured from the 90° elbow flexion position using the humerus as the reference point with the shoulders in 0° of flexion, abduction, and external rotation. In addition, a calibrated dynamometer (Sammons Preston) was used to measure grip strength (position 3) and key pinch strength, and the average of 3 trials of each strength test was calculated. ROM and strength values were calculated as percentages of the contralateral (uninjured) side, as these ratios are more sensitive in detecting clinical changes.¹⁰ A 10% adjustment for dominant hand grip strength in right-handed patients was used for this comparison.¹¹

Union (osseous bridging across fracture site on 2 of 3 views), radial height, radial inclination, and volar tilt were measured on standard posteroanterior and lateral radiographs taken at several points: time of injury, postreduction and/or preoperative, initial postoperative, and final follow-up. All radiographic measurements were independently taken by Dr. Dwyer and Dr. Crosby, who used a digital goniometer and ruler (Siemens Medical Solutions) or, when necessary, manual instruments. Means of the original and independent measurements were used for calculations.

The Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire, the Mayo wrist score, and the patient-rated wrist evaluation were used to assess activities of daily living, pain, and quality of life after surgery. Mayo wrist scores were adjusted for unemployed patients; work status was replaced with return to normal activities. 

Complications of surgical treatment were evaluated. Major complications evaluated were loss of reduction, malunion, nonunion, deep infection, neuropathy, and tendon rupture. Minor complication possibilities were transient extensor tendon irritation, superficial infection, and finger stiffness. Also noted were 1 patient who subsequently required another procedure and 7 patients who were immobilized after external fixation removal.

We compared our study group’s outcomes with those of historical control patients who underwent fixation with internal volar locking plates. The 2 groups had similar demographic characteristics. To obtain the historical controls, we used the key words distal, radi*, volar, and plat* in a PubMed search. From the 169 citations found, we removed biomechanical cadaver studies, studies that focused on patients with demographics and fracture types dissimilar from our patient population’s, and studies that focused on special circumstances, such as complications or patient characteristics. Eight studies remained for historical comparison.

Results

Radiographic Outcomes

On the injury radiographs, mean volar tilt was –16.7° (range, 2° to –42°), mean radial inclination was 14.1° (range, –1° to 44°), and mean radial height was 5.3 mm (range, –2 mm to 11 mm). Minor improvement after reduction was noted. All patients had intraoperative or postoperative radiographs with external fixation in place (Figure 3).

On the final (post-fixation removal) radiographs, mean volar tilt was 3.3° (range, –16° to 21°), mean radial inclination was 20.7° (range, 0° to 31°), and mean radial height was 7.5 mm (range, 0 mm to 13 mm). Comparison of the injury and final means revealed correction of ~20° for volar tilt, 6° for radial inclination, and 2 mm for radial height. All but 5 patients had type C fractures (AO classification).

Clinical Outcomes

Eleven patients underwent clinical evaluation (functional assessment, physical examination). Mean DASH score was 11.4 (SD, 10.5; range, 0-27.3), mean Mayo wrist score was 79.0 (SD, 12.2; range, 65-100), and mean patient-rated wrist evaluation was 12.2 (SD, 11.9; range, 0-25.5). There was no statistical difference in DASH scores between this group and the historical control group (Table 3). ROM was measured under active effort. In our group, mean wrist flexion was 69.3° (86% of contralateral side), and mean extension was 64.0° (94%). Mean radial deviation of the wrist was 47.4° (135% of relative normal for patient), and mean ulnar deviation was 29.2° (101%). Mean (SD) pronation was 84.6° (4.7°), and mean (SD) supination was 82.3° (8.5°), or about 100% of contralateral pronosupination.

For each hand, 3 grip strength values and 3 key pinch strength values were obtained. These values were averaged, and the injury and contralateral sides were compared. Mean grip strength was 49.6 pounds (85% of contralateral), and mean key pinch strength was 14.0 pounds (97%).

Complications

Of the 25 patients, 6 (24%) had a pin-tract infection treated with oral antibiotics. One of these infections resulted in the removal of the entire fixator. One (4%) of the 25 patients reported transient hypoesthesia of the dorsal first webspace, and 3 (12%) reported pain at the pin sites.

Although all fractures achieved complete bony union, 1 patient (4%) had a refracture on the same fracture line after a fall within 6 weeks after fixator removal; this refracture was successfully treated with a cast worn for 6 weeks. Of the 3 patients with complete follow-up (27%) who lost reduction with external fixation in place, 2 had radiographic parameters maintained within acceptable limits, and 1 (9%) had a malunion with –16° volar tilt.

Our study patients had no tendon rupture, tendon irritation, or stiffness. By contrast, fixation with volar locking plates has been associated with extensor tendon and flexor tendon injury, flexor pollicis rupture, carpal tunnel syndrome, complex regional pain syndrome, loss of reduction, and hardware failure.¹⁹ Flexor pollicis longus ruptures that occur after volar plate fixation of DRFs are often attributed to plate positioning.^20-22

Discussion

With volar locking plate internal fixation on the rise, CREF has become less widely used.³ This is especially true for comminuted and intra-articular fractures—most earlier external fixators required either spanning of the wrist or limited fixation in the distal articular fragment. Although many studies have found excellent outcomes of ORIF with volar locking plates in the treatment of unstable DRFs,^4,6 few studies have compared volar locking plate ORIF with nonspanning external fixation for unstable comminuted intra-articular DRFs. Both Gradl and colleagues,⁸ using a nonbridging external fixator with multiplanar K-wires, and Mirza and colleagues,9 using the Cross-Pin Fixation system, found wrist function, quality-of-life, and radiographic outcomes similar to those of volar plate fixation in the treatment of DRFs. A comparative meta-analysis by Margaliot and colleagues¹⁷ revealed no superiority of internal fixation over external fixation for unstable DRFs, given the similarity in wrist function, radiographic, and subjective outcomes.

At a mean follow-up of 12.8 months (range, 6-23 months), our retrospective study found that the functional and radiographic outcomes of treating unstable comminuted DRFs with a nonspanning external fixator were similar to those reported in similarly matched control studies. Although followup of >2 years has been shown to be unnecessary,^23-25 small differences may have been detected with interval results over these 2 years. The effect of selection bias on our study results should be considered in light of patients’ involvement in selecting fixation type. Our results parallel those of the temporal studies of Rozental and colleagues5 and Wei and colleagues¹² (Table 2) while allowing for patients to return to function with limited morbidity and complications, similar to Orbay and Fernandez¹⁵ though with a less invasive procedure.

Although we found patient-rated outcome measure values analogous to those of the volar plate fixation group and bridging external fixator group in the study by Wright and colleagues,⁶ we did not measure intra-articular step-off. Another variable not addressed here was operative time. The nonspanning external fixator treatment that we investigated should undergo further study. A randomized prospective study that includes the additional outcome measures of intra-articular step-off and operative time is warranted. 

We found that our study patients, who had their comminuted intra-articular DRFs treated with a nonspanning external fixator, and similar historical control patients, treated with volar locking plate internal fixation, had similar clinical and radiographic outcomes at final follow-up. There was no statistically significant difference in measured outcomes—wrist flexion and extension, radial deviation, pronation and supination, volar tilt, radial height, radial inclination, DASH scores—between the 2 groups. Compared with the historical control group, the external fixator group had significantly more postoperative ulnar deviation.

Given the functional and radiographic outcomes found at final follow-up in this study, we recommend considering a nonspanning external fixator in the treatment of unstable complex comminuted intra-articular DRFs, particularly those that occur in the elderly.

Take-Home Points

• Clinical and radiographic outcomes of patients treated with non-spanning external fixation are comparable to those treated with open reduction and internal volar locked plate fixation.
• Non-spanning external fixation can lead to satisfactory outcomes based on the following features: fragment specific fixation, subchondral support, fixed angle strength, limited dissection, distraction/length adjustment, joint distraction avoidance, and ability to perform early rehabilitation.
• Non-spanning external fixation should be considered as a treatment option for complicated unstable comminuted intra-articular distal radius fractures, specifically in the elderly.

In the United States, distal radius fractures (DRFs) are among the most common fractures, comprising about 15% of all extremity fractures.¹ With a DRF, the primary treatment goal is anatomical reduction with restoration of radiographic parameters and stable fixation of the fracture to restore wrist function.

This fracture type has a variety of treatment alternatives, including nonoperative closed reduction and casting of stable fractures, open reduction and internal fixation (ORIF) with dorsal or volar locking plates, and external fixation. Optimal surgical management of unstable DRFs remains controversial.² Closed reduction with percutaneous pinning or external fixation has become less common with a trend toward using volar locking plates for internal fixation.³

External fixation of DRFs traditionally has involved either spanning or simple nonspanning devices. Spanning fixation is particularly useful in open or highly comminuted fractures with an unstable soft-tissue envelope. In the past, nonspanning external fixation typically was reserved for fractures with a noncomminuted extra-articular distal fragment to which several large pins or Kirschner wires (K-wires) could be secured. The Non-Bridging External Fixator (NBX; Nutek Orthopaedics) may be used in cases that traditionally might be treated with locked plating or fragment-specific fixation. Specifically, this device is indicated for comminuted intra-articular DRFs in which bone quality may be less than ideal. The NBX, also suitable in open fractures with a stable soft-tissue envelope, can restore and maintain articular alignment by providing subchondral support and stability with fragment-specific fixation. A key advantage of this type of external fixation is that it involves percutaneous fixation and allows for early postoperative range of motion (ROM).

Numerous studies have found excellent outcomes of treating unstable DRFs with ORIF with volar locking plates.^4-6 However, few studies have compared the clinical and radiographic outcomes of ORIF with those of nonspanning external fixation in the treatment of unstable comminuted intra-articular DRFs. Windolf and colleagues⁷ found that, in cadaveric unstable intra-articular DRFs, nonspanning external fixation with multiplanar K-wires had biomechanical characteristics comparable to those of volar locking plates. Other suitable DRF treatment options have been found: an alternative nonbridging external fixator with multiplanar K-wires (Gradl and colleagues⁸) and the Cross-Pin Fixation system (A.M. Surgical) (Mirza and colleagues⁹).

We conducted a study to compare functional and radiographic outcomes of unstable comminuted intra-articular DRFs treated with a nonspanning external fixation device (NBX) with outcomes achieved with volar locking plates in a historical control group.

Materials and Methods

This retrospective case-control study was approved by our Institutional Review Board and conducted at 2 institutions. Included in the study were 25 consecutive patients (2 institutions) who underwent closed reduction and external fixation (CREF) with NBX as treatment for unstable DRFs (diagnosis based on radiographic parameters or inability to maintain acceptable alignment after closed reduction and casting). Of these 25 patients, 11 were available for clinical follow-up and medical records review; the other 14 were not available for followup but had their charts reviewed for radiographic data and treatment details. Six of the 14 patients declined to participate in the study, and the other 8 were lost to follow-up because of nonstandardized follow-up protocols. Patients were excluded from the study if their final follow-up had not occurred, or if it occurred before 6 months. For their participation in clinical follow-up, patients received nominal time compensation and mileage reimbursement through a grant from the NBX manufacturer.

The 25 patients underwent CREF with NBX between November 2008 and March 2013. Indications for external fixation consideration were intra-articular extension or significant comminution in patients with poor soft tissue or in patients who wanted to avoid invasive surgery or a permanent implant. Of the 11 patients who agreed to participate in the study, 7 were women and 4 were men; mean age was 64 years (range, 15-81 years). Of the 14 patients unable to follow up, 11 were women and 3 were men; mean age was 63 years (range, 26-89 years). At the last available follow-up, each of the 25 patients was doing well, was satisfied with treatment received and function regained, and had a healed DRF. In almost every case, the mechanism of injury was a fall onto an outstretched hand; most fractures were type C per AO (Arbeitsgemeinschaft für Osteosynthesefragen) classification (Table 1).

The surgical technique for this nonspanning external fixator involves closed reduction with longitudinal traction using ligamentotaxis to grossly align the fracture fragments, with small adjustments made throughout the procedure. A dorsally placed radiolucent fixator is used with fluoroscopic guidance to percutaneously affix a subchondral raft of smooth bicortical .062-inch K-wires. The fixator’s abundant pin holes allow for each specific distal fragment to be captured by pins that are a part of the external fixation construct. Furthermore, radially based pins that use a side bar allow for a “weave” of fixation. Radial length is then obtained and maintained by attaching the distal complex to proximal pins in the radial diaphysis. After pins are cut and wrist and digits are taken through full ROM to ensure smooth tracking, fluoroscopy is used to confirm final fracture fixation and alignment (Figure 1).

In ideal scenarios with good fixation, patients can begin gentle ROM exercises within 1 week after surgery. This regimen can progress to more aggressive motion exercises and even light strengthening (Figure 2).

The 11 clinical follow-up patients underwent directed clinical examination, including ROM and strength evaluation, by Dr. Dwyer and Dr. Crosby. Follow-up also included completion of questionnaires and review of radiographs.

During the clinical follow-up, a standard goniometer was used to evaluate active ROM (wrist flexion and extension and wrist radial and ulnar deviation, measured down the long axis of the forearm and the index ray), and forearm pronation and supination were measured from the 90° elbow flexion position using the humerus as the reference point with the shoulders in 0° of flexion, abduction, and external rotation. In addition, a calibrated dynamometer (Sammons Preston) was used to measure grip strength (position 3) and key pinch strength, and the average of 3 trials of each strength test was calculated. ROM and strength values were calculated as percentages of the contralateral (uninjured) side, as these ratios are more sensitive in detecting clinical changes.¹⁰ A 10% adjustment for dominant hand grip strength in right-handed patients was used for this comparison.¹¹

Union (osseous bridging across fracture site on 2 of 3 views), radial height, radial inclination, and volar tilt were measured on standard posteroanterior and lateral radiographs taken at several points: time of injury, postreduction and/or preoperative, initial postoperative, and final follow-up. All radiographic measurements were independently taken by Dr. Dwyer and Dr. Crosby, who used a digital goniometer and ruler (Siemens Medical Solutions) or, when necessary, manual instruments. Means of the original and independent measurements were used for calculations.

The Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire, the Mayo wrist score, and the patient-rated wrist evaluation were used to assess activities of daily living, pain, and quality of life after surgery. Mayo wrist scores were adjusted for unemployed patients; work status was replaced with return to normal activities. 

Complications of surgical treatment were evaluated. Major complications evaluated were loss of reduction, malunion, nonunion, deep infection, neuropathy, and tendon rupture. Minor complication possibilities were transient extensor tendon irritation, superficial infection, and finger stiffness. Also noted were 1 patient who subsequently required another procedure and 7 patients who were immobilized after external fixation removal.

We compared our study group’s outcomes with those of historical control patients who underwent fixation with internal volar locking plates. The 2 groups had similar demographic characteristics. To obtain the historical controls, we used the key words distal, radi*, volar, and plat* in a PubMed search. From the 169 citations found, we removed biomechanical cadaver studies, studies that focused on patients with demographics and fracture types dissimilar from our patient population’s, and studies that focused on special circumstances, such as complications or patient characteristics. Eight studies remained for historical comparison.

Results

Radiographic Outcomes

On the injury radiographs, mean volar tilt was –16.7° (range, 2° to –42°), mean radial inclination was 14.1° (range, –1° to 44°), and mean radial height was 5.3 mm (range, –2 mm to 11 mm). Minor improvement after reduction was noted. All patients had intraoperative or postoperative radiographs with external fixation in place (Figure 3).

On the final (post-fixation removal) radiographs, mean volar tilt was 3.3° (range, –16° to 21°), mean radial inclination was 20.7° (range, 0° to 31°), and mean radial height was 7.5 mm (range, 0 mm to 13 mm). Comparison of the injury and final means revealed correction of ~20° for volar tilt, 6° for radial inclination, and 2 mm for radial height. All but 5 patients had type C fractures (AO classification).

Clinical Outcomes

Eleven patients underwent clinical evaluation (functional assessment, physical examination). Mean DASH score was 11.4 (SD, 10.5; range, 0-27.3), mean Mayo wrist score was 79.0 (SD, 12.2; range, 65-100), and mean patient-rated wrist evaluation was 12.2 (SD, 11.9; range, 0-25.5). There was no statistical difference in DASH scores between this group and the historical control group (Table 3). ROM was measured under active effort. In our group, mean wrist flexion was 69.3° (86% of contralateral side), and mean extension was 64.0° (94%). Mean radial deviation of the wrist was 47.4° (135% of relative normal for patient), and mean ulnar deviation was 29.2° (101%). Mean (SD) pronation was 84.6° (4.7°), and mean (SD) supination was 82.3° (8.5°), or about 100% of contralateral pronosupination.

For each hand, 3 grip strength values and 3 key pinch strength values were obtained. These values were averaged, and the injury and contralateral sides were compared. Mean grip strength was 49.6 pounds (85% of contralateral), and mean key pinch strength was 14.0 pounds (97%).

Complications

Of the 25 patients, 6 (24%) had a pin-tract infection treated with oral antibiotics. One of these infections resulted in the removal of the entire fixator. One (4%) of the 25 patients reported transient hypoesthesia of the dorsal first webspace, and 3 (12%) reported pain at the pin sites.

Although all fractures achieved complete bony union, 1 patient (4%) had a refracture on the same fracture line after a fall within 6 weeks after fixator removal; this refracture was successfully treated with a cast worn for 6 weeks. Of the 3 patients with complete follow-up (27%) who lost reduction with external fixation in place, 2 had radiographic parameters maintained within acceptable limits, and 1 (9%) had a malunion with –16° volar tilt.

Our study patients had no tendon rupture, tendon irritation, or stiffness. By contrast, fixation with volar locking plates has been associated with extensor tendon and flexor tendon injury, flexor pollicis rupture, carpal tunnel syndrome, complex regional pain syndrome, loss of reduction, and hardware failure.¹⁹ Flexor pollicis longus ruptures that occur after volar plate fixation of DRFs are often attributed to plate positioning.^20-22

Discussion

With volar locking plate internal fixation on the rise, CREF has become less widely used.³ This is especially true for comminuted and intra-articular fractures—most earlier external fixators required either spanning of the wrist or limited fixation in the distal articular fragment. Although many studies have found excellent outcomes of ORIF with volar locking plates in the treatment of unstable DRFs,^4,6 few studies have compared volar locking plate ORIF with nonspanning external fixation for unstable comminuted intra-articular DRFs. Both Gradl and colleagues,⁸ using a nonbridging external fixator with multiplanar K-wires, and Mirza and colleagues,9 using the Cross-Pin Fixation system, found wrist function, quality-of-life, and radiographic outcomes similar to those of volar plate fixation in the treatment of DRFs. A comparative meta-analysis by Margaliot and colleagues¹⁷ revealed no superiority of internal fixation over external fixation for unstable DRFs, given the similarity in wrist function, radiographic, and subjective outcomes.

At a mean follow-up of 12.8 months (range, 6-23 months), our retrospective study found that the functional and radiographic outcomes of treating unstable comminuted DRFs with a nonspanning external fixator were similar to those reported in similarly matched control studies. Although followup of >2 years has been shown to be unnecessary,^23-25 small differences may have been detected with interval results over these 2 years. The effect of selection bias on our study results should be considered in light of patients’ involvement in selecting fixation type. Our results parallel those of the temporal studies of Rozental and colleagues5 and Wei and colleagues¹² (Table 2) while allowing for patients to return to function with limited morbidity and complications, similar to Orbay and Fernandez¹⁵ though with a less invasive procedure.

Although we found patient-rated outcome measure values analogous to those of the volar plate fixation group and bridging external fixator group in the study by Wright and colleagues,⁶ we did not measure intra-articular step-off. Another variable not addressed here was operative time. The nonspanning external fixator treatment that we investigated should undergo further study. A randomized prospective study that includes the additional outcome measures of intra-articular step-off and operative time is warranted. 

We found that our study patients, who had their comminuted intra-articular DRFs treated with a nonspanning external fixator, and similar historical control patients, treated with volar locking plate internal fixation, had similar clinical and radiographic outcomes at final follow-up. There was no statistically significant difference in measured outcomes—wrist flexion and extension, radial deviation, pronation and supination, volar tilt, radial height, radial inclination, DASH scores—between the 2 groups. Compared with the historical control group, the external fixator group had significantly more postoperative ulnar deviation.

Given the functional and radiographic outcomes found at final follow-up in this study, we recommend considering a nonspanning external fixator in the treatment of unstable complex comminuted intra-articular DRFs, particularly those that occur in the elderly.

Take-Home Points

• Clinical and radiographic outcomes of patients treated with non-spanning external fixation are comparable to those treated with open reduction and internal volar locked plate fixation.
• Non-spanning external fixation can lead to satisfactory outcomes based on the following features: fragment specific fixation, subchondral support, fixed angle strength, limited dissection, distraction/length adjustment, joint distraction avoidance, and ability to perform early rehabilitation.
• Non-spanning external fixation should be considered as a treatment option for complicated unstable comminuted intra-articular distal radius fractures, specifically in the elderly.

In the United States, distal radius fractures (DRFs) are among the most common fractures, comprising about 15% of all extremity fractures.¹ With a DRF, the primary treatment goal is anatomical reduction with restoration of radiographic parameters and stable fixation of the fracture to restore wrist function.

This fracture type has a variety of treatment alternatives, including nonoperative closed reduction and casting of stable fractures, open reduction and internal fixation (ORIF) with dorsal or volar locking plates, and external fixation. Optimal surgical management of unstable DRFs remains controversial.² Closed reduction with percutaneous pinning or external fixation has become less common with a trend toward using volar locking plates for internal fixation.³

External fixation of DRFs traditionally has involved either spanning or simple nonspanning devices. Spanning fixation is particularly useful in open or highly comminuted fractures with an unstable soft-tissue envelope. In the past, nonspanning external fixation typically was reserved for fractures with a noncomminuted extra-articular distal fragment to which several large pins or Kirschner wires (K-wires) could be secured. The Non-Bridging External Fixator (NBX; Nutek Orthopaedics) may be used in cases that traditionally might be treated with locked plating or fragment-specific fixation. Specifically, this device is indicated for comminuted intra-articular DRFs in which bone quality may be less than ideal. The NBX, also suitable in open fractures with a stable soft-tissue envelope, can restore and maintain articular alignment by providing subchondral support and stability with fragment-specific fixation. A key advantage of this type of external fixation is that it involves percutaneous fixation and allows for early postoperative range of motion (ROM).

Numerous studies have found excellent outcomes of treating unstable DRFs with ORIF with volar locking plates.^4-6 However, few studies have compared the clinical and radiographic outcomes of ORIF with those of nonspanning external fixation in the treatment of unstable comminuted intra-articular DRFs. Windolf and colleagues⁷ found that, in cadaveric unstable intra-articular DRFs, nonspanning external fixation with multiplanar K-wires had biomechanical characteristics comparable to those of volar locking plates. Other suitable DRF treatment options have been found: an alternative nonbridging external fixator with multiplanar K-wires (Gradl and colleagues⁸) and the Cross-Pin Fixation system (A.M. Surgical) (Mirza and colleagues⁹).

We conducted a study to compare functional and radiographic outcomes of unstable comminuted intra-articular DRFs treated with a nonspanning external fixation device (NBX) with outcomes achieved with volar locking plates in a historical control group.

Materials and Methods

This retrospective case-control study was approved by our Institutional Review Board and conducted at 2 institutions. Included in the study were 25 consecutive patients (2 institutions) who underwent closed reduction and external fixation (CREF) with NBX as treatment for unstable DRFs (diagnosis based on radiographic parameters or inability to maintain acceptable alignment after closed reduction and casting). Of these 25 patients, 11 were available for clinical follow-up and medical records review; the other 14 were not available for followup but had their charts reviewed for radiographic data and treatment details. Six of the 14 patients declined to participate in the study, and the other 8 were lost to follow-up because of nonstandardized follow-up protocols. Patients were excluded from the study if their final follow-up had not occurred, or if it occurred before 6 months. For their participation in clinical follow-up, patients received nominal time compensation and mileage reimbursement through a grant from the NBX manufacturer.

The 25 patients underwent CREF with NBX between November 2008 and March 2013. Indications for external fixation consideration were intra-articular extension or significant comminution in patients with poor soft tissue or in patients who wanted to avoid invasive surgery or a permanent implant. Of the 11 patients who agreed to participate in the study, 7 were women and 4 were men; mean age was 64 years (range, 15-81 years). Of the 14 patients unable to follow up, 11 were women and 3 were men; mean age was 63 years (range, 26-89 years). At the last available follow-up, each of the 25 patients was doing well, was satisfied with treatment received and function regained, and had a healed DRF. In almost every case, the mechanism of injury was a fall onto an outstretched hand; most fractures were type C per AO (Arbeitsgemeinschaft für Osteosynthesefragen) classification (Table 1).

The surgical technique for this nonspanning external fixator involves closed reduction with longitudinal traction using ligamentotaxis to grossly align the fracture fragments, with small adjustments made throughout the procedure. A dorsally placed radiolucent fixator is used with fluoroscopic guidance to percutaneously affix a subchondral raft of smooth bicortical .062-inch K-wires. The fixator’s abundant pin holes allow for each specific distal fragment to be captured by pins that are a part of the external fixation construct. Furthermore, radially based pins that use a side bar allow for a “weave” of fixation. Radial length is then obtained and maintained by attaching the distal complex to proximal pins in the radial diaphysis. After pins are cut and wrist and digits are taken through full ROM to ensure smooth tracking, fluoroscopy is used to confirm final fracture fixation and alignment (Figure 1).

In ideal scenarios with good fixation, patients can begin gentle ROM exercises within 1 week after surgery. This regimen can progress to more aggressive motion exercises and even light strengthening (Figure 2).

The 11 clinical follow-up patients underwent directed clinical examination, including ROM and strength evaluation, by Dr. Dwyer and Dr. Crosby. Follow-up also included completion of questionnaires and review of radiographs.

During the clinical follow-up, a standard goniometer was used to evaluate active ROM (wrist flexion and extension and wrist radial and ulnar deviation, measured down the long axis of the forearm and the index ray), and forearm pronation and supination were measured from the 90° elbow flexion position using the humerus as the reference point with the shoulders in 0° of flexion, abduction, and external rotation. In addition, a calibrated dynamometer (Sammons Preston) was used to measure grip strength (position 3) and key pinch strength, and the average of 3 trials of each strength test was calculated. ROM and strength values were calculated as percentages of the contralateral (uninjured) side, as these ratios are more sensitive in detecting clinical changes.¹⁰ A 10% adjustment for dominant hand grip strength in right-handed patients was used for this comparison.¹¹

Union (osseous bridging across fracture site on 2 of 3 views), radial height, radial inclination, and volar tilt were measured on standard posteroanterior and lateral radiographs taken at several points: time of injury, postreduction and/or preoperative, initial postoperative, and final follow-up. All radiographic measurements were independently taken by Dr. Dwyer and Dr. Crosby, who used a digital goniometer and ruler (Siemens Medical Solutions) or, when necessary, manual instruments. Means of the original and independent measurements were used for calculations.

The Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire, the Mayo wrist score, and the patient-rated wrist evaluation were used to assess activities of daily living, pain, and quality of life after surgery. Mayo wrist scores were adjusted for unemployed patients; work status was replaced with return to normal activities. 

Complications of surgical treatment were evaluated. Major complications evaluated were loss of reduction, malunion, nonunion, deep infection, neuropathy, and tendon rupture. Minor complication possibilities were transient extensor tendon irritation, superficial infection, and finger stiffness. Also noted were 1 patient who subsequently required another procedure and 7 patients who were immobilized after external fixation removal.

We compared our study group’s outcomes with those of historical control patients who underwent fixation with internal volar locking plates. The 2 groups had similar demographic characteristics. To obtain the historical controls, we used the key words distal, radi*, volar, and plat* in a PubMed search. From the 169 citations found, we removed biomechanical cadaver studies, studies that focused on patients with demographics and fracture types dissimilar from our patient population’s, and studies that focused on special circumstances, such as complications or patient characteristics. Eight studies remained for historical comparison.

Results

Radiographic Outcomes

On the injury radiographs, mean volar tilt was –16.7° (range, 2° to –42°), mean radial inclination was 14.1° (range, –1° to 44°), and mean radial height was 5.3 mm (range, –2 mm to 11 mm). Minor improvement after reduction was noted. All patients had intraoperative or postoperative radiographs with external fixation in place (Figure 3).

On the final (post-fixation removal) radiographs, mean volar tilt was 3.3° (range, –16° to 21°), mean radial inclination was 20.7° (range, 0° to 31°), and mean radial height was 7.5 mm (range, 0 mm to 13 mm). Comparison of the injury and final means revealed correction of ~20° for volar tilt, 6° for radial inclination, and 2 mm for radial height. All but 5 patients had type C fractures (AO classification).

Clinical Outcomes

Eleven patients underwent clinical evaluation (functional assessment, physical examination). Mean DASH score was 11.4 (SD, 10.5; range, 0-27.3), mean Mayo wrist score was 79.0 (SD, 12.2; range, 65-100), and mean patient-rated wrist evaluation was 12.2 (SD, 11.9; range, 0-25.5). There was no statistical difference in DASH scores between this group and the historical control group (Table 3). ROM was measured under active effort. In our group, mean wrist flexion was 69.3° (86% of contralateral side), and mean extension was 64.0° (94%). Mean radial deviation of the wrist was 47.4° (135% of relative normal for patient), and mean ulnar deviation was 29.2° (101%). Mean (SD) pronation was 84.6° (4.7°), and mean (SD) supination was 82.3° (8.5°), or about 100% of contralateral pronosupination.

For each hand, 3 grip strength values and 3 key pinch strength values were obtained. These values were averaged, and the injury and contralateral sides were compared. Mean grip strength was 49.6 pounds (85% of contralateral), and mean key pinch strength was 14.0 pounds (97%).

Complications

Of the 25 patients, 6 (24%) had a pin-tract infection treated with oral antibiotics. One of these infections resulted in the removal of the entire fixator. One (4%) of the 25 patients reported transient hypoesthesia of the dorsal first webspace, and 3 (12%) reported pain at the pin sites.

Although all fractures achieved complete bony union, 1 patient (4%) had a refracture on the same fracture line after a fall within 6 weeks after fixator removal; this refracture was successfully treated with a cast worn for 6 weeks. Of the 3 patients with complete follow-up (27%) who lost reduction with external fixation in place, 2 had radiographic parameters maintained within acceptable limits, and 1 (9%) had a malunion with –16° volar tilt.

Our study patients had no tendon rupture, tendon irritation, or stiffness. By contrast, fixation with volar locking plates has been associated with extensor tendon and flexor tendon injury, flexor pollicis rupture, carpal tunnel syndrome, complex regional pain syndrome, loss of reduction, and hardware failure.¹⁹ Flexor pollicis longus ruptures that occur after volar plate fixation of DRFs are often attributed to plate positioning.^20-22

Discussion

With volar locking plate internal fixation on the rise, CREF has become less widely used.³ This is especially true for comminuted and intra-articular fractures—most earlier external fixators required either spanning of the wrist or limited fixation in the distal articular fragment. Although many studies have found excellent outcomes of ORIF with volar locking plates in the treatment of unstable DRFs,^4,6 few studies have compared volar locking plate ORIF with nonspanning external fixation for unstable comminuted intra-articular DRFs. Both Gradl and colleagues,⁸ using a nonbridging external fixator with multiplanar K-wires, and Mirza and colleagues,9 using the Cross-Pin Fixation system, found wrist function, quality-of-life, and radiographic outcomes similar to those of volar plate fixation in the treatment of DRFs. A comparative meta-analysis by Margaliot and colleagues¹⁷ revealed no superiority of internal fixation over external fixation for unstable DRFs, given the similarity in wrist function, radiographic, and subjective outcomes.

At a mean follow-up of 12.8 months (range, 6-23 months), our retrospective study found that the functional and radiographic outcomes of treating unstable comminuted DRFs with a nonspanning external fixator were similar to those reported in similarly matched control studies. Although followup of >2 years has been shown to be unnecessary,^23-25 small differences may have been detected with interval results over these 2 years. The effect of selection bias on our study results should be considered in light of patients’ involvement in selecting fixation type. Our results parallel those of the temporal studies of Rozental and colleagues5 and Wei and colleagues¹² (Table 2) while allowing for patients to return to function with limited morbidity and complications, similar to Orbay and Fernandez¹⁵ though with a less invasive procedure.

Although we found patient-rated outcome measure values analogous to those of the volar plate fixation group and bridging external fixator group in the study by Wright and colleagues,⁶ we did not measure intra-articular step-off. Another variable not addressed here was operative time. The nonspanning external fixator treatment that we investigated should undergo further study. A randomized prospective study that includes the additional outcome measures of intra-articular step-off and operative time is warranted. 

We found that our study patients, who had their comminuted intra-articular DRFs treated with a nonspanning external fixator, and similar historical control patients, treated with volar locking plate internal fixation, had similar clinical and radiographic outcomes at final follow-up. There was no statistically significant difference in measured outcomes—wrist flexion and extension, radial deviation, pronation and supination, volar tilt, radial height, radial inclination, DASH scores—between the 2 groups. Compared with the historical control group, the external fixator group had significantly more postoperative ulnar deviation.

Given the functional and radiographic outcomes found at final follow-up in this study, we recommend considering a nonspanning external fixator in the treatment of unstable complex comminuted intra-articular DRFs, particularly those that occur in the elderly.

For more than a century, medical rounds have been a cornerstone of patient care and medical education in teaching hospitals. They remain critical activities for exposing generations of trainees to clinical decision making, coordination of care, and patient communication.¹

Despite this established importance within medical education and patient care, there is a relative paucity of research addressing the purpose of medical rounds in the 21st century. Medicine has evolved significantly since Osler’s day, and it is unclear whether the purpose of rounds has evolved along with it. Rounds, to Osler, were an important opportunity for future physicians to learn at the bedside from an attending physician. Increased duty hour restrictions, mandatory adoption of electronic medical records, and increasingly complex care have changed how rounds are performed, making it more difficult to achieve Osler’s ideals.^2,3 While several studies have aimed to quantify the changes to rounds and have demonstrated a significant decline in bedside teaching,^4-6 few studies have explored the purpose of rounds from the perspective of pertinent stakeholders, students, residents, and faculty. The authors have published the results of focus groups of resident stakeholders recently.⁷ We made the decision to combine the student/faculty data and describe it separately from the resident data to allow the most accurate and relevant discussion as it pertained to each group.

The aim of this study was to explore the perceptions of faculty and students of general inpatient rounds on internal medicine and pediatric rotations, and to identify any notable differences between these key stakeholders.

Between April 2014 and June 2014, we conducted 10 semistructured focus groups at 4 teaching hospitals: The University of Chicago Medical Center, Children’s National Health System, Georgetown University Medical Center, and the University of California, San Francisco Medical Center. A sample of eligible 3rd-year medical students and residents on pediatrics and internal medicine hospitalist services as well as hospitalist attendings in pediatrics and internal medicine were invited by e-mail to participate voluntarily without compensation. Identical semistructured focus groups were also conducted with pediatric and internal medicine interns (postgraduate year [PGY1]) and senior residents (PGY2 and PGY3), and those data have been published previously.⁷

Data Collection

Most focus groups had 6 to 8 participants, with 2 groups of 3 and 4. The groups were interviewed separately by training and specialty: 3rd-year medical students who had completed internal medicine and/or pediatrics rotations, hospitalist attendings in pediatrics, and hospitalist attendings in internal medicine. Attendings with training in medicine-pediatrics were included in the department in which they worked most frequently. The focus group script was informed by a literature review and expert input, and we used open-ended questions to explore perspectives on current and ideal purposes of rounds. Interviews were digitally recorded, transcribed, and names of speakers or references to specific patients were removed to preserve confidentiality and anonymity. The focus groups lasted between 30 and 60 minutes. The author (OH) conducted focus groups at 1 site, and trained facilitators conducted focus groups at the remaining 3 sites. The protocol was determined to be exempt by the institutional review boards at all participating sites. Prior to the focus groups, the definition of family-centered rounds was read aloud; after which, participants were asked to fill out a demographic survey.

Data Analysis

The authors employed a grounded theory approach to data collection and analysis,⁸ and data were analyzed by using the constant-comparative method.⁹ There was no a priori hypothesis. Four transcripts were independently reviewed by 2 authors (OH and RR) by using sentences and phrases as the units of data, which were coded with an identifier. The authors discussed initial codes and resolved discrepancies through deliberation and consensus to create codebooks. Themes, made up of multiple codes, were identified inductively and iteratively and were refined to reflect the evolving dataset. One author (OH) independently coded the remaining transcripts by using a revised codebook as a guide. A faculty author (JF) assessed the interrater reliability of the final codebook by reviewing 2 previously coded, randomly selected transcripts with no new codes emerging in the process, with a kappa coefficient of >0.8 indicating significant agreement.

What Do You Perceive the Purpose of Rounds to Be?

With respect to this prompt, we identified 4 themes, which represent 16 codes describing what attendings and medical students believed to be the purpose of rounds (Table 2). These themes are communication, medical education, patient care, and assessment.

Communication

Communication includes all comments addressing the role of rounds as it relates to communication between team members, patients, family members, and all those involved in patient care. There were 4 main codes, including coordination of patient care team, patient/family communication, establishing rapport with patients and/or family, and establishment of roles.

Coordination of patient care team identified rounds as a time “to make sure everyone is on the same page” and “to come together whenever possible,” so that everyone “had the same information of what was going on.” It also included comments related to interdisciplinary communication, with 1 participant describing rounds as “a time when your consulting team, or people with outside expertise, can weigh in on some medical issues.”

Medical Education

The theme of medical education is made up of 6 codes that encompass comments related to teaching and learning during rounds. These 6 codes include delivery of clinical education, exposure to clinical decision making, role modeling, student presentations, establishment of trainee autonomy, and providing a safe learning environment.

Delivery of clinical education included comments identifying rounds as a time for didactic teaching, teachable moments, “clinical pearls,” and bedside teaching of physical exam skills. Exposure to clinical decision making included comments by both medical students and attendings who described the purpose of rounds as a time for learning and teaching, specifically about how best to approach problems and decision making in a systematic manner, with 1 medical student explaining it as a time to “expose [trainees] to the way that people think about problems and how they decided to go about addressing them.”

Role modeling includes comments addressing rounds as a time for attendings to demonstrate appropriate behaviors and skills to trainees. One attending explained that “everybody learns from watching other people present and interact…so everybody has a chance to pick up things that they think, ‘Oh, this works well.’” Student presentations include comments, predominantly from students, that described rounds as an opportunity to practice presentations and receive feedback, with 1 student explaining it was a time “to learn how to present but also to be questioned and challenged.”

Establishing trainee autonomy is a code that identifies rounds as a time to encourage resident and student autonomy in order to achieve rounds that function with minimal input from the attending, with 1 attending describing how they “put resident leadership first as far as priorities… [and] fostering that because I usually let them decide what we’re going to do.”

Providing a safe learning environment identifies the purpose of rounds as being a space in which trainees can feel comfortable learning from their mistakes. One student described rounds as, “…a setting where it’s okay to be wrong and feel comfortable enough to know that it’s about a learning process.”

Assessment

Assessment is a theme composed of comments identifying the purpose of rounds as being related to observation, assessment, and feedback, and it includes 2 codes: attending observation, assessment, and feedback and establishment of expectations. Attending observation, assessment, and feedback includes comments from attendings and students alike who described rounds as a place for observation, evaluation, and provision of feedback regarding the skills and abilities of trainees. One attending explained that rounds gave him an “opportunity to observe trainees interacting with each other, with the patient, the patient’s family, and ancillary staff,” with another commenting it was time used “to assess how med students are gathering information, presenting information, and eventually their assessment and plan.” Establishment of expectations captures comments that describe rounds as a time for the establishment of expectations and goals of the team.

Patient Care

Patient care is a theme comprised of comments identifying the purpose of rounds as being directly related to the formation and delivery of the patient care plan, and it includes 2 codes: formation of the patient care plan and delivery of patient care. Formation of the patient care plan includes comments, which identified rounds as a time for discussing and forming the plan for the day, with an attending stating, “The purpose [of rounds] was to make a plan, a treatment plan, and to include the parents in making the treatment plan.” Delivery of patient care included comments identifying rounds as a means of ensuring timely, safe, and appropriate delivery of patient care occurred. One attending explained, “It can’t be undersold that the priority of rounds is patient care and the more eyes that look over information the less likely there are to be mistakes.”

What Do You Believe the Ideal Purpose of RoundsShould Be?

This study originally sought to compare responses to 2 different questions: “What do you perceive the purpose of rounds to be?” and “What do you believe the ideal purpose of rounds should be?” What became clear during the focus groups was that these were often interpreted to be the same question, and as such, responses to the latter question were truncated or were reiterations of what was previously said: “I think we’ve already discussed that, I think it’s no different than what we already kind of said, patient care, education, and communication,” explained 1 attending. Fifty-four responses to the question regarding the ideal purpose of rounds were coded and did not differ significantly from the previously noted results in terms of the domains represented and the frequency of representation.

Variation Among Respondents

Overall, there is a high level of concordance between the comments from medical students and attendings regarding the purpose of rounds, particularly in the medical education theme. However, medicine and pediatric attendings differ in their comments relating to the theme of communication, with 2 codes primarily accounting for this difference: pediatric attendings place more emphasis on time for patient/family communication and establishing rapport with patients than their internal medicine colleagues. Of note, all of the pediatric attendings involved in the study answered that they conducted family-centered rounds (FCR), compared with 22% of internal medicine attendings.¹⁰

Another notable discrepancy came up during focus groups involving comments from medical students who reiterated that the purpose of rounds was not fixed, but rather dependent on the attending that was running rounds. This theme was only identified in focus groups involving medical students. One student explained, “I think that it depends on the attending and if they actually want to teach,” and another commented that “it’s incredibly dependent on what the attending… is willing to invest.” No attendings identified student or attending variability as an important factor influencing the purpose of rounds.

This qualitative study is one of the first to explore the purpose of rounds from the perspective of both medical students and attendings. Reassuringly, our results indicate that medical student and attending perceptions are largely concordant. The 4 themes of communication, medical education, assessment, and patient care are in line with the findings of previous observational studies of internal medicine and pediatrics rounds.^1,11 The themes are similar to the findings of resident focus groups done at these same sites.⁷

Our results support that both medical students and attendings identify the importance of medical education during rounds. This is in contrast with findings in previous observational time-motion research by Stickrath that describes the focus on patient care related activities and the relative scarcity of education during rounds.¹ This stresses a divide between how medical students and attendings define the purpose of rounds and what other research suggests actually occurs on rounds. This distinction is an important one. It is possible that the way we, and others, define “medical education” and “patient care” may be at least partially responsible for these findings. This is supported by the ambiguous distinction between formal and informal educational activities on rounds and the challenges in characterizing the hidden curriculum and its role in medical student and resident education.¹¹ Attendings role modeling effective patient communication strategies, for example, highlights that patient care, medical education, and communication are frequently indistinguishable.¹² This hybridization of activities and dedication to diverse types of learning is an essential quality of rounds and is suggestive of why they have survived as a preeminent tool within the arsenal of medical education for the past century.

Yet, this finding does not excuse or adequately explain a well-documented disappearance of more formal educational activities during rounds. Recent observational studies have shown that the percentage of rounds dedicated to educational activities fell from 25% to 10% after the implementation of duty hour restrictions,^1,13,14 and a recent ethnographic study of pediatric attending rounds confirmed teaching during rounds, though seen as a pedagogical ideal, occurred infrequently and inconsistently in large part because of time pressures.¹⁵ In our attending focus groups, duty hours and time pressures were frequently cited as actively working against the purpose of rounds, specifically opportunities for teaching, with 1 attending explaining, “I just don’t think we achieve our [teaching] goals like we used to.” Another attending mentioned that, because of time pressures, “I often find myself apologizing. ‘I’m so sorry. I can’t resist. Can I just tell you this one thing? I’m so sorry to do teaching.’” This tension between time pressures and education on rounds is well documented in the literature.^4,16,17

Our results highlight that attendings and medical students still believe that medical education is a primary and important purpose of rounds even in the face of increasing time pressures. As such, efforts should be made to better align the many purposes of rounds with the realities of the modern day rounding environment. Increasing the presence of medical education on rounds need not be at the expense of time given that techniques like the 1-minute preceptor have been rated as both efficient and effective methods of teaching and delivering feedback.¹⁸ This is echoed in research that has found that faculty development with a focus on teaching significantly increased the rate of clinical education and interdisciplinary communication during rounds.¹ Opportunities for faculty development are increasingly accessible,¹⁹ including programs like the Advancing Pediatric Excellence Teaching Program, sponsored by the American Academy of Pediatrics Section on Hospital Medicine and the Academic Pediatric Association, and the Teaching Educators Across the Continuum of Healthcare program, sponsored by the Society for General Internal Medicine.^20,21

A testament to the adaptability of rounds can be seen in our findings that expose the increased emphasis with which pediatric attendings identify communication as a purpose of rounds, particularly within the themes of patient/family communication and establishing rapport with patients. This is likely due to the practice of FCR by 100% of the pediatric attendings in our focus groups, and is supported elsewhere in the literature.²² A key to family-centered rounds is communication, with active participation in the care discussion by patients and families as described and endorsed by a 2012 American Academy of Pediatrics (AAP) policy.^10,23

This emphasis could explain the increased frequency of comments made by pediatric attendings within the themes of patient/family communication and establishing rapport with patients. Furthermore, the AAP policy statement stresses the need to share information in a way that patients and families “effectively participate in care and decision making,” which could explain why pediatric attendings placed greater emphasis on the formation of the patient care plan in the theme of patient care.

As noted, the authors published a related study focusing on resident perceptions regarding the purpose of rounds. We initially undertook a separate analysis of the 3 groups: faculty, residents, and medical students. From that analysis, it became apparent that residents (PGY1-PGY3) viewed rounds differently than faculty and medical students. Where faculty and medical students were more focused on communication and medical education, the residents were more focused on the practical aspects of rounds (eg, “getting work done”). It was also noted that the residents’ focus aligned with the graduate medical education milestones, and framing the results within the milestones made the interpretation far more robust. In addition, the residents discussed their difficulties with patient and family involvement, especially in the context of family centered rounds, which is a topic that was rarely discussed by attendings or medical students.

Our study has a number of limitations. Only 4 university-based hospitals were included in the focus groups. This has the potential to limit the generalizability to the community hospital setting. Within the focus groups, the number of participants varied, and this may have had an impact on the flow and content of conversation. Facilitators were chosen to minimize potential bias and prior relationships with participants; however, this was not always possible, and as such, may have influenced responses. There may be a discrepancy between how people perceive rounds and how rounds actually function. Rounds were not standardized between institutions, departments, or attendings.

Rounds are an appropriate metaphor for medical education at large: they are time consuming, complex, and vary in quality, but are nevertheless essential to the goals of patients and learners alike because of their adaptability and hybridization of purpose. Our results highlight that rounds serve 4 critical purposes, including communication, medical education, patient care, and assessment. Importantly, both attendings and students agree on what they perceive to be the many purposes of rounds. Despite this agreement, a disconnect appears to exist between what people believe are the purposes of rounds and what is perceived to be happening during rounds. The causes of this gap are not well defined, and further efforts should be made to better understand the obstacles facing effective rounding. To improve rounds and adapt them to the needs of 21st century learners, it is critical that we better define the scope of medical education, both formal and informal, that occurs during rounds. In doing so, it will be possible to identify areas of development and training for faculty, residents, and medical students, which will ensure that rounds remain useful and critical tools for the development and education of future physicians.

Acknowledgments

The authors would like to acknowledge the following people who assisted on this project: Meghan Daly from The University of Chicago Pritzker School of Medicine, Shannon Martin, MD, MS, Assistant Professor of Medicine from the Department of Medicine at The University of Chicago, Joyce Campbell, BSN, MS, Senior Quality Manager at the Children’s National Medical Center, Benjamin Colburn from the University of California, San Francisco School of Medicine, Kelly Sanders from the University of California, San Francisco School of Medicine, and Alekist Quach from the University of California, San Francisco School of Medicine.

Disclosure 

The authors report no external funding source for this study. The authors declare no conflict of interest. The protocol was approved by the institutional review board at all participating institutions.

For more than a century, medical rounds have been a cornerstone of patient care and medical education in teaching hospitals. They remain critical activities for exposing generations of trainees to clinical decision making, coordination of care, and patient communication.¹

Despite this established importance within medical education and patient care, there is a relative paucity of research addressing the purpose of medical rounds in the 21st century. Medicine has evolved significantly since Osler’s day, and it is unclear whether the purpose of rounds has evolved along with it. Rounds, to Osler, were an important opportunity for future physicians to learn at the bedside from an attending physician. Increased duty hour restrictions, mandatory adoption of electronic medical records, and increasingly complex care have changed how rounds are performed, making it more difficult to achieve Osler’s ideals.^2,3 While several studies have aimed to quantify the changes to rounds and have demonstrated a significant decline in bedside teaching,^4-6 few studies have explored the purpose of rounds from the perspective of pertinent stakeholders, students, residents, and faculty. The authors have published the results of focus groups of resident stakeholders recently.⁷ We made the decision to combine the student/faculty data and describe it separately from the resident data to allow the most accurate and relevant discussion as it pertained to each group.

The aim of this study was to explore the perceptions of faculty and students of general inpatient rounds on internal medicine and pediatric rotations, and to identify any notable differences between these key stakeholders.

Between April 2014 and June 2014, we conducted 10 semistructured focus groups at 4 teaching hospitals: The University of Chicago Medical Center, Children’s National Health System, Georgetown University Medical Center, and the University of California, San Francisco Medical Center. A sample of eligible 3rd-year medical students and residents on pediatrics and internal medicine hospitalist services as well as hospitalist attendings in pediatrics and internal medicine were invited by e-mail to participate voluntarily without compensation. Identical semistructured focus groups were also conducted with pediatric and internal medicine interns (postgraduate year [PGY1]) and senior residents (PGY2 and PGY3), and those data have been published previously.⁷

Data Collection

Most focus groups had 6 to 8 participants, with 2 groups of 3 and 4. The groups were interviewed separately by training and specialty: 3rd-year medical students who had completed internal medicine and/or pediatrics rotations, hospitalist attendings in pediatrics, and hospitalist attendings in internal medicine. Attendings with training in medicine-pediatrics were included in the department in which they worked most frequently. The focus group script was informed by a literature review and expert input, and we used open-ended questions to explore perspectives on current and ideal purposes of rounds. Interviews were digitally recorded, transcribed, and names of speakers or references to specific patients were removed to preserve confidentiality and anonymity. The focus groups lasted between 30 and 60 minutes. The author (OH) conducted focus groups at 1 site, and trained facilitators conducted focus groups at the remaining 3 sites. The protocol was determined to be exempt by the institutional review boards at all participating sites. Prior to the focus groups, the definition of family-centered rounds was read aloud; after which, participants were asked to fill out a demographic survey.

Data Analysis

The authors employed a grounded theory approach to data collection and analysis,⁸ and data were analyzed by using the constant-comparative method.⁹ There was no a priori hypothesis. Four transcripts were independently reviewed by 2 authors (OH and RR) by using sentences and phrases as the units of data, which were coded with an identifier. The authors discussed initial codes and resolved discrepancies through deliberation and consensus to create codebooks. Themes, made up of multiple codes, were identified inductively and iteratively and were refined to reflect the evolving dataset. One author (OH) independently coded the remaining transcripts by using a revised codebook as a guide. A faculty author (JF) assessed the interrater reliability of the final codebook by reviewing 2 previously coded, randomly selected transcripts with no new codes emerging in the process, with a kappa coefficient of >0.8 indicating significant agreement.

What Do You Perceive the Purpose of Rounds to Be?

With respect to this prompt, we identified 4 themes, which represent 16 codes describing what attendings and medical students believed to be the purpose of rounds (Table 2). These themes are communication, medical education, patient care, and assessment.

Communication

Communication includes all comments addressing the role of rounds as it relates to communication between team members, patients, family members, and all those involved in patient care. There were 4 main codes, including coordination of patient care team, patient/family communication, establishing rapport with patients and/or family, and establishment of roles.

Coordination of patient care team identified rounds as a time “to make sure everyone is on the same page” and “to come together whenever possible,” so that everyone “had the same information of what was going on.” It also included comments related to interdisciplinary communication, with 1 participant describing rounds as “a time when your consulting team, or people with outside expertise, can weigh in on some medical issues.”

Medical Education

The theme of medical education is made up of 6 codes that encompass comments related to teaching and learning during rounds. These 6 codes include delivery of clinical education, exposure to clinical decision making, role modeling, student presentations, establishment of trainee autonomy, and providing a safe learning environment.

Delivery of clinical education included comments identifying rounds as a time for didactic teaching, teachable moments, “clinical pearls,” and bedside teaching of physical exam skills. Exposure to clinical decision making included comments by both medical students and attendings who described the purpose of rounds as a time for learning and teaching, specifically about how best to approach problems and decision making in a systematic manner, with 1 medical student explaining it as a time to “expose [trainees] to the way that people think about problems and how they decided to go about addressing them.”

Role modeling includes comments addressing rounds as a time for attendings to demonstrate appropriate behaviors and skills to trainees. One attending explained that “everybody learns from watching other people present and interact…so everybody has a chance to pick up things that they think, ‘Oh, this works well.’” Student presentations include comments, predominantly from students, that described rounds as an opportunity to practice presentations and receive feedback, with 1 student explaining it was a time “to learn how to present but also to be questioned and challenged.”

Establishing trainee autonomy is a code that identifies rounds as a time to encourage resident and student autonomy in order to achieve rounds that function with minimal input from the attending, with 1 attending describing how they “put resident leadership first as far as priorities… [and] fostering that because I usually let them decide what we’re going to do.”

Providing a safe learning environment identifies the purpose of rounds as being a space in which trainees can feel comfortable learning from their mistakes. One student described rounds as, “…a setting where it’s okay to be wrong and feel comfortable enough to know that it’s about a learning process.”

Assessment

Assessment is a theme composed of comments identifying the purpose of rounds as being related to observation, assessment, and feedback, and it includes 2 codes: attending observation, assessment, and feedback and establishment of expectations. Attending observation, assessment, and feedback includes comments from attendings and students alike who described rounds as a place for observation, evaluation, and provision of feedback regarding the skills and abilities of trainees. One attending explained that rounds gave him an “opportunity to observe trainees interacting with each other, with the patient, the patient’s family, and ancillary staff,” with another commenting it was time used “to assess how med students are gathering information, presenting information, and eventually their assessment and plan.” Establishment of expectations captures comments that describe rounds as a time for the establishment of expectations and goals of the team.

Patient Care

Patient care is a theme comprised of comments identifying the purpose of rounds as being directly related to the formation and delivery of the patient care plan, and it includes 2 codes: formation of the patient care plan and delivery of patient care. Formation of the patient care plan includes comments, which identified rounds as a time for discussing and forming the plan for the day, with an attending stating, “The purpose [of rounds] was to make a plan, a treatment plan, and to include the parents in making the treatment plan.” Delivery of patient care included comments identifying rounds as a means of ensuring timely, safe, and appropriate delivery of patient care occurred. One attending explained, “It can’t be undersold that the priority of rounds is patient care and the more eyes that look over information the less likely there are to be mistakes.”

What Do You Believe the Ideal Purpose of RoundsShould Be?

This study originally sought to compare responses to 2 different questions: “What do you perceive the purpose of rounds to be?” and “What do you believe the ideal purpose of rounds should be?” What became clear during the focus groups was that these were often interpreted to be the same question, and as such, responses to the latter question were truncated or were reiterations of what was previously said: “I think we’ve already discussed that, I think it’s no different than what we already kind of said, patient care, education, and communication,” explained 1 attending. Fifty-four responses to the question regarding the ideal purpose of rounds were coded and did not differ significantly from the previously noted results in terms of the domains represented and the frequency of representation.

Variation Among Respondents

Overall, there is a high level of concordance between the comments from medical students and attendings regarding the purpose of rounds, particularly in the medical education theme. However, medicine and pediatric attendings differ in their comments relating to the theme of communication, with 2 codes primarily accounting for this difference: pediatric attendings place more emphasis on time for patient/family communication and establishing rapport with patients than their internal medicine colleagues. Of note, all of the pediatric attendings involved in the study answered that they conducted family-centered rounds (FCR), compared with 22% of internal medicine attendings.¹⁰

Another notable discrepancy came up during focus groups involving comments from medical students who reiterated that the purpose of rounds was not fixed, but rather dependent on the attending that was running rounds. This theme was only identified in focus groups involving medical students. One student explained, “I think that it depends on the attending and if they actually want to teach,” and another commented that “it’s incredibly dependent on what the attending… is willing to invest.” No attendings identified student or attending variability as an important factor influencing the purpose of rounds.

This qualitative study is one of the first to explore the purpose of rounds from the perspective of both medical students and attendings. Reassuringly, our results indicate that medical student and attending perceptions are largely concordant. The 4 themes of communication, medical education, assessment, and patient care are in line with the findings of previous observational studies of internal medicine and pediatrics rounds.^1,11 The themes are similar to the findings of resident focus groups done at these same sites.⁷

Our results support that both medical students and attendings identify the importance of medical education during rounds. This is in contrast with findings in previous observational time-motion research by Stickrath that describes the focus on patient care related activities and the relative scarcity of education during rounds.¹ This stresses a divide between how medical students and attendings define the purpose of rounds and what other research suggests actually occurs on rounds. This distinction is an important one. It is possible that the way we, and others, define “medical education” and “patient care” may be at least partially responsible for these findings. This is supported by the ambiguous distinction between formal and informal educational activities on rounds and the challenges in characterizing the hidden curriculum and its role in medical student and resident education.¹¹ Attendings role modeling effective patient communication strategies, for example, highlights that patient care, medical education, and communication are frequently indistinguishable.¹² This hybridization of activities and dedication to diverse types of learning is an essential quality of rounds and is suggestive of why they have survived as a preeminent tool within the arsenal of medical education for the past century.

Yet, this finding does not excuse or adequately explain a well-documented disappearance of more formal educational activities during rounds. Recent observational studies have shown that the percentage of rounds dedicated to educational activities fell from 25% to 10% after the implementation of duty hour restrictions,^1,13,14 and a recent ethnographic study of pediatric attending rounds confirmed teaching during rounds, though seen as a pedagogical ideal, occurred infrequently and inconsistently in large part because of time pressures.¹⁵ In our attending focus groups, duty hours and time pressures were frequently cited as actively working against the purpose of rounds, specifically opportunities for teaching, with 1 attending explaining, “I just don’t think we achieve our [teaching] goals like we used to.” Another attending mentioned that, because of time pressures, “I often find myself apologizing. ‘I’m so sorry. I can’t resist. Can I just tell you this one thing? I’m so sorry to do teaching.’” This tension between time pressures and education on rounds is well documented in the literature.^4,16,17

Our results highlight that attendings and medical students still believe that medical education is a primary and important purpose of rounds even in the face of increasing time pressures. As such, efforts should be made to better align the many purposes of rounds with the realities of the modern day rounding environment. Increasing the presence of medical education on rounds need not be at the expense of time given that techniques like the 1-minute preceptor have been rated as both efficient and effective methods of teaching and delivering feedback.¹⁸ This is echoed in research that has found that faculty development with a focus on teaching significantly increased the rate of clinical education and interdisciplinary communication during rounds.¹ Opportunities for faculty development are increasingly accessible,¹⁹ including programs like the Advancing Pediatric Excellence Teaching Program, sponsored by the American Academy of Pediatrics Section on Hospital Medicine and the Academic Pediatric Association, and the Teaching Educators Across the Continuum of Healthcare program, sponsored by the Society for General Internal Medicine.^20,21

A testament to the adaptability of rounds can be seen in our findings that expose the increased emphasis with which pediatric attendings identify communication as a purpose of rounds, particularly within the themes of patient/family communication and establishing rapport with patients. This is likely due to the practice of FCR by 100% of the pediatric attendings in our focus groups, and is supported elsewhere in the literature.²² A key to family-centered rounds is communication, with active participation in the care discussion by patients and families as described and endorsed by a 2012 American Academy of Pediatrics (AAP) policy.^10,23

This emphasis could explain the increased frequency of comments made by pediatric attendings within the themes of patient/family communication and establishing rapport with patients. Furthermore, the AAP policy statement stresses the need to share information in a way that patients and families “effectively participate in care and decision making,” which could explain why pediatric attendings placed greater emphasis on the formation of the patient care plan in the theme of patient care.

As noted, the authors published a related study focusing on resident perceptions regarding the purpose of rounds. We initially undertook a separate analysis of the 3 groups: faculty, residents, and medical students. From that analysis, it became apparent that residents (PGY1-PGY3) viewed rounds differently than faculty and medical students. Where faculty and medical students were more focused on communication and medical education, the residents were more focused on the practical aspects of rounds (eg, “getting work done”). It was also noted that the residents’ focus aligned with the graduate medical education milestones, and framing the results within the milestones made the interpretation far more robust. In addition, the residents discussed their difficulties with patient and family involvement, especially in the context of family centered rounds, which is a topic that was rarely discussed by attendings or medical students.

Our study has a number of limitations. Only 4 university-based hospitals were included in the focus groups. This has the potential to limit the generalizability to the community hospital setting. Within the focus groups, the number of participants varied, and this may have had an impact on the flow and content of conversation. Facilitators were chosen to minimize potential bias and prior relationships with participants; however, this was not always possible, and as such, may have influenced responses. There may be a discrepancy between how people perceive rounds and how rounds actually function. Rounds were not standardized between institutions, departments, or attendings.

Rounds are an appropriate metaphor for medical education at large: they are time consuming, complex, and vary in quality, but are nevertheless essential to the goals of patients and learners alike because of their adaptability and hybridization of purpose. Our results highlight that rounds serve 4 critical purposes, including communication, medical education, patient care, and assessment. Importantly, both attendings and students agree on what they perceive to be the many purposes of rounds. Despite this agreement, a disconnect appears to exist between what people believe are the purposes of rounds and what is perceived to be happening during rounds. The causes of this gap are not well defined, and further efforts should be made to better understand the obstacles facing effective rounding. To improve rounds and adapt them to the needs of 21st century learners, it is critical that we better define the scope of medical education, both formal and informal, that occurs during rounds. In doing so, it will be possible to identify areas of development and training for faculty, residents, and medical students, which will ensure that rounds remain useful and critical tools for the development and education of future physicians.

Acknowledgments

The authors would like to acknowledge the following people who assisted on this project: Meghan Daly from The University of Chicago Pritzker School of Medicine, Shannon Martin, MD, MS, Assistant Professor of Medicine from the Department of Medicine at The University of Chicago, Joyce Campbell, BSN, MS, Senior Quality Manager at the Children’s National Medical Center, Benjamin Colburn from the University of California, San Francisco School of Medicine, Kelly Sanders from the University of California, San Francisco School of Medicine, and Alekist Quach from the University of California, San Francisco School of Medicine.

Disclosure 

The authors report no external funding source for this study. The authors declare no conflict of interest. The protocol was approved by the institutional review board at all participating institutions.

For more than a century, medical rounds have been a cornerstone of patient care and medical education in teaching hospitals. They remain critical activities for exposing generations of trainees to clinical decision making, coordination of care, and patient communication.¹

Despite this established importance within medical education and patient care, there is a relative paucity of research addressing the purpose of medical rounds in the 21st century. Medicine has evolved significantly since Osler’s day, and it is unclear whether the purpose of rounds has evolved along with it. Rounds, to Osler, were an important opportunity for future physicians to learn at the bedside from an attending physician. Increased duty hour restrictions, mandatory adoption of electronic medical records, and increasingly complex care have changed how rounds are performed, making it more difficult to achieve Osler’s ideals.^2,3 While several studies have aimed to quantify the changes to rounds and have demonstrated a significant decline in bedside teaching,^4-6 few studies have explored the purpose of rounds from the perspective of pertinent stakeholders, students, residents, and faculty. The authors have published the results of focus groups of resident stakeholders recently.⁷ We made the decision to combine the student/faculty data and describe it separately from the resident data to allow the most accurate and relevant discussion as it pertained to each group.

The aim of this study was to explore the perceptions of faculty and students of general inpatient rounds on internal medicine and pediatric rotations, and to identify any notable differences between these key stakeholders.

Between April 2014 and June 2014, we conducted 10 semistructured focus groups at 4 teaching hospitals: The University of Chicago Medical Center, Children’s National Health System, Georgetown University Medical Center, and the University of California, San Francisco Medical Center. A sample of eligible 3rd-year medical students and residents on pediatrics and internal medicine hospitalist services as well as hospitalist attendings in pediatrics and internal medicine were invited by e-mail to participate voluntarily without compensation. Identical semistructured focus groups were also conducted with pediatric and internal medicine interns (postgraduate year [PGY1]) and senior residents (PGY2 and PGY3), and those data have been published previously.⁷

Data Collection

Most focus groups had 6 to 8 participants, with 2 groups of 3 and 4. The groups were interviewed separately by training and specialty: 3rd-year medical students who had completed internal medicine and/or pediatrics rotations, hospitalist attendings in pediatrics, and hospitalist attendings in internal medicine. Attendings with training in medicine-pediatrics were included in the department in which they worked most frequently. The focus group script was informed by a literature review and expert input, and we used open-ended questions to explore perspectives on current and ideal purposes of rounds. Interviews were digitally recorded, transcribed, and names of speakers or references to specific patients were removed to preserve confidentiality and anonymity. The focus groups lasted between 30 and 60 minutes. The author (OH) conducted focus groups at 1 site, and trained facilitators conducted focus groups at the remaining 3 sites. The protocol was determined to be exempt by the institutional review boards at all participating sites. Prior to the focus groups, the definition of family-centered rounds was read aloud; after which, participants were asked to fill out a demographic survey.

Data Analysis

The authors employed a grounded theory approach to data collection and analysis,⁸ and data were analyzed by using the constant-comparative method.⁹ There was no a priori hypothesis. Four transcripts were independently reviewed by 2 authors (OH and RR) by using sentences and phrases as the units of data, which were coded with an identifier. The authors discussed initial codes and resolved discrepancies through deliberation and consensus to create codebooks. Themes, made up of multiple codes, were identified inductively and iteratively and were refined to reflect the evolving dataset. One author (OH) independently coded the remaining transcripts by using a revised codebook as a guide. A faculty author (JF) assessed the interrater reliability of the final codebook by reviewing 2 previously coded, randomly selected transcripts with no new codes emerging in the process, with a kappa coefficient of >0.8 indicating significant agreement.

What Do You Perceive the Purpose of Rounds to Be?

With respect to this prompt, we identified 4 themes, which represent 16 codes describing what attendings and medical students believed to be the purpose of rounds (Table 2). These themes are communication, medical education, patient care, and assessment.

Communication

Communication includes all comments addressing the role of rounds as it relates to communication between team members, patients, family members, and all those involved in patient care. There were 4 main codes, including coordination of patient care team, patient/family communication, establishing rapport with patients and/or family, and establishment of roles.

Coordination of patient care team identified rounds as a time “to make sure everyone is on the same page” and “to come together whenever possible,” so that everyone “had the same information of what was going on.” It also included comments related to interdisciplinary communication, with 1 participant describing rounds as “a time when your consulting team, or people with outside expertise, can weigh in on some medical issues.”

Medical Education

The theme of medical education is made up of 6 codes that encompass comments related to teaching and learning during rounds. These 6 codes include delivery of clinical education, exposure to clinical decision making, role modeling, student presentations, establishment of trainee autonomy, and providing a safe learning environment.

Delivery of clinical education included comments identifying rounds as a time for didactic teaching, teachable moments, “clinical pearls,” and bedside teaching of physical exam skills. Exposure to clinical decision making included comments by both medical students and attendings who described the purpose of rounds as a time for learning and teaching, specifically about how best to approach problems and decision making in a systematic manner, with 1 medical student explaining it as a time to “expose [trainees] to the way that people think about problems and how they decided to go about addressing them.”

Role modeling includes comments addressing rounds as a time for attendings to demonstrate appropriate behaviors and skills to trainees. One attending explained that “everybody learns from watching other people present and interact…so everybody has a chance to pick up things that they think, ‘Oh, this works well.’” Student presentations include comments, predominantly from students, that described rounds as an opportunity to practice presentations and receive feedback, with 1 student explaining it was a time “to learn how to present but also to be questioned and challenged.”

Establishing trainee autonomy is a code that identifies rounds as a time to encourage resident and student autonomy in order to achieve rounds that function with minimal input from the attending, with 1 attending describing how they “put resident leadership first as far as priorities… [and] fostering that because I usually let them decide what we’re going to do.”

Providing a safe learning environment identifies the purpose of rounds as being a space in which trainees can feel comfortable learning from their mistakes. One student described rounds as, “…a setting where it’s okay to be wrong and feel comfortable enough to know that it’s about a learning process.”

Assessment

Assessment is a theme composed of comments identifying the purpose of rounds as being related to observation, assessment, and feedback, and it includes 2 codes: attending observation, assessment, and feedback and establishment of expectations. Attending observation, assessment, and feedback includes comments from attendings and students alike who described rounds as a place for observation, evaluation, and provision of feedback regarding the skills and abilities of trainees. One attending explained that rounds gave him an “opportunity to observe trainees interacting with each other, with the patient, the patient’s family, and ancillary staff,” with another commenting it was time used “to assess how med students are gathering information, presenting information, and eventually their assessment and plan.” Establishment of expectations captures comments that describe rounds as a time for the establishment of expectations and goals of the team.

Patient Care

Patient care is a theme comprised of comments identifying the purpose of rounds as being directly related to the formation and delivery of the patient care plan, and it includes 2 codes: formation of the patient care plan and delivery of patient care. Formation of the patient care plan includes comments, which identified rounds as a time for discussing and forming the plan for the day, with an attending stating, “The purpose [of rounds] was to make a plan, a treatment plan, and to include the parents in making the treatment plan.” Delivery of patient care included comments identifying rounds as a means of ensuring timely, safe, and appropriate delivery of patient care occurred. One attending explained, “It can’t be undersold that the priority of rounds is patient care and the more eyes that look over information the less likely there are to be mistakes.”

What Do You Believe the Ideal Purpose of RoundsShould Be?

This study originally sought to compare responses to 2 different questions: “What do you perceive the purpose of rounds to be?” and “What do you believe the ideal purpose of rounds should be?” What became clear during the focus groups was that these were often interpreted to be the same question, and as such, responses to the latter question were truncated or were reiterations of what was previously said: “I think we’ve already discussed that, I think it’s no different than what we already kind of said, patient care, education, and communication,” explained 1 attending. Fifty-four responses to the question regarding the ideal purpose of rounds were coded and did not differ significantly from the previously noted results in terms of the domains represented and the frequency of representation.

Variation Among Respondents

Overall, there is a high level of concordance between the comments from medical students and attendings regarding the purpose of rounds, particularly in the medical education theme. However, medicine and pediatric attendings differ in their comments relating to the theme of communication, with 2 codes primarily accounting for this difference: pediatric attendings place more emphasis on time for patient/family communication and establishing rapport with patients than their internal medicine colleagues. Of note, all of the pediatric attendings involved in the study answered that they conducted family-centered rounds (FCR), compared with 22% of internal medicine attendings.¹⁰

Another notable discrepancy came up during focus groups involving comments from medical students who reiterated that the purpose of rounds was not fixed, but rather dependent on the attending that was running rounds. This theme was only identified in focus groups involving medical students. One student explained, “I think that it depends on the attending and if they actually want to teach,” and another commented that “it’s incredibly dependent on what the attending… is willing to invest.” No attendings identified student or attending variability as an important factor influencing the purpose of rounds.

This qualitative study is one of the first to explore the purpose of rounds from the perspective of both medical students and attendings. Reassuringly, our results indicate that medical student and attending perceptions are largely concordant. The 4 themes of communication, medical education, assessment, and patient care are in line with the findings of previous observational studies of internal medicine and pediatrics rounds.^1,11 The themes are similar to the findings of resident focus groups done at these same sites.⁷

Our results support that both medical students and attendings identify the importance of medical education during rounds. This is in contrast with findings in previous observational time-motion research by Stickrath that describes the focus on patient care related activities and the relative scarcity of education during rounds.¹ This stresses a divide between how medical students and attendings define the purpose of rounds and what other research suggests actually occurs on rounds. This distinction is an important one. It is possible that the way we, and others, define “medical education” and “patient care” may be at least partially responsible for these findings. This is supported by the ambiguous distinction between formal and informal educational activities on rounds and the challenges in characterizing the hidden curriculum and its role in medical student and resident education.¹¹ Attendings role modeling effective patient communication strategies, for example, highlights that patient care, medical education, and communication are frequently indistinguishable.¹² This hybridization of activities and dedication to diverse types of learning is an essential quality of rounds and is suggestive of why they have survived as a preeminent tool within the arsenal of medical education for the past century.

Yet, this finding does not excuse or adequately explain a well-documented disappearance of more formal educational activities during rounds. Recent observational studies have shown that the percentage of rounds dedicated to educational activities fell from 25% to 10% after the implementation of duty hour restrictions,^1,13,14 and a recent ethnographic study of pediatric attending rounds confirmed teaching during rounds, though seen as a pedagogical ideal, occurred infrequently and inconsistently in large part because of time pressures.¹⁵ In our attending focus groups, duty hours and time pressures were frequently cited as actively working against the purpose of rounds, specifically opportunities for teaching, with 1 attending explaining, “I just don’t think we achieve our [teaching] goals like we used to.” Another attending mentioned that, because of time pressures, “I often find myself apologizing. ‘I’m so sorry. I can’t resist. Can I just tell you this one thing? I’m so sorry to do teaching.’” This tension between time pressures and education on rounds is well documented in the literature.^4,16,17

Our results highlight that attendings and medical students still believe that medical education is a primary and important purpose of rounds even in the face of increasing time pressures. As such, efforts should be made to better align the many purposes of rounds with the realities of the modern day rounding environment. Increasing the presence of medical education on rounds need not be at the expense of time given that techniques like the 1-minute preceptor have been rated as both efficient and effective methods of teaching and delivering feedback.¹⁸ This is echoed in research that has found that faculty development with a focus on teaching significantly increased the rate of clinical education and interdisciplinary communication during rounds.¹ Opportunities for faculty development are increasingly accessible,¹⁹ including programs like the Advancing Pediatric Excellence Teaching Program, sponsored by the American Academy of Pediatrics Section on Hospital Medicine and the Academic Pediatric Association, and the Teaching Educators Across the Continuum of Healthcare program, sponsored by the Society for General Internal Medicine.^20,21

A testament to the adaptability of rounds can be seen in our findings that expose the increased emphasis with which pediatric attendings identify communication as a purpose of rounds, particularly within the themes of patient/family communication and establishing rapport with patients. This is likely due to the practice of FCR by 100% of the pediatric attendings in our focus groups, and is supported elsewhere in the literature.²² A key to family-centered rounds is communication, with active participation in the care discussion by patients and families as described and endorsed by a 2012 American Academy of Pediatrics (AAP) policy.^10,23

This emphasis could explain the increased frequency of comments made by pediatric attendings within the themes of patient/family communication and establishing rapport with patients. Furthermore, the AAP policy statement stresses the need to share information in a way that patients and families “effectively participate in care and decision making,” which could explain why pediatric attendings placed greater emphasis on the formation of the patient care plan in the theme of patient care.

As noted, the authors published a related study focusing on resident perceptions regarding the purpose of rounds. We initially undertook a separate analysis of the 3 groups: faculty, residents, and medical students. From that analysis, it became apparent that residents (PGY1-PGY3) viewed rounds differently than faculty and medical students. Where faculty and medical students were more focused on communication and medical education, the residents were more focused on the practical aspects of rounds (eg, “getting work done”). It was also noted that the residents’ focus aligned with the graduate medical education milestones, and framing the results within the milestones made the interpretation far more robust. In addition, the residents discussed their difficulties with patient and family involvement, especially in the context of family centered rounds, which is a topic that was rarely discussed by attendings or medical students.

Our study has a number of limitations. Only 4 university-based hospitals were included in the focus groups. This has the potential to limit the generalizability to the community hospital setting. Within the focus groups, the number of participants varied, and this may have had an impact on the flow and content of conversation. Facilitators were chosen to minimize potential bias and prior relationships with participants; however, this was not always possible, and as such, may have influenced responses. There may be a discrepancy between how people perceive rounds and how rounds actually function. Rounds were not standardized between institutions, departments, or attendings.

Rounds are an appropriate metaphor for medical education at large: they are time consuming, complex, and vary in quality, but are nevertheless essential to the goals of patients and learners alike because of their adaptability and hybridization of purpose. Our results highlight that rounds serve 4 critical purposes, including communication, medical education, patient care, and assessment. Importantly, both attendings and students agree on what they perceive to be the many purposes of rounds. Despite this agreement, a disconnect appears to exist between what people believe are the purposes of rounds and what is perceived to be happening during rounds. The causes of this gap are not well defined, and further efforts should be made to better understand the obstacles facing effective rounding. To improve rounds and adapt them to the needs of 21st century learners, it is critical that we better define the scope of medical education, both formal and informal, that occurs during rounds. In doing so, it will be possible to identify areas of development and training for faculty, residents, and medical students, which will ensure that rounds remain useful and critical tools for the development and education of future physicians.

Acknowledgments

The authors would like to acknowledge the following people who assisted on this project: Meghan Daly from The University of Chicago Pritzker School of Medicine, Shannon Martin, MD, MS, Assistant Professor of Medicine from the Department of Medicine at The University of Chicago, Joyce Campbell, BSN, MS, Senior Quality Manager at the Children’s National Medical Center, Benjamin Colburn from the University of California, San Francisco School of Medicine, Kelly Sanders from the University of California, San Francisco School of Medicine, and Alekist Quach from the University of California, San Francisco School of Medicine.

Disclosure 

The authors report no external funding source for this study. The authors declare no conflict of interest. The protocol was approved by the institutional review board at all participating institutions.

Fatigue is the most common clinical symptom of anemia and is a significant concern to patients.^1,2 In ambulatory patients, lower hemoglobin (Hb) concentration is associated with increased fatigue.^2,3 Accordingly, therapies that treat anemia by increasing Hb concentration, such as erythropoiesis stimulating agents,^4-7 often use fatigue as an outcome measure.

In hospitalized patients, transfusion of red blood cell increases Hb concentration and is the primary treatment for anemia. However, the extent to which transfusion and changes in Hb concentration affect hospitalized patients’ fatigue levels is not well established. Guidelines support transfusing patients with symptoms of anemia, such as fatigue, on the assumption that the increased oxygen delivery will improve the symptoms of anemia. While transfusion studies in hospitalized patients have consistently reported that transfusion at lower or “restrictive” Hb concentrations is safe compared with transfusion at higher Hb concentrations,^8-10 these studies have mainly used cardiac events and mortality as outcomes rather than patient symptoms, such as fatigue. Nevertheless, they have resulted in hospitals increasingly adopting restrictive transfusion policies that discourage transfusion at higher Hb levels.^11,12 Consequently, the rate of transfusion in hospitalized patients has decreased,¹³ raising questions of whether some patients with lower Hb concentrations may experience increased fatigue as a result of restrictive transfusion policies. Fatigue among hospitalized patients is important not only because it is an adverse symptom but because it may result in decreased activity levels, deconditioning, and losses in functional status.^14,15While the effect of alternative transfusion policies on fatigue in hospitalized patients could be answered by a randomized clinical trial using fatigue and functional status as outcomes, an important first step is to assess whether the Hb concentration of hospitalized patients is associated with their fatigue level during hospitalization. Because hospitalized patients often have acute illnesses that can cause fatigue in and of themselves, it is possible that anemia is not associated with fatigue in hospitalized patients despite anemia’s association with fatigue in ambulatory patients. Additionally, Hb concentration varies during hospitalization,¹⁶ raising the question of what measures of Hb during hospitalization might be most associated with anemia-related fatigue.

The objective of this study is to explore multiple Hb measures in hospitalized medical patients with anemia and test whether any of these Hb measures are associated with patients’ fatigue level.

Study Design

We performed a prospective, observational study of hospitalized patients with anemia on the general medicine services at The University of Chicago Medical Center (UCMC). The institutional review board approved the study procedures, and all study subjects provided informed consent.

Study Eligibility

Between April 2014 and June 2015, all general medicine inpatients were approached for written consent for The University of Chicago Hospitalist Project,¹⁷ a research infrastructure at UCMC. Among patients consenting to participate in the Hospitalist Project, patients were eligible if they had Hb <9 g/dL at any point during their hospitalization and were age ≥50 years. Hb concentration of <9 g/dL was chosen to include the range of Hb values covered by most restrictive transfusion policies.^8-10,18 Age ≥50 years was an inclusion criteria because anemia is more strongly associated with poor outcomes, including functional impairment, among older patients compared with younger patients.^14,19-21 If patients were not eligible for inclusion at the time of consent for the Hospitalist Project, their Hb values were reviewed twice daily until hospital discharge to assess if their Hb was <9 g/dL. Proxies were sought to answer questions for patients who failed the Short Portable Mental Status Questionnaire.²²

Patient Demographic Data Collection

Research assistants abstracted patient age and sex from the electronic health record (EHR), and asked patients to self-identify their race. The individual comorbidities included as part of the Charlson Comorbidity Index were identified using International Classification of Diseases, 9th Revision codes from hospital administrative data for each encounter and specifically included the following: myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatic disease, peptic ulcer disease, liver disease, diabetes, hemiplegia and/or paraplegia, renal disease, cancer, and human immunodeficiency virus/acquired immunodeficiency syndrome.²³ We also used Healthcare Cost and Utilization Project (www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp) diagnosis categories to identify whether patients had sickle cell (SC) anemia, gastrointestinal bleeding (GIB), or a depressive disorder (DD) because these conditions are associated with anemia (SC and GIB) and fatigue (DD).²⁴

Measuring Anemia

Hb measures were available only when hospital providers ordered them as part of routine practice. The first Hb concentration <9 g/dL during a patient’s hospitalization, which made them eligible for study participation, was obtained through manual review of the EHR. All additional Hb values during the patient’s hospitalization were obtained from the hospital’s administrative data mart. All Hb values collected for each patient during the hospitalization were used to calculate summary measures of Hb during the hospitalization, including the mean Hb, median Hb, minimum Hb, maximum Hb, admission (first recorded) Hb, and discharge (last recorded) Hb. Hb measures were analyzed both as a continuous variable and as a categorical variable created by dividing the continuous Hb measures into integer ranges of 3 groups of approximately the same size.

Measuring Fatigue

Our primary outcome was patients’ level of fatigue reported during hospitalization, measured using the Functional Assessment of Chronic Illness Therapy (FACIT)-Anemia questionnaire. Fatigue was measured using a 13-question fatigue subscale,^1,2,25 which measures fatigue within the past 7 days. Scores on the fatigue subscale range from 0 to 52, with lower scores reflecting greater levels of fatigue. As soon as patients met the eligibility criteria for study participation during their hospitalization (age ≥50 years and Hb <9 g/dL), they were approached to answer the FACIT questions. Values for missing data in the fatigue subscale for individual subjects were filled in using a prorated score from their answered questions as long as >50% of the items in the fatigue subscale were answered, in accordance with recommendations for addressing missing data in the FACIT.²⁶ Fatigue was analyzed as a continuous variable and as a dichotomous variable created by dividing the sample into high (FACIT <27) and low (FACIT ≥27) levels of fatigue based on the median FACIT score of the population. Previous literature has shown a FACIT fatigue subscale score between 23 and 26 to be associated with an Eastern Cooperative Oncology Group (ECOG)²⁷ C Performance Status rating of 2 to 3³ compared to scores ≥27.

Statistical Analysis

Statistical analysis was performed using Stata statistical software (StataCorp, College Station, Texas). Descriptive statistics were used to characterize patient demographics. Analysis of variance was used to test for differences in the mean fatigue levels across Hb measures. χ² tests were performed to test for associations between high fatigue levels and the Hb measures. Multivariable analysis, including both linear and logistic regression models, were used to test the association of Hb concentration and fatigue. P values <0.05 using a 2-tailed test were deemed statistically significant.

Patient Characteristics

During the study period, 8559 patients were admitted to the general medicine service. Of those, 5073 (59%) consented for participation in the Hospitalist Project, and 3670 (72%) completed the Hospitalist Project inpatient interview. Of these patients, 1292 (35%) had Hb <9 g/dL, and 784 (61%) were 50 years or older and completed the FACIT questionnaire.

Table 1 reports the demographic characteristics and comorbidities for the sample, the mean (standard deviation [SD]) for the 6 Hb measures, and mean (SD) and median FACIT scores.

Bivariate Association of Fatigue and Hb

Categorizing patients into low, middle, or high Hb for each of the 6 Hb measures, minimum Hb was strongly associated with fatigue, with a weaker association for mean Hb and no statistically significant association for the other measures.

Minimum Hb

Patients with a minimum Hb <7 g/dL and patients with Hb 7-8 g/dL had higher fatigue levels (FACIT = 25 for each) than patients with a minimum Hb ≥8 g/dL (FACIT = 29; P < 0.001; Table 2). When excluding patients with SC and/or GIB because their average minimum Hb differed from the average minimum Hb of the full population (P < 0.001), patients with a minimum Hb <7 g/dL or 7-8 g/dL had even higher fatigue levels (FACIT = 23 and FACIT = 24, respectively), with no change in the fatigue level of patients with a minimum Hb ≥8 g/dL (FACIT = 29; P < 0.001; Table 2). Lower minimum Hb continued to be associated with higher fatigue levels when analyzed in 0.5 g/dL increments (Figure).

Mean Hb and Other Measures

Fatigue levels were high for 47% of patients with a mean Hb <8g /dL and 53% of patients with a mean Hb 8-9 g/dL compared with 43% of patients with a mean Hb ≥9 g/dL (P = 0.05). However, the association between high fatigue and mean Hb was not statistically significant when patients with SC and/or GIB were excluded (Table 2). None of the other 4 Hb measures was significantly associated with fatigue.

Linear Regression of Fatigue on Hb

In linear regression models, minimum Hb consistently predicted patient fatigue, mean Hb had a less robust association with fatigue, and the other Hb measures were not associated with patient fatigue. Increases in minimum Hb (analyzed as a continuous variable) were associated with reduced fatigue (higher FACIT score; β = 1.4; P = 0.005). In models in which minimum Hb was a categorical variable, patients with a minimum Hb of <7 g/dL or 7-8 g/dL had greater fatigue (lower FACIT score) than patients whose minimum Hb was ≥8 g/dL (Hb <7 g/dL: β = −4.2; P ≤ 0.001; Hb 7-8 g/dL: β = −4.1; P < 0.001). These results control for patients’ age, sex, individual comorbidities, and whether their minimum Hb occurred before or after the measurement of fatigue during hospitalization (Model 1), and the results are unchanged when also controlling for the number of Hb laboratory draws patients had during their hospitalization (Model 2; Table 3). In a stratified analysis excluding patients with either SC and/or GIB, changes in minimum Hb were associated with larger changes in patient fatigue levels (Supplemental Table 1). We also stratified our analysis to include only patients whose minimum Hb occurred before the measurement of their fatigue level during hospitalization to avoid a spurious association of fatigue with minimum Hb occurring after fatigue was measured. In both Models 1 and 2, minimum Hb remained a predictor of patients’ fatigue levels with similar effect sizes, although in Model 2, the results did not quite reach a statistically significant level, in part due to larger confidence intervals from the smaller sample size of this stratified analysis (Supplemental Table 2a). We further stratified this analysis to include only patients whose transfusion, if they received one, occurred after their minimum Hb and the measurement of their fatigue level to account for the possibility that a transfusion could affect the fatigue level patients report. In this analysis, most of the estimates of the effect of minimum Hb on fatigue were larger than those seen when only analyzing patients whose minimum Hb occurred before the measurement of their fatigue level, although again, the smaller sample size of this additional stratified analysis does produce larger confidence intervals for these estimates (Supplemental Table 2b).

Logistic Regression of High Fatigue Level on Hb

Using logistic regression, minimum Hb analyzed as a categorical variable predicted increased odds of a high fatigue level. Patients with a minimum Hb <7 g/dL were 50% (odds ratio [OR] = 1.5; P = 0.03) more likely to have high fatigue and patients with a minimum Hb 7-8 g/dL were 90% (OR = 1.9; P < 0.001) more likely to have high fatigue compared with patients with a minimum Hb ≥8 g/dL in Model 1. These results were similar in Model 2, although the effect was only statistically significant in the 7-8 g/dL Hb group (Table 3). When excluding SC and/or GIB patients, the odds of having high fatigue as minimum Hb decreased were the same or higher for both models compared to the full population of patients. However, again, in Model 2, the effect was only statistically significant in the 7-8 g/dL Hb group (Supplemental Table 1).

Patients with a mean Hb <8 g/dL were 20% to 30% more likely to have high fatigue and patients with mean Hb 8-9 g/dL were 50% more likely to have high fatigue compared with patients with a mean Hb ≥9 g/dL, but the effects were only statistically significant for patients with a mean Hb 8-9 g/dL in both Models 1 and 2 (Table 3). These results were similar when excluding patients with SC and/or GIB, but they were only significant for patients with a mean Hb 8-9 g/dL in Model 1 and patients with a mean Hb <8 g/dL in the Model 2 (Supplemental Table 3).

These results demonstrate that minimum Hb during hospitalization is associated with fatigue in hospitalized patients age ≥50 years, and the association is stronger among patients without SC and/or GIB as comorbidities. The analysis of Hb as a continuous and categorical variable and the use of both linear and logistic regression models support the robustness of these associations and illuminate their clinical significance. For example, in linear regression with minimum Hb a continuous variable, the coefficient of 1.4 suggests that an increase of 2 g/dL in Hb, as might be expected from transfusion of 2 units of red blood cells, would be associated with about a 3-point improvement in fatigue. Additionally, as a categorical variable, a minimum Hb ≥8 g/dL compared with a minimum Hb <7 g/dL or 7-8 g/dL is associated with a 3- to 4-point improvement in fatigue. Previous literature suggests that a difference of 3 in the FACIT score is the minimum clinically important difference in fatigue,³ and changes in minimum Hb in either model predict changes in fatigue that are in the range of potential clinical significance.

The clinical significance of the findings is also reflected in the results of the logistic regressions, which may be mapped to potential effects on functional status. Specifically, the odds of having a high fatigue level (FACIT <27) increase 90% for persons with a minimum Hb 7–8 g/dL compared with persons with a minimum Hb ≥8 g/dL. For persons with a minimum Hb <7 g/dL, point estimates suggest a smaller (50%) increase in the odds of high fatigue, but the 95% confidence interval overlaps heavily with the estimate of patients whose minimum Hb is 7-8 g/dL. While it might be expected that patients with a minimum Hb <7 g/dL have greater levels of fatigue compared with patients with a minimum Hb 7-8 g/dL, we did not observe such a pattern. One reason may be that the confidence intervals of our estimated effects are wide enough that we cannot exclude such a pattern. Another possible explanation is that in both groups, the fatigue levels are sufficiently severe, such that the difference in their fatigue levels may not be clinically meaningful. For example, a FACIT score of 23 to 26 has been shown to be associated with an ECOG performance status of 2 to 3, requiring bed rest for at least part of the day.³ Therefore, patients with a minimum Hb 7-8 g/dL (mean FACIT score = 24; Table 2) or a minimum Hb of <7 g/dL (mean FACIT score = 23; Table 2) are already functionally limited to the point of being partially bed bound, such that further decreases in their Hb may not produce additional fatigue in part because they reduce their activity sufficiently to prevent an increase in fatigue. In such cases, the potential benefits of increased Hb may be better assessed by measuring fatigue in response to a specific and provoked activity level, a concept known as fatigability.²⁰

That minimum Hb is more strongly associated with fatigue than any other measure of Hb during hospitalization may not be surprising. Mean, median, maximum, and discharge Hb may all be affected by transfusion during hospitalization that could affect fatigue. Admission Hb may not reflect true oxygen-carrying capacity because of hemoconcentration.

The association between Hb and fatigue in hospitalized patients is important because increased fatigue could contribute to slower clinical recovery in hospitalized patients. Additionally, increased fatigue during hospitalization and at hospital discharge could exacerbate the known deleterious consequences of fatigue on patients and their health outcomes^14,15 after hospital discharge. Although one previous study, the Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair (FOCUS)⁸ trial, did not report differences in patients’ fatigue levels at 30 and 60 days postdischarge when transfused at restrictive (8 g/dL) compared with liberal (10 g/dL) Hb thresholds, confidence in the validity of this finding is reduced by the fact that more than half of the patients were lost to follow-up at the 30- and 60-day time points. Further, patients in the restrictive transfusion arm of FOCUS were transfused to maintain Hb levels at or above 8 g/dL. This transfusion threshold of 8 g/dL may have mitigated the high levels of fatigue that are seen in our study when patients’ Hb drops below 8 g/dL, and maintaining a Hb level of 7 g/dL is now the standard of care in stable hospitalized patients. Lastly, FOCUS was limited to postoperative hip fracture patients, and the generalizability of FOCUS to hospitalized medicine patients with anemia is limited.

Therefore, our results support guideline suggestions that practitioners incorporate the presence of patient symptoms such as fatigue into transfusion decisions, particularly if patients’ Hb is <8 g/dL.¹⁸ Though reasonable, the suggestion to incorporate symptoms such as fatigue into transfusion decisions has not been strongly supported by evidence so far, and it may often be neglected in practice. Definitive evidence to support such recommendations would benefit from study through an optimal trial¹⁸ that incorporates symptoms into decision making. Our findings add support for a study of transfusion strategies that incorporates patients’ fatigue level in addition to Hb concentration.

This study has several limitations. Although our sample size is large and includes patients with a range of comorbidities that we believe are representative of hospitalized general medicine patients, as a single-center, observational study, our results may not be generalizable to other centers. Additionally, although these data support a reliable association between hospitalized patients’ minimum Hb and fatigue level, the observational design of this study cannot prove that this relationship is causal. Also, patients’ Hb values were measured at the discretion of their clinician, and therefore, the measures of Hb were not uniformly measured for participating patients. In addition, fatigue was only measured at one time point during a patient’s hospitalization, and it is possible that patients’ fatigue levels change during hospitalization in relation to variables we did not consider. Finally, our study was not designed to assess the association of Hb with longer-term functional outcomes, which may be of greater concern than fatigue.

In hospitalized patients ≥50 years old, minimum Hb is reliably associated with patients’ fatigue level. Patients whose minimum Hb is <8 g/dL experience higher fatigue levels compared to patients whose minimum Hb is ≥8 g/dL. Additional studies are warranted to understand if patients may benefit from improved fatigue levels by correcting their anemia through transfusion.

Disclosure 

Dr. Prochaska is supported by an Agency for Healthcare Research and Quality Patient-Centered Outcomes Research Institutional K12 Award (1K12HS023007-01, principal investigator Meltzer). Dr. Meltzer is supported by a National Institutes of Health Clinical and Translational Science Award (2UL1TR000430-06, principal investigator Solway) and a grant from the Patient-Centered Outcomes Research Network in support of the Chicago Patient-Centered Outcomes Research Network. The authors report no conflicts of interest.

Fatigue is the most common clinical symptom of anemia and is a significant concern to patients.^1,2 In ambulatory patients, lower hemoglobin (Hb) concentration is associated with increased fatigue.^2,3 Accordingly, therapies that treat anemia by increasing Hb concentration, such as erythropoiesis stimulating agents,^4-7 often use fatigue as an outcome measure.

In hospitalized patients, transfusion of red blood cell increases Hb concentration and is the primary treatment for anemia. However, the extent to which transfusion and changes in Hb concentration affect hospitalized patients’ fatigue levels is not well established. Guidelines support transfusing patients with symptoms of anemia, such as fatigue, on the assumption that the increased oxygen delivery will improve the symptoms of anemia. While transfusion studies in hospitalized patients have consistently reported that transfusion at lower or “restrictive” Hb concentrations is safe compared with transfusion at higher Hb concentrations,^8-10 these studies have mainly used cardiac events and mortality as outcomes rather than patient symptoms, such as fatigue. Nevertheless, they have resulted in hospitals increasingly adopting restrictive transfusion policies that discourage transfusion at higher Hb levels.^11,12 Consequently, the rate of transfusion in hospitalized patients has decreased,¹³ raising questions of whether some patients with lower Hb concentrations may experience increased fatigue as a result of restrictive transfusion policies. Fatigue among hospitalized patients is important not only because it is an adverse symptom but because it may result in decreased activity levels, deconditioning, and losses in functional status.^14,15While the effect of alternative transfusion policies on fatigue in hospitalized patients could be answered by a randomized clinical trial using fatigue and functional status as outcomes, an important first step is to assess whether the Hb concentration of hospitalized patients is associated with their fatigue level during hospitalization. Because hospitalized patients often have acute illnesses that can cause fatigue in and of themselves, it is possible that anemia is not associated with fatigue in hospitalized patients despite anemia’s association with fatigue in ambulatory patients. Additionally, Hb concentration varies during hospitalization,¹⁶ raising the question of what measures of Hb during hospitalization might be most associated with anemia-related fatigue.

The objective of this study is to explore multiple Hb measures in hospitalized medical patients with anemia and test whether any of these Hb measures are associated with patients’ fatigue level.

Study Design

We performed a prospective, observational study of hospitalized patients with anemia on the general medicine services at The University of Chicago Medical Center (UCMC). The institutional review board approved the study procedures, and all study subjects provided informed consent.

Study Eligibility

Between April 2014 and June 2015, all general medicine inpatients were approached for written consent for The University of Chicago Hospitalist Project,¹⁷ a research infrastructure at UCMC. Among patients consenting to participate in the Hospitalist Project, patients were eligible if they had Hb <9 g/dL at any point during their hospitalization and were age ≥50 years. Hb concentration of <9 g/dL was chosen to include the range of Hb values covered by most restrictive transfusion policies.^8-10,18 Age ≥50 years was an inclusion criteria because anemia is more strongly associated with poor outcomes, including functional impairment, among older patients compared with younger patients.^14,19-21 If patients were not eligible for inclusion at the time of consent for the Hospitalist Project, their Hb values were reviewed twice daily until hospital discharge to assess if their Hb was <9 g/dL. Proxies were sought to answer questions for patients who failed the Short Portable Mental Status Questionnaire.²²

Patient Demographic Data Collection

Research assistants abstracted patient age and sex from the electronic health record (EHR), and asked patients to self-identify their race. The individual comorbidities included as part of the Charlson Comorbidity Index were identified using International Classification of Diseases, 9th Revision codes from hospital administrative data for each encounter and specifically included the following: myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatic disease, peptic ulcer disease, liver disease, diabetes, hemiplegia and/or paraplegia, renal disease, cancer, and human immunodeficiency virus/acquired immunodeficiency syndrome.²³ We also used Healthcare Cost and Utilization Project (www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp) diagnosis categories to identify whether patients had sickle cell (SC) anemia, gastrointestinal bleeding (GIB), or a depressive disorder (DD) because these conditions are associated with anemia (SC and GIB) and fatigue (DD).²⁴

Measuring Anemia

Hb measures were available only when hospital providers ordered them as part of routine practice. The first Hb concentration <9 g/dL during a patient’s hospitalization, which made them eligible for study participation, was obtained through manual review of the EHR. All additional Hb values during the patient’s hospitalization were obtained from the hospital’s administrative data mart. All Hb values collected for each patient during the hospitalization were used to calculate summary measures of Hb during the hospitalization, including the mean Hb, median Hb, minimum Hb, maximum Hb, admission (first recorded) Hb, and discharge (last recorded) Hb. Hb measures were analyzed both as a continuous variable and as a categorical variable created by dividing the continuous Hb measures into integer ranges of 3 groups of approximately the same size.

Measuring Fatigue

Our primary outcome was patients’ level of fatigue reported during hospitalization, measured using the Functional Assessment of Chronic Illness Therapy (FACIT)-Anemia questionnaire. Fatigue was measured using a 13-question fatigue subscale,^1,2,25 which measures fatigue within the past 7 days. Scores on the fatigue subscale range from 0 to 52, with lower scores reflecting greater levels of fatigue. As soon as patients met the eligibility criteria for study participation during their hospitalization (age ≥50 years and Hb <9 g/dL), they were approached to answer the FACIT questions. Values for missing data in the fatigue subscale for individual subjects were filled in using a prorated score from their answered questions as long as >50% of the items in the fatigue subscale were answered, in accordance with recommendations for addressing missing data in the FACIT.²⁶ Fatigue was analyzed as a continuous variable and as a dichotomous variable created by dividing the sample into high (FACIT <27) and low (FACIT ≥27) levels of fatigue based on the median FACIT score of the population. Previous literature has shown a FACIT fatigue subscale score between 23 and 26 to be associated with an Eastern Cooperative Oncology Group (ECOG)²⁷ C Performance Status rating of 2 to 3³ compared to scores ≥27.

Statistical Analysis

Statistical analysis was performed using Stata statistical software (StataCorp, College Station, Texas). Descriptive statistics were used to characterize patient demographics. Analysis of variance was used to test for differences in the mean fatigue levels across Hb measures. χ² tests were performed to test for associations between high fatigue levels and the Hb measures. Multivariable analysis, including both linear and logistic regression models, were used to test the association of Hb concentration and fatigue. P values <0.05 using a 2-tailed test were deemed statistically significant.

Patient Characteristics

During the study period, 8559 patients were admitted to the general medicine service. Of those, 5073 (59%) consented for participation in the Hospitalist Project, and 3670 (72%) completed the Hospitalist Project inpatient interview. Of these patients, 1292 (35%) had Hb <9 g/dL, and 784 (61%) were 50 years or older and completed the FACIT questionnaire.

Table 1 reports the demographic characteristics and comorbidities for the sample, the mean (standard deviation [SD]) for the 6 Hb measures, and mean (SD) and median FACIT scores.

Bivariate Association of Fatigue and Hb

Categorizing patients into low, middle, or high Hb for each of the 6 Hb measures, minimum Hb was strongly associated with fatigue, with a weaker association for mean Hb and no statistically significant association for the other measures.

Minimum Hb

Patients with a minimum Hb <7 g/dL and patients with Hb 7-8 g/dL had higher fatigue levels (FACIT = 25 for each) than patients with a minimum Hb ≥8 g/dL (FACIT = 29; P < 0.001; Table 2). When excluding patients with SC and/or GIB because their average minimum Hb differed from the average minimum Hb of the full population (P < 0.001), patients with a minimum Hb <7 g/dL or 7-8 g/dL had even higher fatigue levels (FACIT = 23 and FACIT = 24, respectively), with no change in the fatigue level of patients with a minimum Hb ≥8 g/dL (FACIT = 29; P < 0.001; Table 2). Lower minimum Hb continued to be associated with higher fatigue levels when analyzed in 0.5 g/dL increments (Figure).

Mean Hb and Other Measures

Fatigue levels were high for 47% of patients with a mean Hb <8g /dL and 53% of patients with a mean Hb 8-9 g/dL compared with 43% of patients with a mean Hb ≥9 g/dL (P = 0.05). However, the association between high fatigue and mean Hb was not statistically significant when patients with SC and/or GIB were excluded (Table 2). None of the other 4 Hb measures was significantly associated with fatigue.

Linear Regression of Fatigue on Hb

In linear regression models, minimum Hb consistently predicted patient fatigue, mean Hb had a less robust association with fatigue, and the other Hb measures were not associated with patient fatigue. Increases in minimum Hb (analyzed as a continuous variable) were associated with reduced fatigue (higher FACIT score; β = 1.4; P = 0.005). In models in which minimum Hb was a categorical variable, patients with a minimum Hb of <7 g/dL or 7-8 g/dL had greater fatigue (lower FACIT score) than patients whose minimum Hb was ≥8 g/dL (Hb <7 g/dL: β = −4.2; P ≤ 0.001; Hb 7-8 g/dL: β = −4.1; P < 0.001). These results control for patients’ age, sex, individual comorbidities, and whether their minimum Hb occurred before or after the measurement of fatigue during hospitalization (Model 1), and the results are unchanged when also controlling for the number of Hb laboratory draws patients had during their hospitalization (Model 2; Table 3). In a stratified analysis excluding patients with either SC and/or GIB, changes in minimum Hb were associated with larger changes in patient fatigue levels (Supplemental Table 1). We also stratified our analysis to include only patients whose minimum Hb occurred before the measurement of their fatigue level during hospitalization to avoid a spurious association of fatigue with minimum Hb occurring after fatigue was measured. In both Models 1 and 2, minimum Hb remained a predictor of patients’ fatigue levels with similar effect sizes, although in Model 2, the results did not quite reach a statistically significant level, in part due to larger confidence intervals from the smaller sample size of this stratified analysis (Supplemental Table 2a). We further stratified this analysis to include only patients whose transfusion, if they received one, occurred after their minimum Hb and the measurement of their fatigue level to account for the possibility that a transfusion could affect the fatigue level patients report. In this analysis, most of the estimates of the effect of minimum Hb on fatigue were larger than those seen when only analyzing patients whose minimum Hb occurred before the measurement of their fatigue level, although again, the smaller sample size of this additional stratified analysis does produce larger confidence intervals for these estimates (Supplemental Table 2b).

Logistic Regression of High Fatigue Level on Hb

Using logistic regression, minimum Hb analyzed as a categorical variable predicted increased odds of a high fatigue level. Patients with a minimum Hb <7 g/dL were 50% (odds ratio [OR] = 1.5; P = 0.03) more likely to have high fatigue and patients with a minimum Hb 7-8 g/dL were 90% (OR = 1.9; P < 0.001) more likely to have high fatigue compared with patients with a minimum Hb ≥8 g/dL in Model 1. These results were similar in Model 2, although the effect was only statistically significant in the 7-8 g/dL Hb group (Table 3). When excluding SC and/or GIB patients, the odds of having high fatigue as minimum Hb decreased were the same or higher for both models compared to the full population of patients. However, again, in Model 2, the effect was only statistically significant in the 7-8 g/dL Hb group (Supplemental Table 1).

Patients with a mean Hb <8 g/dL were 20% to 30% more likely to have high fatigue and patients with mean Hb 8-9 g/dL were 50% more likely to have high fatigue compared with patients with a mean Hb ≥9 g/dL, but the effects were only statistically significant for patients with a mean Hb 8-9 g/dL in both Models 1 and 2 (Table 3). These results were similar when excluding patients with SC and/or GIB, but they were only significant for patients with a mean Hb 8-9 g/dL in Model 1 and patients with a mean Hb <8 g/dL in the Model 2 (Supplemental Table 3).

These results demonstrate that minimum Hb during hospitalization is associated with fatigue in hospitalized patients age ≥50 years, and the association is stronger among patients without SC and/or GIB as comorbidities. The analysis of Hb as a continuous and categorical variable and the use of both linear and logistic regression models support the robustness of these associations and illuminate their clinical significance. For example, in linear regression with minimum Hb a continuous variable, the coefficient of 1.4 suggests that an increase of 2 g/dL in Hb, as might be expected from transfusion of 2 units of red blood cells, would be associated with about a 3-point improvement in fatigue. Additionally, as a categorical variable, a minimum Hb ≥8 g/dL compared with a minimum Hb <7 g/dL or 7-8 g/dL is associated with a 3- to 4-point improvement in fatigue. Previous literature suggests that a difference of 3 in the FACIT score is the minimum clinically important difference in fatigue,³ and changes in minimum Hb in either model predict changes in fatigue that are in the range of potential clinical significance.

The clinical significance of the findings is also reflected in the results of the logistic regressions, which may be mapped to potential effects on functional status. Specifically, the odds of having a high fatigue level (FACIT <27) increase 90% for persons with a minimum Hb 7–8 g/dL compared with persons with a minimum Hb ≥8 g/dL. For persons with a minimum Hb <7 g/dL, point estimates suggest a smaller (50%) increase in the odds of high fatigue, but the 95% confidence interval overlaps heavily with the estimate of patients whose minimum Hb is 7-8 g/dL. While it might be expected that patients with a minimum Hb <7 g/dL have greater levels of fatigue compared with patients with a minimum Hb 7-8 g/dL, we did not observe such a pattern. One reason may be that the confidence intervals of our estimated effects are wide enough that we cannot exclude such a pattern. Another possible explanation is that in both groups, the fatigue levels are sufficiently severe, such that the difference in their fatigue levels may not be clinically meaningful. For example, a FACIT score of 23 to 26 has been shown to be associated with an ECOG performance status of 2 to 3, requiring bed rest for at least part of the day.³ Therefore, patients with a minimum Hb 7-8 g/dL (mean FACIT score = 24; Table 2) or a minimum Hb of <7 g/dL (mean FACIT score = 23; Table 2) are already functionally limited to the point of being partially bed bound, such that further decreases in their Hb may not produce additional fatigue in part because they reduce their activity sufficiently to prevent an increase in fatigue. In such cases, the potential benefits of increased Hb may be better assessed by measuring fatigue in response to a specific and provoked activity level, a concept known as fatigability.²⁰

That minimum Hb is more strongly associated with fatigue than any other measure of Hb during hospitalization may not be surprising. Mean, median, maximum, and discharge Hb may all be affected by transfusion during hospitalization that could affect fatigue. Admission Hb may not reflect true oxygen-carrying capacity because of hemoconcentration.

The association between Hb and fatigue in hospitalized patients is important because increased fatigue could contribute to slower clinical recovery in hospitalized patients. Additionally, increased fatigue during hospitalization and at hospital discharge could exacerbate the known deleterious consequences of fatigue on patients and their health outcomes^14,15 after hospital discharge. Although one previous study, the Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair (FOCUS)⁸ trial, did not report differences in patients’ fatigue levels at 30 and 60 days postdischarge when transfused at restrictive (8 g/dL) compared with liberal (10 g/dL) Hb thresholds, confidence in the validity of this finding is reduced by the fact that more than half of the patients were lost to follow-up at the 30- and 60-day time points. Further, patients in the restrictive transfusion arm of FOCUS were transfused to maintain Hb levels at or above 8 g/dL. This transfusion threshold of 8 g/dL may have mitigated the high levels of fatigue that are seen in our study when patients’ Hb drops below 8 g/dL, and maintaining a Hb level of 7 g/dL is now the standard of care in stable hospitalized patients. Lastly, FOCUS was limited to postoperative hip fracture patients, and the generalizability of FOCUS to hospitalized medicine patients with anemia is limited.

Therefore, our results support guideline suggestions that practitioners incorporate the presence of patient symptoms such as fatigue into transfusion decisions, particularly if patients’ Hb is <8 g/dL.¹⁸ Though reasonable, the suggestion to incorporate symptoms such as fatigue into transfusion decisions has not been strongly supported by evidence so far, and it may often be neglected in practice. Definitive evidence to support such recommendations would benefit from study through an optimal trial¹⁸ that incorporates symptoms into decision making. Our findings add support for a study of transfusion strategies that incorporates patients’ fatigue level in addition to Hb concentration.

This study has several limitations. Although our sample size is large and includes patients with a range of comorbidities that we believe are representative of hospitalized general medicine patients, as a single-center, observational study, our results may not be generalizable to other centers. Additionally, although these data support a reliable association between hospitalized patients’ minimum Hb and fatigue level, the observational design of this study cannot prove that this relationship is causal. Also, patients’ Hb values were measured at the discretion of their clinician, and therefore, the measures of Hb were not uniformly measured for participating patients. In addition, fatigue was only measured at one time point during a patient’s hospitalization, and it is possible that patients’ fatigue levels change during hospitalization in relation to variables we did not consider. Finally, our study was not designed to assess the association of Hb with longer-term functional outcomes, which may be of greater concern than fatigue.

In hospitalized patients ≥50 years old, minimum Hb is reliably associated with patients’ fatigue level. Patients whose minimum Hb is <8 g/dL experience higher fatigue levels compared to patients whose minimum Hb is ≥8 g/dL. Additional studies are warranted to understand if patients may benefit from improved fatigue levels by correcting their anemia through transfusion.

Disclosure 

Dr. Prochaska is supported by an Agency for Healthcare Research and Quality Patient-Centered Outcomes Research Institutional K12 Award (1K12HS023007-01, principal investigator Meltzer). Dr. Meltzer is supported by a National Institutes of Health Clinical and Translational Science Award (2UL1TR000430-06, principal investigator Solway) and a grant from the Patient-Centered Outcomes Research Network in support of the Chicago Patient-Centered Outcomes Research Network. The authors report no conflicts of interest.

Fatigue is the most common clinical symptom of anemia and is a significant concern to patients.^1,2 In ambulatory patients, lower hemoglobin (Hb) concentration is associated with increased fatigue.^2,3 Accordingly, therapies that treat anemia by increasing Hb concentration, such as erythropoiesis stimulating agents,^4-7 often use fatigue as an outcome measure.

In hospitalized patients, transfusion of red blood cell increases Hb concentration and is the primary treatment for anemia. However, the extent to which transfusion and changes in Hb concentration affect hospitalized patients’ fatigue levels is not well established. Guidelines support transfusing patients with symptoms of anemia, such as fatigue, on the assumption that the increased oxygen delivery will improve the symptoms of anemia. While transfusion studies in hospitalized patients have consistently reported that transfusion at lower or “restrictive” Hb concentrations is safe compared with transfusion at higher Hb concentrations,^8-10 these studies have mainly used cardiac events and mortality as outcomes rather than patient symptoms, such as fatigue. Nevertheless, they have resulted in hospitals increasingly adopting restrictive transfusion policies that discourage transfusion at higher Hb levels.^11,12 Consequently, the rate of transfusion in hospitalized patients has decreased,¹³ raising questions of whether some patients with lower Hb concentrations may experience increased fatigue as a result of restrictive transfusion policies. Fatigue among hospitalized patients is important not only because it is an adverse symptom but because it may result in decreased activity levels, deconditioning, and losses in functional status.^14,15While the effect of alternative transfusion policies on fatigue in hospitalized patients could be answered by a randomized clinical trial using fatigue and functional status as outcomes, an important first step is to assess whether the Hb concentration of hospitalized patients is associated with their fatigue level during hospitalization. Because hospitalized patients often have acute illnesses that can cause fatigue in and of themselves, it is possible that anemia is not associated with fatigue in hospitalized patients despite anemia’s association with fatigue in ambulatory patients. Additionally, Hb concentration varies during hospitalization,¹⁶ raising the question of what measures of Hb during hospitalization might be most associated with anemia-related fatigue.

The objective of this study is to explore multiple Hb measures in hospitalized medical patients with anemia and test whether any of these Hb measures are associated with patients’ fatigue level.

Study Design

We performed a prospective, observational study of hospitalized patients with anemia on the general medicine services at The University of Chicago Medical Center (UCMC). The institutional review board approved the study procedures, and all study subjects provided informed consent.

Study Eligibility

Between April 2014 and June 2015, all general medicine inpatients were approached for written consent for The University of Chicago Hospitalist Project,¹⁷ a research infrastructure at UCMC. Among patients consenting to participate in the Hospitalist Project, patients were eligible if they had Hb <9 g/dL at any point during their hospitalization and were age ≥50 years. Hb concentration of <9 g/dL was chosen to include the range of Hb values covered by most restrictive transfusion policies.^8-10,18 Age ≥50 years was an inclusion criteria because anemia is more strongly associated with poor outcomes, including functional impairment, among older patients compared with younger patients.^14,19-21 If patients were not eligible for inclusion at the time of consent for the Hospitalist Project, their Hb values were reviewed twice daily until hospital discharge to assess if their Hb was <9 g/dL. Proxies were sought to answer questions for patients who failed the Short Portable Mental Status Questionnaire.²²

Patient Demographic Data Collection

Research assistants abstracted patient age and sex from the electronic health record (EHR), and asked patients to self-identify their race. The individual comorbidities included as part of the Charlson Comorbidity Index were identified using International Classification of Diseases, 9th Revision codes from hospital administrative data for each encounter and specifically included the following: myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatic disease, peptic ulcer disease, liver disease, diabetes, hemiplegia and/or paraplegia, renal disease, cancer, and human immunodeficiency virus/acquired immunodeficiency syndrome.²³ We also used Healthcare Cost and Utilization Project (www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp) diagnosis categories to identify whether patients had sickle cell (SC) anemia, gastrointestinal bleeding (GIB), or a depressive disorder (DD) because these conditions are associated with anemia (SC and GIB) and fatigue (DD).²⁴

Measuring Anemia

Hb measures were available only when hospital providers ordered them as part of routine practice. The first Hb concentration <9 g/dL during a patient’s hospitalization, which made them eligible for study participation, was obtained through manual review of the EHR. All additional Hb values during the patient’s hospitalization were obtained from the hospital’s administrative data mart. All Hb values collected for each patient during the hospitalization were used to calculate summary measures of Hb during the hospitalization, including the mean Hb, median Hb, minimum Hb, maximum Hb, admission (first recorded) Hb, and discharge (last recorded) Hb. Hb measures were analyzed both as a continuous variable and as a categorical variable created by dividing the continuous Hb measures into integer ranges of 3 groups of approximately the same size.

Measuring Fatigue

Our primary outcome was patients’ level of fatigue reported during hospitalization, measured using the Functional Assessment of Chronic Illness Therapy (FACIT)-Anemia questionnaire. Fatigue was measured using a 13-question fatigue subscale,^1,2,25 which measures fatigue within the past 7 days. Scores on the fatigue subscale range from 0 to 52, with lower scores reflecting greater levels of fatigue. As soon as patients met the eligibility criteria for study participation during their hospitalization (age ≥50 years and Hb <9 g/dL), they were approached to answer the FACIT questions. Values for missing data in the fatigue subscale for individual subjects were filled in using a prorated score from their answered questions as long as >50% of the items in the fatigue subscale were answered, in accordance with recommendations for addressing missing data in the FACIT.²⁶ Fatigue was analyzed as a continuous variable and as a dichotomous variable created by dividing the sample into high (FACIT <27) and low (FACIT ≥27) levels of fatigue based on the median FACIT score of the population. Previous literature has shown a FACIT fatigue subscale score between 23 and 26 to be associated with an Eastern Cooperative Oncology Group (ECOG)²⁷ C Performance Status rating of 2 to 3³ compared to scores ≥27.

Statistical Analysis

Statistical analysis was performed using Stata statistical software (StataCorp, College Station, Texas). Descriptive statistics were used to characterize patient demographics. Analysis of variance was used to test for differences in the mean fatigue levels across Hb measures. χ² tests were performed to test for associations between high fatigue levels and the Hb measures. Multivariable analysis, including both linear and logistic regression models, were used to test the association of Hb concentration and fatigue. P values <0.05 using a 2-tailed test were deemed statistically significant.

Patient Characteristics

During the study period, 8559 patients were admitted to the general medicine service. Of those, 5073 (59%) consented for participation in the Hospitalist Project, and 3670 (72%) completed the Hospitalist Project inpatient interview. Of these patients, 1292 (35%) had Hb <9 g/dL, and 784 (61%) were 50 years or older and completed the FACIT questionnaire.

Table 1 reports the demographic characteristics and comorbidities for the sample, the mean (standard deviation [SD]) for the 6 Hb measures, and mean (SD) and median FACIT scores.

Bivariate Association of Fatigue and Hb

Categorizing patients into low, middle, or high Hb for each of the 6 Hb measures, minimum Hb was strongly associated with fatigue, with a weaker association for mean Hb and no statistically significant association for the other measures.

Minimum Hb

Patients with a minimum Hb <7 g/dL and patients with Hb 7-8 g/dL had higher fatigue levels (FACIT = 25 for each) than patients with a minimum Hb ≥8 g/dL (FACIT = 29; P < 0.001; Table 2). When excluding patients with SC and/or GIB because their average minimum Hb differed from the average minimum Hb of the full population (P < 0.001), patients with a minimum Hb <7 g/dL or 7-8 g/dL had even higher fatigue levels (FACIT = 23 and FACIT = 24, respectively), with no change in the fatigue level of patients with a minimum Hb ≥8 g/dL (FACIT = 29; P < 0.001; Table 2). Lower minimum Hb continued to be associated with higher fatigue levels when analyzed in 0.5 g/dL increments (Figure).

Mean Hb and Other Measures

Fatigue levels were high for 47% of patients with a mean Hb <8g /dL and 53% of patients with a mean Hb 8-9 g/dL compared with 43% of patients with a mean Hb ≥9 g/dL (P = 0.05). However, the association between high fatigue and mean Hb was not statistically significant when patients with SC and/or GIB were excluded (Table 2). None of the other 4 Hb measures was significantly associated with fatigue.

Linear Regression of Fatigue on Hb

In linear regression models, minimum Hb consistently predicted patient fatigue, mean Hb had a less robust association with fatigue, and the other Hb measures were not associated with patient fatigue. Increases in minimum Hb (analyzed as a continuous variable) were associated with reduced fatigue (higher FACIT score; β = 1.4; P = 0.005). In models in which minimum Hb was a categorical variable, patients with a minimum Hb of <7 g/dL or 7-8 g/dL had greater fatigue (lower FACIT score) than patients whose minimum Hb was ≥8 g/dL (Hb <7 g/dL: β = −4.2; P ≤ 0.001; Hb 7-8 g/dL: β = −4.1; P < 0.001). These results control for patients’ age, sex, individual comorbidities, and whether their minimum Hb occurred before or after the measurement of fatigue during hospitalization (Model 1), and the results are unchanged when also controlling for the number of Hb laboratory draws patients had during their hospitalization (Model 2; Table 3). In a stratified analysis excluding patients with either SC and/or GIB, changes in minimum Hb were associated with larger changes in patient fatigue levels (Supplemental Table 1). We also stratified our analysis to include only patients whose minimum Hb occurred before the measurement of their fatigue level during hospitalization to avoid a spurious association of fatigue with minimum Hb occurring after fatigue was measured. In both Models 1 and 2, minimum Hb remained a predictor of patients’ fatigue levels with similar effect sizes, although in Model 2, the results did not quite reach a statistically significant level, in part due to larger confidence intervals from the smaller sample size of this stratified analysis (Supplemental Table 2a). We further stratified this analysis to include only patients whose transfusion, if they received one, occurred after their minimum Hb and the measurement of their fatigue level to account for the possibility that a transfusion could affect the fatigue level patients report. In this analysis, most of the estimates of the effect of minimum Hb on fatigue were larger than those seen when only analyzing patients whose minimum Hb occurred before the measurement of their fatigue level, although again, the smaller sample size of this additional stratified analysis does produce larger confidence intervals for these estimates (Supplemental Table 2b).

Logistic Regression of High Fatigue Level on Hb

Using logistic regression, minimum Hb analyzed as a categorical variable predicted increased odds of a high fatigue level. Patients with a minimum Hb <7 g/dL were 50% (odds ratio [OR] = 1.5; P = 0.03) more likely to have high fatigue and patients with a minimum Hb 7-8 g/dL were 90% (OR = 1.9; P < 0.001) more likely to have high fatigue compared with patients with a minimum Hb ≥8 g/dL in Model 1. These results were similar in Model 2, although the effect was only statistically significant in the 7-8 g/dL Hb group (Table 3). When excluding SC and/or GIB patients, the odds of having high fatigue as minimum Hb decreased were the same or higher for both models compared to the full population of patients. However, again, in Model 2, the effect was only statistically significant in the 7-8 g/dL Hb group (Supplemental Table 1).

Patients with a mean Hb <8 g/dL were 20% to 30% more likely to have high fatigue and patients with mean Hb 8-9 g/dL were 50% more likely to have high fatigue compared with patients with a mean Hb ≥9 g/dL, but the effects were only statistically significant for patients with a mean Hb 8-9 g/dL in both Models 1 and 2 (Table 3). These results were similar when excluding patients with SC and/or GIB, but they were only significant for patients with a mean Hb 8-9 g/dL in Model 1 and patients with a mean Hb <8 g/dL in the Model 2 (Supplemental Table 3).

These results demonstrate that minimum Hb during hospitalization is associated with fatigue in hospitalized patients age ≥50 years, and the association is stronger among patients without SC and/or GIB as comorbidities. The analysis of Hb as a continuous and categorical variable and the use of both linear and logistic regression models support the robustness of these associations and illuminate their clinical significance. For example, in linear regression with minimum Hb a continuous variable, the coefficient of 1.4 suggests that an increase of 2 g/dL in Hb, as might be expected from transfusion of 2 units of red blood cells, would be associated with about a 3-point improvement in fatigue. Additionally, as a categorical variable, a minimum Hb ≥8 g/dL compared with a minimum Hb <7 g/dL or 7-8 g/dL is associated with a 3- to 4-point improvement in fatigue. Previous literature suggests that a difference of 3 in the FACIT score is the minimum clinically important difference in fatigue,³ and changes in minimum Hb in either model predict changes in fatigue that are in the range of potential clinical significance.

The clinical significance of the findings is also reflected in the results of the logistic regressions, which may be mapped to potential effects on functional status. Specifically, the odds of having a high fatigue level (FACIT <27) increase 90% for persons with a minimum Hb 7–8 g/dL compared with persons with a minimum Hb ≥8 g/dL. For persons with a minimum Hb <7 g/dL, point estimates suggest a smaller (50%) increase in the odds of high fatigue, but the 95% confidence interval overlaps heavily with the estimate of patients whose minimum Hb is 7-8 g/dL. While it might be expected that patients with a minimum Hb <7 g/dL have greater levels of fatigue compared with patients with a minimum Hb 7-8 g/dL, we did not observe such a pattern. One reason may be that the confidence intervals of our estimated effects are wide enough that we cannot exclude such a pattern. Another possible explanation is that in both groups, the fatigue levels are sufficiently severe, such that the difference in their fatigue levels may not be clinically meaningful. For example, a FACIT score of 23 to 26 has been shown to be associated with an ECOG performance status of 2 to 3, requiring bed rest for at least part of the day.³ Therefore, patients with a minimum Hb 7-8 g/dL (mean FACIT score = 24; Table 2) or a minimum Hb of <7 g/dL (mean FACIT score = 23; Table 2) are already functionally limited to the point of being partially bed bound, such that further decreases in their Hb may not produce additional fatigue in part because they reduce their activity sufficiently to prevent an increase in fatigue. In such cases, the potential benefits of increased Hb may be better assessed by measuring fatigue in response to a specific and provoked activity level, a concept known as fatigability.²⁰

That minimum Hb is more strongly associated with fatigue than any other measure of Hb during hospitalization may not be surprising. Mean, median, maximum, and discharge Hb may all be affected by transfusion during hospitalization that could affect fatigue. Admission Hb may not reflect true oxygen-carrying capacity because of hemoconcentration.

The association between Hb and fatigue in hospitalized patients is important because increased fatigue could contribute to slower clinical recovery in hospitalized patients. Additionally, increased fatigue during hospitalization and at hospital discharge could exacerbate the known deleterious consequences of fatigue on patients and their health outcomes^14,15 after hospital discharge. Although one previous study, the Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair (FOCUS)⁸ trial, did not report differences in patients’ fatigue levels at 30 and 60 days postdischarge when transfused at restrictive (8 g/dL) compared with liberal (10 g/dL) Hb thresholds, confidence in the validity of this finding is reduced by the fact that more than half of the patients were lost to follow-up at the 30- and 60-day time points. Further, patients in the restrictive transfusion arm of FOCUS were transfused to maintain Hb levels at or above 8 g/dL. This transfusion threshold of 8 g/dL may have mitigated the high levels of fatigue that are seen in our study when patients’ Hb drops below 8 g/dL, and maintaining a Hb level of 7 g/dL is now the standard of care in stable hospitalized patients. Lastly, FOCUS was limited to postoperative hip fracture patients, and the generalizability of FOCUS to hospitalized medicine patients with anemia is limited.

Therefore, our results support guideline suggestions that practitioners incorporate the presence of patient symptoms such as fatigue into transfusion decisions, particularly if patients’ Hb is <8 g/dL.¹⁸ Though reasonable, the suggestion to incorporate symptoms such as fatigue into transfusion decisions has not been strongly supported by evidence so far, and it may often be neglected in practice. Definitive evidence to support such recommendations would benefit from study through an optimal trial¹⁸ that incorporates symptoms into decision making. Our findings add support for a study of transfusion strategies that incorporates patients’ fatigue level in addition to Hb concentration.

This study has several limitations. Although our sample size is large and includes patients with a range of comorbidities that we believe are representative of hospitalized general medicine patients, as a single-center, observational study, our results may not be generalizable to other centers. Additionally, although these data support a reliable association between hospitalized patients’ minimum Hb and fatigue level, the observational design of this study cannot prove that this relationship is causal. Also, patients’ Hb values were measured at the discretion of their clinician, and therefore, the measures of Hb were not uniformly measured for participating patients. In addition, fatigue was only measured at one time point during a patient’s hospitalization, and it is possible that patients’ fatigue levels change during hospitalization in relation to variables we did not consider. Finally, our study was not designed to assess the association of Hb with longer-term functional outcomes, which may be of greater concern than fatigue.

In hospitalized patients ≥50 years old, minimum Hb is reliably associated with patients’ fatigue level. Patients whose minimum Hb is <8 g/dL experience higher fatigue levels compared to patients whose minimum Hb is ≥8 g/dL. Additional studies are warranted to understand if patients may benefit from improved fatigue levels by correcting their anemia through transfusion.

Disclosure 

Dr. Prochaska is supported by an Agency for Healthcare Research and Quality Patient-Centered Outcomes Research Institutional K12 Award (1K12HS023007-01, principal investigator Meltzer). Dr. Meltzer is supported by a National Institutes of Health Clinical and Translational Science Award (2UL1TR000430-06, principal investigator Solway) and a grant from the Patient-Centered Outcomes Research Network in support of the Chicago Patient-Centered Outcomes Research Network. The authors report no conflicts of interest.

Acute viral bronchiolitis is the most common cause of hospitalization for children less than 1 year of age.¹ Overuse of ineffective therapies has persisted despite the existence of the evidence-based American Academy of Pediatrics (AAP) clinical practice guideline (CPG), which recommends primarily supportive care.^2-8 Adherence to the AAP CPG recommendations for management of bronchiolitis improved significantly through the AAP’s Bronchiolitis Quality Improvement Project (BQIP), a 12-month, multiinstitutional collaborative of community and free-standing children’s hospitals.⁹ This subsequent study investigates if these improvements were sustained after completion of the formal 12-month project.

Published multiinstitutional bronchiolitis quality improvement (QI) work is limited to 1 study⁵ that describes the results of a single intervention season at academic medical centers. Multiyear bronchiolitis QI projects are limited to single-center studies, and results have been mixed.^5,6,8,10-13 One study¹¹ observed continued improvement in bronchodilator use in subsequent seasons, whereas a second study¹⁰ observed a return to baseline bronchodilator use in the following season. Mittal⁶ observed inconsistent improvements in key bronchiolitis measures during postintervention seasons.

Our specific aim was to assess the sustainability of improvements in bronchiolitis management at participating institutions 1 year following completion of the AAP BQIP collaborative.⁹ Because no studies demonstrate the most effective way to support long-term improvement through a QI collaborative, we hypothesized that the initial collaborative activities, which were designed to build the capacity of local interdisciplinary teams while providing standardized evidence-based care pathways, would lead to performance in the subsequent season at levels similar to or better than those observed during the active phase of the collaborative, without additional project interventions.

Study Design and Setting

This was a follow-up study of the AAP Quality Improvement Innovation Networks project entitled “A Quality Collaborative for Improving Hospital Compliance with the AAP Bronchiolitis Guideline” (BQIP).⁹ The AAP Institutional Review Board approved this project.

Twenty-one multidisciplinary, hospital-based teams participated in the BQIP collaborative and provided monthly data during the January through March bronchiolitis season. Teams submitted 2013 baseline data and 2014 intervention data. Nine sites provided 2015 sustainability data following the completion of the collaborative.

Participants

Hospital encounters with a primary diagnosis of acute viral bronchiolitis were eligible for inclusion among patients from 1 month to 2 years of age. Encounters were excluded for prematurity (<35 weeks gestational age), congenital heart disease, bronchopulmonary dysplasia, genetic, congenital or neuromuscular abnormalities, and pediatric intensive-care admission.

Data Collection

Hospital characteristics were collected, including hospital type (academic, community), bed size, location (urban, rural), hospital distributions of race/ethnicity and public payer, cases of bronchiolitis per year, presence of an electronic medical record and a pediatric respiratory therapist, and self-rated QI knowledge of the multidisciplinary team (very knowledgeable, knowledgeable, and somewhat knowledgeable). A trained member at each site collected data through structured chart review in baseline, intervention, and sustainability bronchiolitis seasons for January, February, and March. Site members reviewed the first 20 charts per month that met the inclusion criteria or all charts if there were fewer than 20 eligible encounters. Sites input data about key quality measures into the AAP’s Quality Improvement Data Aggregator, a web-based data repository.

Intervention

The BQIP project was designed as a virtual collaborative consisting of monthly education webinars about QI methods and bronchiolitis management, opportunities for collaboration via teleconference and e-mail listserv, and individual site-coaching by e-mail or telephone.⁹ A change package was shared with sites that included examples of evidence-based pathways, ordersets, a respiratory scoring tool, communication tools for parents and referring physicians, and slide sets for individual site education efforts. Following completion of the collaborative, written resources remained available to participants, although virtual collaboration ceased and no additional project interventions to promote sustainability were introduced.

Bronchiolitis Process and Outcome Measures

Process measures following admission included the following: severity assessment using a respiratory score, respiratory score use to assess response to bronchodilators, bronchodilator use, bronchodilator doses, steroid doses per patient encounter, chest radiographs per encounter, and presence of an order to transition to intermittent pulse oximetry monitoring. Outcome measures included length of stay and readmissions within 72 hours.

Analysis

Changes among baseline-, intervention-, and sustainability-season data were assessed using generalized linear mixed-effects models with random effect for study sites. Negative binomial models were used for count variables to allow for overdispersion. Length of stay was log-transformed to achieve a normal distribution. We also analyzed each site individually to assess whether sustained improvements were the result of broad sustainability across all sites or whether they represented an aggregation of some sites that continued to improve while other sites actually worsened.

To address any bias introduced by the voluntary and incomplete participation of sites in the sustainability season, we planned a priori to conduct 3 additional analyses. First, we compared the characteristics of sites that did participate in the sustainability season with those that did not participate by using Chi-squared tests for differences in proportions and t tests for differences in means. Second, we determined whether the baseline-season process and outcome measures were different between sites that did and did not participate using descriptive statistics. Third, we assessed whether improvements between the baseline and intervention seasons were different between sites that did and did not participate using a linear mixed-effects model for normally distributed outcomes and generalized linear mixed-effects model with site-specific random effects for nonnormally distributed outcomes. All study outcomes were summarized in terms of model-adjusted means along with the corresponding 95% confidence intervals. All P values are 2-sided, and P < 0.05 was used to define statistical significance. Data analyses were conducted using SAS software (SAS Institute Inc., Cary, North Carolina) version 9.4.

To our knowledge, this is the first report of sustained improvements in care achieved through a multiinstitutional QI collaborative of community and academic hospitals focused on bronchiolitis care. We found that overall sites participating in a national bronchiolitis QI project sustained improvements in key bronchiolitis quality measures for 1 year following the project’s completion. For the aggregate group no measures worsened, and one measure, orders for intermittent pulse oximetry monitoring, continued to increase during the sustainability season. Furthermore, the sustained improvements were primarily the result of consistent sustained performance of each individual site, as opposed to averages wherein some sites worsened while others improved (Appendix 1a). These findings suggest that designing a collaborative approach, which provides an evidence-based best-practice toolkit while building the QI capacity of local interdisciplinary teams, can support performance gains that persist beyond the project’s active phase.

There are a number of possible reasons why improvements were sustained following the collaborative. The BQIP requirement for institutional leadership buy-in may have motivated accountability to local leaders in subsequent bronchiolitis seasons at each site. We suspect that culture change such as flattened hierarchies through multidisciplinary teams,¹⁴ which empowered nurse and respiratory therapy staff, may have facilitated consistent use of tools created locally. The synergy of interdisciplinary teams composed of physician, nurse, and respiratory therapy champions may have created accountability to perpetuate the previous year’s efforts.¹⁵ In addition, the sites adopted elements of the evidence-based toolkit, such as pathways,^16,17 forcing function tools^13,18 and order sets that limited management decision options and bronchodilator use contingent on respiratory scores,^9,19 which may have driven desired behaviors.

Moreover, the 2014 AAP CPG for the management of bronchiolitis,²⁰ released prior to the sustainability bronchiolitis season, may have underscored the key concepts of the collaborative. Similarly, national exposure of best practices for bronchiolitis management, including the 3 widespread Choosing Wisely recommendations related to bronchiolitis,²¹ might have been a compelling reason for sites to maintain their improvement efforts and contribute to secular trends toward decreasing interventions in bronchiolitis management nationally.³ Lastly, the mechanisms developed for local data collection may have created opportunities at each site to conduct ongoing evaluation of performance on key bronchiolitis quality measures through data-driven feedback systems.²² Our study highlights the need for additional research in order to understand why improvements are or are not sustained.

Even with substantial, sustained improvements in this initiative, further reduction in unnecessary care may be possible. Findings from previous studies suggest that even multifaceted QI interventions, including provider education, guidelines and use of respiratory scores, may only modestly reduce bronchodilators, steroids, and chest radiograph use.^8,13 To achieve continued improvements in bronchiolitis care, additional active efforts may be needed to develop new interventions that target root causes for areas of overuse at individual sites.

Future multiinstitutional collaboratives might benefit their participants if they include a focus on helping sites develop skills to ensure that local improvement activities continue after the collaborative phases are completed. Proactively scheduling intermittent check-ins with collaborative members to discuss experiences with both sustainability and ongoing improvement may be valuable and likely needs to be incorporated into the initial collaborative planning.

As these sustainability data represent a subset of 9 of the original 21 BQIP sites, there is concern for potential selection bias related to factors that could have motivated sites to participate in the sustainability season’s data collection and simultaneously influenced their performance. These concerns were mitigated to some extent through 3 specific analyses: finding limited differences in hospital characteristics, baseline performance in key bronchiolitis measures, and performance change from baseline to intervention seasons between sites that did and did not participate in the sustainability season.

Notably, sites that participated in the sustainability phase actually had lower baseline respiratory score use and fewer orders for intermittent pulse oximetry at baseline. Theoretically, if participation in the collaborative highlighted this disparity for these sites, it could have been a motivating factor for their continued participation and sustained performance across these measures. Similarly, sites that recognized their higher baseline performance through participation in the collaborative might have felt less motivation to participate in ongoing data collection during the sustainability season. Whether they might have also sustained, declined, or continued improving is not known. Additionally, the magnitude of improvement in the collaborative period might have also motivated ongoing participation during the sustainability phase. For example, although all sites improved in score use during the collaborative, sites participating in the sustainability season demonstrated significantly more improvement in these measures. Sites with a higher magnitude of improvement in collaborative measures might have more enthusiasm about the project, more commitment to the project activities, or feel a sense of obligation to respond to requests for additional data collection.

This work has several limitations. Selection bias may limit generalizability of the results, as sites that did not participate in the sustainability season may have had different results than those that did participate. It is unknown whether sites that regressed toward their baseline were deterred from participating in the sustainability season. The analyses that we were able to preform, however, suggest that the 2 groups were similar in their characteristics as well as in their baseline and improvement performance.

We have limited knowledge of the local improvement work that sites conducted between the completion of the collaborative and the sustainability season. Site-specific factors may have influenced improvement sustainability. For example, qualitative research with the original group found that team engagement had a quantitative association with better performance, but only for the bronchodilator use measure.²³ Sites were responsible for their own data collection, and despite attempts to centralize and standardize the process, data collection inconsistencies may have occurred. For instance, it is unknown how closely that orders for intermittent pulse oximetry correlate with intermittent use at the bedside. Lastly, the absence of a control group limits examination of the causal relationships of interventions and the influence of secular trends.

Improvements gained during the BQIP collaborative were sustained at 1 year following completion of the collaborative. These findings are encouraging, as national QI collaborative efforts are increasingly common. Our findings suggest that opportunities exist to even further reduce unnecessary care in the management of bronchiolitis. Such opportunities highlight the importance of integrating strategies to both measure sustainability and plan for ongoing independent local activities after completion of the collaborative. Future efforts should focus on supporting local sites to continue individual practice-improvement as they transition from collaborative to independent quality initiatives.

Acknowledgments

The authors thank the 21 hospitals that participated in the BQIP collaborative, and in particular the 9 hospital teams that contributed sustainability data for their ongoing dedication. There was no external funding for this manuscript.

Disclosure

The authors report no financial conflicts of interest.

Acute viral bronchiolitis is the most common cause of hospitalization for children less than 1 year of age.¹ Overuse of ineffective therapies has persisted despite the existence of the evidence-based American Academy of Pediatrics (AAP) clinical practice guideline (CPG), which recommends primarily supportive care.^2-8 Adherence to the AAP CPG recommendations for management of bronchiolitis improved significantly through the AAP’s Bronchiolitis Quality Improvement Project (BQIP), a 12-month, multiinstitutional collaborative of community and free-standing children’s hospitals.⁹ This subsequent study investigates if these improvements were sustained after completion of the formal 12-month project.

Published multiinstitutional bronchiolitis quality improvement (QI) work is limited to 1 study⁵ that describes the results of a single intervention season at academic medical centers. Multiyear bronchiolitis QI projects are limited to single-center studies, and results have been mixed.^5,6,8,10-13 One study¹¹ observed continued improvement in bronchodilator use in subsequent seasons, whereas a second study¹⁰ observed a return to baseline bronchodilator use in the following season. Mittal⁶ observed inconsistent improvements in key bronchiolitis measures during postintervention seasons.

Our specific aim was to assess the sustainability of improvements in bronchiolitis management at participating institutions 1 year following completion of the AAP BQIP collaborative.⁹ Because no studies demonstrate the most effective way to support long-term improvement through a QI collaborative, we hypothesized that the initial collaborative activities, which were designed to build the capacity of local interdisciplinary teams while providing standardized evidence-based care pathways, would lead to performance in the subsequent season at levels similar to or better than those observed during the active phase of the collaborative, without additional project interventions.

Study Design and Setting

This was a follow-up study of the AAP Quality Improvement Innovation Networks project entitled “A Quality Collaborative for Improving Hospital Compliance with the AAP Bronchiolitis Guideline” (BQIP).⁹ The AAP Institutional Review Board approved this project.

Twenty-one multidisciplinary, hospital-based teams participated in the BQIP collaborative and provided monthly data during the January through March bronchiolitis season. Teams submitted 2013 baseline data and 2014 intervention data. Nine sites provided 2015 sustainability data following the completion of the collaborative.

Participants

Hospital encounters with a primary diagnosis of acute viral bronchiolitis were eligible for inclusion among patients from 1 month to 2 years of age. Encounters were excluded for prematurity (<35 weeks gestational age), congenital heart disease, bronchopulmonary dysplasia, genetic, congenital or neuromuscular abnormalities, and pediatric intensive-care admission.

Data Collection

Hospital characteristics were collected, including hospital type (academic, community), bed size, location (urban, rural), hospital distributions of race/ethnicity and public payer, cases of bronchiolitis per year, presence of an electronic medical record and a pediatric respiratory therapist, and self-rated QI knowledge of the multidisciplinary team (very knowledgeable, knowledgeable, and somewhat knowledgeable). A trained member at each site collected data through structured chart review in baseline, intervention, and sustainability bronchiolitis seasons for January, February, and March. Site members reviewed the first 20 charts per month that met the inclusion criteria or all charts if there were fewer than 20 eligible encounters. Sites input data about key quality measures into the AAP’s Quality Improvement Data Aggregator, a web-based data repository.

Intervention

The BQIP project was designed as a virtual collaborative consisting of monthly education webinars about QI methods and bronchiolitis management, opportunities for collaboration via teleconference and e-mail listserv, and individual site-coaching by e-mail or telephone.⁹ A change package was shared with sites that included examples of evidence-based pathways, ordersets, a respiratory scoring tool, communication tools for parents and referring physicians, and slide sets for individual site education efforts. Following completion of the collaborative, written resources remained available to participants, although virtual collaboration ceased and no additional project interventions to promote sustainability were introduced.

Bronchiolitis Process and Outcome Measures

Process measures following admission included the following: severity assessment using a respiratory score, respiratory score use to assess response to bronchodilators, bronchodilator use, bronchodilator doses, steroid doses per patient encounter, chest radiographs per encounter, and presence of an order to transition to intermittent pulse oximetry monitoring. Outcome measures included length of stay and readmissions within 72 hours.

Analysis

Changes among baseline-, intervention-, and sustainability-season data were assessed using generalized linear mixed-effects models with random effect for study sites. Negative binomial models were used for count variables to allow for overdispersion. Length of stay was log-transformed to achieve a normal distribution. We also analyzed each site individually to assess whether sustained improvements were the result of broad sustainability across all sites or whether they represented an aggregation of some sites that continued to improve while other sites actually worsened.

To address any bias introduced by the voluntary and incomplete participation of sites in the sustainability season, we planned a priori to conduct 3 additional analyses. First, we compared the characteristics of sites that did participate in the sustainability season with those that did not participate by using Chi-squared tests for differences in proportions and t tests for differences in means. Second, we determined whether the baseline-season process and outcome measures were different between sites that did and did not participate using descriptive statistics. Third, we assessed whether improvements between the baseline and intervention seasons were different between sites that did and did not participate using a linear mixed-effects model for normally distributed outcomes and generalized linear mixed-effects model with site-specific random effects for nonnormally distributed outcomes. All study outcomes were summarized in terms of model-adjusted means along with the corresponding 95% confidence intervals. All P values are 2-sided, and P < 0.05 was used to define statistical significance. Data analyses were conducted using SAS software (SAS Institute Inc., Cary, North Carolina) version 9.4.

To our knowledge, this is the first report of sustained improvements in care achieved through a multiinstitutional QI collaborative of community and academic hospitals focused on bronchiolitis care. We found that overall sites participating in a national bronchiolitis QI project sustained improvements in key bronchiolitis quality measures for 1 year following the project’s completion. For the aggregate group no measures worsened, and one measure, orders for intermittent pulse oximetry monitoring, continued to increase during the sustainability season. Furthermore, the sustained improvements were primarily the result of consistent sustained performance of each individual site, as opposed to averages wherein some sites worsened while others improved (Appendix 1a). These findings suggest that designing a collaborative approach, which provides an evidence-based best-practice toolkit while building the QI capacity of local interdisciplinary teams, can support performance gains that persist beyond the project’s active phase.

There are a number of possible reasons why improvements were sustained following the collaborative. The BQIP requirement for institutional leadership buy-in may have motivated accountability to local leaders in subsequent bronchiolitis seasons at each site. We suspect that culture change such as flattened hierarchies through multidisciplinary teams,¹⁴ which empowered nurse and respiratory therapy staff, may have facilitated consistent use of tools created locally. The synergy of interdisciplinary teams composed of physician, nurse, and respiratory therapy champions may have created accountability to perpetuate the previous year’s efforts.¹⁵ In addition, the sites adopted elements of the evidence-based toolkit, such as pathways,^16,17 forcing function tools^13,18 and order sets that limited management decision options and bronchodilator use contingent on respiratory scores,^9,19 which may have driven desired behaviors.

Moreover, the 2014 AAP CPG for the management of bronchiolitis,²⁰ released prior to the sustainability bronchiolitis season, may have underscored the key concepts of the collaborative. Similarly, national exposure of best practices for bronchiolitis management, including the 3 widespread Choosing Wisely recommendations related to bronchiolitis,²¹ might have been a compelling reason for sites to maintain their improvement efforts and contribute to secular trends toward decreasing interventions in bronchiolitis management nationally.³ Lastly, the mechanisms developed for local data collection may have created opportunities at each site to conduct ongoing evaluation of performance on key bronchiolitis quality measures through data-driven feedback systems.²² Our study highlights the need for additional research in order to understand why improvements are or are not sustained.

Even with substantial, sustained improvements in this initiative, further reduction in unnecessary care may be possible. Findings from previous studies suggest that even multifaceted QI interventions, including provider education, guidelines and use of respiratory scores, may only modestly reduce bronchodilators, steroids, and chest radiograph use.^8,13 To achieve continued improvements in bronchiolitis care, additional active efforts may be needed to develop new interventions that target root causes for areas of overuse at individual sites.

Future multiinstitutional collaboratives might benefit their participants if they include a focus on helping sites develop skills to ensure that local improvement activities continue after the collaborative phases are completed. Proactively scheduling intermittent check-ins with collaborative members to discuss experiences with both sustainability and ongoing improvement may be valuable and likely needs to be incorporated into the initial collaborative planning.

As these sustainability data represent a subset of 9 of the original 21 BQIP sites, there is concern for potential selection bias related to factors that could have motivated sites to participate in the sustainability season’s data collection and simultaneously influenced their performance. These concerns were mitigated to some extent through 3 specific analyses: finding limited differences in hospital characteristics, baseline performance in key bronchiolitis measures, and performance change from baseline to intervention seasons between sites that did and did not participate in the sustainability season.

Notably, sites that participated in the sustainability phase actually had lower baseline respiratory score use and fewer orders for intermittent pulse oximetry at baseline. Theoretically, if participation in the collaborative highlighted this disparity for these sites, it could have been a motivating factor for their continued participation and sustained performance across these measures. Similarly, sites that recognized their higher baseline performance through participation in the collaborative might have felt less motivation to participate in ongoing data collection during the sustainability season. Whether they might have also sustained, declined, or continued improving is not known. Additionally, the magnitude of improvement in the collaborative period might have also motivated ongoing participation during the sustainability phase. For example, although all sites improved in score use during the collaborative, sites participating in the sustainability season demonstrated significantly more improvement in these measures. Sites with a higher magnitude of improvement in collaborative measures might have more enthusiasm about the project, more commitment to the project activities, or feel a sense of obligation to respond to requests for additional data collection.

This work has several limitations. Selection bias may limit generalizability of the results, as sites that did not participate in the sustainability season may have had different results than those that did participate. It is unknown whether sites that regressed toward their baseline were deterred from participating in the sustainability season. The analyses that we were able to preform, however, suggest that the 2 groups were similar in their characteristics as well as in their baseline and improvement performance.

We have limited knowledge of the local improvement work that sites conducted between the completion of the collaborative and the sustainability season. Site-specific factors may have influenced improvement sustainability. For example, qualitative research with the original group found that team engagement had a quantitative association with better performance, but only for the bronchodilator use measure.²³ Sites were responsible for their own data collection, and despite attempts to centralize and standardize the process, data collection inconsistencies may have occurred. For instance, it is unknown how closely that orders for intermittent pulse oximetry correlate with intermittent use at the bedside. Lastly, the absence of a control group limits examination of the causal relationships of interventions and the influence of secular trends.

Improvements gained during the BQIP collaborative were sustained at 1 year following completion of the collaborative. These findings are encouraging, as national QI collaborative efforts are increasingly common. Our findings suggest that opportunities exist to even further reduce unnecessary care in the management of bronchiolitis. Such opportunities highlight the importance of integrating strategies to both measure sustainability and plan for ongoing independent local activities after completion of the collaborative. Future efforts should focus on supporting local sites to continue individual practice-improvement as they transition from collaborative to independent quality initiatives.

Acknowledgments

The authors thank the 21 hospitals that participated in the BQIP collaborative, and in particular the 9 hospital teams that contributed sustainability data for their ongoing dedication. There was no external funding for this manuscript.

Disclosure

The authors report no financial conflicts of interest.

Acute viral bronchiolitis is the most common cause of hospitalization for children less than 1 year of age.¹ Overuse of ineffective therapies has persisted despite the existence of the evidence-based American Academy of Pediatrics (AAP) clinical practice guideline (CPG), which recommends primarily supportive care.^2-8 Adherence to the AAP CPG recommendations for management of bronchiolitis improved significantly through the AAP’s Bronchiolitis Quality Improvement Project (BQIP), a 12-month, multiinstitutional collaborative of community and free-standing children’s hospitals.⁹ This subsequent study investigates if these improvements were sustained after completion of the formal 12-month project.

Published multiinstitutional bronchiolitis quality improvement (QI) work is limited to 1 study⁵ that describes the results of a single intervention season at academic medical centers. Multiyear bronchiolitis QI projects are limited to single-center studies, and results have been mixed.^5,6,8,10-13 One study¹¹ observed continued improvement in bronchodilator use in subsequent seasons, whereas a second study¹⁰ observed a return to baseline bronchodilator use in the following season. Mittal⁶ observed inconsistent improvements in key bronchiolitis measures during postintervention seasons.

Our specific aim was to assess the sustainability of improvements in bronchiolitis management at participating institutions 1 year following completion of the AAP BQIP collaborative.⁹ Because no studies demonstrate the most effective way to support long-term improvement through a QI collaborative, we hypothesized that the initial collaborative activities, which were designed to build the capacity of local interdisciplinary teams while providing standardized evidence-based care pathways, would lead to performance in the subsequent season at levels similar to or better than those observed during the active phase of the collaborative, without additional project interventions.

Study Design and Setting

This was a follow-up study of the AAP Quality Improvement Innovation Networks project entitled “A Quality Collaborative for Improving Hospital Compliance with the AAP Bronchiolitis Guideline” (BQIP).⁹ The AAP Institutional Review Board approved this project.

Twenty-one multidisciplinary, hospital-based teams participated in the BQIP collaborative and provided monthly data during the January through March bronchiolitis season. Teams submitted 2013 baseline data and 2014 intervention data. Nine sites provided 2015 sustainability data following the completion of the collaborative.

Participants

Hospital encounters with a primary diagnosis of acute viral bronchiolitis were eligible for inclusion among patients from 1 month to 2 years of age. Encounters were excluded for prematurity (<35 weeks gestational age), congenital heart disease, bronchopulmonary dysplasia, genetic, congenital or neuromuscular abnormalities, and pediatric intensive-care admission.

Data Collection

Hospital characteristics were collected, including hospital type (academic, community), bed size, location (urban, rural), hospital distributions of race/ethnicity and public payer, cases of bronchiolitis per year, presence of an electronic medical record and a pediatric respiratory therapist, and self-rated QI knowledge of the multidisciplinary team (very knowledgeable, knowledgeable, and somewhat knowledgeable). A trained member at each site collected data through structured chart review in baseline, intervention, and sustainability bronchiolitis seasons for January, February, and March. Site members reviewed the first 20 charts per month that met the inclusion criteria or all charts if there were fewer than 20 eligible encounters. Sites input data about key quality measures into the AAP’s Quality Improvement Data Aggregator, a web-based data repository.

Intervention

The BQIP project was designed as a virtual collaborative consisting of monthly education webinars about QI methods and bronchiolitis management, opportunities for collaboration via teleconference and e-mail listserv, and individual site-coaching by e-mail or telephone.⁹ A change package was shared with sites that included examples of evidence-based pathways, ordersets, a respiratory scoring tool, communication tools for parents and referring physicians, and slide sets for individual site education efforts. Following completion of the collaborative, written resources remained available to participants, although virtual collaboration ceased and no additional project interventions to promote sustainability were introduced.

Bronchiolitis Process and Outcome Measures

Process measures following admission included the following: severity assessment using a respiratory score, respiratory score use to assess response to bronchodilators, bronchodilator use, bronchodilator doses, steroid doses per patient encounter, chest radiographs per encounter, and presence of an order to transition to intermittent pulse oximetry monitoring. Outcome measures included length of stay and readmissions within 72 hours.

Analysis

Changes among baseline-, intervention-, and sustainability-season data were assessed using generalized linear mixed-effects models with random effect for study sites. Negative binomial models were used for count variables to allow for overdispersion. Length of stay was log-transformed to achieve a normal distribution. We also analyzed each site individually to assess whether sustained improvements were the result of broad sustainability across all sites or whether they represented an aggregation of some sites that continued to improve while other sites actually worsened.

To address any bias introduced by the voluntary and incomplete participation of sites in the sustainability season, we planned a priori to conduct 3 additional analyses. First, we compared the characteristics of sites that did participate in the sustainability season with those that did not participate by using Chi-squared tests for differences in proportions and t tests for differences in means. Second, we determined whether the baseline-season process and outcome measures were different between sites that did and did not participate using descriptive statistics. Third, we assessed whether improvements between the baseline and intervention seasons were different between sites that did and did not participate using a linear mixed-effects model for normally distributed outcomes and generalized linear mixed-effects model with site-specific random effects for nonnormally distributed outcomes. All study outcomes were summarized in terms of model-adjusted means along with the corresponding 95% confidence intervals. All P values are 2-sided, and P < 0.05 was used to define statistical significance. Data analyses were conducted using SAS software (SAS Institute Inc., Cary, North Carolina) version 9.4.

To our knowledge, this is the first report of sustained improvements in care achieved through a multiinstitutional QI collaborative of community and academic hospitals focused on bronchiolitis care. We found that overall sites participating in a national bronchiolitis QI project sustained improvements in key bronchiolitis quality measures for 1 year following the project’s completion. For the aggregate group no measures worsened, and one measure, orders for intermittent pulse oximetry monitoring, continued to increase during the sustainability season. Furthermore, the sustained improvements were primarily the result of consistent sustained performance of each individual site, as opposed to averages wherein some sites worsened while others improved (Appendix 1a). These findings suggest that designing a collaborative approach, which provides an evidence-based best-practice toolkit while building the QI capacity of local interdisciplinary teams, can support performance gains that persist beyond the project’s active phase.

There are a number of possible reasons why improvements were sustained following the collaborative. The BQIP requirement for institutional leadership buy-in may have motivated accountability to local leaders in subsequent bronchiolitis seasons at each site. We suspect that culture change such as flattened hierarchies through multidisciplinary teams,¹⁴ which empowered nurse and respiratory therapy staff, may have facilitated consistent use of tools created locally. The synergy of interdisciplinary teams composed of physician, nurse, and respiratory therapy champions may have created accountability to perpetuate the previous year’s efforts.¹⁵ In addition, the sites adopted elements of the evidence-based toolkit, such as pathways,^16,17 forcing function tools^13,18 and order sets that limited management decision options and bronchodilator use contingent on respiratory scores,^9,19 which may have driven desired behaviors.

Moreover, the 2014 AAP CPG for the management of bronchiolitis,²⁰ released prior to the sustainability bronchiolitis season, may have underscored the key concepts of the collaborative. Similarly, national exposure of best practices for bronchiolitis management, including the 3 widespread Choosing Wisely recommendations related to bronchiolitis,²¹ might have been a compelling reason for sites to maintain their improvement efforts and contribute to secular trends toward decreasing interventions in bronchiolitis management nationally.³ Lastly, the mechanisms developed for local data collection may have created opportunities at each site to conduct ongoing evaluation of performance on key bronchiolitis quality measures through data-driven feedback systems.²² Our study highlights the need for additional research in order to understand why improvements are or are not sustained.

Even with substantial, sustained improvements in this initiative, further reduction in unnecessary care may be possible. Findings from previous studies suggest that even multifaceted QI interventions, including provider education, guidelines and use of respiratory scores, may only modestly reduce bronchodilators, steroids, and chest radiograph use.^8,13 To achieve continued improvements in bronchiolitis care, additional active efforts may be needed to develop new interventions that target root causes for areas of overuse at individual sites.

Future multiinstitutional collaboratives might benefit their participants if they include a focus on helping sites develop skills to ensure that local improvement activities continue after the collaborative phases are completed. Proactively scheduling intermittent check-ins with collaborative members to discuss experiences with both sustainability and ongoing improvement may be valuable and likely needs to be incorporated into the initial collaborative planning.

As these sustainability data represent a subset of 9 of the original 21 BQIP sites, there is concern for potential selection bias related to factors that could have motivated sites to participate in the sustainability season’s data collection and simultaneously influenced their performance. These concerns were mitigated to some extent through 3 specific analyses: finding limited differences in hospital characteristics, baseline performance in key bronchiolitis measures, and performance change from baseline to intervention seasons between sites that did and did not participate in the sustainability season.

Notably, sites that participated in the sustainability phase actually had lower baseline respiratory score use and fewer orders for intermittent pulse oximetry at baseline. Theoretically, if participation in the collaborative highlighted this disparity for these sites, it could have been a motivating factor for their continued participation and sustained performance across these measures. Similarly, sites that recognized their higher baseline performance through participation in the collaborative might have felt less motivation to participate in ongoing data collection during the sustainability season. Whether they might have also sustained, declined, or continued improving is not known. Additionally, the magnitude of improvement in the collaborative period might have also motivated ongoing participation during the sustainability phase. For example, although all sites improved in score use during the collaborative, sites participating in the sustainability season demonstrated significantly more improvement in these measures. Sites with a higher magnitude of improvement in collaborative measures might have more enthusiasm about the project, more commitment to the project activities, or feel a sense of obligation to respond to requests for additional data collection.

This work has several limitations. Selection bias may limit generalizability of the results, as sites that did not participate in the sustainability season may have had different results than those that did participate. It is unknown whether sites that regressed toward their baseline were deterred from participating in the sustainability season. The analyses that we were able to preform, however, suggest that the 2 groups were similar in their characteristics as well as in their baseline and improvement performance.

We have limited knowledge of the local improvement work that sites conducted between the completion of the collaborative and the sustainability season. Site-specific factors may have influenced improvement sustainability. For example, qualitative research with the original group found that team engagement had a quantitative association with better performance, but only for the bronchodilator use measure.²³ Sites were responsible for their own data collection, and despite attempts to centralize and standardize the process, data collection inconsistencies may have occurred. For instance, it is unknown how closely that orders for intermittent pulse oximetry correlate with intermittent use at the bedside. Lastly, the absence of a control group limits examination of the causal relationships of interventions and the influence of secular trends.

Improvements gained during the BQIP collaborative were sustained at 1 year following completion of the collaborative. These findings are encouraging, as national QI collaborative efforts are increasingly common. Our findings suggest that opportunities exist to even further reduce unnecessary care in the management of bronchiolitis. Such opportunities highlight the importance of integrating strategies to both measure sustainability and plan for ongoing independent local activities after completion of the collaborative. Future efforts should focus on supporting local sites to continue individual practice-improvement as they transition from collaborative to independent quality initiatives.

Acknowledgments

The authors thank the 21 hospitals that participated in the BQIP collaborative, and in particular the 9 hospital teams that contributed sustainability data for their ongoing dedication. There was no external funding for this manuscript.

Disclosure

The authors report no financial conflicts of interest.

A special concern since the institution of hospital readmission penalties¹ is the transitions in care of a patient from one care setting to another, often at hospital discharge. Burke et al.² proposed a framework for an ideal transition in care (ITC) to study and improve transitions from the hospital to home. The features in the ITC were identified based upon their inclusion in the interventions that improved discharge outcomes.^3-5 Inspired by the ITC and other patient risk tools,⁶ we identified 10 domains of transitional care needs ([TCN] specified below), which we define as patient-centered risk factors that should be addressed to foster a safe and effective transition in care.⁷

One particularly important risk factor in patient self-management at transition points is health literacy, a patient’s ability to obtain, understand, and use basic health information and services. Low health literacy affects approximately 26% to 36% of adults in the United States.^8,9 Health literacy is associated with many factors that may affect successful navigation of care transitions, including doctor-patient communication,^10,11 understanding of the medication regimen,¹² and self-management.^13-15 Research has also demonstrated an association between low health literacy and poor outcomes after hospital discharge, including medication errors,¹⁶ 30-day hospital readmission,¹⁷ and mortality.¹⁸ Transitional care initiatives have begun to incorporate health literacy into patient risk assessments⁶ and provide specific attention to low health literacy in interventions to reduce adverse drug events and readmission.^4,19 Training programs for medical students and nurses advise teaching skills in health literacy as part of fostering effective transitions in care.^20,21

Although low health literacy is generally recognized as a barrier to patient education and self-management, little is known about whether patients with low health literacy are more likely to have other risk factors that could further increase their risk for poor transitions in care. A better understanding of associated risks would inform and improve patient care. We hypothesized that TCNs are more common among patients with low health literacy, as compared with those with adequate health literacy. We also aimed to describe the relationship between low health literacy and specific TCNs in order to guide clinical care and future interventions.

Setting

The present study is a cross-sectional analysis of data from a quality improvement (QI) intervention that was performed at Vanderbilt University Medical Center, a tertiary care facility in Nashville, Tennessee. The QI intervention, My Health Team (MHT), was funded by the Centers for Medicare and Medicaid Services Innovation Award program. The overall MHT program included outpatient care coordination for chronic disease management as well as a transitional care program that was designed to reduce hospital readmission. The latter included an inpatient needs assessment (which provided data for the present analysis), inpatient intervention, and postdischarge phone follow-up. The MHT initiative was reviewed by the institutional review board (IRB), which deemed it a QI program and granted a waiver of informed consent. The present secondary data analysis was reviewed and approved by the IRB.

Sample

Patients were identified for inclusion in the MHT transitions of care program if the presenting problem for hospital admission was pneumonia, chronic obstructive pulmonary disease (COPD) exacerbation, or decompensated heart failure, as determined by the review of clinical documentation by nurse transition care coordinators (TCCs). Adults over the age of 18 years were eligible, though priority was given to patients aged 65 years or older. This study includes the first inpatient encounter between June 2013 and December 31, 2014, for patients having a completed needs assessment and documentation of health literacy data in the medical record.

Data Collection

TCN assessment was developed from published patient risk tools and the ITC framework.^2,6,22 The assessment has 10 domains composed of 49 individual items as follows: (1) caregiver support (caregiver support not sufficient for patient needs), (2) transportation (relies on public or others for transportation and misses medical care because of transportation), (3) health care utilization (no primary care physician, unplanned hospitalization in the last year, emergency department [ED] visit in the last 6 months, or home health services in the last 60 days), (4) high-risk medical comorbidities (malnutrition or body mass index <18.5, renal failure, chronic pain, diabetes, heart failure, COPD, or stroke), (5) medication management provider or caregiver concern (cannot provide medication list, >10 preadmission medications, high-risk medications [eg, insulin, warfarin], poor medication understanding, or adherence issue identified), (6) medical devices (vascular access, urinary catheter, wounds, or home supplemental oxygen), (7) functional status (weakness of extremities, limited extremity range of motion, difficulty with mobility, falls at home, or activities of daily living challenges), (8) mental health comorbidities (over the past month has felt down, depressed, or hopeless or over the past month has felt little interest or pleasure in doing things, high-risk alcohol use, or high-risk substance use), (9) communication (limited English proficiency or at risk for limited health literacy), and (10) financial resources (no health insurance, skips or rations medicines because of cost, misses medical care because of cost, or misses medical care because of job).

The 49 items of the TCN assessment were documented as being present or absent by nurse TCCs at the time patients were enrolled in the transitional care program, based on patient and family interview and chart review, and the items were later extracted for analysis. Patients were determined to have a domain-level need if they reported a need on any individual item within that domain, resulting in a binary score (any need present, absent) for each of the 10 TCN domains.

Health literacy was assessed by using the Brief Health Literacy Screen (BHLS), which is administered routinely by nurses at hospital intake and documented in the medical record, with completion rates of approximately 90%.²³ The BHLS is a 3-question subjective health literacy assessment (scoring range 3-15) that has been validated against longer objective measures²⁴ and shown to predict disease control and mortality.^18,25 To improve the stability of scores (for patients who completed the BHLS more than once because of repeat hospitalizations) and to reduce missing values, we calculated the patient’s mean BHLS score for assessments obtained between January 1, 2013, and December 31, 2014. Patients were then categorized as having inadequate health literacy (BHLS ≤ 9) or adequate health literacy (BHLS > 9).^18,25 Demographic information was extracted from patient records and included age, sex (male/female), marital status (married/without a partner), race (white/nonwhite), and years of education. Income level and primary language were not available for analysis.

Statistical Analysis

Patient characteristics and TCNs were summarized by using the frequency and percentages for categorical variables and the mean and standard deviation (SD) for continuous variables. We compared patient characteristics (age, sex, marital status, race, and education) between health literacy groups (inadequate vs adequate) by using χ² or analysis of variance as appropriate. We assessed Pearson correlations among the 10 TCN domains, and we examined differences in reported needs for each of 10 TCN domains by the level of health literacy by using the χ² test. Because the TCN domain of communication included low health literacy as one of its items, we excluded this domain from subsequent analyses. We then compared differences in the number of TCNs documented (scoring range 0-9) by using an independent samples Student t test.

Multivariate logistic regression models were then constructed to examine the independent association of inadequate health literacy with 8 TCN domains while controlling for age, sex, marital status, race, and education. Patients with incomplete demographic data were excluded from these models. Additionally, these analyses excluded 2 TCN domains: the communication domain for reasons noted above and the high-risk medical comorbidity domain because it ended up being positive in 98.4% of patients. Statistical significance was set at an alpha of 0.05. All analyses were performed by using SPSS Statistics for Mac, version 23.0 (IBM Corp., Armonk, New York)

A structured patient risk factor assessment derived from literature was used to record TCNs in preparation for hospital discharge. On average, patients had needs in about half of the TCN domains (4.6 of 9). The most common areas identified were related to the presence of high-risk comorbidities (98.4%), frequent or prior healthcare utilization (76.6%), medication management (76.3%), functional status (54.9%), and transportation (48.7%). Many of the TCNs were significantly correlated with one another. The prevalence of these needs highlights the importance of using a structured assessment to identify patient concerns so that they may be addressed through discharge planning and follow-up. In addition, using a standardized TCN instrument based on a framework for ITC promotes further research in understanding patient needs and in developing personalized interventions to address them.

As hypothesized, we found that TCNs were more common in patients with inadequate health literacy. After adjustment for demographic factors, inadequate health literacy was significantly associated with transportation barriers and inadequate caregiver support. Analyses also suggested a relationship with needs related to medical devices, functional status, and mental health comorbidities. A review of the literature substantiates a link between inadequate health literacy and these needs and also suggests solutions to address these barriers.

The association with inadequate caregiver support is concerning because there is often a high degree of reliance on caregivers at transitions in care.^3-5 Caregivers are routinely called upon to provide assistance with activities that may be difficult for patients with low health literacy, including medication adherence, provider communication, and self-care activities.^26,27 Our finding that patients with inadequate health literacy are more likely to have inadequate caregiver support indicates additional vulnerability. This may be because of the absence of a caregiver, or in many cases, the presence of a caregiver who is underprepared to assist with care. Prior research has shown that when caregivers are present, up to 33% have low health literacy, even when they are paid nonfamilial caregivers.^26,28 Other studies have noted the inadequacy of information and patient training for caregivers.^29,30 Transitional care programs to improve caregiver understanding have been developed³¹ and have been demonstrated to lower rehospitalization and ED visits.³²

Patients with inadequate health literacy were also more likely to have transportation barriers. Lack of transportation has been recorded as a factor in early hospital readmission in patients with chronic disease,³³ and it has been shown to have a negative effect on a variety of health outcomes.³⁴ A likely link between readmission and lack of transportation is poor follow-up care. Wheeler et al.³⁵ found that 59% of patients expected difficulty keeping postdischarge appointments because of transportation needs. Instead of expecting patients to navigate their own transportation, the Agency for Healthcare Research and Quality recommends identifying community resources for patients with low health literacy.³⁶

In this sample, inadequate health literacy also had near significant associations with TCNs in the use of medical devices, lower functional status, and mental health comorbidities. The use of a medical device, such as home oxygen, is a risk factor for readmission,³⁷ and early reports suggest that interventions, including education related to home oxygen use, can dramatically reduce these readmissions.³⁸ Lower functional capacity and faster functional decline are associated with inadequate health literacy,³⁹ which may have to do with the inability to appropriately utilize health resources.⁴⁰ If so, structured discharge planning could alleviate the known connection between functional impairment and hospital readmissions.⁴¹ A relationship between low health literacy and depression has been demonstrated repeatedly,⁴² with worsened symptoms in those with addiction.⁴³ As has been shown in other domains where health literacy is a factor, literacy-focused interventions provide greater benefits to these depressed patients.⁴⁴

The TCN assessment worked well overall, but certain domains proved less valuable and could be removed in the future. First, it was not useful to separately identify communication barriers, because doing so did not add to information beyond the measurement of health literacy. Second, high-risk comorbidities were ubiquitous within the sample and therefore unhelpful for group comparisons. In hindsight, this is unsurprising because the sample was comprised primarily of elderly patients admitted to medical services. Still, in a younger population or a surgical setting, identifying patients with high-risk medical comorbidities may be more useful.

We acknowledge several limitations of this study. First, the study was performed at a single center, and the TCN assessments were conducted by a small number of registered nurses who received training. Therefore, the results may not generalize to the profile of patient needs at other settings, and the instrument may perform differently when scaled across an organization. Second, the needs assessment was developed for this QI initiative and did not undergo formal validation, although it was developed from published frameworks and similar assessments. Third, for the measure of health literacy, we relied on data collected by nurses as part of their normal workflow. As is often the case with data collected during routine care, the scores are imperfect,⁴⁵ but they have proven to be a valuable and valid indicator of health literacy in our previous research.^18,24,25,46 Fourth, we chose to declare a domain as positive if any item in that domain was positive and to perform a domain-level analysis (for greater clarity). We did not take into account the variable number of items within each domain or attempt to grade their severity, as this would be a subjective exercise and impractical in the discharge planning process. Finally, we were unable to address associations among socioeconomic status,⁴⁷ primary language,⁴⁸ and health literacy, because relevant data were not available for this analysis.

In this sample of hospitalized patients who were administered a structured needs assessment, patients commonly had needs that placed them at a higher risk of adverse outcomes, such as hospital readmission. Patients with low health literacy had more TCNs that extended beyond the areas that we normally associate with low health literacy, namely patient education and self-management. Healthcare professionals should be aware of the greater likelihood of transportation barriers and inadequate caregiver support among patients with low health literacy. Screening for health literacy and TCN at admission or as part of the discharge planning process will elevate such risks, better positioning clinicians and hospitals to address them as a part of the efforts to ensure a quality transition of care.

Disclosure 

This work was funded by the Centers for Medicare and Medicaid Services (1C1CMS330979) and in part by the National Center for Advancing Translational Sciences (2 UL1 TR000445-06). The content is solely the responsibility of the authors and does not necessarily represent official views of the funding agencies, which did not participate in the planning, collection, analysis, or interpretation of data or in the decision to submit for publication.

Dr. Dittus reports personal fees as a board member of the Robert Wood Johnson Foundation Medical Faculty Scholars Program National Advisory Committee; consultancy fees from the University of Virginia, Indiana University, University of Michigan, Northwestern University, Montana State University, and Purdue University; has grants/grants pending from NIH (research grants), PCORI (research grant), CME (innovation award), VA (training grant); payment for lectures including service on speakers bureaus from Corporate Parity (conference organizer) for the Global Hospital Management & Innovation Summit; and other from Medical Decision Making, Inc. (passive owner); all outside the submitted work. Dr. Kripalani has grants from NIH (research grant), PCORI (research grant), and CMS (QI grant); outside the submitted work. All other authors have nothing to disclose.

A special concern since the institution of hospital readmission penalties¹ is the transitions in care of a patient from one care setting to another, often at hospital discharge. Burke et al.² proposed a framework for an ideal transition in care (ITC) to study and improve transitions from the hospital to home. The features in the ITC were identified based upon their inclusion in the interventions that improved discharge outcomes.^3-5 Inspired by the ITC and other patient risk tools,⁶ we identified 10 domains of transitional care needs ([TCN] specified below), which we define as patient-centered risk factors that should be addressed to foster a safe and effective transition in care.⁷

One particularly important risk factor in patient self-management at transition points is health literacy, a patient’s ability to obtain, understand, and use basic health information and services. Low health literacy affects approximately 26% to 36% of adults in the United States.^8,9 Health literacy is associated with many factors that may affect successful navigation of care transitions, including doctor-patient communication,^10,11 understanding of the medication regimen,¹² and self-management.^13-15 Research has also demonstrated an association between low health literacy and poor outcomes after hospital discharge, including medication errors,¹⁶ 30-day hospital readmission,¹⁷ and mortality.¹⁸ Transitional care initiatives have begun to incorporate health literacy into patient risk assessments⁶ and provide specific attention to low health literacy in interventions to reduce adverse drug events and readmission.^4,19 Training programs for medical students and nurses advise teaching skills in health literacy as part of fostering effective transitions in care.^20,21

Although low health literacy is generally recognized as a barrier to patient education and self-management, little is known about whether patients with low health literacy are more likely to have other risk factors that could further increase their risk for poor transitions in care. A better understanding of associated risks would inform and improve patient care. We hypothesized that TCNs are more common among patients with low health literacy, as compared with those with adequate health literacy. We also aimed to describe the relationship between low health literacy and specific TCNs in order to guide clinical care and future interventions.

Setting

The present study is a cross-sectional analysis of data from a quality improvement (QI) intervention that was performed at Vanderbilt University Medical Center, a tertiary care facility in Nashville, Tennessee. The QI intervention, My Health Team (MHT), was funded by the Centers for Medicare and Medicaid Services Innovation Award program. The overall MHT program included outpatient care coordination for chronic disease management as well as a transitional care program that was designed to reduce hospital readmission. The latter included an inpatient needs assessment (which provided data for the present analysis), inpatient intervention, and postdischarge phone follow-up. The MHT initiative was reviewed by the institutional review board (IRB), which deemed it a QI program and granted a waiver of informed consent. The present secondary data analysis was reviewed and approved by the IRB.

Sample

Patients were identified for inclusion in the MHT transitions of care program if the presenting problem for hospital admission was pneumonia, chronic obstructive pulmonary disease (COPD) exacerbation, or decompensated heart failure, as determined by the review of clinical documentation by nurse transition care coordinators (TCCs). Adults over the age of 18 years were eligible, though priority was given to patients aged 65 years or older. This study includes the first inpatient encounter between June 2013 and December 31, 2014, for patients having a completed needs assessment and documentation of health literacy data in the medical record.

Data Collection

TCN assessment was developed from published patient risk tools and the ITC framework.^2,6,22 The assessment has 10 domains composed of 49 individual items as follows: (1) caregiver support (caregiver support not sufficient for patient needs), (2) transportation (relies on public or others for transportation and misses medical care because of transportation), (3) health care utilization (no primary care physician, unplanned hospitalization in the last year, emergency department [ED] visit in the last 6 months, or home health services in the last 60 days), (4) high-risk medical comorbidities (malnutrition or body mass index <18.5, renal failure, chronic pain, diabetes, heart failure, COPD, or stroke), (5) medication management provider or caregiver concern (cannot provide medication list, >10 preadmission medications, high-risk medications [eg, insulin, warfarin], poor medication understanding, or adherence issue identified), (6) medical devices (vascular access, urinary catheter, wounds, or home supplemental oxygen), (7) functional status (weakness of extremities, limited extremity range of motion, difficulty with mobility, falls at home, or activities of daily living challenges), (8) mental health comorbidities (over the past month has felt down, depressed, or hopeless or over the past month has felt little interest or pleasure in doing things, high-risk alcohol use, or high-risk substance use), (9) communication (limited English proficiency or at risk for limited health literacy), and (10) financial resources (no health insurance, skips or rations medicines because of cost, misses medical care because of cost, or misses medical care because of job).

The 49 items of the TCN assessment were documented as being present or absent by nurse TCCs at the time patients were enrolled in the transitional care program, based on patient and family interview and chart review, and the items were later extracted for analysis. Patients were determined to have a domain-level need if they reported a need on any individual item within that domain, resulting in a binary score (any need present, absent) for each of the 10 TCN domains.

Health literacy was assessed by using the Brief Health Literacy Screen (BHLS), which is administered routinely by nurses at hospital intake and documented in the medical record, with completion rates of approximately 90%.²³ The BHLS is a 3-question subjective health literacy assessment (scoring range 3-15) that has been validated against longer objective measures²⁴ and shown to predict disease control and mortality.^18,25 To improve the stability of scores (for patients who completed the BHLS more than once because of repeat hospitalizations) and to reduce missing values, we calculated the patient’s mean BHLS score for assessments obtained between January 1, 2013, and December 31, 2014. Patients were then categorized as having inadequate health literacy (BHLS ≤ 9) or adequate health literacy (BHLS > 9).^18,25 Demographic information was extracted from patient records and included age, sex (male/female), marital status (married/without a partner), race (white/nonwhite), and years of education. Income level and primary language were not available for analysis.

Statistical Analysis

Patient characteristics and TCNs were summarized by using the frequency and percentages for categorical variables and the mean and standard deviation (SD) for continuous variables. We compared patient characteristics (age, sex, marital status, race, and education) between health literacy groups (inadequate vs adequate) by using χ² or analysis of variance as appropriate. We assessed Pearson correlations among the 10 TCN domains, and we examined differences in reported needs for each of 10 TCN domains by the level of health literacy by using the χ² test. Because the TCN domain of communication included low health literacy as one of its items, we excluded this domain from subsequent analyses. We then compared differences in the number of TCNs documented (scoring range 0-9) by using an independent samples Student t test.

Multivariate logistic regression models were then constructed to examine the independent association of inadequate health literacy with 8 TCN domains while controlling for age, sex, marital status, race, and education. Patients with incomplete demographic data were excluded from these models. Additionally, these analyses excluded 2 TCN domains: the communication domain for reasons noted above and the high-risk medical comorbidity domain because it ended up being positive in 98.4% of patients. Statistical significance was set at an alpha of 0.05. All analyses were performed by using SPSS Statistics for Mac, version 23.0 (IBM Corp., Armonk, New York)

A structured patient risk factor assessment derived from literature was used to record TCNs in preparation for hospital discharge. On average, patients had needs in about half of the TCN domains (4.6 of 9). The most common areas identified were related to the presence of high-risk comorbidities (98.4%), frequent or prior healthcare utilization (76.6%), medication management (76.3%), functional status (54.9%), and transportation (48.7%). Many of the TCNs were significantly correlated with one another. The prevalence of these needs highlights the importance of using a structured assessment to identify patient concerns so that they may be addressed through discharge planning and follow-up. In addition, using a standardized TCN instrument based on a framework for ITC promotes further research in understanding patient needs and in developing personalized interventions to address them.

As hypothesized, we found that TCNs were more common in patients with inadequate health literacy. After adjustment for demographic factors, inadequate health literacy was significantly associated with transportation barriers and inadequate caregiver support. Analyses also suggested a relationship with needs related to medical devices, functional status, and mental health comorbidities. A review of the literature substantiates a link between inadequate health literacy and these needs and also suggests solutions to address these barriers.

The association with inadequate caregiver support is concerning because there is often a high degree of reliance on caregivers at transitions in care.^3-5 Caregivers are routinely called upon to provide assistance with activities that may be difficult for patients with low health literacy, including medication adherence, provider communication, and self-care activities.^26,27 Our finding that patients with inadequate health literacy are more likely to have inadequate caregiver support indicates additional vulnerability. This may be because of the absence of a caregiver, or in many cases, the presence of a caregiver who is underprepared to assist with care. Prior research has shown that when caregivers are present, up to 33% have low health literacy, even when they are paid nonfamilial caregivers.^26,28 Other studies have noted the inadequacy of information and patient training for caregivers.^29,30 Transitional care programs to improve caregiver understanding have been developed³¹ and have been demonstrated to lower rehospitalization and ED visits.³²

Patients with inadequate health literacy were also more likely to have transportation barriers. Lack of transportation has been recorded as a factor in early hospital readmission in patients with chronic disease,³³ and it has been shown to have a negative effect on a variety of health outcomes.³⁴ A likely link between readmission and lack of transportation is poor follow-up care. Wheeler et al.³⁵ found that 59% of patients expected difficulty keeping postdischarge appointments because of transportation needs. Instead of expecting patients to navigate their own transportation, the Agency for Healthcare Research and Quality recommends identifying community resources for patients with low health literacy.³⁶

In this sample, inadequate health literacy also had near significant associations with TCNs in the use of medical devices, lower functional status, and mental health comorbidities. The use of a medical device, such as home oxygen, is a risk factor for readmission,³⁷ and early reports suggest that interventions, including education related to home oxygen use, can dramatically reduce these readmissions.³⁸ Lower functional capacity and faster functional decline are associated with inadequate health literacy,³⁹ which may have to do with the inability to appropriately utilize health resources.⁴⁰ If so, structured discharge planning could alleviate the known connection between functional impairment and hospital readmissions.⁴¹ A relationship between low health literacy and depression has been demonstrated repeatedly,⁴² with worsened symptoms in those with addiction.⁴³ As has been shown in other domains where health literacy is a factor, literacy-focused interventions provide greater benefits to these depressed patients.⁴⁴

The TCN assessment worked well overall, but certain domains proved less valuable and could be removed in the future. First, it was not useful to separately identify communication barriers, because doing so did not add to information beyond the measurement of health literacy. Second, high-risk comorbidities were ubiquitous within the sample and therefore unhelpful for group comparisons. In hindsight, this is unsurprising because the sample was comprised primarily of elderly patients admitted to medical services. Still, in a younger population or a surgical setting, identifying patients with high-risk medical comorbidities may be more useful.

We acknowledge several limitations of this study. First, the study was performed at a single center, and the TCN assessments were conducted by a small number of registered nurses who received training. Therefore, the results may not generalize to the profile of patient needs at other settings, and the instrument may perform differently when scaled across an organization. Second, the needs assessment was developed for this QI initiative and did not undergo formal validation, although it was developed from published frameworks and similar assessments. Third, for the measure of health literacy, we relied on data collected by nurses as part of their normal workflow. As is often the case with data collected during routine care, the scores are imperfect,⁴⁵ but they have proven to be a valuable and valid indicator of health literacy in our previous research.^18,24,25,46 Fourth, we chose to declare a domain as positive if any item in that domain was positive and to perform a domain-level analysis (for greater clarity). We did not take into account the variable number of items within each domain or attempt to grade their severity, as this would be a subjective exercise and impractical in the discharge planning process. Finally, we were unable to address associations among socioeconomic status,⁴⁷ primary language,⁴⁸ and health literacy, because relevant data were not available for this analysis.

In this sample of hospitalized patients who were administered a structured needs assessment, patients commonly had needs that placed them at a higher risk of adverse outcomes, such as hospital readmission. Patients with low health literacy had more TCNs that extended beyond the areas that we normally associate with low health literacy, namely patient education and self-management. Healthcare professionals should be aware of the greater likelihood of transportation barriers and inadequate caregiver support among patients with low health literacy. Screening for health literacy and TCN at admission or as part of the discharge planning process will elevate such risks, better positioning clinicians and hospitals to address them as a part of the efforts to ensure a quality transition of care.

Disclosure 

This work was funded by the Centers for Medicare and Medicaid Services (1C1CMS330979) and in part by the National Center for Advancing Translational Sciences (2 UL1 TR000445-06). The content is solely the responsibility of the authors and does not necessarily represent official views of the funding agencies, which did not participate in the planning, collection, analysis, or interpretation of data or in the decision to submit for publication.

Dr. Dittus reports personal fees as a board member of the Robert Wood Johnson Foundation Medical Faculty Scholars Program National Advisory Committee; consultancy fees from the University of Virginia, Indiana University, University of Michigan, Northwestern University, Montana State University, and Purdue University; has grants/grants pending from NIH (research grants), PCORI (research grant), CME (innovation award), VA (training grant); payment for lectures including service on speakers bureaus from Corporate Parity (conference organizer) for the Global Hospital Management & Innovation Summit; and other from Medical Decision Making, Inc. (passive owner); all outside the submitted work. Dr. Kripalani has grants from NIH (research grant), PCORI (research grant), and CMS (QI grant); outside the submitted work. All other authors have nothing to disclose.

A special concern since the institution of hospital readmission penalties¹ is the transitions in care of a patient from one care setting to another, often at hospital discharge. Burke et al.² proposed a framework for an ideal transition in care (ITC) to study and improve transitions from the hospital to home. The features in the ITC were identified based upon their inclusion in the interventions that improved discharge outcomes.^3-5 Inspired by the ITC and other patient risk tools,⁶ we identified 10 domains of transitional care needs ([TCN] specified below), which we define as patient-centered risk factors that should be addressed to foster a safe and effective transition in care.⁷

One particularly important risk factor in patient self-management at transition points is health literacy, a patient’s ability to obtain, understand, and use basic health information and services. Low health literacy affects approximately 26% to 36% of adults in the United States.^8,9 Health literacy is associated with many factors that may affect successful navigation of care transitions, including doctor-patient communication,^10,11 understanding of the medication regimen,¹² and self-management.^13-15 Research has also demonstrated an association between low health literacy and poor outcomes after hospital discharge, including medication errors,¹⁶ 30-day hospital readmission,¹⁷ and mortality.¹⁸ Transitional care initiatives have begun to incorporate health literacy into patient risk assessments⁶ and provide specific attention to low health literacy in interventions to reduce adverse drug events and readmission.^4,19 Training programs for medical students and nurses advise teaching skills in health literacy as part of fostering effective transitions in care.^20,21

Although low health literacy is generally recognized as a barrier to patient education and self-management, little is known about whether patients with low health literacy are more likely to have other risk factors that could further increase their risk for poor transitions in care. A better understanding of associated risks would inform and improve patient care. We hypothesized that TCNs are more common among patients with low health literacy, as compared with those with adequate health literacy. We also aimed to describe the relationship between low health literacy and specific TCNs in order to guide clinical care and future interventions.

Setting

The present study is a cross-sectional analysis of data from a quality improvement (QI) intervention that was performed at Vanderbilt University Medical Center, a tertiary care facility in Nashville, Tennessee. The QI intervention, My Health Team (MHT), was funded by the Centers for Medicare and Medicaid Services Innovation Award program. The overall MHT program included outpatient care coordination for chronic disease management as well as a transitional care program that was designed to reduce hospital readmission. The latter included an inpatient needs assessment (which provided data for the present analysis), inpatient intervention, and postdischarge phone follow-up. The MHT initiative was reviewed by the institutional review board (IRB), which deemed it a QI program and granted a waiver of informed consent. The present secondary data analysis was reviewed and approved by the IRB.

Sample

Patients were identified for inclusion in the MHT transitions of care program if the presenting problem for hospital admission was pneumonia, chronic obstructive pulmonary disease (COPD) exacerbation, or decompensated heart failure, as determined by the review of clinical documentation by nurse transition care coordinators (TCCs). Adults over the age of 18 years were eligible, though priority was given to patients aged 65 years or older. This study includes the first inpatient encounter between June 2013 and December 31, 2014, for patients having a completed needs assessment and documentation of health literacy data in the medical record.

Data Collection

TCN assessment was developed from published patient risk tools and the ITC framework.^2,6,22 The assessment has 10 domains composed of 49 individual items as follows: (1) caregiver support (caregiver support not sufficient for patient needs), (2) transportation (relies on public or others for transportation and misses medical care because of transportation), (3) health care utilization (no primary care physician, unplanned hospitalization in the last year, emergency department [ED] visit in the last 6 months, or home health services in the last 60 days), (4) high-risk medical comorbidities (malnutrition or body mass index <18.5, renal failure, chronic pain, diabetes, heart failure, COPD, or stroke), (5) medication management provider or caregiver concern (cannot provide medication list, >10 preadmission medications, high-risk medications [eg, insulin, warfarin], poor medication understanding, or adherence issue identified), (6) medical devices (vascular access, urinary catheter, wounds, or home supplemental oxygen), (7) functional status (weakness of extremities, limited extremity range of motion, difficulty with mobility, falls at home, or activities of daily living challenges), (8) mental health comorbidities (over the past month has felt down, depressed, or hopeless or over the past month has felt little interest or pleasure in doing things, high-risk alcohol use, or high-risk substance use), (9) communication (limited English proficiency or at risk for limited health literacy), and (10) financial resources (no health insurance, skips or rations medicines because of cost, misses medical care because of cost, or misses medical care because of job).

The 49 items of the TCN assessment were documented as being present or absent by nurse TCCs at the time patients were enrolled in the transitional care program, based on patient and family interview and chart review, and the items were later extracted for analysis. Patients were determined to have a domain-level need if they reported a need on any individual item within that domain, resulting in a binary score (any need present, absent) for each of the 10 TCN domains.

Health literacy was assessed by using the Brief Health Literacy Screen (BHLS), which is administered routinely by nurses at hospital intake and documented in the medical record, with completion rates of approximately 90%.²³ The BHLS is a 3-question subjective health literacy assessment (scoring range 3-15) that has been validated against longer objective measures²⁴ and shown to predict disease control and mortality.^18,25 To improve the stability of scores (for patients who completed the BHLS more than once because of repeat hospitalizations) and to reduce missing values, we calculated the patient’s mean BHLS score for assessments obtained between January 1, 2013, and December 31, 2014. Patients were then categorized as having inadequate health literacy (BHLS ≤ 9) or adequate health literacy (BHLS > 9).^18,25 Demographic information was extracted from patient records and included age, sex (male/female), marital status (married/without a partner), race (white/nonwhite), and years of education. Income level and primary language were not available for analysis.

Statistical Analysis

Patient characteristics and TCNs were summarized by using the frequency and percentages for categorical variables and the mean and standard deviation (SD) for continuous variables. We compared patient characteristics (age, sex, marital status, race, and education) between health literacy groups (inadequate vs adequate) by using χ² or analysis of variance as appropriate. We assessed Pearson correlations among the 10 TCN domains, and we examined differences in reported needs for each of 10 TCN domains by the level of health literacy by using the χ² test. Because the TCN domain of communication included low health literacy as one of its items, we excluded this domain from subsequent analyses. We then compared differences in the number of TCNs documented (scoring range 0-9) by using an independent samples Student t test.

Multivariate logistic regression models were then constructed to examine the independent association of inadequate health literacy with 8 TCN domains while controlling for age, sex, marital status, race, and education. Patients with incomplete demographic data were excluded from these models. Additionally, these analyses excluded 2 TCN domains: the communication domain for reasons noted above and the high-risk medical comorbidity domain because it ended up being positive in 98.4% of patients. Statistical significance was set at an alpha of 0.05. All analyses were performed by using SPSS Statistics for Mac, version 23.0 (IBM Corp., Armonk, New York)

A structured patient risk factor assessment derived from literature was used to record TCNs in preparation for hospital discharge. On average, patients had needs in about half of the TCN domains (4.6 of 9). The most common areas identified were related to the presence of high-risk comorbidities (98.4%), frequent or prior healthcare utilization (76.6%), medication management (76.3%), functional status (54.9%), and transportation (48.7%). Many of the TCNs were significantly correlated with one another. The prevalence of these needs highlights the importance of using a structured assessment to identify patient concerns so that they may be addressed through discharge planning and follow-up. In addition, using a standardized TCN instrument based on a framework for ITC promotes further research in understanding patient needs and in developing personalized interventions to address them.

As hypothesized, we found that TCNs were more common in patients with inadequate health literacy. After adjustment for demographic factors, inadequate health literacy was significantly associated with transportation barriers and inadequate caregiver support. Analyses also suggested a relationship with needs related to medical devices, functional status, and mental health comorbidities. A review of the literature substantiates a link between inadequate health literacy and these needs and also suggests solutions to address these barriers.

The association with inadequate caregiver support is concerning because there is often a high degree of reliance on caregivers at transitions in care.^3-5 Caregivers are routinely called upon to provide assistance with activities that may be difficult for patients with low health literacy, including medication adherence, provider communication, and self-care activities.^26,27 Our finding that patients with inadequate health literacy are more likely to have inadequate caregiver support indicates additional vulnerability. This may be because of the absence of a caregiver, or in many cases, the presence of a caregiver who is underprepared to assist with care. Prior research has shown that when caregivers are present, up to 33% have low health literacy, even when they are paid nonfamilial caregivers.^26,28 Other studies have noted the inadequacy of information and patient training for caregivers.^29,30 Transitional care programs to improve caregiver understanding have been developed³¹ and have been demonstrated to lower rehospitalization and ED visits.³²

Patients with inadequate health literacy were also more likely to have transportation barriers. Lack of transportation has been recorded as a factor in early hospital readmission in patients with chronic disease,³³ and it has been shown to have a negative effect on a variety of health outcomes.³⁴ A likely link between readmission and lack of transportation is poor follow-up care. Wheeler et al.³⁵ found that 59% of patients expected difficulty keeping postdischarge appointments because of transportation needs. Instead of expecting patients to navigate their own transportation, the Agency for Healthcare Research and Quality recommends identifying community resources for patients with low health literacy.³⁶

In this sample, inadequate health literacy also had near significant associations with TCNs in the use of medical devices, lower functional status, and mental health comorbidities. The use of a medical device, such as home oxygen, is a risk factor for readmission,³⁷ and early reports suggest that interventions, including education related to home oxygen use, can dramatically reduce these readmissions.³⁸ Lower functional capacity and faster functional decline are associated with inadequate health literacy,³⁹ which may have to do with the inability to appropriately utilize health resources.⁴⁰ If so, structured discharge planning could alleviate the known connection between functional impairment and hospital readmissions.⁴¹ A relationship between low health literacy and depression has been demonstrated repeatedly,⁴² with worsened symptoms in those with addiction.⁴³ As has been shown in other domains where health literacy is a factor, literacy-focused interventions provide greater benefits to these depressed patients.⁴⁴

The TCN assessment worked well overall, but certain domains proved less valuable and could be removed in the future. First, it was not useful to separately identify communication barriers, because doing so did not add to information beyond the measurement of health literacy. Second, high-risk comorbidities were ubiquitous within the sample and therefore unhelpful for group comparisons. In hindsight, this is unsurprising because the sample was comprised primarily of elderly patients admitted to medical services. Still, in a younger population or a surgical setting, identifying patients with high-risk medical comorbidities may be more useful.

We acknowledge several limitations of this study. First, the study was performed at a single center, and the TCN assessments were conducted by a small number of registered nurses who received training. Therefore, the results may not generalize to the profile of patient needs at other settings, and the instrument may perform differently when scaled across an organization. Second, the needs assessment was developed for this QI initiative and did not undergo formal validation, although it was developed from published frameworks and similar assessments. Third, for the measure of health literacy, we relied on data collected by nurses as part of their normal workflow. As is often the case with data collected during routine care, the scores are imperfect,⁴⁵ but they have proven to be a valuable and valid indicator of health literacy in our previous research.^18,24,25,46 Fourth, we chose to declare a domain as positive if any item in that domain was positive and to perform a domain-level analysis (for greater clarity). We did not take into account the variable number of items within each domain or attempt to grade their severity, as this would be a subjective exercise and impractical in the discharge planning process. Finally, we were unable to address associations among socioeconomic status,⁴⁷ primary language,⁴⁸ and health literacy, because relevant data were not available for this analysis.

In this sample of hospitalized patients who were administered a structured needs assessment, patients commonly had needs that placed them at a higher risk of adverse outcomes, such as hospital readmission. Patients with low health literacy had more TCNs that extended beyond the areas that we normally associate with low health literacy, namely patient education and self-management. Healthcare professionals should be aware of the greater likelihood of transportation barriers and inadequate caregiver support among patients with low health literacy. Screening for health literacy and TCN at admission or as part of the discharge planning process will elevate such risks, better positioning clinicians and hospitals to address them as a part of the efforts to ensure a quality transition of care.

Disclosure 

This work was funded by the Centers for Medicare and Medicaid Services (1C1CMS330979) and in part by the National Center for Advancing Translational Sciences (2 UL1 TR000445-06). The content is solely the responsibility of the authors and does not necessarily represent official views of the funding agencies, which did not participate in the planning, collection, analysis, or interpretation of data or in the decision to submit for publication.

Dr. Dittus reports personal fees as a board member of the Robert Wood Johnson Foundation Medical Faculty Scholars Program National Advisory Committee; consultancy fees from the University of Virginia, Indiana University, University of Michigan, Northwestern University, Montana State University, and Purdue University; has grants/grants pending from NIH (research grants), PCORI (research grant), CME (innovation award), VA (training grant); payment for lectures including service on speakers bureaus from Corporate Parity (conference organizer) for the Global Hospital Management & Innovation Summit; and other from Medical Decision Making, Inc. (passive owner); all outside the submitted work. Dr. Kripalani has grants from NIH (research grant), PCORI (research grant), and CMS (QI grant); outside the submitted work. All other authors have nothing to disclose.

When seniors are discharged from the hospital, many will require additional support and therapy to regain their independence and return safely home.^1,2 Most seniors do not understand what their support needs will entail or the differences between therapy choices.³ To complicate the issue, seniors are often incapacitated and unable to make discharge selections for themselves.

Consequently, discharge planners and social workers often explain options to family members and loved ones, who frequently feel overwhelmed.^4,5 While often balancing jobs, loved ones are divided between wanting to stay with the senior in the hospital and driving to area skilled nursing facilities (SNFs) for consideration. Discharges are delayed waiting for families to make visits and choose an SNF. Longer lengths of stay are detrimental to seniors due to the increased risks of infection, functional loss, and cognitive decline.⁶

Although seniors comprised only 12% of the US population in 2003,⁷ they accounted for one-third of all hospitalizations, over 13.2 million hospital stays.⁸ Hospital stays for seniors resulted in hospital charges totaling nearly $329 billion, or 43.6% of national hospital bills in 2003.⁷ Seniors are also high consumers of postacute care services. By 2050, the number of individuals using long-term care services in any setting (eg, at home, assisted living, or SNFs) will be close to 27 million.^9-11 With the knowledge that many seniors will be hospitalized and subsequently require services thereafter, we sought to assist seniors in planning for their hospital discharge needs before they were hospitalized.

Our team developed PlanYourLifespan.org (PYL) to facilitate this planning for postdischarge needs and fill the knowledge gap in understanding postdischarge options. With funding from the Patient Centered Outcomes Research Institute, we aimed to test the effectiveness of PYL on improving knowledge of hospital discharge resources among seniors.

PlanYourLifespan.org

PlanYourLifespan.org (PYL) educates users on the health crises that often occur with age and connects them to posthospital and home-based resources available locally and nationally. PYL is personalized, dynamic, and adaptable in that all the information can be changed per the senior’s wishes or changing health needs.

Content of PYL

Previously, we conducted focus groups with seniors about current and perceived home needs and aging-in-place. Major themes of what advanced life events (ALEs) would impact aging-in-place were identified as follows: hospitalizations, falls, and Alzheimer’s.¹² We organized PYL around these health-related ALEs. Our multidisciplinary team of researchers, seniors, social workers, caregiver agencies, and Area Agencies on Aging representatives then determined what information and resources should be included.

Each section of PYL starts with a video of a senior discussing their real-life personal experiences, with subsections providing interactive information on what seniors can expect, types of resources available to support home needs, and choices to be made. Descriptions of types of settings for therapy, options available, and links to national/local resources (eg, quality indicators for SNFs) are also included. For example, by entering their zip code, users can identify their neighborhood SNFs, closest Area Agency on Aging, and what home caregiver agencies exist in their area.

Users can save their preferences and revisit their choices at any time. To support communication between seniors and their loved ones, a summary of their choices can be printed or e-mailed to relevant parties. For example, a senior uses PYL and can e-mail these choices to family members, which can stimulate a conversation about future posthospital care expectations.

As inadequate health literacy and cognitive impairment are prevalent among seniors, PYL presents information understandable at all levels of health literacy and sensitive to cognitive load.⁹ There is simplified, large-font, no mouse scrolling and audio available for the visually impaired.

Study Design and Randomization

To test PYL, a 2-armed (attention control [AC] and PYL intervention), parallel, randomized controlled trial was conducted. Participants were randomly assigned to 1 of the 2 conditions via a pregenerated central randomization list using equal (1:1) allocation and random permuted block design to ensure relatively equal allocation throughout the study. The AC condition exposed participants to the National Institute on Aging-sponsored website, Go4Life.nia.nih.gov, an educational website on physical activity for seniors. This website has comparable design and layout to PYL but does not include information about advanced planning. The AC condition controlled for the possibility that regular contact with the study team may improve outcomes in participants randomized to the intervention website.

The trial was conducted from October 2014 to September 2015 in Chicago, Illinois; Fort Wayne, Indiana; and Houston, Texas. Inclusion criteria were as follows: age 65 and older, English-speaking, scoring ≥4 questions correctly on the Brief Cognitive Screen,¹⁴ and current self-reported use of a computer or smartphone with internet. Participants were excluded if they had previously participated in the focus groups or beta testing of the PYL website. Community-based patient partners/stakeholders drove subject recruitment in their communities through word of mouth, e-mail bursts, newsletters, and flyers. At the Area Agency on Aging and community centers where services such as food vouchers and case management are provided, participants were recruited on-site and given study information. At the clinical sites, staff referred potential participants. Study materials such as flyers and information sheets were also located in the clinic waiting rooms. The Villages, nonprofit, grassroots, membership organizations that are redefining aging by being a key resource to community members wishing to age in place, heavily relied on electronic recruitment using their regular e-newsletters and e-mail lists to recruit potential participants. Potential subjects were also recruited by distributing flyers at local senior centers and senior housing buildings. Interested seniors contacted research staff who explained the study and assessed their eligibility. If eligible, subjects were scheduled for a face-to-face interview.

At the face-to-face encounter, all study subjects completed a written consent, answered baseline questions, and were randomized to either arm. Next, research staff introduced all study subjects to the website to which they were randomized and provided instructions on its use. Staff were present to assist with questions as needed on navigation but did not assist with decision making for either website. A minimum of 15 minutes and a maximum of 45 minutes was allotted for navigating either website. After navigating their website, participants were administered an immediate in-person posttest survey. One month and 3 months after the face-to-face encounter, research staff contacted all study participants over the phone to complete a follow-up survey. Staff attempted to reach participants up to 3 times by phone. Data were entered into Research Electronic Data Capture survey software.¹⁵ This study was approved by the Northwestern University Institutional Review Board.

Understanding of Posthospital Discharge and Home Services

As part of the larger trial, which included behavioural outcomes that will be reported elsewhere, we sought to explore the effects of PYL on participants’ knowledge and understanding of posthospital discharge and home services (UHS). Participants were asked to respond to 6 questions at baseline, immediate posttest, and at the 1- and 3-month follow-up time points. Knowledge items were developed by the study team in conjunction with the patient/partner stakeholders. UHS scores were calculated as the sum of the 6 questions (each scored 0 if incorrect and 1 if correct) with a possible range of 0-6.

Covariates

Demographic information, self-reported health, importance of religion, and existence of a power of attorney, living will, advanced directive (eg, Physician Orders for Life-Sustaining Treatment) were obtained via self-report. Participants were asked about their general and social self-efficacy using the validated Self-efficacy Scale¹⁶ and their social support using the Lubben Social Network Scale–6.¹⁷ Health literacy was assessed using the Rapid Estimate of Adult Literacy in Medicine–Short Form.¹⁸ To measure burden of disease, participant comorbidities were measured using a nonvalidated 9-item dichotomous response condition list, which included some items adapted from the Charlson Comorbidity Index and the Elixhauser Comorbidity Index.

Statistical Analysis

Data analysis included all available data in the intention-to-treat dataset. As UHS was collected at multiple time points up to 3 months postintervention, we employed linear- mixed modeling with random participant effect and fixed arm, time, and time-by-arm interaction terms. The time-by-arm interaction allows for comparison of UHS slopes (or trajectories) across arms. Analyses explored multiple potential covariates, including current utilization of services, physical function, comorbidities, social support, health literacy, self-efficacy, and sociodemographics. Those covariates found to have a significant association (P < 0.05) with outcome were considered for inclusion in the overall model selection process. Ultimately, we developed a final parsimonious, adjusted longitudinal model with primary predictors of time, arm, and their interaction, controlling for only significant baseline variables following a manual backward selection method. All analyses were conducted in SAS software (version 9.4, copyright 2012, The SAS Institute Inc., Cary, North Carolina).

The results of our study show that among seniors, PYL improved their understanding of home-based services and the services that may be required following a hospitalization. Educating seniors about what to expect regarding the transition home from a hospital before a hospitalization even occurs may enable seniors and their families to plan ahead instead of reacting to a hospitalization. PYL, a national, publicly available tool with links to local resources may potentially help in advancing transitional discharge care to prior to a hospitalization.

To our knowledge, this is one of the first websites and trials devoted to planning for seniors’ health trajectory as they age into their 70s, 80s, 90s, and 100s. Clinicians regularly discuss code status and powers of attorney during their end-of-life discussions with patients. We encourage clinicians to ask patients, “What about the 10 to 20 years before you die? Have you considered what you will do if you get sick or need help at home?” While not replacing a social worker, the ability of PYL to connect seniors to local resources makes it somewhat of a “virtual social worker.” With many physician practices unable to afford social workers, PYL provides a free-of-charge means of connecting seniors to area resources.

Acknowledgments

This work was supported through a Patient-Centered Outcomes Research Institute Award (IH-12-11-4259). Dr. Lindquist and Dr. Ciolino had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Research reported in this publication was also supported, in part, by the National Institutes of Health’s National Center for Advancing Translational Sciences, Grant Number UL1TR000150.

Disclaimer 

All statements in this manuscript, including its findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors or Methodology Committee, or the National Institutes of Health.

Disclosure: The authors have nothing to report.

When seniors are discharged from the hospital, many will require additional support and therapy to regain their independence and return safely home.^1,2 Most seniors do not understand what their support needs will entail or the differences between therapy choices.³ To complicate the issue, seniors are often incapacitated and unable to make discharge selections for themselves.

Consequently, discharge planners and social workers often explain options to family members and loved ones, who frequently feel overwhelmed.^4,5 While often balancing jobs, loved ones are divided between wanting to stay with the senior in the hospital and driving to area skilled nursing facilities (SNFs) for consideration. Discharges are delayed waiting for families to make visits and choose an SNF. Longer lengths of stay are detrimental to seniors due to the increased risks of infection, functional loss, and cognitive decline.⁶

Although seniors comprised only 12% of the US population in 2003,⁷ they accounted for one-third of all hospitalizations, over 13.2 million hospital stays.⁸ Hospital stays for seniors resulted in hospital charges totaling nearly $329 billion, or 43.6% of national hospital bills in 2003.⁷ Seniors are also high consumers of postacute care services. By 2050, the number of individuals using long-term care services in any setting (eg, at home, assisted living, or SNFs) will be close to 27 million.^9-11 With the knowledge that many seniors will be hospitalized and subsequently require services thereafter, we sought to assist seniors in planning for their hospital discharge needs before they were hospitalized.

Our team developed PlanYourLifespan.org (PYL) to facilitate this planning for postdischarge needs and fill the knowledge gap in understanding postdischarge options. With funding from the Patient Centered Outcomes Research Institute, we aimed to test the effectiveness of PYL on improving knowledge of hospital discharge resources among seniors.

PlanYourLifespan.org

PlanYourLifespan.org (PYL) educates users on the health crises that often occur with age and connects them to posthospital and home-based resources available locally and nationally. PYL is personalized, dynamic, and adaptable in that all the information can be changed per the senior’s wishes or changing health needs.

Content of PYL

Previously, we conducted focus groups with seniors about current and perceived home needs and aging-in-place. Major themes of what advanced life events (ALEs) would impact aging-in-place were identified as follows: hospitalizations, falls, and Alzheimer’s.¹² We organized PYL around these health-related ALEs. Our multidisciplinary team of researchers, seniors, social workers, caregiver agencies, and Area Agencies on Aging representatives then determined what information and resources should be included.

Each section of PYL starts with a video of a senior discussing their real-life personal experiences, with subsections providing interactive information on what seniors can expect, types of resources available to support home needs, and choices to be made. Descriptions of types of settings for therapy, options available, and links to national/local resources (eg, quality indicators for SNFs) are also included. For example, by entering their zip code, users can identify their neighborhood SNFs, closest Area Agency on Aging, and what home caregiver agencies exist in their area.

Users can save their preferences and revisit their choices at any time. To support communication between seniors and their loved ones, a summary of their choices can be printed or e-mailed to relevant parties. For example, a senior uses PYL and can e-mail these choices to family members, which can stimulate a conversation about future posthospital care expectations.

As inadequate health literacy and cognitive impairment are prevalent among seniors, PYL presents information understandable at all levels of health literacy and sensitive to cognitive load.⁹ There is simplified, large-font, no mouse scrolling and audio available for the visually impaired.

Study Design and Randomization

To test PYL, a 2-armed (attention control [AC] and PYL intervention), parallel, randomized controlled trial was conducted. Participants were randomly assigned to 1 of the 2 conditions via a pregenerated central randomization list using equal (1:1) allocation and random permuted block design to ensure relatively equal allocation throughout the study. The AC condition exposed participants to the National Institute on Aging-sponsored website, Go4Life.nia.nih.gov, an educational website on physical activity for seniors. This website has comparable design and layout to PYL but does not include information about advanced planning. The AC condition controlled for the possibility that regular contact with the study team may improve outcomes in participants randomized to the intervention website.

The trial was conducted from October 2014 to September 2015 in Chicago, Illinois; Fort Wayne, Indiana; and Houston, Texas. Inclusion criteria were as follows: age 65 and older, English-speaking, scoring ≥4 questions correctly on the Brief Cognitive Screen,¹⁴ and current self-reported use of a computer or smartphone with internet. Participants were excluded if they had previously participated in the focus groups or beta testing of the PYL website. Community-based patient partners/stakeholders drove subject recruitment in their communities through word of mouth, e-mail bursts, newsletters, and flyers. At the Area Agency on Aging and community centers where services such as food vouchers and case management are provided, participants were recruited on-site and given study information. At the clinical sites, staff referred potential participants. Study materials such as flyers and information sheets were also located in the clinic waiting rooms. The Villages, nonprofit, grassroots, membership organizations that are redefining aging by being a key resource to community members wishing to age in place, heavily relied on electronic recruitment using their regular e-newsletters and e-mail lists to recruit potential participants. Potential subjects were also recruited by distributing flyers at local senior centers and senior housing buildings. Interested seniors contacted research staff who explained the study and assessed their eligibility. If eligible, subjects were scheduled for a face-to-face interview.

At the face-to-face encounter, all study subjects completed a written consent, answered baseline questions, and were randomized to either arm. Next, research staff introduced all study subjects to the website to which they were randomized and provided instructions on its use. Staff were present to assist with questions as needed on navigation but did not assist with decision making for either website. A minimum of 15 minutes and a maximum of 45 minutes was allotted for navigating either website. After navigating their website, participants were administered an immediate in-person posttest survey. One month and 3 months after the face-to-face encounter, research staff contacted all study participants over the phone to complete a follow-up survey. Staff attempted to reach participants up to 3 times by phone. Data were entered into Research Electronic Data Capture survey software.¹⁵ This study was approved by the Northwestern University Institutional Review Board.

Understanding of Posthospital Discharge and Home Services

As part of the larger trial, which included behavioural outcomes that will be reported elsewhere, we sought to explore the effects of PYL on participants’ knowledge and understanding of posthospital discharge and home services (UHS). Participants were asked to respond to 6 questions at baseline, immediate posttest, and at the 1- and 3-month follow-up time points. Knowledge items were developed by the study team in conjunction with the patient/partner stakeholders. UHS scores were calculated as the sum of the 6 questions (each scored 0 if incorrect and 1 if correct) with a possible range of 0-6.

Covariates

Demographic information, self-reported health, importance of religion, and existence of a power of attorney, living will, advanced directive (eg, Physician Orders for Life-Sustaining Treatment) were obtained via self-report. Participants were asked about their general and social self-efficacy using the validated Self-efficacy Scale¹⁶ and their social support using the Lubben Social Network Scale–6.¹⁷ Health literacy was assessed using the Rapid Estimate of Adult Literacy in Medicine–Short Form.¹⁸ To measure burden of disease, participant comorbidities were measured using a nonvalidated 9-item dichotomous response condition list, which included some items adapted from the Charlson Comorbidity Index and the Elixhauser Comorbidity Index.

Statistical Analysis

Data analysis included all available data in the intention-to-treat dataset. As UHS was collected at multiple time points up to 3 months postintervention, we employed linear- mixed modeling with random participant effect and fixed arm, time, and time-by-arm interaction terms. The time-by-arm interaction allows for comparison of UHS slopes (or trajectories) across arms. Analyses explored multiple potential covariates, including current utilization of services, physical function, comorbidities, social support, health literacy, self-efficacy, and sociodemographics. Those covariates found to have a significant association (P < 0.05) with outcome were considered for inclusion in the overall model selection process. Ultimately, we developed a final parsimonious, adjusted longitudinal model with primary predictors of time, arm, and their interaction, controlling for only significant baseline variables following a manual backward selection method. All analyses were conducted in SAS software (version 9.4, copyright 2012, The SAS Institute Inc., Cary, North Carolina).

The results of our study show that among seniors, PYL improved their understanding of home-based services and the services that may be required following a hospitalization. Educating seniors about what to expect regarding the transition home from a hospital before a hospitalization even occurs may enable seniors and their families to plan ahead instead of reacting to a hospitalization. PYL, a national, publicly available tool with links to local resources may potentially help in advancing transitional discharge care to prior to a hospitalization.

To our knowledge, this is one of the first websites and trials devoted to planning for seniors’ health trajectory as they age into their 70s, 80s, 90s, and 100s. Clinicians regularly discuss code status and powers of attorney during their end-of-life discussions with patients. We encourage clinicians to ask patients, “What about the 10 to 20 years before you die? Have you considered what you will do if you get sick or need help at home?” While not replacing a social worker, the ability of PYL to connect seniors to local resources makes it somewhat of a “virtual social worker.” With many physician practices unable to afford social workers, PYL provides a free-of-charge means of connecting seniors to area resources.

Acknowledgments

This work was supported through a Patient-Centered Outcomes Research Institute Award (IH-12-11-4259). Dr. Lindquist and Dr. Ciolino had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Research reported in this publication was also supported, in part, by the National Institutes of Health’s National Center for Advancing Translational Sciences, Grant Number UL1TR000150.

Disclaimer 

All statements in this manuscript, including its findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors or Methodology Committee, or the National Institutes of Health.

Disclosure: The authors have nothing to report.

When seniors are discharged from the hospital, many will require additional support and therapy to regain their independence and return safely home.^1,2 Most seniors do not understand what their support needs will entail or the differences between therapy choices.³ To complicate the issue, seniors are often incapacitated and unable to make discharge selections for themselves.

Consequently, discharge planners and social workers often explain options to family members and loved ones, who frequently feel overwhelmed.^4,5 While often balancing jobs, loved ones are divided between wanting to stay with the senior in the hospital and driving to area skilled nursing facilities (SNFs) for consideration. Discharges are delayed waiting for families to make visits and choose an SNF. Longer lengths of stay are detrimental to seniors due to the increased risks of infection, functional loss, and cognitive decline.⁶

Although seniors comprised only 12% of the US population in 2003,⁷ they accounted for one-third of all hospitalizations, over 13.2 million hospital stays.⁸ Hospital stays for seniors resulted in hospital charges totaling nearly $329 billion, or 43.6% of national hospital bills in 2003.⁷ Seniors are also high consumers of postacute care services. By 2050, the number of individuals using long-term care services in any setting (eg, at home, assisted living, or SNFs) will be close to 27 million.^9-11 With the knowledge that many seniors will be hospitalized and subsequently require services thereafter, we sought to assist seniors in planning for their hospital discharge needs before they were hospitalized.

Our team developed PlanYourLifespan.org (PYL) to facilitate this planning for postdischarge needs and fill the knowledge gap in understanding postdischarge options. With funding from the Patient Centered Outcomes Research Institute, we aimed to test the effectiveness of PYL on improving knowledge of hospital discharge resources among seniors.

PlanYourLifespan.org

PlanYourLifespan.org (PYL) educates users on the health crises that often occur with age and connects them to posthospital and home-based resources available locally and nationally. PYL is personalized, dynamic, and adaptable in that all the information can be changed per the senior’s wishes or changing health needs.

Content of PYL

Previously, we conducted focus groups with seniors about current and perceived home needs and aging-in-place. Major themes of what advanced life events (ALEs) would impact aging-in-place were identified as follows: hospitalizations, falls, and Alzheimer’s.¹² We organized PYL around these health-related ALEs. Our multidisciplinary team of researchers, seniors, social workers, caregiver agencies, and Area Agencies on Aging representatives then determined what information and resources should be included.

Each section of PYL starts with a video of a senior discussing their real-life personal experiences, with subsections providing interactive information on what seniors can expect, types of resources available to support home needs, and choices to be made. Descriptions of types of settings for therapy, options available, and links to national/local resources (eg, quality indicators for SNFs) are also included. For example, by entering their zip code, users can identify their neighborhood SNFs, closest Area Agency on Aging, and what home caregiver agencies exist in their area.

Users can save their preferences and revisit their choices at any time. To support communication between seniors and their loved ones, a summary of their choices can be printed or e-mailed to relevant parties. For example, a senior uses PYL and can e-mail these choices to family members, which can stimulate a conversation about future posthospital care expectations.

As inadequate health literacy and cognitive impairment are prevalent among seniors, PYL presents information understandable at all levels of health literacy and sensitive to cognitive load.⁹ There is simplified, large-font, no mouse scrolling and audio available for the visually impaired.

Study Design and Randomization

To test PYL, a 2-armed (attention control [AC] and PYL intervention), parallel, randomized controlled trial was conducted. Participants were randomly assigned to 1 of the 2 conditions via a pregenerated central randomization list using equal (1:1) allocation and random permuted block design to ensure relatively equal allocation throughout the study. The AC condition exposed participants to the National Institute on Aging-sponsored website, Go4Life.nia.nih.gov, an educational website on physical activity for seniors. This website has comparable design and layout to PYL but does not include information about advanced planning. The AC condition controlled for the possibility that regular contact with the study team may improve outcomes in participants randomized to the intervention website.

The trial was conducted from October 2014 to September 2015 in Chicago, Illinois; Fort Wayne, Indiana; and Houston, Texas. Inclusion criteria were as follows: age 65 and older, English-speaking, scoring ≥4 questions correctly on the Brief Cognitive Screen,¹⁴ and current self-reported use of a computer or smartphone with internet. Participants were excluded if they had previously participated in the focus groups or beta testing of the PYL website. Community-based patient partners/stakeholders drove subject recruitment in their communities through word of mouth, e-mail bursts, newsletters, and flyers. At the Area Agency on Aging and community centers where services such as food vouchers and case management are provided, participants were recruited on-site and given study information. At the clinical sites, staff referred potential participants. Study materials such as flyers and information sheets were also located in the clinic waiting rooms. The Villages, nonprofit, grassroots, membership organizations that are redefining aging by being a key resource to community members wishing to age in place, heavily relied on electronic recruitment using their regular e-newsletters and e-mail lists to recruit potential participants. Potential subjects were also recruited by distributing flyers at local senior centers and senior housing buildings. Interested seniors contacted research staff who explained the study and assessed their eligibility. If eligible, subjects were scheduled for a face-to-face interview.

At the face-to-face encounter, all study subjects completed a written consent, answered baseline questions, and were randomized to either arm. Next, research staff introduced all study subjects to the website to which they were randomized and provided instructions on its use. Staff were present to assist with questions as needed on navigation but did not assist with decision making for either website. A minimum of 15 minutes and a maximum of 45 minutes was allotted for navigating either website. After navigating their website, participants were administered an immediate in-person posttest survey. One month and 3 months after the face-to-face encounter, research staff contacted all study participants over the phone to complete a follow-up survey. Staff attempted to reach participants up to 3 times by phone. Data were entered into Research Electronic Data Capture survey software.¹⁵ This study was approved by the Northwestern University Institutional Review Board.

Understanding of Posthospital Discharge and Home Services

As part of the larger trial, which included behavioural outcomes that will be reported elsewhere, we sought to explore the effects of PYL on participants’ knowledge and understanding of posthospital discharge and home services (UHS). Participants were asked to respond to 6 questions at baseline, immediate posttest, and at the 1- and 3-month follow-up time points. Knowledge items were developed by the study team in conjunction with the patient/partner stakeholders. UHS scores were calculated as the sum of the 6 questions (each scored 0 if incorrect and 1 if correct) with a possible range of 0-6.

Covariates

Demographic information, self-reported health, importance of religion, and existence of a power of attorney, living will, advanced directive (eg, Physician Orders for Life-Sustaining Treatment) were obtained via self-report. Participants were asked about their general and social self-efficacy using the validated Self-efficacy Scale¹⁶ and their social support using the Lubben Social Network Scale–6.¹⁷ Health literacy was assessed using the Rapid Estimate of Adult Literacy in Medicine–Short Form.¹⁸ To measure burden of disease, participant comorbidities were measured using a nonvalidated 9-item dichotomous response condition list, which included some items adapted from the Charlson Comorbidity Index and the Elixhauser Comorbidity Index.

Statistical Analysis

Data analysis included all available data in the intention-to-treat dataset. As UHS was collected at multiple time points up to 3 months postintervention, we employed linear- mixed modeling with random participant effect and fixed arm, time, and time-by-arm interaction terms. The time-by-arm interaction allows for comparison of UHS slopes (or trajectories) across arms. Analyses explored multiple potential covariates, including current utilization of services, physical function, comorbidities, social support, health literacy, self-efficacy, and sociodemographics. Those covariates found to have a significant association (P < 0.05) with outcome were considered for inclusion in the overall model selection process. Ultimately, we developed a final parsimonious, adjusted longitudinal model with primary predictors of time, arm, and their interaction, controlling for only significant baseline variables following a manual backward selection method. All analyses were conducted in SAS software (version 9.4, copyright 2012, The SAS Institute Inc., Cary, North Carolina).

The results of our study show that among seniors, PYL improved their understanding of home-based services and the services that may be required following a hospitalization. Educating seniors about what to expect regarding the transition home from a hospital before a hospitalization even occurs may enable seniors and their families to plan ahead instead of reacting to a hospitalization. PYL, a national, publicly available tool with links to local resources may potentially help in advancing transitional discharge care to prior to a hospitalization.

To our knowledge, this is one of the first websites and trials devoted to planning for seniors’ health trajectory as they age into their 70s, 80s, 90s, and 100s. Clinicians regularly discuss code status and powers of attorney during their end-of-life discussions with patients. We encourage clinicians to ask patients, “What about the 10 to 20 years before you die? Have you considered what you will do if you get sick or need help at home?” While not replacing a social worker, the ability of PYL to connect seniors to local resources makes it somewhat of a “virtual social worker.” With many physician practices unable to afford social workers, PYL provides a free-of-charge means of connecting seniors to area resources.

Acknowledgments

This work was supported through a Patient-Centered Outcomes Research Institute Award (IH-12-11-4259). Dr. Lindquist and Dr. Ciolino had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Research reported in this publication was also supported, in part, by the National Institutes of Health’s National Center for Advancing Translational Sciences, Grant Number UL1TR000150.

Disclaimer 

All statements in this manuscript, including its findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors or Methodology Committee, or the National Institutes of Health.

Disclosure: The authors have nothing to report.

Successful smoking cessation interventions result in substantial gains in health and life expectancy by reducing smoking-related illnesses and preventing premature deaths.^1,2 The Department of Health and Human Services recommends clinicians use hospitalization as “an opportunity to promote smoking cessation’’ and ‘‘to prescribe medications to alleviate withdrawal symptoms”³ because individual readiness to quit may be high during hospitalizations. A meta-analysis of 50 studies (21 from the United States) examining the efficacy of hospital-initiated smoking cessation interventions concluded that smoking cessation support programs that began in the hospital and continued for at least 1 month postdischarge significantly increase the likelihood of patients being smoke-free in the long term.⁴ The most efficacious strategies included counseling and pharmacotherapy rather than counseling alone.³ Most inpatient smoking cessation studies have focused on quit-rates or medical outcomes, while fewer studies have looked at healthcare utilization.

However, previous research has shown that smoking cessation for inpatients has relatively immediate economic and health benefits. Patients who quit smoking during hospitalizations for cardiovascular disease are less likely to be readmitted or to die during follow-up.^5,6 Patients with acute myocardial infarction (AMI), unstable angina, heart failure, and chronic obstructive pulmonary disease who received an inpatient smoking cessation intervention had reductions in inpatient readmission rates.⁷ A 1% reduction in overall smoking rates would lead to an annual reduction of 3,022 hospitalizations for stroke and 1,684 hospitalizations for AMI.⁸ One comprehensive program, the Ottawa Model for Smoking Cessation (OMSC), found that a hospital-initiated intervention increased long-term cessation rates by 15% in cardiac patients and by 11% in general hospital populations.^9,10 The applicability of this result to US healthcare systems is unknown. This paper adds to the existing literature by evaluating the impact of an inpatient smoking cessation program on healthcare utilization among patients hospitalized for any reason, rather than solely focused on those with cardiopulmonary diagnoses.

The current study focuses on an inpatient smoking cessation program at a teaching hospital in the Rocky Mountain region. The hospital implemented a smoking cessation treatment program on July 1, 2013, based on the OMSC. The goal was to identify and support inpatient adult smokers who wanted to make a quit attempt and help them remain smoke-free after discharge. The objective of the current study was to determine the effect of the program on 30-day readmission rates and length of stay (LOS) of the index hospitalization. Although the general cost effectiveness of properly structured smoking cessation programs are well established,^11-13 the healthcare utilization effects of inpatient smoking cessation programs are not well understood.

Data

The study population consists of patients over age 18 who were admitted to the hospital between July 1, 2012, and July 1, 2014. Baseline smoking status was assessed at hospital admission and recorded in Epic (Epic Systems Corporation, Verona, Wisconsin), the electronic medical records system, as a current smoker (every day and some days), former smoker, never smoker, and never assessed. To check the accuracy of recorded smoking status, a random sample of 819 inpatients was selected and contacted via telephone for verification; 93% of Epic-identified smokers confirmed that they were smokers at hospital admission.¹⁴

Intervention

The intervention, which launched July 13, 2014, modified the Epic system to automatically alert providers viewing a tobacco user’s medical record that the patient should receive standardized orders for a bedside consultation with a Tobacco Treatment Specialist (TTS) and a prescription for nicotine replacement therapy (NRT) while in the hospital.¹⁵ Previously, referrals for tobacco treatment were done on an ad-hoc basis by the physician, and NRT was not routinely available. This system-level intervention standardized and automated the referral process. For patients with a bedside consultation order, TTS used a patient-centered approach (motivational interviewing) to explore patients’ motivation to quit smoking and offered NRT to improve comfort and safety while in the hospital. Patients who chose to make a quit attempt received a free 2-week supply of NRT at discharge and 6 months of free follow-up counseling by interactive voice response (IVR) telephone technology that included (a) prerecorded advice keyed to individual patient needs, (b) a warm-transfer option to speak with a live TTS (later dropped), and (c) a collection of patient smoking and cessation treatment measures.¹⁵

Statistical Analysis

We used an intent-to-treat (ITT) framework for the analysis, which considers everyone eligible for the treatment to be in the treatment group. The approach ignores treatment nonacceptance, nonadherence, protocol deviations, withdrawal from treatment, and cessation outcomes, thus providing conservative estimates of outcomes.^16-18

Readmission rates and LOS were estimated by using a “difference-in-differences” model, comparing outcomes between smokers before versus after the introduction of the cessation treatment program with nonsmokers before versus after program introduction. The difference-in-differences method looks at the difference pre-and-post in the exposed group (smokers) and unexposed group (nonsmokers). Subtracting the difference between the 2 groups gives an estimate of the policy effect controlling for background trends.¹⁹ The smoking cessation treatment effect on readmission is measured by the coefficient on the interaction term between the smoking variable and an indicator that the program is operational. The coefficient is the “difference-in-differences.”

Other control variables include demographic factors (gender, age, race), hospitalization payer (Medicare, Medicaid, commercial), and the service line of the admission. We also included a severity of illness variable from the APR-DRG Grouper (3M, Maplewood, Minnesota)²⁰ and the number of days spent in the intensive care unit. For the readmission model, we included LOS as a control variable, because individuals with longer LOS had a better opportunity to access the intervention.

For readmissions, the model was estimated by using a probit model, predicting the effect of each of the intervention variables and the control variables on the marginal probability of a readmission. Because patients can appear in both the pre- and postyears, clustered standard errors were used, which correct for the lack of independence from multiple observations from the same individual.²¹ For LOS, a truncated negative binomial model was used. The negative binomial model is a specification for count models with a mass of observations plus a long right tail. The truncation is because zero and negative values for LOS are not possible. The dependent variable represents the number of days the individual was hospitalized. For both models, the reported coefficients represent the marginal effect of the independent variable on the dependent variable. This was calculated using the “margins” command in Stata version 13 (StataCorp LLC, College Station, Texas).

This study investigated the effect of an inpatient smoking cessation program, based on a successful Canadian model, on inpatient readmission rates and LOS. The program showed no effect on 30-day readmission rates or LOS. We see several potential explanations for the absence of a detectable impact.

First, the ITT approach reflected real-world implementation of smoking cessation services. The ITT approach adopts the hospital’s perspective because the hospital will assess overall effectiveness without regard to programmatic limitations. The intervention group for this analysis included individuals who were offered but declined treatment, individuals who accepted treatment but failed to quit smoking, and individuals who both accepted treatment and quit smoking. If the analysis had focused only on the latter group, an effect would have been more likely to be found. Further analysis of the subset of patients who accepted the intervention and quit smoking is warranted. Nevertheless, hospitals cannot expect all inpatient smokers, or even a majority, to embrace an offer of cessation treatment. This also emphasizes the challenges hospitals will face in offering tobacco cessation programs to smokers in a timely way. Reasons for patients not receiving orders varied but included issues with weekend admissions.

Second, the timeframe of the analysis is limited to the inpatient stay (for LOS) and 30 days (for readmission). A longer-term analysis might have found an effect. However, we examined this from the hospital perspective. For the hospital, LOS is a key cost driver; thus, reductions in LOS would create a strong financial incentive for hospitals to implement smoking cessation programs. Similarly, reducing readmissions is now a priority for hospitals because of new Medicare rules that penalize hospitals for readmissions. Thus, the 2 outcomes we examined are outcomes that are financially important to hospitals.

There are several limitations to our analysis. First, the difference-in-differences model assumes that in the absence of treatment, the average change in the dependent variables would have been the same for both the treatment and control groups, also known as the parallel trends assumption. Specification tests showed this assumption was met for the preperiod. Second, our study relies on electronic health record data to identify smokers. However, 93% of individuals who were identified as smokers confirmed their smoking status upon interview. Finally, we looked at all categories of inpatient admissions. Improvement in LOS and short-term readmission rates may be limited to patients admitted for specific conditions, such as cardiovascular and respiratory conditions.

There are a number of plausible reasons for our null finding. First, the “dose” of intervention may have been too weak; that is, the number of smokers who were offered the treatment, accepted the treatment, and adhered to the treatment may have been too low, leading to too few smokers quitting smoking and, thus, no effect of the intervention on our outcomes. This follows directly from the ITT design of the study.²³ This suggests that hospitals who wish to adopt smoking cessation programs need to focus on ensuring a timely offering of treatment and encouragement of uptake by smokers.

A second reason for the null finding may have been the short duration for the NRT, which was only offered for 2 weeks. Research suggests that use of NRT for less than 4 weeks is associated with a reduced likelihood of smoking cessation.²⁴ However, a review of the literature concludes that the duration of NRT is less important than the dosage and the combination of NRT with other forms of smoking cessation therapy.²⁵ It is important to note that this study used NRT; other treatments such as Chantix could have different effectiveness.^26,27 Further research on different treatment approaches, including longer duration of NRT, would be appropriate.

Disclosure

 The authors have no competing interests or conflicts to report. The study was supported by contract number 15FLA68717 from the Colorado Department of Public Health and Environment.

Successful smoking cessation interventions result in substantial gains in health and life expectancy by reducing smoking-related illnesses and preventing premature deaths.^1,2 The Department of Health and Human Services recommends clinicians use hospitalization as “an opportunity to promote smoking cessation’’ and ‘‘to prescribe medications to alleviate withdrawal symptoms”³ because individual readiness to quit may be high during hospitalizations. A meta-analysis of 50 studies (21 from the United States) examining the efficacy of hospital-initiated smoking cessation interventions concluded that smoking cessation support programs that began in the hospital and continued for at least 1 month postdischarge significantly increase the likelihood of patients being smoke-free in the long term.⁴ The most efficacious strategies included counseling and pharmacotherapy rather than counseling alone.³ Most inpatient smoking cessation studies have focused on quit-rates or medical outcomes, while fewer studies have looked at healthcare utilization.

However, previous research has shown that smoking cessation for inpatients has relatively immediate economic and health benefits. Patients who quit smoking during hospitalizations for cardiovascular disease are less likely to be readmitted or to die during follow-up.^5,6 Patients with acute myocardial infarction (AMI), unstable angina, heart failure, and chronic obstructive pulmonary disease who received an inpatient smoking cessation intervention had reductions in inpatient readmission rates.⁷ A 1% reduction in overall smoking rates would lead to an annual reduction of 3,022 hospitalizations for stroke and 1,684 hospitalizations for AMI.⁸ One comprehensive program, the Ottawa Model for Smoking Cessation (OMSC), found that a hospital-initiated intervention increased long-term cessation rates by 15% in cardiac patients and by 11% in general hospital populations.^9,10 The applicability of this result to US healthcare systems is unknown. This paper adds to the existing literature by evaluating the impact of an inpatient smoking cessation program on healthcare utilization among patients hospitalized for any reason, rather than solely focused on those with cardiopulmonary diagnoses.

The current study focuses on an inpatient smoking cessation program at a teaching hospital in the Rocky Mountain region. The hospital implemented a smoking cessation treatment program on July 1, 2013, based on the OMSC. The goal was to identify and support inpatient adult smokers who wanted to make a quit attempt and help them remain smoke-free after discharge. The objective of the current study was to determine the effect of the program on 30-day readmission rates and length of stay (LOS) of the index hospitalization. Although the general cost effectiveness of properly structured smoking cessation programs are well established,^11-13 the healthcare utilization effects of inpatient smoking cessation programs are not well understood.

Data

The study population consists of patients over age 18 who were admitted to the hospital between July 1, 2012, and July 1, 2014. Baseline smoking status was assessed at hospital admission and recorded in Epic (Epic Systems Corporation, Verona, Wisconsin), the electronic medical records system, as a current smoker (every day and some days), former smoker, never smoker, and never assessed. To check the accuracy of recorded smoking status, a random sample of 819 inpatients was selected and contacted via telephone for verification; 93% of Epic-identified smokers confirmed that they were smokers at hospital admission.¹⁴

Intervention

The intervention, which launched July 13, 2014, modified the Epic system to automatically alert providers viewing a tobacco user’s medical record that the patient should receive standardized orders for a bedside consultation with a Tobacco Treatment Specialist (TTS) and a prescription for nicotine replacement therapy (NRT) while in the hospital.¹⁵ Previously, referrals for tobacco treatment were done on an ad-hoc basis by the physician, and NRT was not routinely available. This system-level intervention standardized and automated the referral process. For patients with a bedside consultation order, TTS used a patient-centered approach (motivational interviewing) to explore patients’ motivation to quit smoking and offered NRT to improve comfort and safety while in the hospital. Patients who chose to make a quit attempt received a free 2-week supply of NRT at discharge and 6 months of free follow-up counseling by interactive voice response (IVR) telephone technology that included (a) prerecorded advice keyed to individual patient needs, (b) a warm-transfer option to speak with a live TTS (later dropped), and (c) a collection of patient smoking and cessation treatment measures.¹⁵

Statistical Analysis

We used an intent-to-treat (ITT) framework for the analysis, which considers everyone eligible for the treatment to be in the treatment group. The approach ignores treatment nonacceptance, nonadherence, protocol deviations, withdrawal from treatment, and cessation outcomes, thus providing conservative estimates of outcomes.^16-18

Readmission rates and LOS were estimated by using a “difference-in-differences” model, comparing outcomes between smokers before versus after the introduction of the cessation treatment program with nonsmokers before versus after program introduction. The difference-in-differences method looks at the difference pre-and-post in the exposed group (smokers) and unexposed group (nonsmokers). Subtracting the difference between the 2 groups gives an estimate of the policy effect controlling for background trends.¹⁹ The smoking cessation treatment effect on readmission is measured by the coefficient on the interaction term between the smoking variable and an indicator that the program is operational. The coefficient is the “difference-in-differences.”

Other control variables include demographic factors (gender, age, race), hospitalization payer (Medicare, Medicaid, commercial), and the service line of the admission. We also included a severity of illness variable from the APR-DRG Grouper (3M, Maplewood, Minnesota)²⁰ and the number of days spent in the intensive care unit. For the readmission model, we included LOS as a control variable, because individuals with longer LOS had a better opportunity to access the intervention.

For readmissions, the model was estimated by using a probit model, predicting the effect of each of the intervention variables and the control variables on the marginal probability of a readmission. Because patients can appear in both the pre- and postyears, clustered standard errors were used, which correct for the lack of independence from multiple observations from the same individual.²¹ For LOS, a truncated negative binomial model was used. The negative binomial model is a specification for count models with a mass of observations plus a long right tail. The truncation is because zero and negative values for LOS are not possible. The dependent variable represents the number of days the individual was hospitalized. For both models, the reported coefficients represent the marginal effect of the independent variable on the dependent variable. This was calculated using the “margins” command in Stata version 13 (StataCorp LLC, College Station, Texas).

This study investigated the effect of an inpatient smoking cessation program, based on a successful Canadian model, on inpatient readmission rates and LOS. The program showed no effect on 30-day readmission rates or LOS. We see several potential explanations for the absence of a detectable impact.

First, the ITT approach reflected real-world implementation of smoking cessation services. The ITT approach adopts the hospital’s perspective because the hospital will assess overall effectiveness without regard to programmatic limitations. The intervention group for this analysis included individuals who were offered but declined treatment, individuals who accepted treatment but failed to quit smoking, and individuals who both accepted treatment and quit smoking. If the analysis had focused only on the latter group, an effect would have been more likely to be found. Further analysis of the subset of patients who accepted the intervention and quit smoking is warranted. Nevertheless, hospitals cannot expect all inpatient smokers, or even a majority, to embrace an offer of cessation treatment. This also emphasizes the challenges hospitals will face in offering tobacco cessation programs to smokers in a timely way. Reasons for patients not receiving orders varied but included issues with weekend admissions.

Second, the timeframe of the analysis is limited to the inpatient stay (for LOS) and 30 days (for readmission). A longer-term analysis might have found an effect. However, we examined this from the hospital perspective. For the hospital, LOS is a key cost driver; thus, reductions in LOS would create a strong financial incentive for hospitals to implement smoking cessation programs. Similarly, reducing readmissions is now a priority for hospitals because of new Medicare rules that penalize hospitals for readmissions. Thus, the 2 outcomes we examined are outcomes that are financially important to hospitals.

There are several limitations to our analysis. First, the difference-in-differences model assumes that in the absence of treatment, the average change in the dependent variables would have been the same for both the treatment and control groups, also known as the parallel trends assumption. Specification tests showed this assumption was met for the preperiod. Second, our study relies on electronic health record data to identify smokers. However, 93% of individuals who were identified as smokers confirmed their smoking status upon interview. Finally, we looked at all categories of inpatient admissions. Improvement in LOS and short-term readmission rates may be limited to patients admitted for specific conditions, such as cardiovascular and respiratory conditions.

There are a number of plausible reasons for our null finding. First, the “dose” of intervention may have been too weak; that is, the number of smokers who were offered the treatment, accepted the treatment, and adhered to the treatment may have been too low, leading to too few smokers quitting smoking and, thus, no effect of the intervention on our outcomes. This follows directly from the ITT design of the study.²³ This suggests that hospitals who wish to adopt smoking cessation programs need to focus on ensuring a timely offering of treatment and encouragement of uptake by smokers.

A second reason for the null finding may have been the short duration for the NRT, which was only offered for 2 weeks. Research suggests that use of NRT for less than 4 weeks is associated with a reduced likelihood of smoking cessation.²⁴ However, a review of the literature concludes that the duration of NRT is less important than the dosage and the combination of NRT with other forms of smoking cessation therapy.²⁵ It is important to note that this study used NRT; other treatments such as Chantix could have different effectiveness.^26,27 Further research on different treatment approaches, including longer duration of NRT, would be appropriate.

Disclosure

 The authors have no competing interests or conflicts to report. The study was supported by contract number 15FLA68717 from the Colorado Department of Public Health and Environment.

Successful smoking cessation interventions result in substantial gains in health and life expectancy by reducing smoking-related illnesses and preventing premature deaths.^1,2 The Department of Health and Human Services recommends clinicians use hospitalization as “an opportunity to promote smoking cessation’’ and ‘‘to prescribe medications to alleviate withdrawal symptoms”³ because individual readiness to quit may be high during hospitalizations. A meta-analysis of 50 studies (21 from the United States) examining the efficacy of hospital-initiated smoking cessation interventions concluded that smoking cessation support programs that began in the hospital and continued for at least 1 month postdischarge significantly increase the likelihood of patients being smoke-free in the long term.⁴ The most efficacious strategies included counseling and pharmacotherapy rather than counseling alone.³ Most inpatient smoking cessation studies have focused on quit-rates or medical outcomes, while fewer studies have looked at healthcare utilization.

However, previous research has shown that smoking cessation for inpatients has relatively immediate economic and health benefits. Patients who quit smoking during hospitalizations for cardiovascular disease are less likely to be readmitted or to die during follow-up.^5,6 Patients with acute myocardial infarction (AMI), unstable angina, heart failure, and chronic obstructive pulmonary disease who received an inpatient smoking cessation intervention had reductions in inpatient readmission rates.⁷ A 1% reduction in overall smoking rates would lead to an annual reduction of 3,022 hospitalizations for stroke and 1,684 hospitalizations for AMI.⁸ One comprehensive program, the Ottawa Model for Smoking Cessation (OMSC), found that a hospital-initiated intervention increased long-term cessation rates by 15% in cardiac patients and by 11% in general hospital populations.^9,10 The applicability of this result to US healthcare systems is unknown. This paper adds to the existing literature by evaluating the impact of an inpatient smoking cessation program on healthcare utilization among patients hospitalized for any reason, rather than solely focused on those with cardiopulmonary diagnoses.

The current study focuses on an inpatient smoking cessation program at a teaching hospital in the Rocky Mountain region. The hospital implemented a smoking cessation treatment program on July 1, 2013, based on the OMSC. The goal was to identify and support inpatient adult smokers who wanted to make a quit attempt and help them remain smoke-free after discharge. The objective of the current study was to determine the effect of the program on 30-day readmission rates and length of stay (LOS) of the index hospitalization. Although the general cost effectiveness of properly structured smoking cessation programs are well established,^11-13 the healthcare utilization effects of inpatient smoking cessation programs are not well understood.

Data

The study population consists of patients over age 18 who were admitted to the hospital between July 1, 2012, and July 1, 2014. Baseline smoking status was assessed at hospital admission and recorded in Epic (Epic Systems Corporation, Verona, Wisconsin), the electronic medical records system, as a current smoker (every day and some days), former smoker, never smoker, and never assessed. To check the accuracy of recorded smoking status, a random sample of 819 inpatients was selected and contacted via telephone for verification; 93% of Epic-identified smokers confirmed that they were smokers at hospital admission.¹⁴

Intervention

The intervention, which launched July 13, 2014, modified the Epic system to automatically alert providers viewing a tobacco user’s medical record that the patient should receive standardized orders for a bedside consultation with a Tobacco Treatment Specialist (TTS) and a prescription for nicotine replacement therapy (NRT) while in the hospital.¹⁵ Previously, referrals for tobacco treatment were done on an ad-hoc basis by the physician, and NRT was not routinely available. This system-level intervention standardized and automated the referral process. For patients with a bedside consultation order, TTS used a patient-centered approach (motivational interviewing) to explore patients’ motivation to quit smoking and offered NRT to improve comfort and safety while in the hospital. Patients who chose to make a quit attempt received a free 2-week supply of NRT at discharge and 6 months of free follow-up counseling by interactive voice response (IVR) telephone technology that included (a) prerecorded advice keyed to individual patient needs, (b) a warm-transfer option to speak with a live TTS (later dropped), and (c) a collection of patient smoking and cessation treatment measures.¹⁵

Statistical Analysis

We used an intent-to-treat (ITT) framework for the analysis, which considers everyone eligible for the treatment to be in the treatment group. The approach ignores treatment nonacceptance, nonadherence, protocol deviations, withdrawal from treatment, and cessation outcomes, thus providing conservative estimates of outcomes.^16-18

Readmission rates and LOS were estimated by using a “difference-in-differences” model, comparing outcomes between smokers before versus after the introduction of the cessation treatment program with nonsmokers before versus after program introduction. The difference-in-differences method looks at the difference pre-and-post in the exposed group (smokers) and unexposed group (nonsmokers). Subtracting the difference between the 2 groups gives an estimate of the policy effect controlling for background trends.¹⁹ The smoking cessation treatment effect on readmission is measured by the coefficient on the interaction term between the smoking variable and an indicator that the program is operational. The coefficient is the “difference-in-differences.”

Other control variables include demographic factors (gender, age, race), hospitalization payer (Medicare, Medicaid, commercial), and the service line of the admission. We also included a severity of illness variable from the APR-DRG Grouper (3M, Maplewood, Minnesota)²⁰ and the number of days spent in the intensive care unit. For the readmission model, we included LOS as a control variable, because individuals with longer LOS had a better opportunity to access the intervention.

For readmissions, the model was estimated by using a probit model, predicting the effect of each of the intervention variables and the control variables on the marginal probability of a readmission. Because patients can appear in both the pre- and postyears, clustered standard errors were used, which correct for the lack of independence from multiple observations from the same individual.²¹ For LOS, a truncated negative binomial model was used. The negative binomial model is a specification for count models with a mass of observations plus a long right tail. The truncation is because zero and negative values for LOS are not possible. The dependent variable represents the number of days the individual was hospitalized. For both models, the reported coefficients represent the marginal effect of the independent variable on the dependent variable. This was calculated using the “margins” command in Stata version 13 (StataCorp LLC, College Station, Texas).

This study investigated the effect of an inpatient smoking cessation program, based on a successful Canadian model, on inpatient readmission rates and LOS. The program showed no effect on 30-day readmission rates or LOS. We see several potential explanations for the absence of a detectable impact.

First, the ITT approach reflected real-world implementation of smoking cessation services. The ITT approach adopts the hospital’s perspective because the hospital will assess overall effectiveness without regard to programmatic limitations. The intervention group for this analysis included individuals who were offered but declined treatment, individuals who accepted treatment but failed to quit smoking, and individuals who both accepted treatment and quit smoking. If the analysis had focused only on the latter group, an effect would have been more likely to be found. Further analysis of the subset of patients who accepted the intervention and quit smoking is warranted. Nevertheless, hospitals cannot expect all inpatient smokers, or even a majority, to embrace an offer of cessation treatment. This also emphasizes the challenges hospitals will face in offering tobacco cessation programs to smokers in a timely way. Reasons for patients not receiving orders varied but included issues with weekend admissions.

Second, the timeframe of the analysis is limited to the inpatient stay (for LOS) and 30 days (for readmission). A longer-term analysis might have found an effect. However, we examined this from the hospital perspective. For the hospital, LOS is a key cost driver; thus, reductions in LOS would create a strong financial incentive for hospitals to implement smoking cessation programs. Similarly, reducing readmissions is now a priority for hospitals because of new Medicare rules that penalize hospitals for readmissions. Thus, the 2 outcomes we examined are outcomes that are financially important to hospitals.

There are several limitations to our analysis. First, the difference-in-differences model assumes that in the absence of treatment, the average change in the dependent variables would have been the same for both the treatment and control groups, also known as the parallel trends assumption. Specification tests showed this assumption was met for the preperiod. Second, our study relies on electronic health record data to identify smokers. However, 93% of individuals who were identified as smokers confirmed their smoking status upon interview. Finally, we looked at all categories of inpatient admissions. Improvement in LOS and short-term readmission rates may be limited to patients admitted for specific conditions, such as cardiovascular and respiratory conditions.

There are a number of plausible reasons for our null finding. First, the “dose” of intervention may have been too weak; that is, the number of smokers who were offered the treatment, accepted the treatment, and adhered to the treatment may have been too low, leading to too few smokers quitting smoking and, thus, no effect of the intervention on our outcomes. This follows directly from the ITT design of the study.²³ This suggests that hospitals who wish to adopt smoking cessation programs need to focus on ensuring a timely offering of treatment and encouragement of uptake by smokers.

A second reason for the null finding may have been the short duration for the NRT, which was only offered for 2 weeks. Research suggests that use of NRT for less than 4 weeks is associated with a reduced likelihood of smoking cessation.²⁴ However, a review of the literature concludes that the duration of NRT is less important than the dosage and the combination of NRT with other forms of smoking cessation therapy.²⁵ It is important to note that this study used NRT; other treatments such as Chantix could have different effectiveness.^26,27 Further research on different treatment approaches, including longer duration of NRT, would be appropriate.

Disclosure

 The authors have no competing interests or conflicts to report. The study was supported by contract number 15FLA68717 from the Colorado Department of Public Health and Environment.

Bacterial pneumonia remains an important cause of morbidity and mortality in the United States, and is the 8^th leading cause of death with 55,227 deaths among adults annually.¹ In 2005, the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) collaborated to update guidelines for hospital-acquired pneumonia (HAP), ventilator-associated pneumonia, and healthcare-associated pneumonia (HCAP).² This broad document outlines an evidence-based approach to diagnostic testing and antibiotic management based on the epidemiology and risk factors for these conditions. The guideline specifies the following criteria for HCAP: hospitalization in the past 90 days, residence in a skilled nursing facility (SNF), home infusion therapy, hemodialysis, home wound care, family members with multidrug resistant organisms (MDRO), and immunosuppressive diseases or medications, with the presumption that these patients are more likely to be harboring MDRO and should thus be treated empirically with broad-spectrum antibiotic therapy. Prior studies have shown that patients with HCAP have a more severe illness, are more likely to have MDRO, are more likely to be inadequately treated, and are at a higher risk for mortality than patients with community-acquired pneumonia (CAP).^3,4

These guidelines are controversial, especially in regard to the recommendations to empirically treat broadly with 2 antibiotics targeting Pseudomonas species, whether patients with HCAP merit broader spectrum coverage than patients with CAP, and whether the criteria for defining HCAP are adequate to predict which patients are harboring MDRO. It has subsequently been proposed that HCAP is more related to CAP than to HAP, and a recent update to the guideline removed recommendations for treatment of HCAP and will be placing HCAP into the guidelines for CAP instead.⁵ We sought to investigate the degree of uptake of the ATS and IDSA guideline recommendations by physicians over time, and whether this led to a change in outcomes among patients who met the criteria for HCAP.

Setting and Patients

We identified patients discharged between July 1, 2007, and November 30, 2011, from 488 US hospitals that participated in the Premier database (Premier Inc., Charlotte, North Carolina), an inpatient database developed for measuring quality and healthcare utilization. The database is frequently used for healthcare research and has been described previously.⁶ Member hospitals are in all regions of the US and are generally reflective of US hospitals. This database contains multiple data elements, including sociodemographic information, International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) diagnosis and procedure codes, hospital and physician information, source of admission, and discharge status. It also includes a date-stamped log of all billed items and services, including diagnostic tests, medications, and other treatments. Because the data do not contain identifiable information, the institutional review board at our medical center determined that this study did not constitute human subjects research.

We included all patients aged ≥18 years with a principal diagnosis of pneumonia or with a secondary diagnosis of pneumonia paired with a principal diagnosis of respiratory failure, acute respiratory distress syndrome, respiratory arrest, sepsis, or influenza. Patients were excluded if they were transferred to or from another acute care institution, had a length of stay of 1 day or less, had cystic fibrosis, did not have a chest radiograph, or did not receive antibiotics within 48 hours of admission.

For each patient, we extracted age, gender, principal diagnosis, comorbidities, and the specialty of the attending physician. Comorbidities were identified from ICD-9-CM secondary diagnosis codes and Diagnosis Related Groups by using Healthcare Cost and Utilization Project Comorbidity Software, version 3.1, based on the work of Elixhauser (Agency for Healthcare Research and Quality, Rockville, Maryland).⁷ In order to ensure that patients had HCAP, we required the presence of ≥1 HCAP criteria, including hospitalization in the past 90 days, hemodialysis, admission from an SNF, or immune suppression (which was derived from either a secondary diagnosis for neutropenia, hematological malignancy, organ transplant, acquired immunodeficiency virus, or receiving immunosuppressant drugs or corticosteroids [equivalent to ≥20 mg/day of prednisone]).

Definitions of Guideline-Concordant and Discordant Antibiotic Therapy

The ATS and IDSA guidelines recommended the following antibiotic combinations for HCAP: an antipseudomonal cephalosporin or carbapenem or a beta-lactam/lactamase inhibitor, plus an antipseudomonal quinolone or aminoglycoside, plus an antibiotic with activity versus methicillin resistant Staphylococcus aureus (MRSA), such as vancomycin or linezolid. Based on these guidelines, we defined the receipt of fully guideline-concordant antibiotics as 2 recommended antibiotics for Pseudomonas species plus 1 for MRSA administered by the second day of admission. Partially guideline-concordant antibiotics were defined as 1 recommended antibiotic for Pseudomonas species plus 1 for MRSA by the second day of hospitalization. Guideline-discordant antibiotics were defined as all other combinations.

Statistical Analysis

Descriptive statistics on patient characteristics are presented as frequency, proportions for categorical factors, and median with interquartile range (IQR) for continuous variables for the full cohort and by treatment group, defined as fully or partially guideline-concordant antibiotic therapy or discordant therapy. Hospital rates of fully guideline-concordant treatment are presented overall and by hospital characteristics. The association of hospital characteristics with rates of fully guideline-concordant therapy were assessed by using 1-way analysis of variance tests.

To assess trends across hospitals for the association between the use of guideline-concordant therapy and mortality, progression to respiratory failure as measured by the late initiation of invasive mechanical ventilation (day 3 or later), and the length of stay among survivors, we divided the 4.5-year study period into 9 intervals of 6 months each; 292 hospitals that submitted data for all 9 time points were examined in this analysis. Based on the distribution of length of stay in the first time period, we created an indicator variable for extended length of stay with length of stay at or above the 75th percentile, defined as extended. For each hospital at each 6-month interval, we then computed risk-standardized guideline-concordant treatment (RS-treatment) rates and risk-standardized in-hospital outcome rates similar to methods used by the Centers for Medicare and Medicaid Services for public reporting.⁸ For each hospital at each time interval, we estimated a predicted rate of guideline-concordant treatment as the sum of predicted probabilities of guideline-concordant treatment from patient factors and the random intercept for the hospital in which they were admitted. We then calculated the expected rate of guideline-concordant treatment as the sum of expected probabilities of treatment received from patient factors only. RS-treatment was then calculated as the ratio of predicted to expected rates multiplied by the overall unadjusted mean treatment rate from all patients.⁹ We repeated the same modeling strategy to calculate risk-standardized outcome (RS-outcome) rates for each hospital across all time points. All models were adjusted for patient demographics and comorbidities. Similar models using administrative data have moderate discrimination for mortality.¹⁰

We then fit mixed-effects linear models with random hospital intercept and slope across time for the RS-treatment and outcome rates, respectively. From these models, we estimated the mean slope for RS-treatment and for RS-outcome over time. In addition, we estimated a slope or trend over time for each hospital for treatment and for outcome and evaluated the correlation between the treatment and outcome trends.

All analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and STATA release 13 (StataCorp, LLC, College Station, Texas).

In this large, retrospective cohort study, we found that there was a substantial gap between the empiric antibiotics recommended by the ATS and IDSA guidelines and the empiric antibiotics that patients actually received. Over the study period, we saw an increased adherence to guidelines, in spite of growing evidence that HCAP risk factors do not adequately predict which patients are at risk for infection with an MDRO.¹¹ We used this change in antibiotic prescribing behavior over time to determine if there was a clinical impact on patient outcomes and found that at the hospital level, there were no improvements in mortality, excess length of stay, or progression to respiratory failure despite a doubling in guideline-concordant antibiotic use.

At least 2 other large studies have assessed the association between guideline-concordant therapy and outcomes in HCAP.^12,13 Both found that guideline-concordant therapy was associated with increased mortality, despite propensity matching. Both were conducted at the individual patient level by using administrative data, and results were likely affected by unmeasured clinical confounders, with sicker patients being more likely to receive guideline-concordant therapy. Our focus on the outcomes at the hospital level avoids this selection bias because the overall severity of illness of patients at any given hospital would not be expected to change over the study period, while physician uptake of antibiotic prescribing guidelines would be expected to increase over time. Determining the correlation between increases in guideline adherence and changes in patient outcome may offer a better assessment of the impact of guideline adherence. In this regard, our results are similar to those achieved by 1 quality improvement collaborative that was aimed at increasing guideline concordant therapy in ICUs. Despite an increase in guideline concordance from 33% to 47% of patients, they found no change in overall mortality.¹⁴

There were several limitations to our study. We did not have access to microbiologic data, so we were unable to determine which patients had MDRO infection or determine antibiotic-pathogen matching. However, the treating physicians in our study population presumably did not have access to this data at the time of treatment either because the time period we examined was within the first 48 hours of hospitalization, the interval during which cultures are incubating and the patients are being treated empirically. In addition, there may have been HCAP patients that we failed to identify, such as patients who were admitted in the past 90 days to a hospital that does not submit data to Premier. However, it is unlikely that prescribing for such patients should differ systematically from what we observed. While the database draws from 488 hospitals nationwide, it is possible that practices may be different at facilities that are not contained within the Premier database, such as Veterans Administration Hospitals. Similarly, we did not have readings for chest x-rays; hence, there could be some patients in the dataset who did not have pneumonia. However, we tried to overcome this by including only those patients with a principal diagnosis of pneumonia or sepsis with a secondary pneumonia diagnosis, a chest x-ray, and antibiotics administered within the first 48 hours of admission.

There are likely several reasons why so few HCAP patients in our study received guideline-concordant antibiotics. A lack of knowledge about the ATS and IDSA guidelines may have impacted the physicians in our study population. El-Solh et al.¹⁵ surveyed physicians about the ATS-IDSA guidelines 4 years after publication and found that only 45% were familiar with the document. We found that the rate of prescribing at least partially guideline-concordant antibiotics rose steadily over time, supporting the idea that the newness of the guidelines was 1 barrier. Additionally, prior studies have shown that many physicians may not agree with or choose to follow guidelines, with only 20% of physicians indicating that guidelines have a major impact on their clinical decision making,¹⁶ and the majority do not choose HCAP guideline-concordant antibiotics when tested.¹⁷ Alternatively, clinicians may not follow the guidelines because of a belief that the HCAP criteria do not adequately indicate patients who are at risk for MDRO. Previous studies have demonstrated the relative inability of HCAP risk factors to predict patients who harbor MDRO¹⁸ and suggest that better tools such as clinical scoring systems, which include not only the traditional HCAP risk factors but also prior exposure to antibiotics, prior culture data, and a cumulative assessment of both intrinsic and extrinsic factors, could more accurately predict MDRO and lead to a more judicious use of broad-spectrum antimicrobial agents.^19-25 Indeed, these collective findings have led the authors of the recently updated guidelines to remove HCAP as a clinical entity from the hospital-acquired or ventilator-associated pneumonia guidelines and place them instead in the upcoming updated guidelines on the management of CAP.⁵ Of these 3 explanations, the lack of familiarity fits best with our observation that guideline-concordant therapy increased steadily over time with no evidence of reaching a plateau. Ironically, as consensus was building that HCAP is a poor marker for MDROs, routine empiric treatment with vancomycin and piperacillin-tazobactam (“vanco and zosyn”) have become routine in many hospitals. Additional studies are needed to know if this trend has stabilized or reversed.

In conclusion, clinicians in our large, nationally representative sample treated the majority of HCAP patients as though they had CAP. Although there was an increase in the administration of guideline-concordant therapy over time, this increase was not associated with improved outcomes. This study supports the growing consensus that HCAP criteria do not accurately predict which patients benefit from broad-spectrum antibiotics for pneumonia, and most patients fare well with antibiotics targeting common community-acquired organisms.

Disclosure

 This work was supported by grant # R01HS018723 from the Agency for Healthcare Research and Quality. Dr. Lagu is also supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. Dr. Lindenauer is supported by grant K24HL132008 from the National Heart, Lung, and Blood Institute. The funding agency had no role in the data acquisition, analysis, or manuscript preparation for this study. Drs. Haessler and Rothberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Haessler, Lagu, Lindenauer, Skiest, Zilberberg, Higgins, and Rothberg conceived of the study and analyzed and interpreted the data. Dr. Lindenauer acquired the data. Dr. Pekow and Ms. Priya carried out the statistical analyses. Dr. Haessler drafted the manuscript. All authors critically reviewed the manuscript for accuracy and integrity. All authors certify no potential conflicts of interest. Preliminary results from this study were presented in oral and poster format at IDWeek in 2012 and 2013.

Bacterial pneumonia remains an important cause of morbidity and mortality in the United States, and is the 8^th leading cause of death with 55,227 deaths among adults annually.¹ In 2005, the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) collaborated to update guidelines for hospital-acquired pneumonia (HAP), ventilator-associated pneumonia, and healthcare-associated pneumonia (HCAP).² This broad document outlines an evidence-based approach to diagnostic testing and antibiotic management based on the epidemiology and risk factors for these conditions. The guideline specifies the following criteria for HCAP: hospitalization in the past 90 days, residence in a skilled nursing facility (SNF), home infusion therapy, hemodialysis, home wound care, family members with multidrug resistant organisms (MDRO), and immunosuppressive diseases or medications, with the presumption that these patients are more likely to be harboring MDRO and should thus be treated empirically with broad-spectrum antibiotic therapy. Prior studies have shown that patients with HCAP have a more severe illness, are more likely to have MDRO, are more likely to be inadequately treated, and are at a higher risk for mortality than patients with community-acquired pneumonia (CAP).^3,4

These guidelines are controversial, especially in regard to the recommendations to empirically treat broadly with 2 antibiotics targeting Pseudomonas species, whether patients with HCAP merit broader spectrum coverage than patients with CAP, and whether the criteria for defining HCAP are adequate to predict which patients are harboring MDRO. It has subsequently been proposed that HCAP is more related to CAP than to HAP, and a recent update to the guideline removed recommendations for treatment of HCAP and will be placing HCAP into the guidelines for CAP instead.⁵ We sought to investigate the degree of uptake of the ATS and IDSA guideline recommendations by physicians over time, and whether this led to a change in outcomes among patients who met the criteria for HCAP.

Setting and Patients

We identified patients discharged between July 1, 2007, and November 30, 2011, from 488 US hospitals that participated in the Premier database (Premier Inc., Charlotte, North Carolina), an inpatient database developed for measuring quality and healthcare utilization. The database is frequently used for healthcare research and has been described previously.⁶ Member hospitals are in all regions of the US and are generally reflective of US hospitals. This database contains multiple data elements, including sociodemographic information, International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) diagnosis and procedure codes, hospital and physician information, source of admission, and discharge status. It also includes a date-stamped log of all billed items and services, including diagnostic tests, medications, and other treatments. Because the data do not contain identifiable information, the institutional review board at our medical center determined that this study did not constitute human subjects research.

We included all patients aged ≥18 years with a principal diagnosis of pneumonia or with a secondary diagnosis of pneumonia paired with a principal diagnosis of respiratory failure, acute respiratory distress syndrome, respiratory arrest, sepsis, or influenza. Patients were excluded if they were transferred to or from another acute care institution, had a length of stay of 1 day or less, had cystic fibrosis, did not have a chest radiograph, or did not receive antibiotics within 48 hours of admission.

For each patient, we extracted age, gender, principal diagnosis, comorbidities, and the specialty of the attending physician. Comorbidities were identified from ICD-9-CM secondary diagnosis codes and Diagnosis Related Groups by using Healthcare Cost and Utilization Project Comorbidity Software, version 3.1, based on the work of Elixhauser (Agency for Healthcare Research and Quality, Rockville, Maryland).⁷ In order to ensure that patients had HCAP, we required the presence of ≥1 HCAP criteria, including hospitalization in the past 90 days, hemodialysis, admission from an SNF, or immune suppression (which was derived from either a secondary diagnosis for neutropenia, hematological malignancy, organ transplant, acquired immunodeficiency virus, or receiving immunosuppressant drugs or corticosteroids [equivalent to ≥20 mg/day of prednisone]).

Definitions of Guideline-Concordant and Discordant Antibiotic Therapy

The ATS and IDSA guidelines recommended the following antibiotic combinations for HCAP: an antipseudomonal cephalosporin or carbapenem or a beta-lactam/lactamase inhibitor, plus an antipseudomonal quinolone or aminoglycoside, plus an antibiotic with activity versus methicillin resistant Staphylococcus aureus (MRSA), such as vancomycin or linezolid. Based on these guidelines, we defined the receipt of fully guideline-concordant antibiotics as 2 recommended antibiotics for Pseudomonas species plus 1 for MRSA administered by the second day of admission. Partially guideline-concordant antibiotics were defined as 1 recommended antibiotic for Pseudomonas species plus 1 for MRSA by the second day of hospitalization. Guideline-discordant antibiotics were defined as all other combinations.

Statistical Analysis

Descriptive statistics on patient characteristics are presented as frequency, proportions for categorical factors, and median with interquartile range (IQR) for continuous variables for the full cohort and by treatment group, defined as fully or partially guideline-concordant antibiotic therapy or discordant therapy. Hospital rates of fully guideline-concordant treatment are presented overall and by hospital characteristics. The association of hospital characteristics with rates of fully guideline-concordant therapy were assessed by using 1-way analysis of variance tests.

To assess trends across hospitals for the association between the use of guideline-concordant therapy and mortality, progression to respiratory failure as measured by the late initiation of invasive mechanical ventilation (day 3 or later), and the length of stay among survivors, we divided the 4.5-year study period into 9 intervals of 6 months each; 292 hospitals that submitted data for all 9 time points were examined in this analysis. Based on the distribution of length of stay in the first time period, we created an indicator variable for extended length of stay with length of stay at or above the 75th percentile, defined as extended. For each hospital at each 6-month interval, we then computed risk-standardized guideline-concordant treatment (RS-treatment) rates and risk-standardized in-hospital outcome rates similar to methods used by the Centers for Medicare and Medicaid Services for public reporting.⁸ For each hospital at each time interval, we estimated a predicted rate of guideline-concordant treatment as the sum of predicted probabilities of guideline-concordant treatment from patient factors and the random intercept for the hospital in which they were admitted. We then calculated the expected rate of guideline-concordant treatment as the sum of expected probabilities of treatment received from patient factors only. RS-treatment was then calculated as the ratio of predicted to expected rates multiplied by the overall unadjusted mean treatment rate from all patients.⁹ We repeated the same modeling strategy to calculate risk-standardized outcome (RS-outcome) rates for each hospital across all time points. All models were adjusted for patient demographics and comorbidities. Similar models using administrative data have moderate discrimination for mortality.¹⁰

We then fit mixed-effects linear models with random hospital intercept and slope across time for the RS-treatment and outcome rates, respectively. From these models, we estimated the mean slope for RS-treatment and for RS-outcome over time. In addition, we estimated a slope or trend over time for each hospital for treatment and for outcome and evaluated the correlation between the treatment and outcome trends.

All analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and STATA release 13 (StataCorp, LLC, College Station, Texas).

In this large, retrospective cohort study, we found that there was a substantial gap between the empiric antibiotics recommended by the ATS and IDSA guidelines and the empiric antibiotics that patients actually received. Over the study period, we saw an increased adherence to guidelines, in spite of growing evidence that HCAP risk factors do not adequately predict which patients are at risk for infection with an MDRO.¹¹ We used this change in antibiotic prescribing behavior over time to determine if there was a clinical impact on patient outcomes and found that at the hospital level, there were no improvements in mortality, excess length of stay, or progression to respiratory failure despite a doubling in guideline-concordant antibiotic use.

At least 2 other large studies have assessed the association between guideline-concordant therapy and outcomes in HCAP.^12,13 Both found that guideline-concordant therapy was associated with increased mortality, despite propensity matching. Both were conducted at the individual patient level by using administrative data, and results were likely affected by unmeasured clinical confounders, with sicker patients being more likely to receive guideline-concordant therapy. Our focus on the outcomes at the hospital level avoids this selection bias because the overall severity of illness of patients at any given hospital would not be expected to change over the study period, while physician uptake of antibiotic prescribing guidelines would be expected to increase over time. Determining the correlation between increases in guideline adherence and changes in patient outcome may offer a better assessment of the impact of guideline adherence. In this regard, our results are similar to those achieved by 1 quality improvement collaborative that was aimed at increasing guideline concordant therapy in ICUs. Despite an increase in guideline concordance from 33% to 47% of patients, they found no change in overall mortality.¹⁴

There were several limitations to our study. We did not have access to microbiologic data, so we were unable to determine which patients had MDRO infection or determine antibiotic-pathogen matching. However, the treating physicians in our study population presumably did not have access to this data at the time of treatment either because the time period we examined was within the first 48 hours of hospitalization, the interval during which cultures are incubating and the patients are being treated empirically. In addition, there may have been HCAP patients that we failed to identify, such as patients who were admitted in the past 90 days to a hospital that does not submit data to Premier. However, it is unlikely that prescribing for such patients should differ systematically from what we observed. While the database draws from 488 hospitals nationwide, it is possible that practices may be different at facilities that are not contained within the Premier database, such as Veterans Administration Hospitals. Similarly, we did not have readings for chest x-rays; hence, there could be some patients in the dataset who did not have pneumonia. However, we tried to overcome this by including only those patients with a principal diagnosis of pneumonia or sepsis with a secondary pneumonia diagnosis, a chest x-ray, and antibiotics administered within the first 48 hours of admission.

There are likely several reasons why so few HCAP patients in our study received guideline-concordant antibiotics. A lack of knowledge about the ATS and IDSA guidelines may have impacted the physicians in our study population. El-Solh et al.¹⁵ surveyed physicians about the ATS-IDSA guidelines 4 years after publication and found that only 45% were familiar with the document. We found that the rate of prescribing at least partially guideline-concordant antibiotics rose steadily over time, supporting the idea that the newness of the guidelines was 1 barrier. Additionally, prior studies have shown that many physicians may not agree with or choose to follow guidelines, with only 20% of physicians indicating that guidelines have a major impact on their clinical decision making,¹⁶ and the majority do not choose HCAP guideline-concordant antibiotics when tested.¹⁷ Alternatively, clinicians may not follow the guidelines because of a belief that the HCAP criteria do not adequately indicate patients who are at risk for MDRO. Previous studies have demonstrated the relative inability of HCAP risk factors to predict patients who harbor MDRO¹⁸ and suggest that better tools such as clinical scoring systems, which include not only the traditional HCAP risk factors but also prior exposure to antibiotics, prior culture data, and a cumulative assessment of both intrinsic and extrinsic factors, could more accurately predict MDRO and lead to a more judicious use of broad-spectrum antimicrobial agents.^19-25 Indeed, these collective findings have led the authors of the recently updated guidelines to remove HCAP as a clinical entity from the hospital-acquired or ventilator-associated pneumonia guidelines and place them instead in the upcoming updated guidelines on the management of CAP.⁵ Of these 3 explanations, the lack of familiarity fits best with our observation that guideline-concordant therapy increased steadily over time with no evidence of reaching a plateau. Ironically, as consensus was building that HCAP is a poor marker for MDROs, routine empiric treatment with vancomycin and piperacillin-tazobactam (“vanco and zosyn”) have become routine in many hospitals. Additional studies are needed to know if this trend has stabilized or reversed.

In conclusion, clinicians in our large, nationally representative sample treated the majority of HCAP patients as though they had CAP. Although there was an increase in the administration of guideline-concordant therapy over time, this increase was not associated with improved outcomes. This study supports the growing consensus that HCAP criteria do not accurately predict which patients benefit from broad-spectrum antibiotics for pneumonia, and most patients fare well with antibiotics targeting common community-acquired organisms.

Disclosure

 This work was supported by grant # R01HS018723 from the Agency for Healthcare Research and Quality. Dr. Lagu is also supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. Dr. Lindenauer is supported by grant K24HL132008 from the National Heart, Lung, and Blood Institute. The funding agency had no role in the data acquisition, analysis, or manuscript preparation for this study. Drs. Haessler and Rothberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Haessler, Lagu, Lindenauer, Skiest, Zilberberg, Higgins, and Rothberg conceived of the study and analyzed and interpreted the data. Dr. Lindenauer acquired the data. Dr. Pekow and Ms. Priya carried out the statistical analyses. Dr. Haessler drafted the manuscript. All authors critically reviewed the manuscript for accuracy and integrity. All authors certify no potential conflicts of interest. Preliminary results from this study were presented in oral and poster format at IDWeek in 2012 and 2013.

Bacterial pneumonia remains an important cause of morbidity and mortality in the United States, and is the 8^th leading cause of death with 55,227 deaths among adults annually.¹ In 2005, the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) collaborated to update guidelines for hospital-acquired pneumonia (HAP), ventilator-associated pneumonia, and healthcare-associated pneumonia (HCAP).² This broad document outlines an evidence-based approach to diagnostic testing and antibiotic management based on the epidemiology and risk factors for these conditions. The guideline specifies the following criteria for HCAP: hospitalization in the past 90 days, residence in a skilled nursing facility (SNF), home infusion therapy, hemodialysis, home wound care, family members with multidrug resistant organisms (MDRO), and immunosuppressive diseases or medications, with the presumption that these patients are more likely to be harboring MDRO and should thus be treated empirically with broad-spectrum antibiotic therapy. Prior studies have shown that patients with HCAP have a more severe illness, are more likely to have MDRO, are more likely to be inadequately treated, and are at a higher risk for mortality than patients with community-acquired pneumonia (CAP).^3,4

These guidelines are controversial, especially in regard to the recommendations to empirically treat broadly with 2 antibiotics targeting Pseudomonas species, whether patients with HCAP merit broader spectrum coverage than patients with CAP, and whether the criteria for defining HCAP are adequate to predict which patients are harboring MDRO. It has subsequently been proposed that HCAP is more related to CAP than to HAP, and a recent update to the guideline removed recommendations for treatment of HCAP and will be placing HCAP into the guidelines for CAP instead.⁵ We sought to investigate the degree of uptake of the ATS and IDSA guideline recommendations by physicians over time, and whether this led to a change in outcomes among patients who met the criteria for HCAP.

Setting and Patients

We identified patients discharged between July 1, 2007, and November 30, 2011, from 488 US hospitals that participated in the Premier database (Premier Inc., Charlotte, North Carolina), an inpatient database developed for measuring quality and healthcare utilization. The database is frequently used for healthcare research and has been described previously.⁶ Member hospitals are in all regions of the US and are generally reflective of US hospitals. This database contains multiple data elements, including sociodemographic information, International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) diagnosis and procedure codes, hospital and physician information, source of admission, and discharge status. It also includes a date-stamped log of all billed items and services, including diagnostic tests, medications, and other treatments. Because the data do not contain identifiable information, the institutional review board at our medical center determined that this study did not constitute human subjects research.

We included all patients aged ≥18 years with a principal diagnosis of pneumonia or with a secondary diagnosis of pneumonia paired with a principal diagnosis of respiratory failure, acute respiratory distress syndrome, respiratory arrest, sepsis, or influenza. Patients were excluded if they were transferred to or from another acute care institution, had a length of stay of 1 day or less, had cystic fibrosis, did not have a chest radiograph, or did not receive antibiotics within 48 hours of admission.

For each patient, we extracted age, gender, principal diagnosis, comorbidities, and the specialty of the attending physician. Comorbidities were identified from ICD-9-CM secondary diagnosis codes and Diagnosis Related Groups by using Healthcare Cost and Utilization Project Comorbidity Software, version 3.1, based on the work of Elixhauser (Agency for Healthcare Research and Quality, Rockville, Maryland).⁷ In order to ensure that patients had HCAP, we required the presence of ≥1 HCAP criteria, including hospitalization in the past 90 days, hemodialysis, admission from an SNF, or immune suppression (which was derived from either a secondary diagnosis for neutropenia, hematological malignancy, organ transplant, acquired immunodeficiency virus, or receiving immunosuppressant drugs or corticosteroids [equivalent to ≥20 mg/day of prednisone]).

Definitions of Guideline-Concordant and Discordant Antibiotic Therapy

The ATS and IDSA guidelines recommended the following antibiotic combinations for HCAP: an antipseudomonal cephalosporin or carbapenem or a beta-lactam/lactamase inhibitor, plus an antipseudomonal quinolone or aminoglycoside, plus an antibiotic with activity versus methicillin resistant Staphylococcus aureus (MRSA), such as vancomycin or linezolid. Based on these guidelines, we defined the receipt of fully guideline-concordant antibiotics as 2 recommended antibiotics for Pseudomonas species plus 1 for MRSA administered by the second day of admission. Partially guideline-concordant antibiotics were defined as 1 recommended antibiotic for Pseudomonas species plus 1 for MRSA by the second day of hospitalization. Guideline-discordant antibiotics were defined as all other combinations.

Statistical Analysis

Descriptive statistics on patient characteristics are presented as frequency, proportions for categorical factors, and median with interquartile range (IQR) for continuous variables for the full cohort and by treatment group, defined as fully or partially guideline-concordant antibiotic therapy or discordant therapy. Hospital rates of fully guideline-concordant treatment are presented overall and by hospital characteristics. The association of hospital characteristics with rates of fully guideline-concordant therapy were assessed by using 1-way analysis of variance tests.

To assess trends across hospitals for the association between the use of guideline-concordant therapy and mortality, progression to respiratory failure as measured by the late initiation of invasive mechanical ventilation (day 3 or later), and the length of stay among survivors, we divided the 4.5-year study period into 9 intervals of 6 months each; 292 hospitals that submitted data for all 9 time points were examined in this analysis. Based on the distribution of length of stay in the first time period, we created an indicator variable for extended length of stay with length of stay at or above the 75th percentile, defined as extended. For each hospital at each 6-month interval, we then computed risk-standardized guideline-concordant treatment (RS-treatment) rates and risk-standardized in-hospital outcome rates similar to methods used by the Centers for Medicare and Medicaid Services for public reporting.⁸ For each hospital at each time interval, we estimated a predicted rate of guideline-concordant treatment as the sum of predicted probabilities of guideline-concordant treatment from patient factors and the random intercept for the hospital in which they were admitted. We then calculated the expected rate of guideline-concordant treatment as the sum of expected probabilities of treatment received from patient factors only. RS-treatment was then calculated as the ratio of predicted to expected rates multiplied by the overall unadjusted mean treatment rate from all patients.⁹ We repeated the same modeling strategy to calculate risk-standardized outcome (RS-outcome) rates for each hospital across all time points. All models were adjusted for patient demographics and comorbidities. Similar models using administrative data have moderate discrimination for mortality.¹⁰

We then fit mixed-effects linear models with random hospital intercept and slope across time for the RS-treatment and outcome rates, respectively. From these models, we estimated the mean slope for RS-treatment and for RS-outcome over time. In addition, we estimated a slope or trend over time for each hospital for treatment and for outcome and evaluated the correlation between the treatment and outcome trends.

All analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and STATA release 13 (StataCorp, LLC, College Station, Texas).

In this large, retrospective cohort study, we found that there was a substantial gap between the empiric antibiotics recommended by the ATS and IDSA guidelines and the empiric antibiotics that patients actually received. Over the study period, we saw an increased adherence to guidelines, in spite of growing evidence that HCAP risk factors do not adequately predict which patients are at risk for infection with an MDRO.¹¹ We used this change in antibiotic prescribing behavior over time to determine if there was a clinical impact on patient outcomes and found that at the hospital level, there were no improvements in mortality, excess length of stay, or progression to respiratory failure despite a doubling in guideline-concordant antibiotic use.

At least 2 other large studies have assessed the association between guideline-concordant therapy and outcomes in HCAP.^12,13 Both found that guideline-concordant therapy was associated with increased mortality, despite propensity matching. Both were conducted at the individual patient level by using administrative data, and results were likely affected by unmeasured clinical confounders, with sicker patients being more likely to receive guideline-concordant therapy. Our focus on the outcomes at the hospital level avoids this selection bias because the overall severity of illness of patients at any given hospital would not be expected to change over the study period, while physician uptake of antibiotic prescribing guidelines would be expected to increase over time. Determining the correlation between increases in guideline adherence and changes in patient outcome may offer a better assessment of the impact of guideline adherence. In this regard, our results are similar to those achieved by 1 quality improvement collaborative that was aimed at increasing guideline concordant therapy in ICUs. Despite an increase in guideline concordance from 33% to 47% of patients, they found no change in overall mortality.¹⁴

There were several limitations to our study. We did not have access to microbiologic data, so we were unable to determine which patients had MDRO infection or determine antibiotic-pathogen matching. However, the treating physicians in our study population presumably did not have access to this data at the time of treatment either because the time period we examined was within the first 48 hours of hospitalization, the interval during which cultures are incubating and the patients are being treated empirically. In addition, there may have been HCAP patients that we failed to identify, such as patients who were admitted in the past 90 days to a hospital that does not submit data to Premier. However, it is unlikely that prescribing for such patients should differ systematically from what we observed. While the database draws from 488 hospitals nationwide, it is possible that practices may be different at facilities that are not contained within the Premier database, such as Veterans Administration Hospitals. Similarly, we did not have readings for chest x-rays; hence, there could be some patients in the dataset who did not have pneumonia. However, we tried to overcome this by including only those patients with a principal diagnosis of pneumonia or sepsis with a secondary pneumonia diagnosis, a chest x-ray, and antibiotics administered within the first 48 hours of admission.

There are likely several reasons why so few HCAP patients in our study received guideline-concordant antibiotics. A lack of knowledge about the ATS and IDSA guidelines may have impacted the physicians in our study population. El-Solh et al.¹⁵ surveyed physicians about the ATS-IDSA guidelines 4 years after publication and found that only 45% were familiar with the document. We found that the rate of prescribing at least partially guideline-concordant antibiotics rose steadily over time, supporting the idea that the newness of the guidelines was 1 barrier. Additionally, prior studies have shown that many physicians may not agree with or choose to follow guidelines, with only 20% of physicians indicating that guidelines have a major impact on their clinical decision making,¹⁶ and the majority do not choose HCAP guideline-concordant antibiotics when tested.¹⁷ Alternatively, clinicians may not follow the guidelines because of a belief that the HCAP criteria do not adequately indicate patients who are at risk for MDRO. Previous studies have demonstrated the relative inability of HCAP risk factors to predict patients who harbor MDRO¹⁸ and suggest that better tools such as clinical scoring systems, which include not only the traditional HCAP risk factors but also prior exposure to antibiotics, prior culture data, and a cumulative assessment of both intrinsic and extrinsic factors, could more accurately predict MDRO and lead to a more judicious use of broad-spectrum antimicrobial agents.^19-25 Indeed, these collective findings have led the authors of the recently updated guidelines to remove HCAP as a clinical entity from the hospital-acquired or ventilator-associated pneumonia guidelines and place them instead in the upcoming updated guidelines on the management of CAP.⁵ Of these 3 explanations, the lack of familiarity fits best with our observation that guideline-concordant therapy increased steadily over time with no evidence of reaching a plateau. Ironically, as consensus was building that HCAP is a poor marker for MDROs, routine empiric treatment with vancomycin and piperacillin-tazobactam (“vanco and zosyn”) have become routine in many hospitals. Additional studies are needed to know if this trend has stabilized or reversed.

In conclusion, clinicians in our large, nationally representative sample treated the majority of HCAP patients as though they had CAP. Although there was an increase in the administration of guideline-concordant therapy over time, this increase was not associated with improved outcomes. This study supports the growing consensus that HCAP criteria do not accurately predict which patients benefit from broad-spectrum antibiotics for pneumonia, and most patients fare well with antibiotics targeting common community-acquired organisms.

Disclosure

 This work was supported by grant # R01HS018723 from the Agency for Healthcare Research and Quality. Dr. Lagu is also supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. Dr. Lindenauer is supported by grant K24HL132008 from the National Heart, Lung, and Blood Institute. The funding agency had no role in the data acquisition, analysis, or manuscript preparation for this study. Drs. Haessler and Rothberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Haessler, Lagu, Lindenauer, Skiest, Zilberberg, Higgins, and Rothberg conceived of the study and analyzed and interpreted the data. Dr. Lindenauer acquired the data. Dr. Pekow and Ms. Priya carried out the statistical analyses. Dr. Haessler drafted the manuscript. All authors critically reviewed the manuscript for accuracy and integrity. All authors certify no potential conflicts of interest. Preliminary results from this study were presented in oral and poster format at IDWeek in 2012 and 2013.

Peripheral intravenous catheter (PIV) insertion is the fastest, simplest, and most cost-effective method to gain vascular access, and it is used for short-term intravenous (IV) fluids, medications, blood products, and contrast media.¹ It is the most common invasive device in hospitalized patients,² with up to 70% of hospital patients receiving a PIV.³ Unacceptable PIV failure rates have been reported as high as 69%.^4-7 Failure is most frequently due to phlebitis (vein wall irritation/inflammation), occlusion (blockage), infiltration or extravasation (IV fluids/vesicant therapy entering surrounding tissue), partial dislodgement or accidental removal, leakage, and infection.^4,6,8 These failures have important implications for patients, who endure the discomfort of PIV complications and catheter replacements, and healthcare staff and budgets.

To reduce the incidence of catheter failure and avoid preventable PIV replacements, a clear understanding of why catheters fail is required. Previous research has identified that catheter gauge,^9-11 insertion site,^12-14 and inserter skill^10,15 have an impact on PIV failure. Limitations of existing research are small study sizes,^16-18 retrospective design,¹⁹ or secondary analysis of an existing data set; all potentially introduce sampling bias.^10,20

To overcome these potential biases, we developed a data collection instrument based on the catheter-associated risk factors described in the literature,^9-11,13 and other potential insertion and maintenance risks for PIV failure (eg, multiple insertion attempts, medications administered), with data collected prospectively. The study aim was to improve patient outcomes by identifying PIV insertion and maintenance risk factors amenable to modification through education or alternative clinical interventions, such as catheter gauge selection or insertion site.

Study Design and Participants

We conducted this prospective cohort study in a large tertiary hospital in Queensland, Australia. Ethics committee approval was obtained from the hospital (HREC/14/QRBW/76) and Griffith University (NRS/26/14/HREC). The study was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12615000738527). Patients in medical and surgical wards were screened Monday, Wednesday, and Friday between October 2014 and December 2015. Patients over 18 years with a PIV (BD Insyte^TM Autoguard^TM BC; Becton Dickinson, Franklin Lakes, NJ) inserted within 24 hours, and who were able to provide written informed consent, were eligible and recruited sequentially. Patients classified as palliative by the treating clinical team were excluded.

Sample Size Calculation

The “10 events per variable” rule was used to determine the sample size required to study 50 potential risk factors.^21,22 This determined that 1000 patients, with an average of 1.5 PIVs each and an expected PIV failure of 30% (500 events), were required.

Data Collection

At recruitment, baseline patient information was collected by a research nurse (ReNs) (demographics, admitting diagnosis, comorbidities, skin type,²³ and vein condition) and entered into an electronic data platform supported by Research Electronic Data Capture (REDCap).²⁴ Baseline data also included catheter variables (eg, gauge, insertion site, catheterized vein) and insertion details (eg, department of insertion, inserting clinician, number of insertion attempts). We included every PIV the participant had during their admission until hospital discharge or insertion of a central venous access device. PIV sites were reviewed Monday, Wednesday, and Friday by ReNs for site complications (eg, redness, pain, swelling, palpable cord). Potential risk factors for failure were also recorded (eg, infusates and additives, antibiotic type and dosage, flushing regimen, number of times the PIV was accessed each day for administration of IV medications or fluids, dressing type and condition, securement method for the catheter and tubing, presence of extension tubing or 3-way taps, patient mobility status, and delirium). A project manager trained and supervised ReNs for protocol compliance and audited study data quality. We considered PIV failure to have occurred if the catheter had complications at removal identified by the ReNs assessment, from medical charts, or by speaking to the patient and beside nurse. We grouped the failures in 1 of 3 types: (1) occlusion or infiltration, defined as blockage, IV fluids moving into surrounding tissue, induration, or swelling greater than 1 cm from the insertion site at or within 24 hours of removal; (2) phlebitis, defined as per clinicians’ definitions or one or more of the following signs and symptoms: pain or tenderness scored at 2 or more on a 1 to 10 increasing severity pain scale, or redness or a palpable cord (either extending greater than 1 cm from the insertion site) at or within 24 hours of PIV removal; and (3) dislodgement (partial or complete). If multiple complications were present, all were recorded.

Statistical Analysis

Data were downloaded from REDcap to Stata 14.2 (StataCorp., College Station, TX) for data management and analysis. Missing data were not imputed. Nominal data observations were collapsed into a single observation per device. Patient and device variables were described as frequencies and proportions, means and standard deviations, or medians and interquartile ranges. Failure incidence rates were calculated, and a Kaplan-Meier survival curve was plotted. In general, Cox proportional hazards models were fitted (Efron method) to handle tied failures (clustering by patient). Variables significant at P < 0.20 on univariable analyses were subjected to multivariable regression. Generally, the largest category was set as referent. Correlations between variables were checked (Spearman’s rank for binary variables, R-squared value of linear regressions for continuous/categorical or continuous/continuous variables). Correlations were considered significant if r > 0.5 and the lower bound of the 95% confidence interval (CI) was >0.5 (where calculated). Covariate interactions were explored, and effects at P < 0.05 noted. The 4 steps of multivariable model building were (1) baseline covariates only with manual stepwise removal of covariates at P ≥ 0.05, (2) treatment covariates only with manual stepwise removal of covariates at P ≥ 0.05, (3) a combination of the derived models from (1) and (2) and manual stepwise removal of covariates at P ≥ 0.05, and (4) manual stepwise addition and removal (at P ≥ 0.05) of variables dropped during the previous steps and interaction testing. Final models were checked as follows: global proportional-hazards assumption test, concordance probability (that predictions and outcomes were in agreement), and Nelson-Aalen cumulative hazard function plotted against the Cox-Snell residuals.

Patient Characteristics

In total, 1000 patients with 1578 PIVs were recruited. The average age was 54 years and the majority were surgical patients (673; 67%). Almost half of patients (455; 46%) had 2 or more comorbidities, and 334 (33%) were obese (body mass index greater than 30). Sample characteristics are shown by the type of catheter failure in Table 1.

PIV Characteristics

All 1578 PIVs were followed until removal, with only 7 PIVs (0.44%) having missing data for the 3 outcomes of interest (these were coded as nonfailures for analysis). Sixty percent of participants had more than 1 PIV followed in the study. Doctors and physicians inserted 1278 (83%) catheters. A total of 550 (35%) were placed in the ward, with 428 (28%) inserted in the emergency department or ambulance. A third of the catheters (540; 34%) were 18-gauge or larger in diameter, and 1000 (64%) were located in the cubital fossa or hand. Multiple insertion attempts were required to place 315 (23%) PIVs. No PIVs were inserted with ultrasound, as this is rarely used in this hospital. The flushing policy was for the administration of 9% sodium chloride every 8 hours if no IV medications or fluids were ordered. Table 2 contains further details of device-related characteristics. Although the hospital policy was for catheter removal by 72 hours, dwell time ranged from <1 to 14 days, with an average of 2.4 days.

PIV Complications

Catheter failure (any cause) occurred in 512 (32%) catheters, which is a failure rate of 136 per 1000 catheter days (95% CI, 125-148). A total of 346 patients out of 1000 (35%) had at least 1 failed PIV during the study. Failures were 267 phlebitis (17%), 228 occlusion/infiltration (14%), and/or 154 dislodgement (10%; Figure), with some PIVs exhibiting multiple concurrent complications (Table 2).

Multivariable AnalysisOcclusion/Infiltration

The multivariable analysis (Table 3) showed occlusion or infiltration was statistically significantly associated with female patients (hazard ratio [HR], 1.48; 95% CI, 1.10-2.00), with a 22-gauge catheter (HR, 1.43; 95% CI, 1.02-2.00), IV flucloxacillin (HR, 1.98; 95% CI, 1.19-3.31), and with frequent PIV access (HR, 1.12; 95% CI, 1.04-1.21; ie, with each increase of 1 in the mean medications/fluids administrations per day, relative PIV failure increased 112%). Less occlusion and infiltration were statistically significantly associated with securement by using additional nonsterile tape (HR, 0.46; 95% CI, 0.33-0.63), elasticized tubular bandages (HR, 0.49; 95% CI, 0.35-0.70 ), or other types of additional securement for the PIV (HR, 0.35; 95% CI, 0.26-0.47).

Phlebitis

Phlebitis was statistically significantly associated with female patients (HR, 1.81; 95% CI, 1.40-2.35), bruising at the insertion site (HR, 2.16; 95% CI, 1.26-3.71), insertion in patients’ dominant side (HR, 1.39; 95% CI, 1.09-1.77), IV flucloxicillin (HR, 2.01; 95% CI, 1.26-3.21), or with frequent PIV access (HR, 1.14; 95% CI, 1.08-1.21). Older age, (HR, 0.99; 95% CI, 0.98-0.99; ie, each year older was associated with 1% less phlebitis), securement with additional nonsterile tape (HR, 0.63; 95% CI, 0.48-0.82) or with any other additional securement (HR, 0.53; 95% CI, 0.39-0.70), or the administration of IV cephazolin (HR, 0.63; 95% CI, 0.44-0.89) were associated with lower phlebitis risk.

Dislodgement

Statistically significant predictors associated with an increased risk of PIV dislodgement included paramedic insertion (HR, 1.78; 95% CI, 1.03-3.06) and frequent PIV access (HR, 1.11; 95% CI, 1.03-1.20). A decreased risk was associated with the additional securement of the PIV, including nonsterile tape (HR, 0.44; 95% CI, 0.31-0.63) or other forms of additional securement (HR, 0.32; 95% CI, 0.22-0.46).

One in 3 PIVs failed in this study, with phlebitis as the most common cause of PIV failure. The 17% phlebitis rate reflected clinician-reported phlebitis or phlebitis observed by research staff using a 1-criteria definition because any sign or symptom can trigger PIV removal (eg, pain), even if other signs or symptoms are not present. Reported phlebitis rates are lower if definitions require 2 signs or symptoms.^4,6 With over 71 different phlebitis assessment scales in use, and none well validated, the best method for diagnosing phlebitis remains unclear and explains the variation in reported rates.²⁵ Occlusion/infiltration and dislodgement were also highly prevalent forms of PIV failure at 14% and 10%, respectively. Occlusion and infiltration were combined because clinical staff use these terms interchangeably, and differential diagnostic tools are not used in practice. Both result in the same outcome (therapy interruption and PIV removal), and this combination of outcomes has been used previously.²³ No PIV-associated bloodstream infections occurred, despite the heightened awareness of these infections in the literature.³

Females had significantly more occlusion/infiltration and phlebitis than males, in keeping with previous studies.^7,9,10 This could be because of females’ smaller vein caliber, although the effect remained after adjustment for PIV gauge.^7,26 The effect of aging on vascular endothelium and structural integrity may explain the observed decrease in phlebitis of 1% with each older year of age.²⁷ However, gender and age effects could be explained by psychosocial factors (eg, older people may be less likely to admit pain, or we may question them less sympathetically), but, regardless, women and younger patients should be monitored more closely.

We found 22-gauge catheters were more likely to fail from occlusion/infiltration than other sizes. This confirms similar findings from Abolfotouh et al.⁹ PIV gauge selection for this study was made at the inserter’s discretion and may be confounded by smaller vein size, which was not measured. In addition, risk may be because of smaller gauge alone or also more influenced by the shorter length of the studied 22-gauge (25 mm) than the <20-gauge catheters (30 mm). These results question international guidelines, which currently recommend the smallest gauge peripheral catheter possible,^28,29 and randomized trials are needed. Although practice varies between inserters, some preferentially cannulate the nondominant limb. We are not aware of previous studies on this practice; however, our results support this approach.

Flucloxacillin was associated with a 2-fold increase in occlusion/infiltration and phlebitis. Although multiple studies have reported IV medications^9,11 and IV antibiotics^10,30,31 as risk factors for PIV failure, none have identified flucloxacillin as an independent risk factor. IV flucloxacillin is recommended for reconstitution as 1 g in 15 mL to 20 mL of sterile water, and injection over 3 to 4 minutes, although this may not be adhered to in practice. Alternative administration regimes or improved adherence to current policy may be needed. An exception to the relationship between IV antibiotics and catheter failure was IV cephazolin, associated with 40% relatively less phlebitis. This may be a spurious finding because the administration, pH, and osmolality of cephazolin are similar to other IV antibiotics.

The more PIVs that were accessed per day, whether for infusions or medications, the more failure occurred from occlusion/infiltration, phlebitis, and dislodgement. This suggests that peripheral veins are easily damaged and/or inflamed by the influx of fluids or medications. Lower injection pressures or the timely transfer to oral medications may limit this problem. Flushing regimens may also assist because practice varies greatly, and questions on whether slow continuous flush infusion or intermittent manual flushing are more vein-protective, and the optimal flush volume, frequency, and technique (eg, pulsatile) remain.^32,33 Manual handling for frequent access may loosen dressings and securement, thus explaining the observed association between frequent access and catheter dislodgement. Finally, the association between use and failure may indicate that many of these patients were not suitable for a PIV, and different approaches (eg, ultrasound-guided insertion) or a midline may have been a superior option. There is growing emphasis on the need for better preinsertion assessment and selection of the most appropriate device for the patient and the IV treatment required.³⁴

Suboptimal dressings or securements are not unusual in hospitals.³⁵ Despite our policy of PIV securement with bordered transparent dressings, we found 4 dressing types in use. In addition, we found almost 50% of PIVs had an additional (secondary) securement, and this was associated with significantly less PIV failure of all 3 types. This suggests that 1 or more of nonsterile tape, elasticized tubular bandages, or other securement (eg, bandage or second transparent dressing) can reduce PIV failure, although a randomized trial is lacking.³⁶ Whether the dressing was failing and required reinforcement or hospital staff lacked confidence in the dressing and placed additional securement preventatively is unclear. Both PIV failure and PIV dressing failure are common, and further research into superior PIV products and practices is urgently needed. Paramedic insertions had a higher risk of dislodgement, suggesting that the increased emphasis on securement should start in the prehospital setting.

While multiple or difficult insertion attempts were not associated with PIV failure, insertions were not directly observed, and clinicians may have underreported attempts. In contrast, insertion-related bruising (a surrogate for difficult insertion) was associated with more than double the incidence of phlebitis. The long-term implications of multiple insertion attempts on patient’s vasculature are unclear, but we believe first time PIV insertion is important to patients and of interest to clinicians. A recent systematic review of strategies associated with first attempt PIV insertion success in an emergency department found little evidence for effective strategies and recommended further research.³⁷

The overall PIV failure rate in our study was 32%, lower than the 35% to 40% failure observed in our previous randomized controlled trials, which had more stringent inclusion and exclusion criteria (eg, longer predicted duration of therapy).^6,38 The implications for patients and costs to the organization of frequent catheter replacement demonstrate urgent need for further research in this area of practice.³⁹ A strength of this study is that all PIVs, regardless of the expected length of dwell time or reason for insertion, were eligible for inclusion, providing more generalizable results. The PIV failure rate of 32% is concerning because these failures trigger treatment delays and replacement insertions, with significant increased labor and equipment costs. The mean cost of PIV replacement has been costed at AUD $69.30 or US $51.92 (as per 2010 $ value) per episode of IV treatment.⁴⁰ For our hospital, which uses 200,000 PIVs per year, the current level of PIV failure suggests almost AU $5.5 (US $4.1) million in waste annually at this site alone.

The additional strengths of this study include the extensive information collected prospectively about PIV insertion and maintenance, including information on who inserted the PIV, IV medications administered, and PIV dressings used. Limitations were the population of surgical and medical patients in 1 tertiary hospital, which may not be generalizable to other settings.

Our study confirms the high rate of catheter failure in acute care hospitals, validates existing evidence related to PIV failure, and identifies new, potentially modifiable risk factors to improve PIV insertion and management. Implications for future research were also identified.

Acknowledgments

The researchers acknowledge and thank the nurses and patients involved in this study. The authors would also like to acknowledge Becton Dickinson for partly funding this study in the form of an unrestricted grant-in-aid paid to Griffith University. Becton Dickinson did not design the study protocol, collect or analyze data, and did not prepare or review the manuscript.

Disclosure

 On behalf of NM and CMR, Griffith University has received unrestricted educational and research grants and consultancy payment for lectures from 3M and Becton Dickinson. On behalf of NM, MC, and CMR, Griffith University has received unrestricted investigator-initiated research grants from Centurion Medical Products and Entrotech Lifesciences (manufacturers of PIV dressings) and Becton Dickinson (manufacturer of PIVs). On behalf of MC, Griffith University has received a consultancy payment to develop education material from Baxter. On behalf of CMR, Griffith University has received unrestricted donations or investigator initiated research grants unrelated to this research from Adhezion, Angiodynamics, Baxter, Carefusion, Cook Medical, Hospira, Mayo, Smiths Medical, and Vygon. On behalf of CMR, Griffith University has received consultancy payments for educational lectures or professional opinion from B. Braun, Bard, Carefusion, Mayo, ResQDevices, and Smiths Medical. On behalf of EL, Griffith University has received consultancy payments for educational lecture from 3M. On behalf of MC, Griffith University has received a consultancy payment to develop education material from Baxter. As this was an observational study, no products were trialed in this study. JW and GM have no conflicts of interest.

Peripheral intravenous catheter (PIV) insertion is the fastest, simplest, and most cost-effective method to gain vascular access, and it is used for short-term intravenous (IV) fluids, medications, blood products, and contrast media.¹ It is the most common invasive device in hospitalized patients,² with up to 70% of hospital patients receiving a PIV.³ Unacceptable PIV failure rates have been reported as high as 69%.^4-7 Failure is most frequently due to phlebitis (vein wall irritation/inflammation), occlusion (blockage), infiltration or extravasation (IV fluids/vesicant therapy entering surrounding tissue), partial dislodgement or accidental removal, leakage, and infection.^4,6,8 These failures have important implications for patients, who endure the discomfort of PIV complications and catheter replacements, and healthcare staff and budgets.

To reduce the incidence of catheter failure and avoid preventable PIV replacements, a clear understanding of why catheters fail is required. Previous research has identified that catheter gauge,^9-11 insertion site,^12-14 and inserter skill^10,15 have an impact on PIV failure. Limitations of existing research are small study sizes,^16-18 retrospective design,¹⁹ or secondary analysis of an existing data set; all potentially introduce sampling bias.^10,20

To overcome these potential biases, we developed a data collection instrument based on the catheter-associated risk factors described in the literature,^9-11,13 and other potential insertion and maintenance risks for PIV failure (eg, multiple insertion attempts, medications administered), with data collected prospectively. The study aim was to improve patient outcomes by identifying PIV insertion and maintenance risk factors amenable to modification through education or alternative clinical interventions, such as catheter gauge selection or insertion site.

Study Design and Participants

We conducted this prospective cohort study in a large tertiary hospital in Queensland, Australia. Ethics committee approval was obtained from the hospital (HREC/14/QRBW/76) and Griffith University (NRS/26/14/HREC). The study was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12615000738527). Patients in medical and surgical wards were screened Monday, Wednesday, and Friday between October 2014 and December 2015. Patients over 18 years with a PIV (BD Insyte^TM Autoguard^TM BC; Becton Dickinson, Franklin Lakes, NJ) inserted within 24 hours, and who were able to provide written informed consent, were eligible and recruited sequentially. Patients classified as palliative by the treating clinical team were excluded.

Sample Size Calculation

The “10 events per variable” rule was used to determine the sample size required to study 50 potential risk factors.^21,22 This determined that 1000 patients, with an average of 1.5 PIVs each and an expected PIV failure of 30% (500 events), were required.

Data Collection

At recruitment, baseline patient information was collected by a research nurse (ReNs) (demographics, admitting diagnosis, comorbidities, skin type,²³ and vein condition) and entered into an electronic data platform supported by Research Electronic Data Capture (REDCap).²⁴ Baseline data also included catheter variables (eg, gauge, insertion site, catheterized vein) and insertion details (eg, department of insertion, inserting clinician, number of insertion attempts). We included every PIV the participant had during their admission until hospital discharge or insertion of a central venous access device. PIV sites were reviewed Monday, Wednesday, and Friday by ReNs for site complications (eg, redness, pain, swelling, palpable cord). Potential risk factors for failure were also recorded (eg, infusates and additives, antibiotic type and dosage, flushing regimen, number of times the PIV was accessed each day for administration of IV medications or fluids, dressing type and condition, securement method for the catheter and tubing, presence of extension tubing or 3-way taps, patient mobility status, and delirium). A project manager trained and supervised ReNs for protocol compliance and audited study data quality. We considered PIV failure to have occurred if the catheter had complications at removal identified by the ReNs assessment, from medical charts, or by speaking to the patient and beside nurse. We grouped the failures in 1 of 3 types: (1) occlusion or infiltration, defined as blockage, IV fluids moving into surrounding tissue, induration, or swelling greater than 1 cm from the insertion site at or within 24 hours of removal; (2) phlebitis, defined as per clinicians’ definitions or one or more of the following signs and symptoms: pain or tenderness scored at 2 or more on a 1 to 10 increasing severity pain scale, or redness or a palpable cord (either extending greater than 1 cm from the insertion site) at or within 24 hours of PIV removal; and (3) dislodgement (partial or complete). If multiple complications were present, all were recorded.

Statistical Analysis

Data were downloaded from REDcap to Stata 14.2 (StataCorp., College Station, TX) for data management and analysis. Missing data were not imputed. Nominal data observations were collapsed into a single observation per device. Patient and device variables were described as frequencies and proportions, means and standard deviations, or medians and interquartile ranges. Failure incidence rates were calculated, and a Kaplan-Meier survival curve was plotted. In general, Cox proportional hazards models were fitted (Efron method) to handle tied failures (clustering by patient). Variables significant at P < 0.20 on univariable analyses were subjected to multivariable regression. Generally, the largest category was set as referent. Correlations between variables were checked (Spearman’s rank for binary variables, R-squared value of linear regressions for continuous/categorical or continuous/continuous variables). Correlations were considered significant if r > 0.5 and the lower bound of the 95% confidence interval (CI) was >0.5 (where calculated). Covariate interactions were explored, and effects at P < 0.05 noted. The 4 steps of multivariable model building were (1) baseline covariates only with manual stepwise removal of covariates at P ≥ 0.05, (2) treatment covariates only with manual stepwise removal of covariates at P ≥ 0.05, (3) a combination of the derived models from (1) and (2) and manual stepwise removal of covariates at P ≥ 0.05, and (4) manual stepwise addition and removal (at P ≥ 0.05) of variables dropped during the previous steps and interaction testing. Final models were checked as follows: global proportional-hazards assumption test, concordance probability (that predictions and outcomes were in agreement), and Nelson-Aalen cumulative hazard function plotted against the Cox-Snell residuals.

Patient Characteristics

In total, 1000 patients with 1578 PIVs were recruited. The average age was 54 years and the majority were surgical patients (673; 67%). Almost half of patients (455; 46%) had 2 or more comorbidities, and 334 (33%) were obese (body mass index greater than 30). Sample characteristics are shown by the type of catheter failure in Table 1.

PIV Characteristics

All 1578 PIVs were followed until removal, with only 7 PIVs (0.44%) having missing data for the 3 outcomes of interest (these were coded as nonfailures for analysis). Sixty percent of participants had more than 1 PIV followed in the study. Doctors and physicians inserted 1278 (83%) catheters. A total of 550 (35%) were placed in the ward, with 428 (28%) inserted in the emergency department or ambulance. A third of the catheters (540; 34%) were 18-gauge or larger in diameter, and 1000 (64%) were located in the cubital fossa or hand. Multiple insertion attempts were required to place 315 (23%) PIVs. No PIVs were inserted with ultrasound, as this is rarely used in this hospital. The flushing policy was for the administration of 9% sodium chloride every 8 hours if no IV medications or fluids were ordered. Table 2 contains further details of device-related characteristics. Although the hospital policy was for catheter removal by 72 hours, dwell time ranged from <1 to 14 days, with an average of 2.4 days.

PIV Complications

Catheter failure (any cause) occurred in 512 (32%) catheters, which is a failure rate of 136 per 1000 catheter days (95% CI, 125-148). A total of 346 patients out of 1000 (35%) had at least 1 failed PIV during the study. Failures were 267 phlebitis (17%), 228 occlusion/infiltration (14%), and/or 154 dislodgement (10%; Figure), with some PIVs exhibiting multiple concurrent complications (Table 2).

Multivariable AnalysisOcclusion/Infiltration

The multivariable analysis (Table 3) showed occlusion or infiltration was statistically significantly associated with female patients (hazard ratio [HR], 1.48; 95% CI, 1.10-2.00), with a 22-gauge catheter (HR, 1.43; 95% CI, 1.02-2.00), IV flucloxacillin (HR, 1.98; 95% CI, 1.19-3.31), and with frequent PIV access (HR, 1.12; 95% CI, 1.04-1.21; ie, with each increase of 1 in the mean medications/fluids administrations per day, relative PIV failure increased 112%). Less occlusion and infiltration were statistically significantly associated with securement by using additional nonsterile tape (HR, 0.46; 95% CI, 0.33-0.63), elasticized tubular bandages (HR, 0.49; 95% CI, 0.35-0.70 ), or other types of additional securement for the PIV (HR, 0.35; 95% CI, 0.26-0.47).

Phlebitis

Phlebitis was statistically significantly associated with female patients (HR, 1.81; 95% CI, 1.40-2.35), bruising at the insertion site (HR, 2.16; 95% CI, 1.26-3.71), insertion in patients’ dominant side (HR, 1.39; 95% CI, 1.09-1.77), IV flucloxicillin (HR, 2.01; 95% CI, 1.26-3.21), or with frequent PIV access (HR, 1.14; 95% CI, 1.08-1.21). Older age, (HR, 0.99; 95% CI, 0.98-0.99; ie, each year older was associated with 1% less phlebitis), securement with additional nonsterile tape (HR, 0.63; 95% CI, 0.48-0.82) or with any other additional securement (HR, 0.53; 95% CI, 0.39-0.70), or the administration of IV cephazolin (HR, 0.63; 95% CI, 0.44-0.89) were associated with lower phlebitis risk.

Dislodgement

Statistically significant predictors associated with an increased risk of PIV dislodgement included paramedic insertion (HR, 1.78; 95% CI, 1.03-3.06) and frequent PIV access (HR, 1.11; 95% CI, 1.03-1.20). A decreased risk was associated with the additional securement of the PIV, including nonsterile tape (HR, 0.44; 95% CI, 0.31-0.63) or other forms of additional securement (HR, 0.32; 95% CI, 0.22-0.46).

One in 3 PIVs failed in this study, with phlebitis as the most common cause of PIV failure. The 17% phlebitis rate reflected clinician-reported phlebitis or phlebitis observed by research staff using a 1-criteria definition because any sign or symptom can trigger PIV removal (eg, pain), even if other signs or symptoms are not present. Reported phlebitis rates are lower if definitions require 2 signs or symptoms.^4,6 With over 71 different phlebitis assessment scales in use, and none well validated, the best method for diagnosing phlebitis remains unclear and explains the variation in reported rates.²⁵ Occlusion/infiltration and dislodgement were also highly prevalent forms of PIV failure at 14% and 10%, respectively. Occlusion and infiltration were combined because clinical staff use these terms interchangeably, and differential diagnostic tools are not used in practice. Both result in the same outcome (therapy interruption and PIV removal), and this combination of outcomes has been used previously.²³ No PIV-associated bloodstream infections occurred, despite the heightened awareness of these infections in the literature.³

Females had significantly more occlusion/infiltration and phlebitis than males, in keeping with previous studies.^7,9,10 This could be because of females’ smaller vein caliber, although the effect remained after adjustment for PIV gauge.^7,26 The effect of aging on vascular endothelium and structural integrity may explain the observed decrease in phlebitis of 1% with each older year of age.²⁷ However, gender and age effects could be explained by psychosocial factors (eg, older people may be less likely to admit pain, or we may question them less sympathetically), but, regardless, women and younger patients should be monitored more closely.

We found 22-gauge catheters were more likely to fail from occlusion/infiltration than other sizes. This confirms similar findings from Abolfotouh et al.⁹ PIV gauge selection for this study was made at the inserter’s discretion and may be confounded by smaller vein size, which was not measured. In addition, risk may be because of smaller gauge alone or also more influenced by the shorter length of the studied 22-gauge (25 mm) than the <20-gauge catheters (30 mm). These results question international guidelines, which currently recommend the smallest gauge peripheral catheter possible,^28,29 and randomized trials are needed. Although practice varies between inserters, some preferentially cannulate the nondominant limb. We are not aware of previous studies on this practice; however, our results support this approach.

Flucloxacillin was associated with a 2-fold increase in occlusion/infiltration and phlebitis. Although multiple studies have reported IV medications^9,11 and IV antibiotics^10,30,31 as risk factors for PIV failure, none have identified flucloxacillin as an independent risk factor. IV flucloxacillin is recommended for reconstitution as 1 g in 15 mL to 20 mL of sterile water, and injection over 3 to 4 minutes, although this may not be adhered to in practice. Alternative administration regimes or improved adherence to current policy may be needed. An exception to the relationship between IV antibiotics and catheter failure was IV cephazolin, associated with 40% relatively less phlebitis. This may be a spurious finding because the administration, pH, and osmolality of cephazolin are similar to other IV antibiotics.

The more PIVs that were accessed per day, whether for infusions or medications, the more failure occurred from occlusion/infiltration, phlebitis, and dislodgement. This suggests that peripheral veins are easily damaged and/or inflamed by the influx of fluids or medications. Lower injection pressures or the timely transfer to oral medications may limit this problem. Flushing regimens may also assist because practice varies greatly, and questions on whether slow continuous flush infusion or intermittent manual flushing are more vein-protective, and the optimal flush volume, frequency, and technique (eg, pulsatile) remain.^32,33 Manual handling for frequent access may loosen dressings and securement, thus explaining the observed association between frequent access and catheter dislodgement. Finally, the association between use and failure may indicate that many of these patients were not suitable for a PIV, and different approaches (eg, ultrasound-guided insertion) or a midline may have been a superior option. There is growing emphasis on the need for better preinsertion assessment and selection of the most appropriate device for the patient and the IV treatment required.³⁴

Suboptimal dressings or securements are not unusual in hospitals.³⁵ Despite our policy of PIV securement with bordered transparent dressings, we found 4 dressing types in use. In addition, we found almost 50% of PIVs had an additional (secondary) securement, and this was associated with significantly less PIV failure of all 3 types. This suggests that 1 or more of nonsterile tape, elasticized tubular bandages, or other securement (eg, bandage or second transparent dressing) can reduce PIV failure, although a randomized trial is lacking.³⁶ Whether the dressing was failing and required reinforcement or hospital staff lacked confidence in the dressing and placed additional securement preventatively is unclear. Both PIV failure and PIV dressing failure are common, and further research into superior PIV products and practices is urgently needed. Paramedic insertions had a higher risk of dislodgement, suggesting that the increased emphasis on securement should start in the prehospital setting.

While multiple or difficult insertion attempts were not associated with PIV failure, insertions were not directly observed, and clinicians may have underreported attempts. In contrast, insertion-related bruising (a surrogate for difficult insertion) was associated with more than double the incidence of phlebitis. The long-term implications of multiple insertion attempts on patient’s vasculature are unclear, but we believe first time PIV insertion is important to patients and of interest to clinicians. A recent systematic review of strategies associated with first attempt PIV insertion success in an emergency department found little evidence for effective strategies and recommended further research.³⁷

The overall PIV failure rate in our study was 32%, lower than the 35% to 40% failure observed in our previous randomized controlled trials, which had more stringent inclusion and exclusion criteria (eg, longer predicted duration of therapy).^6,38 The implications for patients and costs to the organization of frequent catheter replacement demonstrate urgent need for further research in this area of practice.³⁹ A strength of this study is that all PIVs, regardless of the expected length of dwell time or reason for insertion, were eligible for inclusion, providing more generalizable results. The PIV failure rate of 32% is concerning because these failures trigger treatment delays and replacement insertions, with significant increased labor and equipment costs. The mean cost of PIV replacement has been costed at AUD $69.30 or US $51.92 (as per 2010 $ value) per episode of IV treatment.⁴⁰ For our hospital, which uses 200,000 PIVs per year, the current level of PIV failure suggests almost AU $5.5 (US $4.1) million in waste annually at this site alone.

The additional strengths of this study include the extensive information collected prospectively about PIV insertion and maintenance, including information on who inserted the PIV, IV medications administered, and PIV dressings used. Limitations were the population of surgical and medical patients in 1 tertiary hospital, which may not be generalizable to other settings.

Our study confirms the high rate of catheter failure in acute care hospitals, validates existing evidence related to PIV failure, and identifies new, potentially modifiable risk factors to improve PIV insertion and management. Implications for future research were also identified.

Acknowledgments

The researchers acknowledge and thank the nurses and patients involved in this study. The authors would also like to acknowledge Becton Dickinson for partly funding this study in the form of an unrestricted grant-in-aid paid to Griffith University. Becton Dickinson did not design the study protocol, collect or analyze data, and did not prepare or review the manuscript.

Disclosure

 On behalf of NM and CMR, Griffith University has received unrestricted educational and research grants and consultancy payment for lectures from 3M and Becton Dickinson. On behalf of NM, MC, and CMR, Griffith University has received unrestricted investigator-initiated research grants from Centurion Medical Products and Entrotech Lifesciences (manufacturers of PIV dressings) and Becton Dickinson (manufacturer of PIVs). On behalf of MC, Griffith University has received a consultancy payment to develop education material from Baxter. On behalf of CMR, Griffith University has received unrestricted donations or investigator initiated research grants unrelated to this research from Adhezion, Angiodynamics, Baxter, Carefusion, Cook Medical, Hospira, Mayo, Smiths Medical, and Vygon. On behalf of CMR, Griffith University has received consultancy payments for educational lectures or professional opinion from B. Braun, Bard, Carefusion, Mayo, ResQDevices, and Smiths Medical. On behalf of EL, Griffith University has received consultancy payments for educational lecture from 3M. On behalf of MC, Griffith University has received a consultancy payment to develop education material from Baxter. As this was an observational study, no products were trialed in this study. JW and GM have no conflicts of interest.

Peripheral intravenous catheter (PIV) insertion is the fastest, simplest, and most cost-effective method to gain vascular access, and it is used for short-term intravenous (IV) fluids, medications, blood products, and contrast media.¹ It is the most common invasive device in hospitalized patients,² with up to 70% of hospital patients receiving a PIV.³ Unacceptable PIV failure rates have been reported as high as 69%.^4-7 Failure is most frequently due to phlebitis (vein wall irritation/inflammation), occlusion (blockage), infiltration or extravasation (IV fluids/vesicant therapy entering surrounding tissue), partial dislodgement or accidental removal, leakage, and infection.^4,6,8 These failures have important implications for patients, who endure the discomfort of PIV complications and catheter replacements, and healthcare staff and budgets.

To reduce the incidence of catheter failure and avoid preventable PIV replacements, a clear understanding of why catheters fail is required. Previous research has identified that catheter gauge,^9-11 insertion site,^12-14 and inserter skill^10,15 have an impact on PIV failure. Limitations of existing research are small study sizes,^16-18 retrospective design,¹⁹ or secondary analysis of an existing data set; all potentially introduce sampling bias.^10,20

To overcome these potential biases, we developed a data collection instrument based on the catheter-associated risk factors described in the literature,^9-11,13 and other potential insertion and maintenance risks for PIV failure (eg, multiple insertion attempts, medications administered), with data collected prospectively. The study aim was to improve patient outcomes by identifying PIV insertion and maintenance risk factors amenable to modification through education or alternative clinical interventions, such as catheter gauge selection or insertion site.

Study Design and Participants

We conducted this prospective cohort study in a large tertiary hospital in Queensland, Australia. Ethics committee approval was obtained from the hospital (HREC/14/QRBW/76) and Griffith University (NRS/26/14/HREC). The study was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12615000738527). Patients in medical and surgical wards were screened Monday, Wednesday, and Friday between October 2014 and December 2015. Patients over 18 years with a PIV (BD Insyte^TM Autoguard^TM BC; Becton Dickinson, Franklin Lakes, NJ) inserted within 24 hours, and who were able to provide written informed consent, were eligible and recruited sequentially. Patients classified as palliative by the treating clinical team were excluded.

Sample Size Calculation

The “10 events per variable” rule was used to determine the sample size required to study 50 potential risk factors.^21,22 This determined that 1000 patients, with an average of 1.5 PIVs each and an expected PIV failure of 30% (500 events), were required.

Data Collection

At recruitment, baseline patient information was collected by a research nurse (ReNs) (demographics, admitting diagnosis, comorbidities, skin type,²³ and vein condition) and entered into an electronic data platform supported by Research Electronic Data Capture (REDCap).²⁴ Baseline data also included catheter variables (eg, gauge, insertion site, catheterized vein) and insertion details (eg, department of insertion, inserting clinician, number of insertion attempts). We included every PIV the participant had during their admission until hospital discharge or insertion of a central venous access device. PIV sites were reviewed Monday, Wednesday, and Friday by ReNs for site complications (eg, redness, pain, swelling, palpable cord). Potential risk factors for failure were also recorded (eg, infusates and additives, antibiotic type and dosage, flushing regimen, number of times the PIV was accessed each day for administration of IV medications or fluids, dressing type and condition, securement method for the catheter and tubing, presence of extension tubing or 3-way taps, patient mobility status, and delirium). A project manager trained and supervised ReNs for protocol compliance and audited study data quality. We considered PIV failure to have occurred if the catheter had complications at removal identified by the ReNs assessment, from medical charts, or by speaking to the patient and beside nurse. We grouped the failures in 1 of 3 types: (1) occlusion or infiltration, defined as blockage, IV fluids moving into surrounding tissue, induration, or swelling greater than 1 cm from the insertion site at or within 24 hours of removal; (2) phlebitis, defined as per clinicians’ definitions or one or more of the following signs and symptoms: pain or tenderness scored at 2 or more on a 1 to 10 increasing severity pain scale, or redness or a palpable cord (either extending greater than 1 cm from the insertion site) at or within 24 hours of PIV removal; and (3) dislodgement (partial or complete). If multiple complications were present, all were recorded.

Statistical Analysis

Data were downloaded from REDcap to Stata 14.2 (StataCorp., College Station, TX) for data management and analysis. Missing data were not imputed. Nominal data observations were collapsed into a single observation per device. Patient and device variables were described as frequencies and proportions, means and standard deviations, or medians and interquartile ranges. Failure incidence rates were calculated, and a Kaplan-Meier survival curve was plotted. In general, Cox proportional hazards models were fitted (Efron method) to handle tied failures (clustering by patient). Variables significant at P < 0.20 on univariable analyses were subjected to multivariable regression. Generally, the largest category was set as referent. Correlations between variables were checked (Spearman’s rank for binary variables, R-squared value of linear regressions for continuous/categorical or continuous/continuous variables). Correlations were considered significant if r > 0.5 and the lower bound of the 95% confidence interval (CI) was >0.5 (where calculated). Covariate interactions were explored, and effects at P < 0.05 noted. The 4 steps of multivariable model building were (1) baseline covariates only with manual stepwise removal of covariates at P ≥ 0.05, (2) treatment covariates only with manual stepwise removal of covariates at P ≥ 0.05, (3) a combination of the derived models from (1) and (2) and manual stepwise removal of covariates at P ≥ 0.05, and (4) manual stepwise addition and removal (at P ≥ 0.05) of variables dropped during the previous steps and interaction testing. Final models were checked as follows: global proportional-hazards assumption test, concordance probability (that predictions and outcomes were in agreement), and Nelson-Aalen cumulative hazard function plotted against the Cox-Snell residuals.

Patient Characteristics

In total, 1000 patients with 1578 PIVs were recruited. The average age was 54 years and the majority were surgical patients (673; 67%). Almost half of patients (455; 46%) had 2 or more comorbidities, and 334 (33%) were obese (body mass index greater than 30). Sample characteristics are shown by the type of catheter failure in Table 1.

PIV Characteristics

All 1578 PIVs were followed until removal, with only 7 PIVs (0.44%) having missing data for the 3 outcomes of interest (these were coded as nonfailures for analysis). Sixty percent of participants had more than 1 PIV followed in the study. Doctors and physicians inserted 1278 (83%) catheters. A total of 550 (35%) were placed in the ward, with 428 (28%) inserted in the emergency department or ambulance. A third of the catheters (540; 34%) were 18-gauge or larger in diameter, and 1000 (64%) were located in the cubital fossa or hand. Multiple insertion attempts were required to place 315 (23%) PIVs. No PIVs were inserted with ultrasound, as this is rarely used in this hospital. The flushing policy was for the administration of 9% sodium chloride every 8 hours if no IV medications or fluids were ordered. Table 2 contains further details of device-related characteristics. Although the hospital policy was for catheter removal by 72 hours, dwell time ranged from <1 to 14 days, with an average of 2.4 days.

PIV Complications

Catheter failure (any cause) occurred in 512 (32%) catheters, which is a failure rate of 136 per 1000 catheter days (95% CI, 125-148). A total of 346 patients out of 1000 (35%) had at least 1 failed PIV during the study. Failures were 267 phlebitis (17%), 228 occlusion/infiltration (14%), and/or 154 dislodgement (10%; Figure), with some PIVs exhibiting multiple concurrent complications (Table 2).

Multivariable AnalysisOcclusion/Infiltration

The multivariable analysis (Table 3) showed occlusion or infiltration was statistically significantly associated with female patients (hazard ratio [HR], 1.48; 95% CI, 1.10-2.00), with a 22-gauge catheter (HR, 1.43; 95% CI, 1.02-2.00), IV flucloxacillin (HR, 1.98; 95% CI, 1.19-3.31), and with frequent PIV access (HR, 1.12; 95% CI, 1.04-1.21; ie, with each increase of 1 in the mean medications/fluids administrations per day, relative PIV failure increased 112%). Less occlusion and infiltration were statistically significantly associated with securement by using additional nonsterile tape (HR, 0.46; 95% CI, 0.33-0.63), elasticized tubular bandages (HR, 0.49; 95% CI, 0.35-0.70 ), or other types of additional securement for the PIV (HR, 0.35; 95% CI, 0.26-0.47).

Phlebitis

Phlebitis was statistically significantly associated with female patients (HR, 1.81; 95% CI, 1.40-2.35), bruising at the insertion site (HR, 2.16; 95% CI, 1.26-3.71), insertion in patients’ dominant side (HR, 1.39; 95% CI, 1.09-1.77), IV flucloxicillin (HR, 2.01; 95% CI, 1.26-3.21), or with frequent PIV access (HR, 1.14; 95% CI, 1.08-1.21). Older age, (HR, 0.99; 95% CI, 0.98-0.99; ie, each year older was associated with 1% less phlebitis), securement with additional nonsterile tape (HR, 0.63; 95% CI, 0.48-0.82) or with any other additional securement (HR, 0.53; 95% CI, 0.39-0.70), or the administration of IV cephazolin (HR, 0.63; 95% CI, 0.44-0.89) were associated with lower phlebitis risk.

Dislodgement

Statistically significant predictors associated with an increased risk of PIV dislodgement included paramedic insertion (HR, 1.78; 95% CI, 1.03-3.06) and frequent PIV access (HR, 1.11; 95% CI, 1.03-1.20). A decreased risk was associated with the additional securement of the PIV, including nonsterile tape (HR, 0.44; 95% CI, 0.31-0.63) or other forms of additional securement (HR, 0.32; 95% CI, 0.22-0.46).

One in 3 PIVs failed in this study, with phlebitis as the most common cause of PIV failure. The 17% phlebitis rate reflected clinician-reported phlebitis or phlebitis observed by research staff using a 1-criteria definition because any sign or symptom can trigger PIV removal (eg, pain), even if other signs or symptoms are not present. Reported phlebitis rates are lower if definitions require 2 signs or symptoms.^4,6 With over 71 different phlebitis assessment scales in use, and none well validated, the best method for diagnosing phlebitis remains unclear and explains the variation in reported rates.²⁵ Occlusion/infiltration and dislodgement were also highly prevalent forms of PIV failure at 14% and 10%, respectively. Occlusion and infiltration were combined because clinical staff use these terms interchangeably, and differential diagnostic tools are not used in practice. Both result in the same outcome (therapy interruption and PIV removal), and this combination of outcomes has been used previously.²³ No PIV-associated bloodstream infections occurred, despite the heightened awareness of these infections in the literature.³

Females had significantly more occlusion/infiltration and phlebitis than males, in keeping with previous studies.^7,9,10 This could be because of females’ smaller vein caliber, although the effect remained after adjustment for PIV gauge.^7,26 The effect of aging on vascular endothelium and structural integrity may explain the observed decrease in phlebitis of 1% with each older year of age.²⁷ However, gender and age effects could be explained by psychosocial factors (eg, older people may be less likely to admit pain, or we may question them less sympathetically), but, regardless, women and younger patients should be monitored more closely.

We found 22-gauge catheters were more likely to fail from occlusion/infiltration than other sizes. This confirms similar findings from Abolfotouh et al.⁹ PIV gauge selection for this study was made at the inserter’s discretion and may be confounded by smaller vein size, which was not measured. In addition, risk may be because of smaller gauge alone or also more influenced by the shorter length of the studied 22-gauge (25 mm) than the <20-gauge catheters (30 mm). These results question international guidelines, which currently recommend the smallest gauge peripheral catheter possible,^28,29 and randomized trials are needed. Although practice varies between inserters, some preferentially cannulate the nondominant limb. We are not aware of previous studies on this practice; however, our results support this approach.

Flucloxacillin was associated with a 2-fold increase in occlusion/infiltration and phlebitis. Although multiple studies have reported IV medications^9,11 and IV antibiotics^10,30,31 as risk factors for PIV failure, none have identified flucloxacillin as an independent risk factor. IV flucloxacillin is recommended for reconstitution as 1 g in 15 mL to 20 mL of sterile water, and injection over 3 to 4 minutes, although this may not be adhered to in practice. Alternative administration regimes or improved adherence to current policy may be needed. An exception to the relationship between IV antibiotics and catheter failure was IV cephazolin, associated with 40% relatively less phlebitis. This may be a spurious finding because the administration, pH, and osmolality of cephazolin are similar to other IV antibiotics.

The more PIVs that were accessed per day, whether for infusions or medications, the more failure occurred from occlusion/infiltration, phlebitis, and dislodgement. This suggests that peripheral veins are easily damaged and/or inflamed by the influx of fluids or medications. Lower injection pressures or the timely transfer to oral medications may limit this problem. Flushing regimens may also assist because practice varies greatly, and questions on whether slow continuous flush infusion or intermittent manual flushing are more vein-protective, and the optimal flush volume, frequency, and technique (eg, pulsatile) remain.^32,33 Manual handling for frequent access may loosen dressings and securement, thus explaining the observed association between frequent access and catheter dislodgement. Finally, the association between use and failure may indicate that many of these patients were not suitable for a PIV, and different approaches (eg, ultrasound-guided insertion) or a midline may have been a superior option. There is growing emphasis on the need for better preinsertion assessment and selection of the most appropriate device for the patient and the IV treatment required.³⁴

Suboptimal dressings or securements are not unusual in hospitals.³⁵ Despite our policy of PIV securement with bordered transparent dressings, we found 4 dressing types in use. In addition, we found almost 50% of PIVs had an additional (secondary) securement, and this was associated with significantly less PIV failure of all 3 types. This suggests that 1 or more of nonsterile tape, elasticized tubular bandages, or other securement (eg, bandage or second transparent dressing) can reduce PIV failure, although a randomized trial is lacking.³⁶ Whether the dressing was failing and required reinforcement or hospital staff lacked confidence in the dressing and placed additional securement preventatively is unclear. Both PIV failure and PIV dressing failure are common, and further research into superior PIV products and practices is urgently needed. Paramedic insertions had a higher risk of dislodgement, suggesting that the increased emphasis on securement should start in the prehospital setting.

While multiple or difficult insertion attempts were not associated with PIV failure, insertions were not directly observed, and clinicians may have underreported attempts. In contrast, insertion-related bruising (a surrogate for difficult insertion) was associated with more than double the incidence of phlebitis. The long-term implications of multiple insertion attempts on patient’s vasculature are unclear, but we believe first time PIV insertion is important to patients and of interest to clinicians. A recent systematic review of strategies associated with first attempt PIV insertion success in an emergency department found little evidence for effective strategies and recommended further research.³⁷

The overall PIV failure rate in our study was 32%, lower than the 35% to 40% failure observed in our previous randomized controlled trials, which had more stringent inclusion and exclusion criteria (eg, longer predicted duration of therapy).^6,38 The implications for patients and costs to the organization of frequent catheter replacement demonstrate urgent need for further research in this area of practice.³⁹ A strength of this study is that all PIVs, regardless of the expected length of dwell time or reason for insertion, were eligible for inclusion, providing more generalizable results. The PIV failure rate of 32% is concerning because these failures trigger treatment delays and replacement insertions, with significant increased labor and equipment costs. The mean cost of PIV replacement has been costed at AUD $69.30 or US $51.92 (as per 2010 $ value) per episode of IV treatment.⁴⁰ For our hospital, which uses 200,000 PIVs per year, the current level of PIV failure suggests almost AU $5.5 (US $4.1) million in waste annually at this site alone.

The additional strengths of this study include the extensive information collected prospectively about PIV insertion and maintenance, including information on who inserted the PIV, IV medications administered, and PIV dressings used. Limitations were the population of surgical and medical patients in 1 tertiary hospital, which may not be generalizable to other settings.

Our study confirms the high rate of catheter failure in acute care hospitals, validates existing evidence related to PIV failure, and identifies new, potentially modifiable risk factors to improve PIV insertion and management. Implications for future research were also identified.

Acknowledgments

The researchers acknowledge and thank the nurses and patients involved in this study. The authors would also like to acknowledge Becton Dickinson for partly funding this study in the form of an unrestricted grant-in-aid paid to Griffith University. Becton Dickinson did not design the study protocol, collect or analyze data, and did not prepare or review the manuscript.

Disclosure

 On behalf of NM and CMR, Griffith University has received unrestricted educational and research grants and consultancy payment for lectures from 3M and Becton Dickinson. On behalf of NM, MC, and CMR, Griffith University has received unrestricted investigator-initiated research grants from Centurion Medical Products and Entrotech Lifesciences (manufacturers of PIV dressings) and Becton Dickinson (manufacturer of PIVs). On behalf of MC, Griffith University has received a consultancy payment to develop education material from Baxter. On behalf of CMR, Griffith University has received unrestricted donations or investigator initiated research grants unrelated to this research from Adhezion, Angiodynamics, Baxter, Carefusion, Cook Medical, Hospira, Mayo, Smiths Medical, and Vygon. On behalf of CMR, Griffith University has received consultancy payments for educational lectures or professional opinion from B. Braun, Bard, Carefusion, Mayo, ResQDevices, and Smiths Medical. On behalf of EL, Griffith University has received consultancy payments for educational lecture from 3M. On behalf of MC, Griffith University has received a consultancy payment to develop education material from Baxter. As this was an observational study, no products were trialed in this study. JW and GM have no conflicts of interest.

Transthoracic echocardiography (TTE) is one of the most commonly ordered diagnostic tests in healthcare. Studies of Medicare beneficiaries, for example, have shown that each year, approximately 20% undergo at least 1 TTE, including 4% who have 2 or more.¹ TTE utilization rates increased dramatically in the 1990s and early 2000s. Between 1999 and 2008, for example, the rate of use of TTE per Medicare beneficiary nearly doubled.² In 2014, echocardiography accounted for 10% of all Medicare spending for imaging services, or approximately $930 million.³ In response to concerns about the possible unnecessary use of TTE, the American Heart Association and American Society of Echocardiography developed Appropriate Use Criteria (AUC) in 2007 and 2011, which describe appropriate versus inappropriate indications for TTE.⁴ Subsequent studies have shown that rather than rooting out inappropriate studies, the vast majority of ordered studies appear to be appropriate according to the AUC criteria.⁵ The AUC criteria have also been criticized for being based on expert opinion rather than clinical evidence.⁶ Repeat TTE, defined as TTE done within 1 year of a prior TTE, represents 24% to 42% of all studies,^7-9 and 31% of all Medicare beneficiaries who have a TTE get a repeat TTE within 1 year.¹⁰ In the present study, we reviewed all inpatient TTE performed over 1 year and described the group that have had a prior TTE within the past year (“repeat TTE”). We then derived a clinical prediction model to predict unchanged repeat TTE, with the goal of defining a subset of studies that are potentially unnecessary.

The West Haven Connecticut Veteran’s Administration Hospital (WHVA), located outside New Haven, Connecticut, is a 228-bed tertiary care center affiliated with Yale University School of Medicine. Potential subjects were identified from review of the electronic medical records of all inpatients who had an inpatient echocardiogram between October 1, 2013, and September 30, 2014. Patient’s records were reviewed by using a standardized data extraction form for demographics, comorbidity, cardiovascular risk factors, service ordering the TTE, intensive care unit (ICU) location, prior TTE abnormalities, TTE indication, AUC category, time between TTEs, technical quality of TTE, electrocardiogram (ECG) abnormalities, history of intervening acute coronary syndrome, cardiac surgery, and revascularization. Candidate predictors included any variables suspected by the authors as being potentially associated with the primary outcome of changed repeat TTE. All patients who had an inpatient TTE and a prior TTE within the Veterans Affairs (VA) system within the past year were included in the study. Repeat studies from the same admission were only counted as 1 TTE and patients had to have had a prior TTE from a different admission or a prior outpatient TTE to be included. Patients who did not have a prior TTE within the past year or who had only a transesophageal echocardiogram or stress echocardiography were excluded. Suboptimal studies were included but noted as limited quality. The study was approved by the WHVA Institutional Review Board. The Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis statement was used in planning and reporting this study.¹¹

TTEs were classified as normal, mildly abnormal, or with a major abnormality based on previously published definitions.^12-14 Any abnormality was defined as any left ventricle (LV) dysfunction (left ventricular ejection fraction [LVEF] <55%), any aortic or mitral valve stenosis, any regional wall motion abnormality, any right ventricular dysfunction, any pulmonary hypertension, mild or greater valvular regurgitation, any diastolic dysfunction, moderate or greater pericardial effusion, any ventricular hypertrophy, or any other significant abnormality including thrombus, vegetation, or tamponade. Major abnormality was defined as moderate or greater LV dysfunction (LVEF <45%), moderate or greater valvular regurgitation, moderate or greater valvular stenosis (aortic or mitral valve area <1.5 cm²), any regional wall motion abnormality, right ventricular dysfunction, moderate or greater pulmonary hypertension, moderate or greater diastolic dysfunction, moderate or greater pericardial effusion, or any other major abnormality including thrombus, vegetation, tumor, or tamponade. Repeat TTEs were classified as changed or unchanged. Changed TTEs were divided into any new abnormality or improvement or a new major abnormality or improvement. Any new abnormality or improvement was defined as any new TTE abnormality that had not previously been described or in which there was a change of at least 1 severity grade from a prior TTE, including improvement by 1 grade. A new major TTE abnormality or improvement was defined as any new major TTE abnormality that had previously been normal, or if there had been a prior abnormality, a change in at least 1 severity grade for LVEF or 2 severity grades for abnormal valvular, pericardial, or prior pulmonary hypertension, including improvement by 2 severity grades. A change from mild to moderate mitral regurgitation therefore was classified as a nonmajor change, whereas a change from mild to severe was classified as major. All TTE classifications were based on the electronic TTE reports and were reviewed by 2 independent investigators (CG and JC) blinded to the patients’ other clinical characteristics. For TTE studies in which the investigators agreed, that determination was the final classification. Disagreements were reviewed and the final classification was determined by consensus.

In an analogous manner, ECGs were classified as normal, mildly abnormal, or with a major abnormality based on previous definitions in the literature.¹⁵ Major abnormality was defined as atrial fibrillation or flutter, high-degree atrioventricular blocks, left bundle-branch block, right bundle-branch block, indeterminate conduction delay, q-wave myocardial infarction, isolated ischemic abnormalities, left ventricular hypertrophy with ST-T abnormalities, other arrhythmias including supraventricular tachycardia (SVT) or ventricular tachycardia (VT), low voltage (peak-to-peak QRS amplitude of <5 mm in the limb leads and/or <10 mm in the precordial leads), paced rhythm, sinus tachycardia (heart rate [HR] >100) or bradycardia (HR <50). Mild ECG abnormality was defined as low-grade atrioventricular blocks, borderline prolonged ventricular excitation, prolonged ventricular repolarization, isolated minor Q and ST-T abnormalities, left ventricular hypertrophy without ST-T abnormalities, left atrial enlargement, atrial or ventricular premature beats, or fascicular blocks. New major ECG abnormalities were any of the listed major ECG abnormalities that were not present on ECGs prior to the admission during which the repeat TTE was performed.

Other study definitions included intervening acute myocardial infarction (AMI), which was defined by any intervening history of elevated troponins, regardless of symptoms or ECG changes and including demand ischemia. Chronic kidney disease (CKD) was defined as an abnormal serum creatinine on 2 or more occasions 3 months apart. Active cancer was defined as receiving chemotherapy or palliative care for advanced cancer. Valvular heart disease was defined as prior moderate or severe valvular stenosis or regurgitation.

For analysis, we first compared patients with repeat TTE with major changes with those without major changes. For comparison of dichotomous variables, χ² or Fisher exact tests were used. For continuous variables, Student t test or the Mann-Whitney U test were performed. Because many of the frequencies of individual AUC criteria were small, related AUC criteria were grouped for analysis as grouped by the tables of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Society of Echocardiography, American Heart Association, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Critical Care Medicine, Society of Cardiovascular Computed Tomography, and Society for Cardiovascular Magnetic Resonance (ACCF/ASE/AHA) Guideline.⁴ Criteria groupings that were significantly less likely to have major TTE changes on analysis were classified as low risk and criteria that were significantly more likely were classified as high risk. Criteria groupings that were not significantly associated with TTE change were classified as average risk. All variables with P values less than 0.05 on bivariate analysis were then entered into a multivariate logistic regression analysis with major TTE change as the dependent variable, using backward stepwise variable selection with entry and exit criteria of P < 0.05 and P > 0.10, respectively. Scores were derived by converting the regression coefficients of independently predictive variables in the logistic regression model into corresponding integers. A total score was calculated for each patient by summing up the points for each independently significant variable. Model performance was described by calculating a C statistic by creation of a receiver operating characteristic curve to assess discrimination, and by performing the Hosmer and Lemeshow test to assess calibration. Internal validation was assessed by calculating the C statistic using the statistical method of bootstrapping in which the data were resampled multiple times (n = 200) and the average resultant C statistic reported. The bootstrap analysis was performed using R version 3.1 (R Foundation for Statistical Computing, Vienna, Austria). All other analyses were performed using SPSS version 21.0 (IBM, Armonk, New York). P values <0.05 were considered significant.

During the 1-year study period, there were 3944 medical/surgical admissions for 3266 patients and 845 inpatient TTEs obtained on 601 patients. Of all patients who were admitted, 601/3266 (18.4%) had at least 1 inpatient TTE. Of these 601 TTEs, 211 (35%) had a TTE within the VA system during the prior year. Of the 211 repeat TTEs, 67 (32%) were unchanged, 66 (31%) had minor changes, and 78 (37%) had major changes. The kappa statistic for agreement between extractors for “major TTE change” was 0.91, P < 0.001. The 10 most common AUC indications for TTE, which accounted for 72% of all studies, are listed in Table 1. The initial AUCs assigned by reviewers were the same in 187 of 211 TTEs (kappa 0.86, P < 0.001). Most indications were not associated with TTE outcome, although studies ordered for AUC indications 1 and 2 were less likely be associated with major changes and AUC indications 22 and 47 were more likely to be associated with major changes. Table 2 shows the comparison of the 78 patients that had repeat TTE with major changes compared with the 133 patients that did not. Nine variables were significantly different between the 2 groups; repeat TTEs with major changes were more likely to have dementia, be ordered by the surgery service, be located in an ICU, have major new ECG changes, have had prior valvular heart disease, have had an intervening AMI or cardiac surgery, or be in a high-risk AUC category. Patients with CKD were less likely to have major changes. Table 3 shows the results of the multivariate analysis; CKD, intervening AMI, prior valvular heart disease, major new ECG changes, and intervening cardiac surgery all independently predicted major changes on repeat TTE. Based on the β-coefficient for each variable, a point system was assigned to each variable and a total score calculated for each patient. Most variables had β-coefficients close to 1 and were therefore assigned a score of 1. CKD was associated with a lower risk of major TTE abnormality and was assigned a negative score. Intervening AMI was associated with a β-coefficient of 2.2 and was assigned a score of 2. Based on the points assigned to each variable and its presence or absence for each patient, a total score, which we named the CAVES score, was calculated. The acronym CAVES stands for CKD, AMI, valvular disease, ECG changes, and surgery (cardiac). Table 4 shows the frequencies of each score for each patient, ranging from patients with CKD and no other risk factors who scored −1 to patients without CKD who had all 4 of the other variables who scored 5. The prevalence of major TTE change for the full cohort was 37%. For the group with a CAVES score of −1, the probability was only 5.6%; for the group with a score of 0, the probability was 17.7%; and for the group with a score ≥1, the probability was 55.3%.

In this retrospective study, we found that approximately 18% of all patients admitted to the hospital had an inpatient TTE performed, and that approximately 35% of this group had a prior TTE within the past year. Of the group with prior TTEs within the past year, 37% had a major new change and 63% had either minor or no changes. Prior studies have reported similar high rates of repeat TTE^7-9 and of major changes on repeat TTE.^8,14,16 On multivariate analysis, we found that 5 variables were independent predictors of new changes on TTE—absence of CKD, intervening AMI, intervening cardiac surgery, history of valvular heart disease, and major new ECG changes. We developed and internally validated a risk score based on these 5 variables, which was found to have good overall accuracy as measured by the bootstrap corrected C statistic. The simplified version of the score divides patients into low, intermediate, and high risk for major changes on TTE. The low-risk group, defined as the group with no risk factors, had an approximately 6% risk of a major TTE change; the intermediate risk group, defined as a score of 0, had an 18% risk of major TTE change; and the high-risk group, defined as a score of 1 or greater, had a 55% chance of major TTE change. We believe that this risk score, if further validated, will potentially allow hospital-based clinicians to estimate the chance of a major change on TTE prior to ordering the study. For the low-risk group, this may indicate that the study is unnecessary. Conversely, for patients at high risk, this may offer further evidence that it will be useful to obtain a repeat TTE.

Disclosure 

The authors indicated no conflicts of interest.

Transthoracic echocardiography (TTE) is one of the most commonly ordered diagnostic tests in healthcare. Studies of Medicare beneficiaries, for example, have shown that each year, approximately 20% undergo at least 1 TTE, including 4% who have 2 or more.¹ TTE utilization rates increased dramatically in the 1990s and early 2000s. Between 1999 and 2008, for example, the rate of use of TTE per Medicare beneficiary nearly doubled.² In 2014, echocardiography accounted for 10% of all Medicare spending for imaging services, or approximately $930 million.³ In response to concerns about the possible unnecessary use of TTE, the American Heart Association and American Society of Echocardiography developed Appropriate Use Criteria (AUC) in 2007 and 2011, which describe appropriate versus inappropriate indications for TTE.⁴ Subsequent studies have shown that rather than rooting out inappropriate studies, the vast majority of ordered studies appear to be appropriate according to the AUC criteria.⁵ The AUC criteria have also been criticized for being based on expert opinion rather than clinical evidence.⁶ Repeat TTE, defined as TTE done within 1 year of a prior TTE, represents 24% to 42% of all studies,^7-9 and 31% of all Medicare beneficiaries who have a TTE get a repeat TTE within 1 year.¹⁰ In the present study, we reviewed all inpatient TTE performed over 1 year and described the group that have had a prior TTE within the past year (“repeat TTE”). We then derived a clinical prediction model to predict unchanged repeat TTE, with the goal of defining a subset of studies that are potentially unnecessary.

The West Haven Connecticut Veteran’s Administration Hospital (WHVA), located outside New Haven, Connecticut, is a 228-bed tertiary care center affiliated with Yale University School of Medicine. Potential subjects were identified from review of the electronic medical records of all inpatients who had an inpatient echocardiogram between October 1, 2013, and September 30, 2014. Patient’s records were reviewed by using a standardized data extraction form for demographics, comorbidity, cardiovascular risk factors, service ordering the TTE, intensive care unit (ICU) location, prior TTE abnormalities, TTE indication, AUC category, time between TTEs, technical quality of TTE, electrocardiogram (ECG) abnormalities, history of intervening acute coronary syndrome, cardiac surgery, and revascularization. Candidate predictors included any variables suspected by the authors as being potentially associated with the primary outcome of changed repeat TTE. All patients who had an inpatient TTE and a prior TTE within the Veterans Affairs (VA) system within the past year were included in the study. Repeat studies from the same admission were only counted as 1 TTE and patients had to have had a prior TTE from a different admission or a prior outpatient TTE to be included. Patients who did not have a prior TTE within the past year or who had only a transesophageal echocardiogram or stress echocardiography were excluded. Suboptimal studies were included but noted as limited quality. The study was approved by the WHVA Institutional Review Board. The Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis statement was used in planning and reporting this study.¹¹

TTEs were classified as normal, mildly abnormal, or with a major abnormality based on previously published definitions.^12-14 Any abnormality was defined as any left ventricle (LV) dysfunction (left ventricular ejection fraction [LVEF] <55%), any aortic or mitral valve stenosis, any regional wall motion abnormality, any right ventricular dysfunction, any pulmonary hypertension, mild or greater valvular regurgitation, any diastolic dysfunction, moderate or greater pericardial effusion, any ventricular hypertrophy, or any other significant abnormality including thrombus, vegetation, or tamponade. Major abnormality was defined as moderate or greater LV dysfunction (LVEF <45%), moderate or greater valvular regurgitation, moderate or greater valvular stenosis (aortic or mitral valve area <1.5 cm²), any regional wall motion abnormality, right ventricular dysfunction, moderate or greater pulmonary hypertension, moderate or greater diastolic dysfunction, moderate or greater pericardial effusion, or any other major abnormality including thrombus, vegetation, tumor, or tamponade. Repeat TTEs were classified as changed or unchanged. Changed TTEs were divided into any new abnormality or improvement or a new major abnormality or improvement. Any new abnormality or improvement was defined as any new TTE abnormality that had not previously been described or in which there was a change of at least 1 severity grade from a prior TTE, including improvement by 1 grade. A new major TTE abnormality or improvement was defined as any new major TTE abnormality that had previously been normal, or if there had been a prior abnormality, a change in at least 1 severity grade for LVEF or 2 severity grades for abnormal valvular, pericardial, or prior pulmonary hypertension, including improvement by 2 severity grades. A change from mild to moderate mitral regurgitation therefore was classified as a nonmajor change, whereas a change from mild to severe was classified as major. All TTE classifications were based on the electronic TTE reports and were reviewed by 2 independent investigators (CG and JC) blinded to the patients’ other clinical characteristics. For TTE studies in which the investigators agreed, that determination was the final classification. Disagreements were reviewed and the final classification was determined by consensus.

In an analogous manner, ECGs were classified as normal, mildly abnormal, or with a major abnormality based on previous definitions in the literature.¹⁵ Major abnormality was defined as atrial fibrillation or flutter, high-degree atrioventricular blocks, left bundle-branch block, right bundle-branch block, indeterminate conduction delay, q-wave myocardial infarction, isolated ischemic abnormalities, left ventricular hypertrophy with ST-T abnormalities, other arrhythmias including supraventricular tachycardia (SVT) or ventricular tachycardia (VT), low voltage (peak-to-peak QRS amplitude of <5 mm in the limb leads and/or <10 mm in the precordial leads), paced rhythm, sinus tachycardia (heart rate [HR] >100) or bradycardia (HR <50). Mild ECG abnormality was defined as low-grade atrioventricular blocks, borderline prolonged ventricular excitation, prolonged ventricular repolarization, isolated minor Q and ST-T abnormalities, left ventricular hypertrophy without ST-T abnormalities, left atrial enlargement, atrial or ventricular premature beats, or fascicular blocks. New major ECG abnormalities were any of the listed major ECG abnormalities that were not present on ECGs prior to the admission during which the repeat TTE was performed.

Other study definitions included intervening acute myocardial infarction (AMI), which was defined by any intervening history of elevated troponins, regardless of symptoms or ECG changes and including demand ischemia. Chronic kidney disease (CKD) was defined as an abnormal serum creatinine on 2 or more occasions 3 months apart. Active cancer was defined as receiving chemotherapy or palliative care for advanced cancer. Valvular heart disease was defined as prior moderate or severe valvular stenosis or regurgitation.

For analysis, we first compared patients with repeat TTE with major changes with those without major changes. For comparison of dichotomous variables, χ² or Fisher exact tests were used. For continuous variables, Student t test or the Mann-Whitney U test were performed. Because many of the frequencies of individual AUC criteria were small, related AUC criteria were grouped for analysis as grouped by the tables of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Society of Echocardiography, American Heart Association, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Critical Care Medicine, Society of Cardiovascular Computed Tomography, and Society for Cardiovascular Magnetic Resonance (ACCF/ASE/AHA) Guideline.⁴ Criteria groupings that were significantly less likely to have major TTE changes on analysis were classified as low risk and criteria that were significantly more likely were classified as high risk. Criteria groupings that were not significantly associated with TTE change were classified as average risk. All variables with P values less than 0.05 on bivariate analysis were then entered into a multivariate logistic regression analysis with major TTE change as the dependent variable, using backward stepwise variable selection with entry and exit criteria of P < 0.05 and P > 0.10, respectively. Scores were derived by converting the regression coefficients of independently predictive variables in the logistic regression model into corresponding integers. A total score was calculated for each patient by summing up the points for each independently significant variable. Model performance was described by calculating a C statistic by creation of a receiver operating characteristic curve to assess discrimination, and by performing the Hosmer and Lemeshow test to assess calibration. Internal validation was assessed by calculating the C statistic using the statistical method of bootstrapping in which the data were resampled multiple times (n = 200) and the average resultant C statistic reported. The bootstrap analysis was performed using R version 3.1 (R Foundation for Statistical Computing, Vienna, Austria). All other analyses were performed using SPSS version 21.0 (IBM, Armonk, New York). P values <0.05 were considered significant.

During the 1-year study period, there were 3944 medical/surgical admissions for 3266 patients and 845 inpatient TTEs obtained on 601 patients. Of all patients who were admitted, 601/3266 (18.4%) had at least 1 inpatient TTE. Of these 601 TTEs, 211 (35%) had a TTE within the VA system during the prior year. Of the 211 repeat TTEs, 67 (32%) were unchanged, 66 (31%) had minor changes, and 78 (37%) had major changes. The kappa statistic for agreement between extractors for “major TTE change” was 0.91, P < 0.001. The 10 most common AUC indications for TTE, which accounted for 72% of all studies, are listed in Table 1. The initial AUCs assigned by reviewers were the same in 187 of 211 TTEs (kappa 0.86, P < 0.001). Most indications were not associated with TTE outcome, although studies ordered for AUC indications 1 and 2 were less likely be associated with major changes and AUC indications 22 and 47 were more likely to be associated with major changes. Table 2 shows the comparison of the 78 patients that had repeat TTE with major changes compared with the 133 patients that did not. Nine variables were significantly different between the 2 groups; repeat TTEs with major changes were more likely to have dementia, be ordered by the surgery service, be located in an ICU, have major new ECG changes, have had prior valvular heart disease, have had an intervening AMI or cardiac surgery, or be in a high-risk AUC category. Patients with CKD were less likely to have major changes. Table 3 shows the results of the multivariate analysis; CKD, intervening AMI, prior valvular heart disease, major new ECG changes, and intervening cardiac surgery all independently predicted major changes on repeat TTE. Based on the β-coefficient for each variable, a point system was assigned to each variable and a total score calculated for each patient. Most variables had β-coefficients close to 1 and were therefore assigned a score of 1. CKD was associated with a lower risk of major TTE abnormality and was assigned a negative score. Intervening AMI was associated with a β-coefficient of 2.2 and was assigned a score of 2. Based on the points assigned to each variable and its presence or absence for each patient, a total score, which we named the CAVES score, was calculated. The acronym CAVES stands for CKD, AMI, valvular disease, ECG changes, and surgery (cardiac). Table 4 shows the frequencies of each score for each patient, ranging from patients with CKD and no other risk factors who scored −1 to patients without CKD who had all 4 of the other variables who scored 5. The prevalence of major TTE change for the full cohort was 37%. For the group with a CAVES score of −1, the probability was only 5.6%; for the group with a score of 0, the probability was 17.7%; and for the group with a score ≥1, the probability was 55.3%.

In this retrospective study, we found that approximately 18% of all patients admitted to the hospital had an inpatient TTE performed, and that approximately 35% of this group had a prior TTE within the past year. Of the group with prior TTEs within the past year, 37% had a major new change and 63% had either minor or no changes. Prior studies have reported similar high rates of repeat TTE^7-9 and of major changes on repeat TTE.^8,14,16 On multivariate analysis, we found that 5 variables were independent predictors of new changes on TTE—absence of CKD, intervening AMI, intervening cardiac surgery, history of valvular heart disease, and major new ECG changes. We developed and internally validated a risk score based on these 5 variables, which was found to have good overall accuracy as measured by the bootstrap corrected C statistic. The simplified version of the score divides patients into low, intermediate, and high risk for major changes on TTE. The low-risk group, defined as the group with no risk factors, had an approximately 6% risk of a major TTE change; the intermediate risk group, defined as a score of 0, had an 18% risk of major TTE change; and the high-risk group, defined as a score of 1 or greater, had a 55% chance of major TTE change. We believe that this risk score, if further validated, will potentially allow hospital-based clinicians to estimate the chance of a major change on TTE prior to ordering the study. For the low-risk group, this may indicate that the study is unnecessary. Conversely, for patients at high risk, this may offer further evidence that it will be useful to obtain a repeat TTE.

Disclosure 

The authors indicated no conflicts of interest.

Transthoracic echocardiography (TTE) is one of the most commonly ordered diagnostic tests in healthcare. Studies of Medicare beneficiaries, for example, have shown that each year, approximately 20% undergo at least 1 TTE, including 4% who have 2 or more.¹ TTE utilization rates increased dramatically in the 1990s and early 2000s. Between 1999 and 2008, for example, the rate of use of TTE per Medicare beneficiary nearly doubled.² In 2014, echocardiography accounted for 10% of all Medicare spending for imaging services, or approximately $930 million.³ In response to concerns about the possible unnecessary use of TTE, the American Heart Association and American Society of Echocardiography developed Appropriate Use Criteria (AUC) in 2007 and 2011, which describe appropriate versus inappropriate indications for TTE.⁴ Subsequent studies have shown that rather than rooting out inappropriate studies, the vast majority of ordered studies appear to be appropriate according to the AUC criteria.⁵ The AUC criteria have also been criticized for being based on expert opinion rather than clinical evidence.⁶ Repeat TTE, defined as TTE done within 1 year of a prior TTE, represents 24% to 42% of all studies,^7-9 and 31% of all Medicare beneficiaries who have a TTE get a repeat TTE within 1 year.¹⁰ In the present study, we reviewed all inpatient TTE performed over 1 year and described the group that have had a prior TTE within the past year (“repeat TTE”). We then derived a clinical prediction model to predict unchanged repeat TTE, with the goal of defining a subset of studies that are potentially unnecessary.

The West Haven Connecticut Veteran’s Administration Hospital (WHVA), located outside New Haven, Connecticut, is a 228-bed tertiary care center affiliated with Yale University School of Medicine. Potential subjects were identified from review of the electronic medical records of all inpatients who had an inpatient echocardiogram between October 1, 2013, and September 30, 2014. Patient’s records were reviewed by using a standardized data extraction form for demographics, comorbidity, cardiovascular risk factors, service ordering the TTE, intensive care unit (ICU) location, prior TTE abnormalities, TTE indication, AUC category, time between TTEs, technical quality of TTE, electrocardiogram (ECG) abnormalities, history of intervening acute coronary syndrome, cardiac surgery, and revascularization. Candidate predictors included any variables suspected by the authors as being potentially associated with the primary outcome of changed repeat TTE. All patients who had an inpatient TTE and a prior TTE within the Veterans Affairs (VA) system within the past year were included in the study. Repeat studies from the same admission were only counted as 1 TTE and patients had to have had a prior TTE from a different admission or a prior outpatient TTE to be included. Patients who did not have a prior TTE within the past year or who had only a transesophageal echocardiogram or stress echocardiography were excluded. Suboptimal studies were included but noted as limited quality. The study was approved by the WHVA Institutional Review Board. The Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis statement was used in planning and reporting this study.¹¹

TTEs were classified as normal, mildly abnormal, or with a major abnormality based on previously published definitions.^12-14 Any abnormality was defined as any left ventricle (LV) dysfunction (left ventricular ejection fraction [LVEF] <55%), any aortic or mitral valve stenosis, any regional wall motion abnormality, any right ventricular dysfunction, any pulmonary hypertension, mild or greater valvular regurgitation, any diastolic dysfunction, moderate or greater pericardial effusion, any ventricular hypertrophy, or any other significant abnormality including thrombus, vegetation, or tamponade. Major abnormality was defined as moderate or greater LV dysfunction (LVEF <45%), moderate or greater valvular regurgitation, moderate or greater valvular stenosis (aortic or mitral valve area <1.5 cm²), any regional wall motion abnormality, right ventricular dysfunction, moderate or greater pulmonary hypertension, moderate or greater diastolic dysfunction, moderate or greater pericardial effusion, or any other major abnormality including thrombus, vegetation, tumor, or tamponade. Repeat TTEs were classified as changed or unchanged. Changed TTEs were divided into any new abnormality or improvement or a new major abnormality or improvement. Any new abnormality or improvement was defined as any new TTE abnormality that had not previously been described or in which there was a change of at least 1 severity grade from a prior TTE, including improvement by 1 grade. A new major TTE abnormality or improvement was defined as any new major TTE abnormality that had previously been normal, or if there had been a prior abnormality, a change in at least 1 severity grade for LVEF or 2 severity grades for abnormal valvular, pericardial, or prior pulmonary hypertension, including improvement by 2 severity grades. A change from mild to moderate mitral regurgitation therefore was classified as a nonmajor change, whereas a change from mild to severe was classified as major. All TTE classifications were based on the electronic TTE reports and were reviewed by 2 independent investigators (CG and JC) blinded to the patients’ other clinical characteristics. For TTE studies in which the investigators agreed, that determination was the final classification. Disagreements were reviewed and the final classification was determined by consensus.

In an analogous manner, ECGs were classified as normal, mildly abnormal, or with a major abnormality based on previous definitions in the literature.¹⁵ Major abnormality was defined as atrial fibrillation or flutter, high-degree atrioventricular blocks, left bundle-branch block, right bundle-branch block, indeterminate conduction delay, q-wave myocardial infarction, isolated ischemic abnormalities, left ventricular hypertrophy with ST-T abnormalities, other arrhythmias including supraventricular tachycardia (SVT) or ventricular tachycardia (VT), low voltage (peak-to-peak QRS amplitude of <5 mm in the limb leads and/or <10 mm in the precordial leads), paced rhythm, sinus tachycardia (heart rate [HR] >100) or bradycardia (HR <50). Mild ECG abnormality was defined as low-grade atrioventricular blocks, borderline prolonged ventricular excitation, prolonged ventricular repolarization, isolated minor Q and ST-T abnormalities, left ventricular hypertrophy without ST-T abnormalities, left atrial enlargement, atrial or ventricular premature beats, or fascicular blocks. New major ECG abnormalities were any of the listed major ECG abnormalities that were not present on ECGs prior to the admission during which the repeat TTE was performed.

Other study definitions included intervening acute myocardial infarction (AMI), which was defined by any intervening history of elevated troponins, regardless of symptoms or ECG changes and including demand ischemia. Chronic kidney disease (CKD) was defined as an abnormal serum creatinine on 2 or more occasions 3 months apart. Active cancer was defined as receiving chemotherapy or palliative care for advanced cancer. Valvular heart disease was defined as prior moderate or severe valvular stenosis or regurgitation.

For analysis, we first compared patients with repeat TTE with major changes with those without major changes. For comparison of dichotomous variables, χ² or Fisher exact tests were used. For continuous variables, Student t test or the Mann-Whitney U test were performed. Because many of the frequencies of individual AUC criteria were small, related AUC criteria were grouped for analysis as grouped by the tables of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Society of Echocardiography, American Heart Association, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Critical Care Medicine, Society of Cardiovascular Computed Tomography, and Society for Cardiovascular Magnetic Resonance (ACCF/ASE/AHA) Guideline.⁴ Criteria groupings that were significantly less likely to have major TTE changes on analysis were classified as low risk and criteria that were significantly more likely were classified as high risk. Criteria groupings that were not significantly associated with TTE change were classified as average risk. All variables with P values less than 0.05 on bivariate analysis were then entered into a multivariate logistic regression analysis with major TTE change as the dependent variable, using backward stepwise variable selection with entry and exit criteria of P < 0.05 and P > 0.10, respectively. Scores were derived by converting the regression coefficients of independently predictive variables in the logistic regression model into corresponding integers. A total score was calculated for each patient by summing up the points for each independently significant variable. Model performance was described by calculating a C statistic by creation of a receiver operating characteristic curve to assess discrimination, and by performing the Hosmer and Lemeshow test to assess calibration. Internal validation was assessed by calculating the C statistic using the statistical method of bootstrapping in which the data were resampled multiple times (n = 200) and the average resultant C statistic reported. The bootstrap analysis was performed using R version 3.1 (R Foundation for Statistical Computing, Vienna, Austria). All other analyses were performed using SPSS version 21.0 (IBM, Armonk, New York). P values <0.05 were considered significant.

During the 1-year study period, there were 3944 medical/surgical admissions for 3266 patients and 845 inpatient TTEs obtained on 601 patients. Of all patients who were admitted, 601/3266 (18.4%) had at least 1 inpatient TTE. Of these 601 TTEs, 211 (35%) had a TTE within the VA system during the prior year. Of the 211 repeat TTEs, 67 (32%) were unchanged, 66 (31%) had minor changes, and 78 (37%) had major changes. The kappa statistic for agreement between extractors for “major TTE change” was 0.91, P < 0.001. The 10 most common AUC indications for TTE, which accounted for 72% of all studies, are listed in Table 1. The initial AUCs assigned by reviewers were the same in 187 of 211 TTEs (kappa 0.86, P < 0.001). Most indications were not associated with TTE outcome, although studies ordered for AUC indications 1 and 2 were less likely be associated with major changes and AUC indications 22 and 47 were more likely to be associated with major changes. Table 2 shows the comparison of the 78 patients that had repeat TTE with major changes compared with the 133 patients that did not. Nine variables were significantly different between the 2 groups; repeat TTEs with major changes were more likely to have dementia, be ordered by the surgery service, be located in an ICU, have major new ECG changes, have had prior valvular heart disease, have had an intervening AMI or cardiac surgery, or be in a high-risk AUC category. Patients with CKD were less likely to have major changes. Table 3 shows the results of the multivariate analysis; CKD, intervening AMI, prior valvular heart disease, major new ECG changes, and intervening cardiac surgery all independently predicted major changes on repeat TTE. Based on the β-coefficient for each variable, a point system was assigned to each variable and a total score calculated for each patient. Most variables had β-coefficients close to 1 and were therefore assigned a score of 1. CKD was associated with a lower risk of major TTE abnormality and was assigned a negative score. Intervening AMI was associated with a β-coefficient of 2.2 and was assigned a score of 2. Based on the points assigned to each variable and its presence or absence for each patient, a total score, which we named the CAVES score, was calculated. The acronym CAVES stands for CKD, AMI, valvular disease, ECG changes, and surgery (cardiac). Table 4 shows the frequencies of each score for each patient, ranging from patients with CKD and no other risk factors who scored −1 to patients without CKD who had all 4 of the other variables who scored 5. The prevalence of major TTE change for the full cohort was 37%. For the group with a CAVES score of −1, the probability was only 5.6%; for the group with a score of 0, the probability was 17.7%; and for the group with a score ≥1, the probability was 55.3%.

In this retrospective study, we found that approximately 18% of all patients admitted to the hospital had an inpatient TTE performed, and that approximately 35% of this group had a prior TTE within the past year. Of the group with prior TTEs within the past year, 37% had a major new change and 63% had either minor or no changes. Prior studies have reported similar high rates of repeat TTE^7-9 and of major changes on repeat TTE.^8,14,16 On multivariate analysis, we found that 5 variables were independent predictors of new changes on TTE—absence of CKD, intervening AMI, intervening cardiac surgery, history of valvular heart disease, and major new ECG changes. We developed and internally validated a risk score based on these 5 variables, which was found to have good overall accuracy as measured by the bootstrap corrected C statistic. The simplified version of the score divides patients into low, intermediate, and high risk for major changes on TTE. The low-risk group, defined as the group with no risk factors, had an approximately 6% risk of a major TTE change; the intermediate risk group, defined as a score of 0, had an 18% risk of major TTE change; and the high-risk group, defined as a score of 1 or greater, had a 55% chance of major TTE change. We believe that this risk score, if further validated, will potentially allow hospital-based clinicians to estimate the chance of a major change on TTE prior to ordering the study. For the low-risk group, this may indicate that the study is unnecessary. Conversely, for patients at high risk, this may offer further evidence that it will be useful to obtain a repeat TTE.

Disclosure 

The authors indicated no conflicts of interest.