Original Research

ABSTRACT

Patient satisfaction has become a topic of interest within orthopedics as the landscape of provider reimbursement continues to evolve to reward value of care. Online internet physician rating sites are becoming increasingly popular ways for patients to subjectively express their provider experience. Understanding what patients value during their episode of care is important in the modern healthcare environment. The purpose of this study is to determine what preferences, if any, patients have when selecting their physician and how they experience care in an outpatient orthopedic setting. A prospective survey was electronically administered to 212 patients in an adult reconstruction clinic. One hundred ninety-six patients (92.5%) completed the survey. Demographic questions regarding age, sex, ethnicity, and prior adult reconstruction surgical history were obtained. When patients were asked how much time they would like the doctor to spend with them on a routine visit, the most common answer was 10 to 15 minutes (41.3%), with only 10.2% patients desiring >20 minutes. The majority of patients (83.1%) believe ≥30 minutes is too long to wait to see their surgeon. Less than half of patients (41.8%) stated that they would feel as though they were receiving below average care if seen only by a nurse practitioner or physician’s assistant at a postoperative visit. Patients reported no significant age, gender, or ethnicity preferences for their physician. Recommendations from friends or other physicians was the most common (66.4%) way for patients to find their physicians, while 12.2% utilized online rating sites during their search. Optimizing patient experiences in the office may include keeping wait times to <30 minutes and educating patients on the roles of physician extenders. More work needs to be done to further elucidate variables influencing the subjective patient experience with their orthopedic care.

Continue to: Patient satisfaction...

Patient satisfaction has become an important focus in the rapidly changing healthcare environment due to the significant impact it has on healthcare delivery, healthcare economics, assessment of the quality of care, development of patient-care models, and quality improvement initiatives.^1-4 Historically, the quality of care was measured by objective metrics such as complication rates, range-of-motion, and the provider’s expert opinion on the outcome. While those metrics are still impactful variables when defining a successful outcome, the medical community is now increasingly recognizing the importance of patients’ perspectives when defining successful treatments. Patient satisfaction is now highly regarded by clinicians and the government when considering outcomes and is even being incorporated into determining the value of care. Under the Affordable Care Act, patients assumed a more active role in clinical decision-making as well as in creating quality and efficiency initiatives.^5,6 By 2017, 2% of the United States government’s Medicare payments will be redistributed among hospitals and physicians based on their quality and efficiency metrics, which are largely determined by patients’ evaluations of care.⁷ As a result, there has been significant interest in identifying variables influencing patient satisfaction and subjective outcomes.^8,9

Patient satisfaction is related to both the outcomes of care and the process of care. As first described by Donabedian,¹⁰patients may be satisfied with the successful outcome of their care, but dissatisfied with how they received their care. The process of care is complex and considers many aspects of healthcare delivery, including time, cost, healthcare provider interactions, and burdens faced. While patient satisfaction with outcomes and process of care are heavily related, they should be regarded separately. It is essential that providers understand what variables are important to patients with regards to how they experience healthcare and choose their provider, especially surrounding elective procedures such as hip and knee arthroplasty.^11,12

Within orthopedic surgery, patient satisfaction scores are beginning to be incorporated as part of the standard-of-care quality metrics obtained along with patient-reported outcome measures (PROMs) at defined time points postoperatively. Furthermore, PROMs and patient satisfaction data are becoming an increasingly important component of medical decision-making.^13-16 Several authors have reported that increased patient satisfaction is correlated with increased compliance, improved treatment outcomes across numerous medical settings, including orthopedics, decreased risk of litigation, and higher patient ratings of the quality of care.^17,18 Various factors, including meeting of expectations, staff politeness, the communication skills of the surgeon, and waiting times, have been suggested to influence eventual patient satisfaction within the surgical literature.^19-21 However, within orthopedic surgery there is a paucity of investigations evaluating how patients determine preferences and satisfaction with the process of care.

The purpose of this study is to determine what preferences, if any, patients have when selecting their physician and how they experience care in an outpatient orthopedic setting. The authors hypothesize that the majority of patients find their physicians through online rating sites or recommendations from family and friends. The authors believe that patients expect to be seen in <30 minutes and will be unsatisfied overall with the amount of time that they spend with their physician.

Continue to: METHODS...

 

 

METHODS

The senior author (BRL) and a research team created a 15-question survey to evaluate patient preferences regarding the demographic characteristics (eg, age, gender, ethnicity) of their physician, wait times in a waiting room, time spent with the physician, care received from physician extenders (eg, nurse practitioners, physician assistants), and how they learned of their physician (Appendix). An a priori power analysis was conducted to determine that approximately 200 patients were needed for inclusion.^11,22 Following Institutional Review Board approval (ORA 15051104), the survey was administered to 212 patients in a single-surgeon, adult reconstruction clinic. The survey was digitally administered on a touch-screen tablet using an electronic independent third party survey center (SurveyMonkey Inc) devoid of any identifying data. The survey was offered to all patients >21 years of age who were English-speaking and in the common area as patients waiting to be seen, from June 2015 to March 2016. A research assistant approached patients in the waiting room and asked if they would like to participate in a short survey regarding what factors influence the patient-physician relationship from the patient’s perspective.

Appendix 1

1. Do you wish to partake in this 3-minute survey?
   1.  
   2.  

 

1. Have you had a prior knee or hip replacement?
   1.  
   2.  

 

1. What is your age?
   1. 30-40 years
   2. 40-50 years
   3. 50-60 years
   4. 60-70 years
   5. 70-80 years
   6. 80+ years

 

1. What is your gender?
   1.  
   2.  

 

1. Which of the following best represents your racial or ethnic heritage?
   1. African American
   2.  
   3.  
   4.  
   5.  

 

1. How much time would you like the doctor to spend talking to you on a routine visit?
   1. 0-5 minutes
   2. 5-10 minutes
   3. 10-15 minutes
   4. 15-20 minutes
   5. 20-30 minutes
   6. >30 minutes

 

1. How long is too long to wait to see the doctor?
   1. 10 minutes
   2. 20 minutes
   3. 30 minutes
   4. 40 minutes
   5. 50 minutes
   6. An hour or more

 

1. If you were to only see a physician’s assistant or nurse practitioner at your follow-up visit and not the doctor, would you feel like you were getting below average care?
   1.  
   2.  

 

1. Overall I am satisfied with the time the doctor spends with me.
   1.  
   2.  

 

1. If you were to need a major surgery, would you want the physician to tell you what he or she would do if they were in your shoes?
   1.  
   2.  

 

1. Would you prefer your doctor to be the same race/ethnicity as you?
   1. No
   2.  
   3. No Preference

 

1. Would you feel more comfortable with a male as opposed to a female orthopedic surgeon?
   1.  
   2.  

 

1. Would you feel more comfortable with a female as opposed to a male orthopedic surgeon?
   1.  
   2.  

 

1. What age would you like your physician to be?
   1. 25-35 years old
   2. 35-45 years old
   3. 45-55 years old
   4. 55-65 years old
   5. 65 years and older
   6. No preference

 

1. How do you usually find your physician?
   1.  
   2. Friends’ recommendations
   3. Healthcare provider’s recommendations
   4. Insurance plans
   5. Online research/ratings
   6. Other

 

Descriptive statistics were used to analyze subject demographics and survey responses. Chi-square analyses and multinomial logistic regressions were utilized to compare responses. All statistical analyses were conducted using SPSS version 24.0 software (SPSS Inc). Statistical significance was set at P < 0.05.

RESULTS

Of the 212 patients who were invited to participate, 196 patients (92.4%) agreed and completed the survey. Demographic and surgical history information can be found in Table 1. The majority of patients were female (62%) and above the age of 50 years (92.4%). Almost half (48.5%) of patients had a prior hip or knee replacement.

Table 1. Survey Respondent Demographics

	Number	Percent
Age Range
30-40 years	4	2.0%
40-50 years	11	5.6%
50-60 years	47	24.0%
60-70 years	84	42.9%
70-80 years	41	20.9%
>80 years	9	4.6%
Gender
Male	74	37.8%
Female	122	62.2%
Ethnicity
African American	39	19.9%
Asian	3	1.5%
Caucasian	140	71.4%
Hispanic	10	5.1%
Other	4	2.0%
Prior knee or hip replacement
Yes	95	48.5%
No	55	28.1%
No Response	46	23.5%

 

When asked how long is too long to wait to see the doctor, 30 minutes (39.8%) was most commonly selected, followed by 40 minutes (24.5%) (Figure 1). When asked how much time patients would like the doctor to spend with them during an office visit, the majority (68.9%) selected either 10 to 15 minutes (41.3%) or 15 to 20 minutes (27.6%) (Figure 2). The majority of patients (92.3%) were satisfied with the amount of time the doctor spent with them. In addition, 94.9% of respondents would want their doctor to tell them what they would do if they were in the patient’s shoes when making decisions regarding their medical care (Table 2). Less than half of respondents (41.8%) believe that seeing a physician extender (eg, nurse practitioner or physician assistant) at a postoperative visit would result in a lower quality of care (Table 2).

Table 2. Responses to Survey Questions

If you were to only see a physician's assistant or nurse practitioner at your follow-up visit and not the doctor, would you feel like you were getting below average care?
Answer choices	Number	Percent
No	114	58.2%
Yes	82	41.8%
If you were to need a major surgery would you want the physician to tell you what he or she would do if they were in your shoes?
Answer choices	Number	Percent
No	10	5.1%
Yes	186	94.9%
Would you prefer your doctor to be the same race/ethnicity as you?
Answer choices	Number	Percent
No	29	14.8%
Yes	3	1.5%
No Preference	164	83.7%

 

When asked if patients preferred a doctor of the same race/ethnicity, the vast majority (83.7%) had no preference (Table 2). There was no significant difference found between male and female respondents when asked if they would feel more comfortable with a male as opposed to a female orthopedic surgeon (P = .118) and vice versa (P = .604) (Table 3). Most patients preferred a physician between the ages of 45 and 55 years (39.3%), followed by 35 and 45 years (23.0%); however, this preference was not statistically significant (Table 4). Older patients were more likely to prefer younger physicians (odds ratio, 4.612 for 25-35 years of age; odds ratio, 1.328 for 35-45 years of age). Only 12.2% of patients reported online research/rating sites as the main resource utilized when selecting their physician (Figure 3). The majority (68.4%) reported that recommendations from either friends (35.2%) or healthcare providers (33.2%) were the main avenues through which they found their physicians.

Table 3. Overall Responses to Questions Regarding Male and Female Orthopedic Surgeons^a

Would you feel more comfortable with a male as opposed to a female orthopedic surgeon?
Answer choices	Number	Percent	Female responses	Male responses	P value
No	164	83.7%	106 (86.9%)	58 (78.4%)	0.118
Yes	32	16.3%	16 (13.1%)	16 (21.6%)
Would you feel more comfortable with a female as opposed to a male orthopedic surgeon?
Answer choices	Number	Percent	Female responses	Male responses	P value
No	186	94.9%	115 (94.3%)	71 (95.9%)	0.604
Yes	10	5.1%	7 (5.7%)	3 (4.1%)

^aResponses were broken down by gender and compared utilizing a 2 x 2 chi-square analysis to test for significant differences in respondents’ gender preferences for their orthopedic surgeon.

 

Table 4. Patient Preferences Regarding Physician Age

What age would you like your physician to be?					95% Confidence Interval
Answer Choices	Number or Responses	Percent	P value	Exp(β)	Lower Bound	Upper Bound
25-35 years	1	0.5%	0.217	4.612	0.407	52.283
35-45 years	45	23.0%	0.161	1.328	0.893	1.975
45-55 years	77	39.3%	0.159	1.276	0.909	1.791
55-65 years	9	4.6%	0.483	1.302	0.624	2.717
≥65 years	2	1.0%	0.272	0.491	0.138	1.748
No preferencea	62	31.6%	Reference

^aNo preference was used as the reference category for the answer choices, while the age bracket “>80 years” was used as the reference for the age of respondent variable.

 

 

Continue to: DISCUSSION... 

 

 

DISCUSSION

The results of this study demonstrate that patients have several expectations and preferences with regards to the care they receive from physicians in the office. Patients prefer to wait <30 minutes before seeing their provider and desire only 10 to 20 minutes with their doctor. Patients do not have specific preferences with regards to the gender or ethnicity of their physician but would prefer a physician in the middle of their career, aged 45 to 55 years. Ultimately, patients do believe that seeing a physician at a postoperative visit is important, as just under half of patients thought that seeing a physician extender alone at a postoperative visit resulted in a lower quality of care.

While these results were obtained in a population specifically seeking the care of an orthopedic adult reconstruction surgeon, the results demonstrate that patients do not necessarily desire an unreasonable amount of time with their doctor. Patients simply want to be seen in a timely fashion and receive the full undivided attention of their doctor for approximately 20 minutes. Similarly, Patterson and colleagues²² found, in their series of 182 patients who presented to an orthopedic surgeon, that there was a significant correlation between time spent with the surgeon and overall patient satisfaction. Interestingly, the authors reported that patient satisfaction was not correlated with education level, sex, marital status, whether the patients were evaluated by a resident physician before seeing the attending surgeon, self-reported mental status, tobacco usage, the type of clinic visit, or the waiting time to see the surgeon (average, about 40 minutes for this cohort).²² Similarly, Teunis and colleagues²³ reported an average 32-minute wait time in 81 patients presenting for care at an orthopedic hand clinic and demonstrated that a longer wait time was associated with decreased patient satisfaction. These results corroborate the findings of this study that a short wait time is important to patients when evaluating the process of care. Additionally, patients do not have unreasonable expectations with regards to the amount of time they would like to spend with the physician. A physician who has a clinic for 9 hours a day would thus be able to see 54 patients and still spend at least 10 minutes with each patient. The quality of the physician-patient interaction is likely more important than the actual amount of time spent; however, based on this study, patients do have certain expectations about how much time physicians should spend with them.

There were no significant sex, age, or ethnicity preferences in our specific patient cohort. However, a sizable percentage of respondents, 41.8%, believed that they were receiving inferior care if they only saw a physician extender at a routine follow-up visit. Many orthopedic surgeons rely on the care provided by physician extenders to enable them to see additional patients. Physician extenders are well trained to provide high-quality care, including at routine postoperative visits. The results of this study, that many patients believe physician extenders provide lower-quality care, may be a result of inadequate patient education regarding the extensive training and education physician extenders undergo. Physician extenders are qualified, licensed healthcare professionals who are playing increasingly important roles within orthopedics and medicine as a whole. As the demand for orthopedic surgeons to see more patients increases, so does the role of physician extenders. Future research is warranted into educating the public regarding the importance of these healthcare providers and the adequacy of their training.

While many practices now routinely obtain patient satisfaction scores, another modality through which patients can express their satisfaction and experiences with healthcare providers is through online internet physician rating sites (IPRS). These sites have exploded in number and popularity in recent years and, according to some studies, have a very real effect on provider selection.²⁴ Interestingly, a low percentage of patients in this study utilized IPRS reviews to find their doctors. In a recent prospective survey study of 1000 consecutive patients presenting for care at the Mayo Clinic, Burkle and Keegan²⁴ reported that 27% of patients would choose not to see a physician based on a negative IPRS review. Interestingly, only 1.0% of patients reported finding their doctor through advertising. Numerous authors have recently addressed advertising in orthopedic surgery, specifically direct-to-consumer marking, including the influence of physician self-promotion on patients.^25,26 Specifically, Halawi and Barsoum²⁶ discussed how direct-to-consumer marketing is commonly disseminated to the public through television and print advertisements, which are modalities more commonly utilized by older generations. However, many advertising agencies are moving toward internet-based advertising, especially through orthopedic group and individual surgeon websites for self-promoting advertisement, as approximately 75% of Americans use the internet for health-related information.^25,27 The fact that many patients in this study did not utilize IPRS reviews or advertising (much of which is electronic) may be a result of the older, less internet-centric demographic that is often seen in an adult reconstruction clinic. Future research is warranted to determine what demographic of patients value IPRS reviews and how those reviews influence physician selection and the patient experience. 

There are several limitations to this study. First, the majority of the surveyed population was Caucasian, and our results may not be equally reflective of diverse ethnic backgrounds. Second, the cohort size, while based on previous studies conducted in a similar fashion, may be underpowered to detect significant differences for 1 or more of these questions. In addition, having a question regarding the patient’s medical background or experiences may have provided further insight as to why patients selected the answers that they did. Furthermore, questions regarding the patient’s education level, religious background, and income brackets may have provided further context in which to evaluate their responses. These questions were omitted in an effort to keep the questionnaire at a length that would maximize enrollment and prevent survey fatigue. Future research is warranted to determine what patient-specific, injury/symptom-specific, and treatment-specific variables influence the subjective patient experience.

CONCLUSION

The vast majority of patients desire only 10 to 20 minutes with their doctor and are highly satisfied with the amount of time their surgeon spends with them. Patients reported no significant gender- or ethnicity-based preferences for their doctor. The majority of patients believe that a wait time exceeding 30 minutes is too long. A greater effort needs to be made to educate patients and the public about the significant and effective roles nurse practitioners and physician assistants can play within the healthcare system. While this cohort did not report notable utilization of IPRS reviews, it remains essential to understand what factors influence patients’ subjective experiences with their providers to ensure that patients achieve their desired outcomes, and report as such on these websites as they continue to gain popularity. Diminishing clinic wait times and understanding patient preferences may lead to a greater percentage of “satisfied” patients. While the majority of focus has been and will likely continue to be on improving patients’ satisfaction with their outcomes, more work needs to be done focusing specifically on the process through which outcomes are achieved.

ABSTRACT

Patient satisfaction has become a topic of interest within orthopedics as the landscape of provider reimbursement continues to evolve to reward value of care. Online internet physician rating sites are becoming increasingly popular ways for patients to subjectively express their provider experience. Understanding what patients value during their episode of care is important in the modern healthcare environment. The purpose of this study is to determine what preferences, if any, patients have when selecting their physician and how they experience care in an outpatient orthopedic setting. A prospective survey was electronically administered to 212 patients in an adult reconstruction clinic. One hundred ninety-six patients (92.5%) completed the survey. Demographic questions regarding age, sex, ethnicity, and prior adult reconstruction surgical history were obtained. When patients were asked how much time they would like the doctor to spend with them on a routine visit, the most common answer was 10 to 15 minutes (41.3%), with only 10.2% patients desiring >20 minutes. The majority of patients (83.1%) believe ≥30 minutes is too long to wait to see their surgeon. Less than half of patients (41.8%) stated that they would feel as though they were receiving below average care if seen only by a nurse practitioner or physician’s assistant at a postoperative visit. Patients reported no significant age, gender, or ethnicity preferences for their physician. Recommendations from friends or other physicians was the most common (66.4%) way for patients to find their physicians, while 12.2% utilized online rating sites during their search. Optimizing patient experiences in the office may include keeping wait times to <30 minutes and educating patients on the roles of physician extenders. More work needs to be done to further elucidate variables influencing the subjective patient experience with their orthopedic care.

Continue to: Patient satisfaction...

Patient satisfaction has become an important focus in the rapidly changing healthcare environment due to the significant impact it has on healthcare delivery, healthcare economics, assessment of the quality of care, development of patient-care models, and quality improvement initiatives.^1-4 Historically, the quality of care was measured by objective metrics such as complication rates, range-of-motion, and the provider’s expert opinion on the outcome. While those metrics are still impactful variables when defining a successful outcome, the medical community is now increasingly recognizing the importance of patients’ perspectives when defining successful treatments. Patient satisfaction is now highly regarded by clinicians and the government when considering outcomes and is even being incorporated into determining the value of care. Under the Affordable Care Act, patients assumed a more active role in clinical decision-making as well as in creating quality and efficiency initiatives.^5,6 By 2017, 2% of the United States government’s Medicare payments will be redistributed among hospitals and physicians based on their quality and efficiency metrics, which are largely determined by patients’ evaluations of care.⁷ As a result, there has been significant interest in identifying variables influencing patient satisfaction and subjective outcomes.^8,9

Patient satisfaction is related to both the outcomes of care and the process of care. As first described by Donabedian,¹⁰patients may be satisfied with the successful outcome of their care, but dissatisfied with how they received their care. The process of care is complex and considers many aspects of healthcare delivery, including time, cost, healthcare provider interactions, and burdens faced. While patient satisfaction with outcomes and process of care are heavily related, they should be regarded separately. It is essential that providers understand what variables are important to patients with regards to how they experience healthcare and choose their provider, especially surrounding elective procedures such as hip and knee arthroplasty.^11,12

Within orthopedic surgery, patient satisfaction scores are beginning to be incorporated as part of the standard-of-care quality metrics obtained along with patient-reported outcome measures (PROMs) at defined time points postoperatively. Furthermore, PROMs and patient satisfaction data are becoming an increasingly important component of medical decision-making.^13-16 Several authors have reported that increased patient satisfaction is correlated with increased compliance, improved treatment outcomes across numerous medical settings, including orthopedics, decreased risk of litigation, and higher patient ratings of the quality of care.^17,18 Various factors, including meeting of expectations, staff politeness, the communication skills of the surgeon, and waiting times, have been suggested to influence eventual patient satisfaction within the surgical literature.^19-21 However, within orthopedic surgery there is a paucity of investigations evaluating how patients determine preferences and satisfaction with the process of care.

The purpose of this study is to determine what preferences, if any, patients have when selecting their physician and how they experience care in an outpatient orthopedic setting. The authors hypothesize that the majority of patients find their physicians through online rating sites or recommendations from family and friends. The authors believe that patients expect to be seen in <30 minutes and will be unsatisfied overall with the amount of time that they spend with their physician.

Continue to: METHODS...

 

 

METHODS

The senior author (BRL) and a research team created a 15-question survey to evaluate patient preferences regarding the demographic characteristics (eg, age, gender, ethnicity) of their physician, wait times in a waiting room, time spent with the physician, care received from physician extenders (eg, nurse practitioners, physician assistants), and how they learned of their physician (Appendix). An a priori power analysis was conducted to determine that approximately 200 patients were needed for inclusion.^11,22 Following Institutional Review Board approval (ORA 15051104), the survey was administered to 212 patients in a single-surgeon, adult reconstruction clinic. The survey was digitally administered on a touch-screen tablet using an electronic independent third party survey center (SurveyMonkey Inc) devoid of any identifying data. The survey was offered to all patients >21 years of age who were English-speaking and in the common area as patients waiting to be seen, from June 2015 to March 2016. A research assistant approached patients in the waiting room and asked if they would like to participate in a short survey regarding what factors influence the patient-physician relationship from the patient’s perspective.

Appendix 1

1. Do you wish to partake in this 3-minute survey?
   1.  
   2.  

 

1. Have you had a prior knee or hip replacement?
   1.  
   2.  

 

1. What is your age?
   1. 30-40 years
   2. 40-50 years
   3. 50-60 years
   4. 60-70 years
   5. 70-80 years
   6. 80+ years

 

1. What is your gender?
   1.  
   2.  

 

1. Which of the following best represents your racial or ethnic heritage?
   1. African American
   2.  
   3.  
   4.  
   5.  

 

1. How much time would you like the doctor to spend talking to you on a routine visit?
   1. 0-5 minutes
   2. 5-10 minutes
   3. 10-15 minutes
   4. 15-20 minutes
   5. 20-30 minutes
   6. >30 minutes

 

1. How long is too long to wait to see the doctor?
   1. 10 minutes
   2. 20 minutes
   3. 30 minutes
   4. 40 minutes
   5. 50 minutes
   6. An hour or more

 

1. If you were to only see a physician’s assistant or nurse practitioner at your follow-up visit and not the doctor, would you feel like you were getting below average care?
   1.  
   2.  

 

1. Overall I am satisfied with the time the doctor spends with me.
   1.  
   2.  

 

1. If you were to need a major surgery, would you want the physician to tell you what he or she would do if they were in your shoes?
   1.  
   2.  

 

1. Would you prefer your doctor to be the same race/ethnicity as you?
   1. No
   2.  
   3. No Preference

 

1. Would you feel more comfortable with a male as opposed to a female orthopedic surgeon?
   1.  
   2.  

 

1. Would you feel more comfortable with a female as opposed to a male orthopedic surgeon?
   1.  
   2.  

 

1. What age would you like your physician to be?
   1. 25-35 years old
   2. 35-45 years old
   3. 45-55 years old
   4. 55-65 years old
   5. 65 years and older
   6. No preference

 

1. How do you usually find your physician?
   1.  
   2. Friends’ recommendations
   3. Healthcare provider’s recommendations
   4. Insurance plans
   5. Online research/ratings
   6. Other

 

Descriptive statistics were used to analyze subject demographics and survey responses. Chi-square analyses and multinomial logistic regressions were utilized to compare responses. All statistical analyses were conducted using SPSS version 24.0 software (SPSS Inc). Statistical significance was set at P < 0.05.

RESULTS

Of the 212 patients who were invited to participate, 196 patients (92.4%) agreed and completed the survey. Demographic and surgical history information can be found in Table 1. The majority of patients were female (62%) and above the age of 50 years (92.4%). Almost half (48.5%) of patients had a prior hip or knee replacement.

Table 1. Survey Respondent Demographics

	Number	Percent
Age Range
30-40 years	4	2.0%
40-50 years	11	5.6%
50-60 years	47	24.0%
60-70 years	84	42.9%
70-80 years	41	20.9%
>80 years	9	4.6%
Gender
Male	74	37.8%
Female	122	62.2%
Ethnicity
African American	39	19.9%
Asian	3	1.5%
Caucasian	140	71.4%
Hispanic	10	5.1%
Other	4	2.0%
Prior knee or hip replacement
Yes	95	48.5%
No	55	28.1%
No Response	46	23.5%

 

When asked how long is too long to wait to see the doctor, 30 minutes (39.8%) was most commonly selected, followed by 40 minutes (24.5%) (Figure 1). When asked how much time patients would like the doctor to spend with them during an office visit, the majority (68.9%) selected either 10 to 15 minutes (41.3%) or 15 to 20 minutes (27.6%) (Figure 2). The majority of patients (92.3%) were satisfied with the amount of time the doctor spent with them. In addition, 94.9% of respondents would want their doctor to tell them what they would do if they were in the patient’s shoes when making decisions regarding their medical care (Table 2). Less than half of respondents (41.8%) believe that seeing a physician extender (eg, nurse practitioner or physician assistant) at a postoperative visit would result in a lower quality of care (Table 2).

Table 2. Responses to Survey Questions

If you were to only see a physician's assistant or nurse practitioner at your follow-up visit and not the doctor, would you feel like you were getting below average care?
Answer choices	Number	Percent
No	114	58.2%
Yes	82	41.8%
If you were to need a major surgery would you want the physician to tell you what he or she would do if they were in your shoes?
Answer choices	Number	Percent
No	10	5.1%
Yes	186	94.9%
Would you prefer your doctor to be the same race/ethnicity as you?
Answer choices	Number	Percent
No	29	14.8%
Yes	3	1.5%
No Preference	164	83.7%

 

When asked if patients preferred a doctor of the same race/ethnicity, the vast majority (83.7%) had no preference (Table 2). There was no significant difference found between male and female respondents when asked if they would feel more comfortable with a male as opposed to a female orthopedic surgeon (P = .118) and vice versa (P = .604) (Table 3). Most patients preferred a physician between the ages of 45 and 55 years (39.3%), followed by 35 and 45 years (23.0%); however, this preference was not statistically significant (Table 4). Older patients were more likely to prefer younger physicians (odds ratio, 4.612 for 25-35 years of age; odds ratio, 1.328 for 35-45 years of age). Only 12.2% of patients reported online research/rating sites as the main resource utilized when selecting their physician (Figure 3). The majority (68.4%) reported that recommendations from either friends (35.2%) or healthcare providers (33.2%) were the main avenues through which they found their physicians.

Table 3. Overall Responses to Questions Regarding Male and Female Orthopedic Surgeons^a

Would you feel more comfortable with a male as opposed to a female orthopedic surgeon?
Answer choices	Number	Percent	Female responses	Male responses	P value
No	164	83.7%	106 (86.9%)	58 (78.4%)	0.118
Yes	32	16.3%	16 (13.1%)	16 (21.6%)
Would you feel more comfortable with a female as opposed to a male orthopedic surgeon?
Answer choices	Number	Percent	Female responses	Male responses	P value
No	186	94.9%	115 (94.3%)	71 (95.9%)	0.604
Yes	10	5.1%	7 (5.7%)	3 (4.1%)

^aResponses were broken down by gender and compared utilizing a 2 x 2 chi-square analysis to test for significant differences in respondents’ gender preferences for their orthopedic surgeon.

 

Table 4. Patient Preferences Regarding Physician Age

What age would you like your physician to be?					95% Confidence Interval
Answer Choices	Number or Responses	Percent	P value	Exp(β)	Lower Bound	Upper Bound
25-35 years	1	0.5%	0.217	4.612	0.407	52.283
35-45 years	45	23.0%	0.161	1.328	0.893	1.975
45-55 years	77	39.3%	0.159	1.276	0.909	1.791
55-65 years	9	4.6%	0.483	1.302	0.624	2.717
≥65 years	2	1.0%	0.272	0.491	0.138	1.748
No preferencea	62	31.6%	Reference

^aNo preference was used as the reference category for the answer choices, while the age bracket “>80 years” was used as the reference for the age of respondent variable.

 

 

Continue to: DISCUSSION... 

 

 

DISCUSSION

The results of this study demonstrate that patients have several expectations and preferences with regards to the care they receive from physicians in the office. Patients prefer to wait <30 minutes before seeing their provider and desire only 10 to 20 minutes with their doctor. Patients do not have specific preferences with regards to the gender or ethnicity of their physician but would prefer a physician in the middle of their career, aged 45 to 55 years. Ultimately, patients do believe that seeing a physician at a postoperative visit is important, as just under half of patients thought that seeing a physician extender alone at a postoperative visit resulted in a lower quality of care.

While these results were obtained in a population specifically seeking the care of an orthopedic adult reconstruction surgeon, the results demonstrate that patients do not necessarily desire an unreasonable amount of time with their doctor. Patients simply want to be seen in a timely fashion and receive the full undivided attention of their doctor for approximately 20 minutes. Similarly, Patterson and colleagues²² found, in their series of 182 patients who presented to an orthopedic surgeon, that there was a significant correlation between time spent with the surgeon and overall patient satisfaction. Interestingly, the authors reported that patient satisfaction was not correlated with education level, sex, marital status, whether the patients were evaluated by a resident physician before seeing the attending surgeon, self-reported mental status, tobacco usage, the type of clinic visit, or the waiting time to see the surgeon (average, about 40 minutes for this cohort).²² Similarly, Teunis and colleagues²³ reported an average 32-minute wait time in 81 patients presenting for care at an orthopedic hand clinic and demonstrated that a longer wait time was associated with decreased patient satisfaction. These results corroborate the findings of this study that a short wait time is important to patients when evaluating the process of care. Additionally, patients do not have unreasonable expectations with regards to the amount of time they would like to spend with the physician. A physician who has a clinic for 9 hours a day would thus be able to see 54 patients and still spend at least 10 minutes with each patient. The quality of the physician-patient interaction is likely more important than the actual amount of time spent; however, based on this study, patients do have certain expectations about how much time physicians should spend with them.

There were no significant sex, age, or ethnicity preferences in our specific patient cohort. However, a sizable percentage of respondents, 41.8%, believed that they were receiving inferior care if they only saw a physician extender at a routine follow-up visit. Many orthopedic surgeons rely on the care provided by physician extenders to enable them to see additional patients. Physician extenders are well trained to provide high-quality care, including at routine postoperative visits. The results of this study, that many patients believe physician extenders provide lower-quality care, may be a result of inadequate patient education regarding the extensive training and education physician extenders undergo. Physician extenders are qualified, licensed healthcare professionals who are playing increasingly important roles within orthopedics and medicine as a whole. As the demand for orthopedic surgeons to see more patients increases, so does the role of physician extenders. Future research is warranted into educating the public regarding the importance of these healthcare providers and the adequacy of their training.

While many practices now routinely obtain patient satisfaction scores, another modality through which patients can express their satisfaction and experiences with healthcare providers is through online internet physician rating sites (IPRS). These sites have exploded in number and popularity in recent years and, according to some studies, have a very real effect on provider selection.²⁴ Interestingly, a low percentage of patients in this study utilized IPRS reviews to find their doctors. In a recent prospective survey study of 1000 consecutive patients presenting for care at the Mayo Clinic, Burkle and Keegan²⁴ reported that 27% of patients would choose not to see a physician based on a negative IPRS review. Interestingly, only 1.0% of patients reported finding their doctor through advertising. Numerous authors have recently addressed advertising in orthopedic surgery, specifically direct-to-consumer marking, including the influence of physician self-promotion on patients.^25,26 Specifically, Halawi and Barsoum²⁶ discussed how direct-to-consumer marketing is commonly disseminated to the public through television and print advertisements, which are modalities more commonly utilized by older generations. However, many advertising agencies are moving toward internet-based advertising, especially through orthopedic group and individual surgeon websites for self-promoting advertisement, as approximately 75% of Americans use the internet for health-related information.^25,27 The fact that many patients in this study did not utilize IPRS reviews or advertising (much of which is electronic) may be a result of the older, less internet-centric demographic that is often seen in an adult reconstruction clinic. Future research is warranted to determine what demographic of patients value IPRS reviews and how those reviews influence physician selection and the patient experience. 

There are several limitations to this study. First, the majority of the surveyed population was Caucasian, and our results may not be equally reflective of diverse ethnic backgrounds. Second, the cohort size, while based on previous studies conducted in a similar fashion, may be underpowered to detect significant differences for 1 or more of these questions. In addition, having a question regarding the patient’s medical background or experiences may have provided further insight as to why patients selected the answers that they did. Furthermore, questions regarding the patient’s education level, religious background, and income brackets may have provided further context in which to evaluate their responses. These questions were omitted in an effort to keep the questionnaire at a length that would maximize enrollment and prevent survey fatigue. Future research is warranted to determine what patient-specific, injury/symptom-specific, and treatment-specific variables influence the subjective patient experience.

CONCLUSION

The vast majority of patients desire only 10 to 20 minutes with their doctor and are highly satisfied with the amount of time their surgeon spends with them. Patients reported no significant gender- or ethnicity-based preferences for their doctor. The majority of patients believe that a wait time exceeding 30 minutes is too long. A greater effort needs to be made to educate patients and the public about the significant and effective roles nurse practitioners and physician assistants can play within the healthcare system. While this cohort did not report notable utilization of IPRS reviews, it remains essential to understand what factors influence patients’ subjective experiences with their providers to ensure that patients achieve their desired outcomes, and report as such on these websites as they continue to gain popularity. Diminishing clinic wait times and understanding patient preferences may lead to a greater percentage of “satisfied” patients. While the majority of focus has been and will likely continue to be on improving patients’ satisfaction with their outcomes, more work needs to be done focusing specifically on the process through which outcomes are achieved.

ABSTRACT

Patient satisfaction has become a topic of interest within orthopedics as the landscape of provider reimbursement continues to evolve to reward value of care. Online internet physician rating sites are becoming increasingly popular ways for patients to subjectively express their provider experience. Understanding what patients value during their episode of care is important in the modern healthcare environment. The purpose of this study is to determine what preferences, if any, patients have when selecting their physician and how they experience care in an outpatient orthopedic setting. A prospective survey was electronically administered to 212 patients in an adult reconstruction clinic. One hundred ninety-six patients (92.5%) completed the survey. Demographic questions regarding age, sex, ethnicity, and prior adult reconstruction surgical history were obtained. When patients were asked how much time they would like the doctor to spend with them on a routine visit, the most common answer was 10 to 15 minutes (41.3%), with only 10.2% patients desiring >20 minutes. The majority of patients (83.1%) believe ≥30 minutes is too long to wait to see their surgeon. Less than half of patients (41.8%) stated that they would feel as though they were receiving below average care if seen only by a nurse practitioner or physician’s assistant at a postoperative visit. Patients reported no significant age, gender, or ethnicity preferences for their physician. Recommendations from friends or other physicians was the most common (66.4%) way for patients to find their physicians, while 12.2% utilized online rating sites during their search. Optimizing patient experiences in the office may include keeping wait times to <30 minutes and educating patients on the roles of physician extenders. More work needs to be done to further elucidate variables influencing the subjective patient experience with their orthopedic care.

Continue to: Patient satisfaction...

Patient satisfaction has become an important focus in the rapidly changing healthcare environment due to the significant impact it has on healthcare delivery, healthcare economics, assessment of the quality of care, development of patient-care models, and quality improvement initiatives.^1-4 Historically, the quality of care was measured by objective metrics such as complication rates, range-of-motion, and the provider’s expert opinion on the outcome. While those metrics are still impactful variables when defining a successful outcome, the medical community is now increasingly recognizing the importance of patients’ perspectives when defining successful treatments. Patient satisfaction is now highly regarded by clinicians and the government when considering outcomes and is even being incorporated into determining the value of care. Under the Affordable Care Act, patients assumed a more active role in clinical decision-making as well as in creating quality and efficiency initiatives.^5,6 By 2017, 2% of the United States government’s Medicare payments will be redistributed among hospitals and physicians based on their quality and efficiency metrics, which are largely determined by patients’ evaluations of care.⁷ As a result, there has been significant interest in identifying variables influencing patient satisfaction and subjective outcomes.^8,9

Patient satisfaction is related to both the outcomes of care and the process of care. As first described by Donabedian,¹⁰patients may be satisfied with the successful outcome of their care, but dissatisfied with how they received their care. The process of care is complex and considers many aspects of healthcare delivery, including time, cost, healthcare provider interactions, and burdens faced. While patient satisfaction with outcomes and process of care are heavily related, they should be regarded separately. It is essential that providers understand what variables are important to patients with regards to how they experience healthcare and choose their provider, especially surrounding elective procedures such as hip and knee arthroplasty.^11,12

Within orthopedic surgery, patient satisfaction scores are beginning to be incorporated as part of the standard-of-care quality metrics obtained along with patient-reported outcome measures (PROMs) at defined time points postoperatively. Furthermore, PROMs and patient satisfaction data are becoming an increasingly important component of medical decision-making.^13-16 Several authors have reported that increased patient satisfaction is correlated with increased compliance, improved treatment outcomes across numerous medical settings, including orthopedics, decreased risk of litigation, and higher patient ratings of the quality of care.^17,18 Various factors, including meeting of expectations, staff politeness, the communication skills of the surgeon, and waiting times, have been suggested to influence eventual patient satisfaction within the surgical literature.^19-21 However, within orthopedic surgery there is a paucity of investigations evaluating how patients determine preferences and satisfaction with the process of care.

The purpose of this study is to determine what preferences, if any, patients have when selecting their physician and how they experience care in an outpatient orthopedic setting. The authors hypothesize that the majority of patients find their physicians through online rating sites or recommendations from family and friends. The authors believe that patients expect to be seen in <30 minutes and will be unsatisfied overall with the amount of time that they spend with their physician.

Continue to: METHODS...

 

 

METHODS

The senior author (BRL) and a research team created a 15-question survey to evaluate patient preferences regarding the demographic characteristics (eg, age, gender, ethnicity) of their physician, wait times in a waiting room, time spent with the physician, care received from physician extenders (eg, nurse practitioners, physician assistants), and how they learned of their physician (Appendix). An a priori power analysis was conducted to determine that approximately 200 patients were needed for inclusion.^11,22 Following Institutional Review Board approval (ORA 15051104), the survey was administered to 212 patients in a single-surgeon, adult reconstruction clinic. The survey was digitally administered on a touch-screen tablet using an electronic independent third party survey center (SurveyMonkey Inc) devoid of any identifying data. The survey was offered to all patients >21 years of age who were English-speaking and in the common area as patients waiting to be seen, from June 2015 to March 2016. A research assistant approached patients in the waiting room and asked if they would like to participate in a short survey regarding what factors influence the patient-physician relationship from the patient’s perspective.

Appendix 1

1. Do you wish to partake in this 3-minute survey?
   1.  
   2.  

 

1. Have you had a prior knee or hip replacement?
   1.  
   2.  

 

1. What is your age?
   1. 30-40 years
   2. 40-50 years
   3. 50-60 years
   4. 60-70 years
   5. 70-80 years
   6. 80+ years

 

1. What is your gender?
   1.  
   2.  

 

1. Which of the following best represents your racial or ethnic heritage?
   1. African American
   2.  
   3.  
   4.  
   5.  

 

1. How much time would you like the doctor to spend talking to you on a routine visit?
   1. 0-5 minutes
   2. 5-10 minutes
   3. 10-15 minutes
   4. 15-20 minutes
   5. 20-30 minutes
   6. >30 minutes

 

1. How long is too long to wait to see the doctor?
   1. 10 minutes
   2. 20 minutes
   3. 30 minutes
   4. 40 minutes
   5. 50 minutes
   6. An hour or more

 

1. If you were to only see a physician’s assistant or nurse practitioner at your follow-up visit and not the doctor, would you feel like you were getting below average care?
   1.  
   2.  

 

1. Overall I am satisfied with the time the doctor spends with me.
   1.  
   2.  

 

1. If you were to need a major surgery, would you want the physician to tell you what he or she would do if they were in your shoes?
   1.  
   2.  

 

1. Would you prefer your doctor to be the same race/ethnicity as you?
   1. No
   2.  
   3. No Preference

 

1. Would you feel more comfortable with a male as opposed to a female orthopedic surgeon?
   1.  
   2.  

 

1. Would you feel more comfortable with a female as opposed to a male orthopedic surgeon?
   1.  
   2.  

 

1. What age would you like your physician to be?
   1. 25-35 years old
   2. 35-45 years old
   3. 45-55 years old
   4. 55-65 years old
   5. 65 years and older
   6. No preference

 

1. How do you usually find your physician?
   1.  
   2. Friends’ recommendations
   3. Healthcare provider’s recommendations
   4. Insurance plans
   5. Online research/ratings
   6. Other

 

Descriptive statistics were used to analyze subject demographics and survey responses. Chi-square analyses and multinomial logistic regressions were utilized to compare responses. All statistical analyses were conducted using SPSS version 24.0 software (SPSS Inc). Statistical significance was set at P < 0.05.

RESULTS

Of the 212 patients who were invited to participate, 196 patients (92.4%) agreed and completed the survey. Demographic and surgical history information can be found in Table 1. The majority of patients were female (62%) and above the age of 50 years (92.4%). Almost half (48.5%) of patients had a prior hip or knee replacement.

Table 1. Survey Respondent Demographics

	Number	Percent
Age Range
30-40 years	4	2.0%
40-50 years	11	5.6%
50-60 years	47	24.0%
60-70 years	84	42.9%
70-80 years	41	20.9%
>80 years	9	4.6%
Gender
Male	74	37.8%
Female	122	62.2%
Ethnicity
African American	39	19.9%
Asian	3	1.5%
Caucasian	140	71.4%
Hispanic	10	5.1%
Other	4	2.0%
Prior knee or hip replacement
Yes	95	48.5%
No	55	28.1%
No Response	46	23.5%

 

When asked how long is too long to wait to see the doctor, 30 minutes (39.8%) was most commonly selected, followed by 40 minutes (24.5%) (Figure 1). When asked how much time patients would like the doctor to spend with them during an office visit, the majority (68.9%) selected either 10 to 15 minutes (41.3%) or 15 to 20 minutes (27.6%) (Figure 2). The majority of patients (92.3%) were satisfied with the amount of time the doctor spent with them. In addition, 94.9% of respondents would want their doctor to tell them what they would do if they were in the patient’s shoes when making decisions regarding their medical care (Table 2). Less than half of respondents (41.8%) believe that seeing a physician extender (eg, nurse practitioner or physician assistant) at a postoperative visit would result in a lower quality of care (Table 2).

Table 2. Responses to Survey Questions

If you were to only see a physician's assistant or nurse practitioner at your follow-up visit and not the doctor, would you feel like you were getting below average care?
Answer choices	Number	Percent
No	114	58.2%
Yes	82	41.8%
If you were to need a major surgery would you want the physician to tell you what he or she would do if they were in your shoes?
Answer choices	Number	Percent
No	10	5.1%
Yes	186	94.9%
Would you prefer your doctor to be the same race/ethnicity as you?
Answer choices	Number	Percent
No	29	14.8%
Yes	3	1.5%
No Preference	164	83.7%

 

When asked if patients preferred a doctor of the same race/ethnicity, the vast majority (83.7%) had no preference (Table 2). There was no significant difference found between male and female respondents when asked if they would feel more comfortable with a male as opposed to a female orthopedic surgeon (P = .118) and vice versa (P = .604) (Table 3). Most patients preferred a physician between the ages of 45 and 55 years (39.3%), followed by 35 and 45 years (23.0%); however, this preference was not statistically significant (Table 4). Older patients were more likely to prefer younger physicians (odds ratio, 4.612 for 25-35 years of age; odds ratio, 1.328 for 35-45 years of age). Only 12.2% of patients reported online research/rating sites as the main resource utilized when selecting their physician (Figure 3). The majority (68.4%) reported that recommendations from either friends (35.2%) or healthcare providers (33.2%) were the main avenues through which they found their physicians.

Table 3. Overall Responses to Questions Regarding Male and Female Orthopedic Surgeons^a

Would you feel more comfortable with a male as opposed to a female orthopedic surgeon?
Answer choices	Number	Percent	Female responses	Male responses	P value
No	164	83.7%	106 (86.9%)	58 (78.4%)	0.118
Yes	32	16.3%	16 (13.1%)	16 (21.6%)
Would you feel more comfortable with a female as opposed to a male orthopedic surgeon?
Answer choices	Number	Percent	Female responses	Male responses	P value
No	186	94.9%	115 (94.3%)	71 (95.9%)	0.604
Yes	10	5.1%	7 (5.7%)	3 (4.1%)

^aResponses were broken down by gender and compared utilizing a 2 x 2 chi-square analysis to test for significant differences in respondents’ gender preferences for their orthopedic surgeon.

 

Table 4. Patient Preferences Regarding Physician Age

What age would you like your physician to be?					95% Confidence Interval
Answer Choices	Number or Responses	Percent	P value	Exp(β)	Lower Bound	Upper Bound
25-35 years	1	0.5%	0.217	4.612	0.407	52.283
35-45 years	45	23.0%	0.161	1.328	0.893	1.975
45-55 years	77	39.3%	0.159	1.276	0.909	1.791
55-65 years	9	4.6%	0.483	1.302	0.624	2.717
≥65 years	2	1.0%	0.272	0.491	0.138	1.748
No preferencea	62	31.6%	Reference

^aNo preference was used as the reference category for the answer choices, while the age bracket “>80 years” was used as the reference for the age of respondent variable.

 

 

Continue to: DISCUSSION... 

 

 

DISCUSSION

The results of this study demonstrate that patients have several expectations and preferences with regards to the care they receive from physicians in the office. Patients prefer to wait <30 minutes before seeing their provider and desire only 10 to 20 minutes with their doctor. Patients do not have specific preferences with regards to the gender or ethnicity of their physician but would prefer a physician in the middle of their career, aged 45 to 55 years. Ultimately, patients do believe that seeing a physician at a postoperative visit is important, as just under half of patients thought that seeing a physician extender alone at a postoperative visit resulted in a lower quality of care.

While these results were obtained in a population specifically seeking the care of an orthopedic adult reconstruction surgeon, the results demonstrate that patients do not necessarily desire an unreasonable amount of time with their doctor. Patients simply want to be seen in a timely fashion and receive the full undivided attention of their doctor for approximately 20 minutes. Similarly, Patterson and colleagues²² found, in their series of 182 patients who presented to an orthopedic surgeon, that there was a significant correlation between time spent with the surgeon and overall patient satisfaction. Interestingly, the authors reported that patient satisfaction was not correlated with education level, sex, marital status, whether the patients were evaluated by a resident physician before seeing the attending surgeon, self-reported mental status, tobacco usage, the type of clinic visit, or the waiting time to see the surgeon (average, about 40 minutes for this cohort).²² Similarly, Teunis and colleagues²³ reported an average 32-minute wait time in 81 patients presenting for care at an orthopedic hand clinic and demonstrated that a longer wait time was associated with decreased patient satisfaction. These results corroborate the findings of this study that a short wait time is important to patients when evaluating the process of care. Additionally, patients do not have unreasonable expectations with regards to the amount of time they would like to spend with the physician. A physician who has a clinic for 9 hours a day would thus be able to see 54 patients and still spend at least 10 minutes with each patient. The quality of the physician-patient interaction is likely more important than the actual amount of time spent; however, based on this study, patients do have certain expectations about how much time physicians should spend with them.

There were no significant sex, age, or ethnicity preferences in our specific patient cohort. However, a sizable percentage of respondents, 41.8%, believed that they were receiving inferior care if they only saw a physician extender at a routine follow-up visit. Many orthopedic surgeons rely on the care provided by physician extenders to enable them to see additional patients. Physician extenders are well trained to provide high-quality care, including at routine postoperative visits. The results of this study, that many patients believe physician extenders provide lower-quality care, may be a result of inadequate patient education regarding the extensive training and education physician extenders undergo. Physician extenders are qualified, licensed healthcare professionals who are playing increasingly important roles within orthopedics and medicine as a whole. As the demand for orthopedic surgeons to see more patients increases, so does the role of physician extenders. Future research is warranted into educating the public regarding the importance of these healthcare providers and the adequacy of their training.

While many practices now routinely obtain patient satisfaction scores, another modality through which patients can express their satisfaction and experiences with healthcare providers is through online internet physician rating sites (IPRS). These sites have exploded in number and popularity in recent years and, according to some studies, have a very real effect on provider selection.²⁴ Interestingly, a low percentage of patients in this study utilized IPRS reviews to find their doctors. In a recent prospective survey study of 1000 consecutive patients presenting for care at the Mayo Clinic, Burkle and Keegan²⁴ reported that 27% of patients would choose not to see a physician based on a negative IPRS review. Interestingly, only 1.0% of patients reported finding their doctor through advertising. Numerous authors have recently addressed advertising in orthopedic surgery, specifically direct-to-consumer marking, including the influence of physician self-promotion on patients.^25,26 Specifically, Halawi and Barsoum²⁶ discussed how direct-to-consumer marketing is commonly disseminated to the public through television and print advertisements, which are modalities more commonly utilized by older generations. However, many advertising agencies are moving toward internet-based advertising, especially through orthopedic group and individual surgeon websites for self-promoting advertisement, as approximately 75% of Americans use the internet for health-related information.^25,27 The fact that many patients in this study did not utilize IPRS reviews or advertising (much of which is electronic) may be a result of the older, less internet-centric demographic that is often seen in an adult reconstruction clinic. Future research is warranted to determine what demographic of patients value IPRS reviews and how those reviews influence physician selection and the patient experience. 

There are several limitations to this study. First, the majority of the surveyed population was Caucasian, and our results may not be equally reflective of diverse ethnic backgrounds. Second, the cohort size, while based on previous studies conducted in a similar fashion, may be underpowered to detect significant differences for 1 or more of these questions. In addition, having a question regarding the patient’s medical background or experiences may have provided further insight as to why patients selected the answers that they did. Furthermore, questions regarding the patient’s education level, religious background, and income brackets may have provided further context in which to evaluate their responses. These questions were omitted in an effort to keep the questionnaire at a length that would maximize enrollment and prevent survey fatigue. Future research is warranted to determine what patient-specific, injury/symptom-specific, and treatment-specific variables influence the subjective patient experience.

CONCLUSION

The vast majority of patients desire only 10 to 20 minutes with their doctor and are highly satisfied with the amount of time their surgeon spends with them. Patients reported no significant gender- or ethnicity-based preferences for their doctor. The majority of patients believe that a wait time exceeding 30 minutes is too long. A greater effort needs to be made to educate patients and the public about the significant and effective roles nurse practitioners and physician assistants can play within the healthcare system. While this cohort did not report notable utilization of IPRS reviews, it remains essential to understand what factors influence patients’ subjective experiences with their providers to ensure that patients achieve their desired outcomes, and report as such on these websites as they continue to gain popularity. Diminishing clinic wait times and understanding patient preferences may lead to a greater percentage of “satisfied” patients. While the majority of focus has been and will likely continue to be on improving patients’ satisfaction with their outcomes, more work needs to be done focusing specifically on the process through which outcomes are achieved.

ABSTRACT

There is a paucity of evidence describing the types and rates of postoperative complications following total shoulder arthroplasty (TSA). We sought to analyze the complications following TSA and determine their effects on described outcome measures.

Using discharge data from the weighted Nationwide Inpatient Sample from 2006 to 2010, patients who underwent primary TSA were identified. The prevalence of specific complications was identified using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes. The data from this database represent events occurring during admission, prior to discharge. The associations between patient characteristics, complications, and outcomes of TSA were evaluated. The specific outcomes analyzed in this study were mortality and length of stay (LOS).

A total of 125,766 patients were identified. The rate of complication after TSA was 6.7% (8457 patients). The most frequent complications were respiratory, renal, and cardiac, occurring in 2.9%, 0.8%, and 0.8% of cases, respectively. Increasing age and total number of preoperative comorbidities significantly increased the likelihood of having a complication. The prevalence of postoperative shock and central nervous system, cardiac, vascular, and respiratory complications was significantly higher in patients who suffered postoperative mortality (88 patients; 0.07% mortality rate) than in those who survived surgery (P < 0.0001). In terms of LOS, shock and infectious and vascular complications most significantly increased the length of hospitalization.

Postoperative complications following TSA are not uncommon and occur in >6% of patients. Older patients and certain comorbidities are associated with complications after surgery. These complications are associated with postoperative mortality and increased LOS.

Continue to: Total shoulder arthroplasty...

Total shoulder arthroplasty (TSA) provides a predictably high level of satisfaction with survival as high as 92% at 15 years.¹ As implant instrumentation and surgical technique and understanding have improved, the frequency of TSAs being performed has also increased.² Although there are enough data on long-term surgical complications following TSA,^1,3-6 there is a paucity of evidence delineating the incidence and types of postoperative complications during hospitalization. Several current issues motivate the improved understanding of TSA, including the increasing number of TSAs being performed, the desire to improve quality of care, and the desire to create financially efficient healthcare.

The purpose of this study is to detail the postoperative complications that occur following TSA using a large national database. Specifically, our goals are to determine the incidence and types of complications after shoulder arthroplasty, determine the patient factors that are associated with these complications, and evaluate the effects of these complications on postoperative in-hospital mortality and length of stay (LOS). Our hypothesis is that there would be a correlation between specific patient factors and complications and that these complications would adversely correlate to patient postoperative outcomes.

METHODS

DESIGN

We conducted a retrospective analysis of TSAs captured by the Nationwide Inpatient Sample (NIS) database between 2006 and 2010. The NIS is the largest all-payer inpatient database that is currently available to the public in the United States.⁷

The NIS is a part of the Healthcare Cost and Utilization Project funded by the Agency for Healthcare Research and Quality (AHRQ) and the US Department of Health and Human Services. The NIS database is designed to approximate a 20% sample of US hospitals and the patients they serve, including community, academic, general, and specialty-specific hospitals such as orthopedic hospitals.⁷ The 2010 update of the NIS database contains discharge data from 1051 hospitals across 45 states, with a representative sample of >39 million inpatient hospital stays.⁷ The NIS database and its data sources have been independently validated and assessed for quality each year since 1988.⁸Furthermore, comparative analysis of multiple database elements and distributions has been validated against standard norms, including the National Hospital Discharge Survey.⁹ The NIS database has been used in numerous published studies.^2,10,11

PATIENT SELECTION

The yearly NIS databases from 2006 to 2010 were compiled. Patients aged ≥40 years who underwent a TSA were identified using the International Classification of Diseases, 9th Revision (ICD-9), procedural code 81.80. Exclusion criteria were patients with a primary or a secondary diagnosis of humeral or scapular fracture, chronic osteomyelitis, rheumatologic diseases, or evidence of concurrent malignancy (Figure 1).

Native to NIS are patient demographics, including age, sex, and race. Patient comorbidities as described by Elixhauser and colleagues¹² are also included in the database.

Continue to: OUTCOMES...

 

 

OUTCOMES

The primary outcome of this study was a description of the type and frequency of postoperative complications of TSA. To conduct this analysis, we queried the TSA cohort for specific ICD-9 codes representing acute cardiac, central nervous system, infectious, gastrointestinal, genitourinary, postoperative shock, renal, respiratory, surgical, vascular, and wound complications. The ICD-9 codes used to identify complications were modeled according to previous literature on various surgical applications and were further parsed to reflect only acute postoperative diagnoses^13-15(see the Appendix for the comprehensive list of ICD-9 codes).

Two additional outcomes were analyzed, including postoperative mortality and LOS. Postoperative mortality was defined as death occurring prior to discharge. We calculated the average LOS among the complication and the noncomplication cohort.

STATISTICAL ANALYSIS

Patient demographics and target outcomes of the study were analyzed by frequency distribution. Where applicable, the chi-square and the Student’s t tests were used to confirm the statistical difference for dichotomous and continuous variables, respectively. Multivariate regressions were performed after controlling for possible clustering of the data using a generalized estimating equation following a previous analytical methodology.^16-20 The results are reported with odds ratios and 95% confidence intervals where applicable, all statistical tests with P ≤ 0.05 were considered to be significant, and all statistical tests were two-sided. We conducted all analyses using SAS, version 9.2 (SAS Institute).

RESULTS

From 2006 to 2010, a weighted sample of 141,973 patients was found to undergo a TSA. After applying our inclusion and exclusion criteria, our study cohort consisted of 125,766 patients (Figure 1).

Continue to: OVERALL TSA COHORT DEMOGRAPHICS...

 

 

OVERALL TSA COHORT DEMOGRAPHICS

The average age of the TSA cohort was 69.4 years (standard deviation [SD], 21.20), and 54.1% were females. The cohort had significant comorbidities, with 83.3% of them having at least 1 comorbidity at the time of surgery. Specifically, 31.3% of the patients had 1 comorbidity, 26.5% had 2 comorbidities, and 25.4% had ≥3 comorbidities. Hypertension was the most common comorbidity present in 66.2% of patients, and diabetes was the second most common comorbidity with a prevalence of 16.8%.

COMPLICATION COHORT DEMOGRAPHICS

An overall postoperative complication rate of 6.7% (weighted sample of 8457 patients) was noted in the overall TSA cohort. The TSA cohort was dichotomized into patients who suffered at least 1 complication (weighted, n = 8457) and patients undergoing routine TSAs (weighted, n = 117,308). The average age was significantly higher in the complication vs routine cohort (71.38 vs 69.27 years, P < 0.0001). Similarly, there were significantly more comorbidities (2.51 vs 1.71, P < 0.0001) in the complication cohort.

COMPLICATIONS

We noted a complication rate of 6.7% (weighted sample of 8457 patients). A single complication was noted in 5% of these patients, whereas 1.3% and 0.4% of the patients had 2 and ≥3 complications, respectively. Respiratory abnormalities (2.9%), acute renal failure (0.8%), and cardiac complications (0.8%) were the most prevalent complications after TSA. The list of complications is detailed in Figure 2. Logistic regression analysis of patient characteristics predicting complications showed that advanced age (odds ratio [OR], 2.1 in those aged ≥85 years) and increasing number of comorbidities (≥3; OR, 3.5) were most significant in predicting complications (all P < 0.0001) (Figure 3). Despite the ubiquity of hypertension in this patient population, it was not a significant predictor of complication (OR, 0.9); in contrast, pulmonary disorders (OR, 5.1) and fluid and electrolyte disorders (4.0) were most strongly associated with the development of a postoperative complication after surgery (Figure 4).

EFFECT OF COMPLICATIONS ON LOS

The average length of hospitalization was 2.3 days (95% confidence interval, 2.22-2.25) among the entire cohort. The average LOS was longer in the complication cohort (3.9 days) than in patients who did not have a complication (2.1 days, P < 0.0001). Of the specific complications noted, hemodynamic shock (11.8 days); infectious, most commonly pneumonia (7.6 days); and vascular complications (6.9 days) were associated with the longest hospitalizations. This result is summarized in Figure 5.

MORTALITY

An overall postoperative (in-house) mortality rate of 0.07% was noted (weighted, n = 88). Comparison between the patient cohort that died vs those who survived TSA resulted in significant differences in the rates of complications. Complications that were most significantly different between the cohorts included cardiac (60.47% vs 0.75%, P < 0.0001), postoperative shock (26.61% vs 0.04%, P < 0.0001), and respiratory complications (43.1% vs 2.8%, P < 0.0001). It is important to note that the overall rate of postoperative shock was exceedingly low in the TSA cohort, but it was highly prevalent in the mortality cohort, occurring in 26.61% of patients. A summary of the mortality statistics is presented in Figure 6.

Continue to: DISCUSSION...

 

 

DISCUSSION

TSA continues to be associated with high levels of satisfaction;¹ as a result, its incidence is increasing.² As our understanding and efficiency improves nationally, it is imperative that we determine the short-term and longer-term outcomes and complications. In addition, the factors that may affect prognosis must be elucidated to provide a more individualized and effective standard of care. To date, most of the outcome studies of TSA have evaluated long-term outcomes and specific implant-related complications.^1,5,6,21,22 Our intent was to evaluate the complications that occur in the postoperative period and their effect on unique “patient care” outcomes. With knowledge of these complications and the predisposing factors, we can better assess patients, risk-stratify, and provide appropriate guidelines.

We noted that complications occurring after TSA are not uncommon, with >6% of patients suffering a postoperative complication. In this study, the number of complications noted was associated with worse patient outcomes. In addition, we noted that patients undergoing a TSA have a significant burden of comorbidities; however, hematologic and fluid disorders (eg, iron deficiency anemia, pulmonary circulatory disorders, and fluid imbalances) were most important in predicting postoperative complications.

Increased LOS in the hospital after TSA was associated with the occurrence of complications. Of all noted complications, shock and infectious and vascular complications led to the longest hospitalizations. Hospital-acquired pneumonia was the most common infectious etiology, while pulmonary embolism and deep vein thrombosis were the most consistent vascular complications. Although seldom studied in the TSA population, a similar finding has been noted in patients after THA. O’Malley and colleagues,²³ using the American College of Surgeon’s National Surgical Quality Improvement Program database, identified independent factors that were associated with complications and average prolonged LOS. They noted that the occurrence of major complications was associated with a prolonged LOS. Some, but not all the major complications, included organ space infection, cardiac events, pneumonia, and venous thromboembolic events.²³ Therefore, attempts to limit the amount of time spent in hospitals and control the associated costs must focus on managing the incidence of complications.

Postoperative mortality after TSA was uncommon, occurring in 0.07% of the patients in this study. The low incidence of mortality noted in this study is probably related to the fact that our data represent mortality, whereas in the hospital and, unlike most mortality studies, it does not account for patient demise that may occur in the months after surgery. Other reports have noted that mortality occurs in <1.5% of these patients.^24-28 Singh and colleagues²⁵ observed in their evaluation of perioperative mortality after TSA a mortality rate of 0.8% with 90 days after 4380 shoulder replacements performed at their institution. Using multivariate analysis, they were able to identify associations between mortality and increasing American Society of Anesthesiology (ASA) class and Charlson Comorbidity Index. These results in relation to ours would indicate that the majority of patients who die after shoulder arthroplasty do so after initial discharge. Although we could not determine a causal relationship between mortality and patient comorbidities, we noted that certain complications strongly correlated with mortality. In patients who died, there was a relatively high incidence of cardiac (60.5%) and respiratory (43.1%) complications. Similarly, although postoperative shock was almost nonexistent in the patients who survived surgery (0.04%), it was much more common in the patients who suffered mortality (26.6%).

This study is not without limitations. Data were extracted from a national database, therefore precluding the inclusion of specific details of surgery and functional assessment. Inherent to ICD-9 coding, we were unable to assess the exact detail and severity of complications. For instance, we cannot be certain what criteria were used to define “acute renal failure” for each patient. This study is retrospective in nature and therefore adequate randomization and standardization of patients is not possible. Similarly, the nature of the database may not allow for exacting our inclusion and exclusion criteria. However, the large sample size of the patient population lessens the chance of potential biases and type 2 errors. Prior to October 2010, reverse shoulder arthroplasty was coded under the ICD-9procedural code 81.80 as TSA. Therefore, there is some overlap between TSA and reverse shoulder arthroplasty in our data. Reverse shoulder arthroplasty is now coded under ICD-9 procedural code 81.88. It is possible that results may differ if reverse shoulder arthroplasty were excluded from our patient cohort. This can be an area of future research.

CONCLUSION

Although much is known about the long-term hardware and functional complications after TSA, in this study, we have attempted to broaden the understanding of perioperative complications and the associated sequelae. Complications are common after TSA surgery and are related to adverse outcomes. In the setting of healthcare changes, the surgeon and the patient must understand the cause, types, incidence, and outcomes of medical and surgical complications after surgery. This allows for more accurate “standard of care” metrics. Further large-volume multicenter studies are needed to gain further insight into the short- and long-term outcomes of TSA.

ABSTRACT

There is a paucity of evidence describing the types and rates of postoperative complications following total shoulder arthroplasty (TSA). We sought to analyze the complications following TSA and determine their effects on described outcome measures.

Using discharge data from the weighted Nationwide Inpatient Sample from 2006 to 2010, patients who underwent primary TSA were identified. The prevalence of specific complications was identified using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes. The data from this database represent events occurring during admission, prior to discharge. The associations between patient characteristics, complications, and outcomes of TSA were evaluated. The specific outcomes analyzed in this study were mortality and length of stay (LOS).

A total of 125,766 patients were identified. The rate of complication after TSA was 6.7% (8457 patients). The most frequent complications were respiratory, renal, and cardiac, occurring in 2.9%, 0.8%, and 0.8% of cases, respectively. Increasing age and total number of preoperative comorbidities significantly increased the likelihood of having a complication. The prevalence of postoperative shock and central nervous system, cardiac, vascular, and respiratory complications was significantly higher in patients who suffered postoperative mortality (88 patients; 0.07% mortality rate) than in those who survived surgery (P < 0.0001). In terms of LOS, shock and infectious and vascular complications most significantly increased the length of hospitalization.

Postoperative complications following TSA are not uncommon and occur in >6% of patients. Older patients and certain comorbidities are associated with complications after surgery. These complications are associated with postoperative mortality and increased LOS.

Continue to: Total shoulder arthroplasty...

Total shoulder arthroplasty (TSA) provides a predictably high level of satisfaction with survival as high as 92% at 15 years.¹ As implant instrumentation and surgical technique and understanding have improved, the frequency of TSAs being performed has also increased.² Although there are enough data on long-term surgical complications following TSA,^1,3-6 there is a paucity of evidence delineating the incidence and types of postoperative complications during hospitalization. Several current issues motivate the improved understanding of TSA, including the increasing number of TSAs being performed, the desire to improve quality of care, and the desire to create financially efficient healthcare.

The purpose of this study is to detail the postoperative complications that occur following TSA using a large national database. Specifically, our goals are to determine the incidence and types of complications after shoulder arthroplasty, determine the patient factors that are associated with these complications, and evaluate the effects of these complications on postoperative in-hospital mortality and length of stay (LOS). Our hypothesis is that there would be a correlation between specific patient factors and complications and that these complications would adversely correlate to patient postoperative outcomes.

METHODS

DESIGN

We conducted a retrospective analysis of TSAs captured by the Nationwide Inpatient Sample (NIS) database between 2006 and 2010. The NIS is the largest all-payer inpatient database that is currently available to the public in the United States.⁷

The NIS is a part of the Healthcare Cost and Utilization Project funded by the Agency for Healthcare Research and Quality (AHRQ) and the US Department of Health and Human Services. The NIS database is designed to approximate a 20% sample of US hospitals and the patients they serve, including community, academic, general, and specialty-specific hospitals such as orthopedic hospitals.⁷ The 2010 update of the NIS database contains discharge data from 1051 hospitals across 45 states, with a representative sample of >39 million inpatient hospital stays.⁷ The NIS database and its data sources have been independently validated and assessed for quality each year since 1988.⁸Furthermore, comparative analysis of multiple database elements and distributions has been validated against standard norms, including the National Hospital Discharge Survey.⁹ The NIS database has been used in numerous published studies.^2,10,11

PATIENT SELECTION

The yearly NIS databases from 2006 to 2010 were compiled. Patients aged ≥40 years who underwent a TSA were identified using the International Classification of Diseases, 9th Revision (ICD-9), procedural code 81.80. Exclusion criteria were patients with a primary or a secondary diagnosis of humeral or scapular fracture, chronic osteomyelitis, rheumatologic diseases, or evidence of concurrent malignancy (Figure 1).

Native to NIS are patient demographics, including age, sex, and race. Patient comorbidities as described by Elixhauser and colleagues¹² are also included in the database.

Continue to: OUTCOMES...

 

 

OUTCOMES

The primary outcome of this study was a description of the type and frequency of postoperative complications of TSA. To conduct this analysis, we queried the TSA cohort for specific ICD-9 codes representing acute cardiac, central nervous system, infectious, gastrointestinal, genitourinary, postoperative shock, renal, respiratory, surgical, vascular, and wound complications. The ICD-9 codes used to identify complications were modeled according to previous literature on various surgical applications and were further parsed to reflect only acute postoperative diagnoses^13-15(see the Appendix for the comprehensive list of ICD-9 codes).

Two additional outcomes were analyzed, including postoperative mortality and LOS. Postoperative mortality was defined as death occurring prior to discharge. We calculated the average LOS among the complication and the noncomplication cohort.

STATISTICAL ANALYSIS

Patient demographics and target outcomes of the study were analyzed by frequency distribution. Where applicable, the chi-square and the Student’s t tests were used to confirm the statistical difference for dichotomous and continuous variables, respectively. Multivariate regressions were performed after controlling for possible clustering of the data using a generalized estimating equation following a previous analytical methodology.^16-20 The results are reported with odds ratios and 95% confidence intervals where applicable, all statistical tests with P ≤ 0.05 were considered to be significant, and all statistical tests were two-sided. We conducted all analyses using SAS, version 9.2 (SAS Institute).

RESULTS

From 2006 to 2010, a weighted sample of 141,973 patients was found to undergo a TSA. After applying our inclusion and exclusion criteria, our study cohort consisted of 125,766 patients (Figure 1).

Continue to: OVERALL TSA COHORT DEMOGRAPHICS...

 

 

OVERALL TSA COHORT DEMOGRAPHICS

The average age of the TSA cohort was 69.4 years (standard deviation [SD], 21.20), and 54.1% were females. The cohort had significant comorbidities, with 83.3% of them having at least 1 comorbidity at the time of surgery. Specifically, 31.3% of the patients had 1 comorbidity, 26.5% had 2 comorbidities, and 25.4% had ≥3 comorbidities. Hypertension was the most common comorbidity present in 66.2% of patients, and diabetes was the second most common comorbidity with a prevalence of 16.8%.

COMPLICATION COHORT DEMOGRAPHICS

An overall postoperative complication rate of 6.7% (weighted sample of 8457 patients) was noted in the overall TSA cohort. The TSA cohort was dichotomized into patients who suffered at least 1 complication (weighted, n = 8457) and patients undergoing routine TSAs (weighted, n = 117,308). The average age was significantly higher in the complication vs routine cohort (71.38 vs 69.27 years, P < 0.0001). Similarly, there were significantly more comorbidities (2.51 vs 1.71, P < 0.0001) in the complication cohort.

COMPLICATIONS

We noted a complication rate of 6.7% (weighted sample of 8457 patients). A single complication was noted in 5% of these patients, whereas 1.3% and 0.4% of the patients had 2 and ≥3 complications, respectively. Respiratory abnormalities (2.9%), acute renal failure (0.8%), and cardiac complications (0.8%) were the most prevalent complications after TSA. The list of complications is detailed in Figure 2. Logistic regression analysis of patient characteristics predicting complications showed that advanced age (odds ratio [OR], 2.1 in those aged ≥85 years) and increasing number of comorbidities (≥3; OR, 3.5) were most significant in predicting complications (all P < 0.0001) (Figure 3). Despite the ubiquity of hypertension in this patient population, it was not a significant predictor of complication (OR, 0.9); in contrast, pulmonary disorders (OR, 5.1) and fluid and electrolyte disorders (4.0) were most strongly associated with the development of a postoperative complication after surgery (Figure 4).

EFFECT OF COMPLICATIONS ON LOS

The average length of hospitalization was 2.3 days (95% confidence interval, 2.22-2.25) among the entire cohort. The average LOS was longer in the complication cohort (3.9 days) than in patients who did not have a complication (2.1 days, P < 0.0001). Of the specific complications noted, hemodynamic shock (11.8 days); infectious, most commonly pneumonia (7.6 days); and vascular complications (6.9 days) were associated with the longest hospitalizations. This result is summarized in Figure 5.

MORTALITY

An overall postoperative (in-house) mortality rate of 0.07% was noted (weighted, n = 88). Comparison between the patient cohort that died vs those who survived TSA resulted in significant differences in the rates of complications. Complications that were most significantly different between the cohorts included cardiac (60.47% vs 0.75%, P < 0.0001), postoperative shock (26.61% vs 0.04%, P < 0.0001), and respiratory complications (43.1% vs 2.8%, P < 0.0001). It is important to note that the overall rate of postoperative shock was exceedingly low in the TSA cohort, but it was highly prevalent in the mortality cohort, occurring in 26.61% of patients. A summary of the mortality statistics is presented in Figure 6.

Continue to: DISCUSSION...

 

 

DISCUSSION

TSA continues to be associated with high levels of satisfaction;¹ as a result, its incidence is increasing.² As our understanding and efficiency improves nationally, it is imperative that we determine the short-term and longer-term outcomes and complications. In addition, the factors that may affect prognosis must be elucidated to provide a more individualized and effective standard of care. To date, most of the outcome studies of TSA have evaluated long-term outcomes and specific implant-related complications.^1,5,6,21,22 Our intent was to evaluate the complications that occur in the postoperative period and their effect on unique “patient care” outcomes. With knowledge of these complications and the predisposing factors, we can better assess patients, risk-stratify, and provide appropriate guidelines.

We noted that complications occurring after TSA are not uncommon, with >6% of patients suffering a postoperative complication. In this study, the number of complications noted was associated with worse patient outcomes. In addition, we noted that patients undergoing a TSA have a significant burden of comorbidities; however, hematologic and fluid disorders (eg, iron deficiency anemia, pulmonary circulatory disorders, and fluid imbalances) were most important in predicting postoperative complications.

Increased LOS in the hospital after TSA was associated with the occurrence of complications. Of all noted complications, shock and infectious and vascular complications led to the longest hospitalizations. Hospital-acquired pneumonia was the most common infectious etiology, while pulmonary embolism and deep vein thrombosis were the most consistent vascular complications. Although seldom studied in the TSA population, a similar finding has been noted in patients after THA. O’Malley and colleagues,²³ using the American College of Surgeon’s National Surgical Quality Improvement Program database, identified independent factors that were associated with complications and average prolonged LOS. They noted that the occurrence of major complications was associated with a prolonged LOS. Some, but not all the major complications, included organ space infection, cardiac events, pneumonia, and venous thromboembolic events.²³ Therefore, attempts to limit the amount of time spent in hospitals and control the associated costs must focus on managing the incidence of complications.

Postoperative mortality after TSA was uncommon, occurring in 0.07% of the patients in this study. The low incidence of mortality noted in this study is probably related to the fact that our data represent mortality, whereas in the hospital and, unlike most mortality studies, it does not account for patient demise that may occur in the months after surgery. Other reports have noted that mortality occurs in <1.5% of these patients.^24-28 Singh and colleagues²⁵ observed in their evaluation of perioperative mortality after TSA a mortality rate of 0.8% with 90 days after 4380 shoulder replacements performed at their institution. Using multivariate analysis, they were able to identify associations between mortality and increasing American Society of Anesthesiology (ASA) class and Charlson Comorbidity Index. These results in relation to ours would indicate that the majority of patients who die after shoulder arthroplasty do so after initial discharge. Although we could not determine a causal relationship between mortality and patient comorbidities, we noted that certain complications strongly correlated with mortality. In patients who died, there was a relatively high incidence of cardiac (60.5%) and respiratory (43.1%) complications. Similarly, although postoperative shock was almost nonexistent in the patients who survived surgery (0.04%), it was much more common in the patients who suffered mortality (26.6%).

This study is not without limitations. Data were extracted from a national database, therefore precluding the inclusion of specific details of surgery and functional assessment. Inherent to ICD-9 coding, we were unable to assess the exact detail and severity of complications. For instance, we cannot be certain what criteria were used to define “acute renal failure” for each patient. This study is retrospective in nature and therefore adequate randomization and standardization of patients is not possible. Similarly, the nature of the database may not allow for exacting our inclusion and exclusion criteria. However, the large sample size of the patient population lessens the chance of potential biases and type 2 errors. Prior to October 2010, reverse shoulder arthroplasty was coded under the ICD-9procedural code 81.80 as TSA. Therefore, there is some overlap between TSA and reverse shoulder arthroplasty in our data. Reverse shoulder arthroplasty is now coded under ICD-9 procedural code 81.88. It is possible that results may differ if reverse shoulder arthroplasty were excluded from our patient cohort. This can be an area of future research.

CONCLUSION

Although much is known about the long-term hardware and functional complications after TSA, in this study, we have attempted to broaden the understanding of perioperative complications and the associated sequelae. Complications are common after TSA surgery and are related to adverse outcomes. In the setting of healthcare changes, the surgeon and the patient must understand the cause, types, incidence, and outcomes of medical and surgical complications after surgery. This allows for more accurate “standard of care” metrics. Further large-volume multicenter studies are needed to gain further insight into the short- and long-term outcomes of TSA.

ABSTRACT

There is a paucity of evidence describing the types and rates of postoperative complications following total shoulder arthroplasty (TSA). We sought to analyze the complications following TSA and determine their effects on described outcome measures.

Using discharge data from the weighted Nationwide Inpatient Sample from 2006 to 2010, patients who underwent primary TSA were identified. The prevalence of specific complications was identified using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes. The data from this database represent events occurring during admission, prior to discharge. The associations between patient characteristics, complications, and outcomes of TSA were evaluated. The specific outcomes analyzed in this study were mortality and length of stay (LOS).

A total of 125,766 patients were identified. The rate of complication after TSA was 6.7% (8457 patients). The most frequent complications were respiratory, renal, and cardiac, occurring in 2.9%, 0.8%, and 0.8% of cases, respectively. Increasing age and total number of preoperative comorbidities significantly increased the likelihood of having a complication. The prevalence of postoperative shock and central nervous system, cardiac, vascular, and respiratory complications was significantly higher in patients who suffered postoperative mortality (88 patients; 0.07% mortality rate) than in those who survived surgery (P < 0.0001). In terms of LOS, shock and infectious and vascular complications most significantly increased the length of hospitalization.

Postoperative complications following TSA are not uncommon and occur in >6% of patients. Older patients and certain comorbidities are associated with complications after surgery. These complications are associated with postoperative mortality and increased LOS.

Continue to: Total shoulder arthroplasty...

Total shoulder arthroplasty (TSA) provides a predictably high level of satisfaction with survival as high as 92% at 15 years.¹ As implant instrumentation and surgical technique and understanding have improved, the frequency of TSAs being performed has also increased.² Although there are enough data on long-term surgical complications following TSA,^1,3-6 there is a paucity of evidence delineating the incidence and types of postoperative complications during hospitalization. Several current issues motivate the improved understanding of TSA, including the increasing number of TSAs being performed, the desire to improve quality of care, and the desire to create financially efficient healthcare.

The purpose of this study is to detail the postoperative complications that occur following TSA using a large national database. Specifically, our goals are to determine the incidence and types of complications after shoulder arthroplasty, determine the patient factors that are associated with these complications, and evaluate the effects of these complications on postoperative in-hospital mortality and length of stay (LOS). Our hypothesis is that there would be a correlation between specific patient factors and complications and that these complications would adversely correlate to patient postoperative outcomes.

METHODS

DESIGN

We conducted a retrospective analysis of TSAs captured by the Nationwide Inpatient Sample (NIS) database between 2006 and 2010. The NIS is the largest all-payer inpatient database that is currently available to the public in the United States.⁷

The NIS is a part of the Healthcare Cost and Utilization Project funded by the Agency for Healthcare Research and Quality (AHRQ) and the US Department of Health and Human Services. The NIS database is designed to approximate a 20% sample of US hospitals and the patients they serve, including community, academic, general, and specialty-specific hospitals such as orthopedic hospitals.⁷ The 2010 update of the NIS database contains discharge data from 1051 hospitals across 45 states, with a representative sample of >39 million inpatient hospital stays.⁷ The NIS database and its data sources have been independently validated and assessed for quality each year since 1988.⁸Furthermore, comparative analysis of multiple database elements and distributions has been validated against standard norms, including the National Hospital Discharge Survey.⁹ The NIS database has been used in numerous published studies.^2,10,11

PATIENT SELECTION

The yearly NIS databases from 2006 to 2010 were compiled. Patients aged ≥40 years who underwent a TSA were identified using the International Classification of Diseases, 9th Revision (ICD-9), procedural code 81.80. Exclusion criteria were patients with a primary or a secondary diagnosis of humeral or scapular fracture, chronic osteomyelitis, rheumatologic diseases, or evidence of concurrent malignancy (Figure 1).

Native to NIS are patient demographics, including age, sex, and race. Patient comorbidities as described by Elixhauser and colleagues¹² are also included in the database.

Continue to: OUTCOMES...

 

 

OUTCOMES

The primary outcome of this study was a description of the type and frequency of postoperative complications of TSA. To conduct this analysis, we queried the TSA cohort for specific ICD-9 codes representing acute cardiac, central nervous system, infectious, gastrointestinal, genitourinary, postoperative shock, renal, respiratory, surgical, vascular, and wound complications. The ICD-9 codes used to identify complications were modeled according to previous literature on various surgical applications and were further parsed to reflect only acute postoperative diagnoses^13-15(see the Appendix for the comprehensive list of ICD-9 codes).

Two additional outcomes were analyzed, including postoperative mortality and LOS. Postoperative mortality was defined as death occurring prior to discharge. We calculated the average LOS among the complication and the noncomplication cohort.

STATISTICAL ANALYSIS

Patient demographics and target outcomes of the study were analyzed by frequency distribution. Where applicable, the chi-square and the Student’s t tests were used to confirm the statistical difference for dichotomous and continuous variables, respectively. Multivariate regressions were performed after controlling for possible clustering of the data using a generalized estimating equation following a previous analytical methodology.^16-20 The results are reported with odds ratios and 95% confidence intervals where applicable, all statistical tests with P ≤ 0.05 were considered to be significant, and all statistical tests were two-sided. We conducted all analyses using SAS, version 9.2 (SAS Institute).

RESULTS

From 2006 to 2010, a weighted sample of 141,973 patients was found to undergo a TSA. After applying our inclusion and exclusion criteria, our study cohort consisted of 125,766 patients (Figure 1).

Continue to: OVERALL TSA COHORT DEMOGRAPHICS...

 

 

OVERALL TSA COHORT DEMOGRAPHICS

The average age of the TSA cohort was 69.4 years (standard deviation [SD], 21.20), and 54.1% were females. The cohort had significant comorbidities, with 83.3% of them having at least 1 comorbidity at the time of surgery. Specifically, 31.3% of the patients had 1 comorbidity, 26.5% had 2 comorbidities, and 25.4% had ≥3 comorbidities. Hypertension was the most common comorbidity present in 66.2% of patients, and diabetes was the second most common comorbidity with a prevalence of 16.8%.

COMPLICATION COHORT DEMOGRAPHICS

An overall postoperative complication rate of 6.7% (weighted sample of 8457 patients) was noted in the overall TSA cohort. The TSA cohort was dichotomized into patients who suffered at least 1 complication (weighted, n = 8457) and patients undergoing routine TSAs (weighted, n = 117,308). The average age was significantly higher in the complication vs routine cohort (71.38 vs 69.27 years, P < 0.0001). Similarly, there were significantly more comorbidities (2.51 vs 1.71, P < 0.0001) in the complication cohort.

COMPLICATIONS

We noted a complication rate of 6.7% (weighted sample of 8457 patients). A single complication was noted in 5% of these patients, whereas 1.3% and 0.4% of the patients had 2 and ≥3 complications, respectively. Respiratory abnormalities (2.9%), acute renal failure (0.8%), and cardiac complications (0.8%) were the most prevalent complications after TSA. The list of complications is detailed in Figure 2. Logistic regression analysis of patient characteristics predicting complications showed that advanced age (odds ratio [OR], 2.1 in those aged ≥85 years) and increasing number of comorbidities (≥3; OR, 3.5) were most significant in predicting complications (all P < 0.0001) (Figure 3). Despite the ubiquity of hypertension in this patient population, it was not a significant predictor of complication (OR, 0.9); in contrast, pulmonary disorders (OR, 5.1) and fluid and electrolyte disorders (4.0) were most strongly associated with the development of a postoperative complication after surgery (Figure 4).

EFFECT OF COMPLICATIONS ON LOS

The average length of hospitalization was 2.3 days (95% confidence interval, 2.22-2.25) among the entire cohort. The average LOS was longer in the complication cohort (3.9 days) than in patients who did not have a complication (2.1 days, P < 0.0001). Of the specific complications noted, hemodynamic shock (11.8 days); infectious, most commonly pneumonia (7.6 days); and vascular complications (6.9 days) were associated with the longest hospitalizations. This result is summarized in Figure 5.

MORTALITY

An overall postoperative (in-house) mortality rate of 0.07% was noted (weighted, n = 88). Comparison between the patient cohort that died vs those who survived TSA resulted in significant differences in the rates of complications. Complications that were most significantly different between the cohorts included cardiac (60.47% vs 0.75%, P < 0.0001), postoperative shock (26.61% vs 0.04%, P < 0.0001), and respiratory complications (43.1% vs 2.8%, P < 0.0001). It is important to note that the overall rate of postoperative shock was exceedingly low in the TSA cohort, but it was highly prevalent in the mortality cohort, occurring in 26.61% of patients. A summary of the mortality statistics is presented in Figure 6.

Continue to: DISCUSSION...

 

 

DISCUSSION

TSA continues to be associated with high levels of satisfaction;¹ as a result, its incidence is increasing.² As our understanding and efficiency improves nationally, it is imperative that we determine the short-term and longer-term outcomes and complications. In addition, the factors that may affect prognosis must be elucidated to provide a more individualized and effective standard of care. To date, most of the outcome studies of TSA have evaluated long-term outcomes and specific implant-related complications.^1,5,6,21,22 Our intent was to evaluate the complications that occur in the postoperative period and their effect on unique “patient care” outcomes. With knowledge of these complications and the predisposing factors, we can better assess patients, risk-stratify, and provide appropriate guidelines.

We noted that complications occurring after TSA are not uncommon, with >6% of patients suffering a postoperative complication. In this study, the number of complications noted was associated with worse patient outcomes. In addition, we noted that patients undergoing a TSA have a significant burden of comorbidities; however, hematologic and fluid disorders (eg, iron deficiency anemia, pulmonary circulatory disorders, and fluid imbalances) were most important in predicting postoperative complications.

Increased LOS in the hospital after TSA was associated with the occurrence of complications. Of all noted complications, shock and infectious and vascular complications led to the longest hospitalizations. Hospital-acquired pneumonia was the most common infectious etiology, while pulmonary embolism and deep vein thrombosis were the most consistent vascular complications. Although seldom studied in the TSA population, a similar finding has been noted in patients after THA. O’Malley and colleagues,²³ using the American College of Surgeon’s National Surgical Quality Improvement Program database, identified independent factors that were associated with complications and average prolonged LOS. They noted that the occurrence of major complications was associated with a prolonged LOS. Some, but not all the major complications, included organ space infection, cardiac events, pneumonia, and venous thromboembolic events.²³ Therefore, attempts to limit the amount of time spent in hospitals and control the associated costs must focus on managing the incidence of complications.

Postoperative mortality after TSA was uncommon, occurring in 0.07% of the patients in this study. The low incidence of mortality noted in this study is probably related to the fact that our data represent mortality, whereas in the hospital and, unlike most mortality studies, it does not account for patient demise that may occur in the months after surgery. Other reports have noted that mortality occurs in <1.5% of these patients.^24-28 Singh and colleagues²⁵ observed in their evaluation of perioperative mortality after TSA a mortality rate of 0.8% with 90 days after 4380 shoulder replacements performed at their institution. Using multivariate analysis, they were able to identify associations between mortality and increasing American Society of Anesthesiology (ASA) class and Charlson Comorbidity Index. These results in relation to ours would indicate that the majority of patients who die after shoulder arthroplasty do so after initial discharge. Although we could not determine a causal relationship between mortality and patient comorbidities, we noted that certain complications strongly correlated with mortality. In patients who died, there was a relatively high incidence of cardiac (60.5%) and respiratory (43.1%) complications. Similarly, although postoperative shock was almost nonexistent in the patients who survived surgery (0.04%), it was much more common in the patients who suffered mortality (26.6%).

This study is not without limitations. Data were extracted from a national database, therefore precluding the inclusion of specific details of surgery and functional assessment. Inherent to ICD-9 coding, we were unable to assess the exact detail and severity of complications. For instance, we cannot be certain what criteria were used to define “acute renal failure” for each patient. This study is retrospective in nature and therefore adequate randomization and standardization of patients is not possible. Similarly, the nature of the database may not allow for exacting our inclusion and exclusion criteria. However, the large sample size of the patient population lessens the chance of potential biases and type 2 errors. Prior to October 2010, reverse shoulder arthroplasty was coded under the ICD-9procedural code 81.80 as TSA. Therefore, there is some overlap between TSA and reverse shoulder arthroplasty in our data. Reverse shoulder arthroplasty is now coded under ICD-9 procedural code 81.88. It is possible that results may differ if reverse shoulder arthroplasty were excluded from our patient cohort. This can be an area of future research.

CONCLUSION

Although much is known about the long-term hardware and functional complications after TSA, in this study, we have attempted to broaden the understanding of perioperative complications and the associated sequelae. Complications are common after TSA surgery and are related to adverse outcomes. In the setting of healthcare changes, the surgeon and the patient must understand the cause, types, incidence, and outcomes of medical and surgical complications after surgery. This allows for more accurate “standard of care” metrics. Further large-volume multicenter studies are needed to gain further insight into the short- and long-term outcomes of TSA.

The Centers for Disease Control and Prevention lists cardiovascular-related diseases as a leading cause of mortality.¹ The medication class of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors, more commonly known as statins, is first-line therapy to prevent negative cardiovascular outcomes and reduce premature death.² Additional hyperlipidemia medications, such as fish oil, can be added for potential cardiovascular benefit.

Yokoyama and colleagues demonstrated that fish oil is a promising treatment for the prevention of major coronary events in patients with hypercholesterolemia.³ Furthermore, Macchia and colleagues found reductions in cardiovascular outcomes and all-cause mortality in postmyocardial infarction patients treated with fish oil and statin combination therapy.⁴ In contrast, the Outcomes Reduction with an Initial Glargine Intervention (ORIGIN) trial found glucose intolerant and patients with diabetes mellitus did not have improved cardiovascular outcomes with fish oil therapy.⁵ Likewise The Risk and Prevention Study Collaborative Group found fish oil supplementation provided no benefit for primary prevention in patients with multiple cardiovascular risk factors.⁶ These studies demonstrate fish oil therapy can cause diverse cardiovascular outcomes in different patient populations.

Currently, there are no studies examining the impact of fish oil and statin combination therapy on the US veteran population. The research of Yokoyama, Macchia, and The Risk and Prevention Study Collaborative Group took place in Japanese and Italian populations, which impacts their external validity.^3,4,6 Furthermore, these studies had higher rates of female subjects when compared with the US veteran population. For example, 68% of female subjects in the Yokoyama study received fish oil therapy.³ Also, the ORIGIN trial subjects were restricted to patients with diabetes mellitus or who were glucose intolerant, which is not reflective of the entire veteran population.⁵ These differences can make it difficult to define the role of fish oil and statin combination therapy in treating cardiovascular disease and reducing mortality in the veteran population.

This study aims to help the US Department of Veterans Affairs (VA) primary care providers and clinical pharmacists address the role of fish oil and statin combination therapy in the prevention of cardiovascular disease and all-cause mortality in the veteran population. The addition of fish oil to statin therapy was compared with an established standard of care, statin monotherapy, in veterans at the Fargo Veterans Affairs Health Care System (FVAHCS).

Methods

A retrospective chart review was conducted using the FVAHCS Computerized Patient Record System (CPRS). The institution’s review board and VA medical center approved the study. Eligible veterans with prescriptions for fish oil or statin therapy between January 1, 2000 and September 30, 2015 were randomly selected, reviewed, and sorted based on inclusion and exclusion criteria. 

The primary outcome was time to aggregate cardiovascular events, specifically myocardial infarction (MI), stroke, transient ischemic attack, coronary artery bypass graft, and percutaneous intervention. Adverse cardiovascular event data were obtained from the veterans’ International Classification of Disease (ICD) 9 codes. Furthermore, the secondary outcome—time to all-cause mortality—was gathered by death records in CPRS. Time to these events was compared in veterans on fish oil and statin combination therapy or statin monotherapy. The date of the cardiovascular event or death was recorded for each outcome and was obtained by reviewing provider notes that documented the incidence. If a specific day or month of incidence was not documented, July 1 was selected as the default date for the adverse cardiovascular event.

Demographics, medication adherence, diagnoses, lab values within 90 days of initiation of therapy, and primary and secondary outcomes were collected. Demographics that included age, race, and sex all were obtained via chart review. Diagnoses were gathered using ICD 9 codes. Refill history was retrieved to assess adherence. Adherence was calculated by the total days of medication therapy divided by the total days within the study. Total days in the study was calculated by the duration of therapy days between therapy initiation date and a terminating factor. Terminating factors included an adverse cardiovascular event, death, or the study termination date.

Statistics

Demographic and other cohort characterization variables were compared either by a t test, rank sum test, or Fisher exact test given the character of the variable. Kaplan Meier analysis was used to evaluate time to aggregate cardiovascular events and all-cause mortality. VA Informatics and Computing Infrastructure (VINCI) R 3.4.3 was used for data analysis. One of the few combination studies by Macchia and colleagues gave an estimate of unadjusted incident rates for patients treated with statin monotherapy vs fish oil and statin combination therapy in patients having a recent MI.⁴ Based on this information, a power analysis determined that a 2% difference in incidence rate of adverse cardiovascular events could be detected between the 2 cohorts with 1,000 veterans in the statin cohort, and 500 veterans in the fish oil and statin cohort assuming a time to event interval of about 7.5 years. An α level of 0.05 was set to determine statistical significance.

Results

A total of 3,940 veterans with prescriptions for fish oil or statin therapy were randomly reviewed and sorted based on inclusion and exclusion criteria. This inclusion criteria resulted in 2,575 fish oil and statin combination patients and 1,365 statin monotherapy patients. Exclusion criteria produced a final total of 437 fish oil and statin combination patients and 559 statin monotherapy patients. Patient demographics are presented in Table 3. 

All baseline laboratory data were collected within 90 days of therapy initiation (Table 4). Statin monotherapy patients had lower triglyceride levels compared with those of the fish oil and statin combination patients. However, both high-density lipoprotein (HDL) and low-density lipoprotein (LDL) levels were higher in the statin monotherapy patients. As seen in Table 5, diagnosis of heart failure, hypertension, hypothyroidism, and dyslipidemia were higher in the statin monotherapy cohort, while tobacco use and pancreatitis were more prevalent in the fish oil and statin combination cohort. 

Kaplan Meier curves of the primary outcome, time to aggregate adverse cardiovascular event, and the secondary outcome, time to all-cause mortality are shown in Figure 1 and Figure 2, respectively.  This shows adverse cardiovascular events and all-cause mortality for approximately 4,500 days for the fish oil and statin cohort and approximately 6,000 days for the statin monotherapy cohort.

Discussion

Analysis of this study failed to detect a statistically significant difference for time to aggregate adverse cardiovascular events or all-cause mortality. This may be due to fewer adverse cardiovascular events and mortality in the 2 cohorts than was anticipated.

There are no studies examining fish oil and statin therapy in the veteran population and only limited studies comparing statin and fish oil combination therapy vs statin monotherapy for adverse cardiovascular outcomes and all-cause mortality. One of the few comparison studies was by Macchia and colleagues and consisted of 7,924 post MI patients in Italy. Over a 4-year period, researchers found a slight improvement in the adjusted paired-matched population for all-cause mortality in the fish oil and statin therapy cohort vs statin monotherapy (8.6% vs 13.6% P < .001).³ A benefit also was seen in the fish oil and statin cohort vs statin monotherapy in the adjusted paired-matched population for death or stroke (16.7% vs 11.5% P < .001).³

In contrast, this study did not address postmyocardial infarction patients exclusively. Rather, patients in this study had lower morbidity, which resulted in fewer adverse cardiovascular outcomes and a greater difficulty to detect a difference in this healthier population. These healthier patients may derive less benefit from primary or secondary prevention with statin and fish oil combination therapy.

In this study, there were extensive inclusion and exclusion criteria to assess the relationship between the cohorts for adverse cardiovascular events caused by atherosclerotic disease. Veterans were required to take fish oil and statin therapy or statin monotherapy for at least 1 year. Other literature has only examined clinical impact on adverse cardiovascular event outcomes if therapy was a year or longer.⁷ Therefore, to prevent confounders from other medications, veterans who used any hyperlipidemia agent other than fish oil and statin therapy for longer than 1 year were excluded. Extensive exclusion criteria eliminated many veterans. However, the robust exclusion of clotting disorders, arrhythmias, chronic anticoagulation other than aspirin, hormonal medication use, or illegal substance abuse prevented the potential confounder of nonatherosclerotic adverse cardiovascular events, for example, a stroke due to poorly controlled atrial fibrillation.

Comparison of demographic data showed both cohorts were of similar age, sex, and race. Of note the Fargo veteran population was primarily white (> 80% in both cohorts). This is slightly higher than the percentage of whites for all US veterans. The slight difference most likely had a minimal clinical impact. Laboratory values recorded within 90 days of initiation of therapy were largely clinically similar except for triglycerides being significantly higher in the fish oil and statin combination cohort (Table 4). This may reflect selection bias, where providers may be more likely to add fish oil therapy for the potential to further control triglycerides.

Diagnoses of hypertension, heart failure, and dyslipidemia were higher in the statin monotherapy cohort. However, body mass index, tobacco use, and pancreatitis were statistically higher in fish oil and statin combination cohort. Even though there was a statistically significant difference in disease diagnoses, this likely created a minor clinical difference between the groups. This is further illustrated by the similarity of the Charlson Comorbidity Index of 1.6 for fish oil and statin cohort and 1.4 in statin monotherapy cohort.

Strengths

A strength of this study was its adherence rates. Adherence rates were high in both cohorts (Table 6). Fish oil and statin cohort did have slightly lower adherence compared with that of statin monotherapy. This may demonstrate extra pill burden influencing adherence. Overall demographics, laboratory values, disease, and adherence rates were clinically similar in both cohorts, thus reducing the potential for confounders.

Limitations

Limitations of this study include its retrospective chart review design. This design is susceptible to incorrect recording of events. The primary outcome, aggregate adverse cardiovascular events, may have been incorrectly recorded in the medical record as other diseases, such as coronary artery disease or heart disease, and therefore not captured by ICD 9 code retrieval. Also, important information, such as laboratory data, disease, and medication adherence, may not have been documented for all patients. Of note 1 patient in the fish oil and statin combination cohort did not have any recorded laboratory data, disease, or adherence data.

Another limitation is lack of access to medical notes from non-VA providers, which can result in missed data collection. To reduce this limitation, the study excluded veterans that received non-VA fish oil, statins, or other hyperlipidemia medications for > 1 year. Veterans were included only if they used VA-provided fish oil or statins. This inclusion and exclusion criteria reduced the chance of missing data from other facilities because it favored inclusion of only subjects that received care exclusively through the VA.

Last, on study initiation it was not realized that fish oil was not provided by the health care system until about the year 2004. This resulted in less risk days for the fish oil and statin cohort. However, Kaplan Meier analysis lessens this issue from being a confounder. Time to event rates for both the primary and secondary outcomes were similar and most likely would have continued to trend together with the same therapy duration.

Conclusion

Fish oil and statin combination therapy when compared with statin monotherapy failed to show that a statistically significant difference exists in the rates of MI, stroke, transient ischemic attack, coronary artery bypass graft, and percutaneous intervention. The clinical difference of fish oil and statin combination therapy vs statin monotherapy is most likely small or nonexistent. From our literature search, this is the only study concerning the use of fish oil and statin combination therapy in the veteran population. It is most likely that fish oil and statin combination therapy and statin monotherapy are similar for the reduction of time to aggregate adverse cardiovascular events and all-cause mortality in the veteran population.

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Fargo VA Healthcare System.

The Centers for Disease Control and Prevention lists cardiovascular-related diseases as a leading cause of mortality.¹ The medication class of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors, more commonly known as statins, is first-line therapy to prevent negative cardiovascular outcomes and reduce premature death.² Additional hyperlipidemia medications, such as fish oil, can be added for potential cardiovascular benefit.

Yokoyama and colleagues demonstrated that fish oil is a promising treatment for the prevention of major coronary events in patients with hypercholesterolemia.³ Furthermore, Macchia and colleagues found reductions in cardiovascular outcomes and all-cause mortality in postmyocardial infarction patients treated with fish oil and statin combination therapy.⁴ In contrast, the Outcomes Reduction with an Initial Glargine Intervention (ORIGIN) trial found glucose intolerant and patients with diabetes mellitus did not have improved cardiovascular outcomes with fish oil therapy.⁵ Likewise The Risk and Prevention Study Collaborative Group found fish oil supplementation provided no benefit for primary prevention in patients with multiple cardiovascular risk factors.⁶ These studies demonstrate fish oil therapy can cause diverse cardiovascular outcomes in different patient populations.

Currently, there are no studies examining the impact of fish oil and statin combination therapy on the US veteran population. The research of Yokoyama, Macchia, and The Risk and Prevention Study Collaborative Group took place in Japanese and Italian populations, which impacts their external validity.^3,4,6 Furthermore, these studies had higher rates of female subjects when compared with the US veteran population. For example, 68% of female subjects in the Yokoyama study received fish oil therapy.³ Also, the ORIGIN trial subjects were restricted to patients with diabetes mellitus or who were glucose intolerant, which is not reflective of the entire veteran population.⁵ These differences can make it difficult to define the role of fish oil and statin combination therapy in treating cardiovascular disease and reducing mortality in the veteran population.

This study aims to help the US Department of Veterans Affairs (VA) primary care providers and clinical pharmacists address the role of fish oil and statin combination therapy in the prevention of cardiovascular disease and all-cause mortality in the veteran population. The addition of fish oil to statin therapy was compared with an established standard of care, statin monotherapy, in veterans at the Fargo Veterans Affairs Health Care System (FVAHCS).

Methods

A retrospective chart review was conducted using the FVAHCS Computerized Patient Record System (CPRS). The institution’s review board and VA medical center approved the study. Eligible veterans with prescriptions for fish oil or statin therapy between January 1, 2000 and September 30, 2015 were randomly selected, reviewed, and sorted based on inclusion and exclusion criteria. 

The primary outcome was time to aggregate cardiovascular events, specifically myocardial infarction (MI), stroke, transient ischemic attack, coronary artery bypass graft, and percutaneous intervention. Adverse cardiovascular event data were obtained from the veterans’ International Classification of Disease (ICD) 9 codes. Furthermore, the secondary outcome—time to all-cause mortality—was gathered by death records in CPRS. Time to these events was compared in veterans on fish oil and statin combination therapy or statin monotherapy. The date of the cardiovascular event or death was recorded for each outcome and was obtained by reviewing provider notes that documented the incidence. If a specific day or month of incidence was not documented, July 1 was selected as the default date for the adverse cardiovascular event.

Demographics, medication adherence, diagnoses, lab values within 90 days of initiation of therapy, and primary and secondary outcomes were collected. Demographics that included age, race, and sex all were obtained via chart review. Diagnoses were gathered using ICD 9 codes. Refill history was retrieved to assess adherence. Adherence was calculated by the total days of medication therapy divided by the total days within the study. Total days in the study was calculated by the duration of therapy days between therapy initiation date and a terminating factor. Terminating factors included an adverse cardiovascular event, death, or the study termination date.

Statistics

Demographic and other cohort characterization variables were compared either by a t test, rank sum test, or Fisher exact test given the character of the variable. Kaplan Meier analysis was used to evaluate time to aggregate cardiovascular events and all-cause mortality. VA Informatics and Computing Infrastructure (VINCI) R 3.4.3 was used for data analysis. One of the few combination studies by Macchia and colleagues gave an estimate of unadjusted incident rates for patients treated with statin monotherapy vs fish oil and statin combination therapy in patients having a recent MI.⁴ Based on this information, a power analysis determined that a 2% difference in incidence rate of adverse cardiovascular events could be detected between the 2 cohorts with 1,000 veterans in the statin cohort, and 500 veterans in the fish oil and statin cohort assuming a time to event interval of about 7.5 years. An α level of 0.05 was set to determine statistical significance.

Results

A total of 3,940 veterans with prescriptions for fish oil or statin therapy were randomly reviewed and sorted based on inclusion and exclusion criteria. This inclusion criteria resulted in 2,575 fish oil and statin combination patients and 1,365 statin monotherapy patients. Exclusion criteria produced a final total of 437 fish oil and statin combination patients and 559 statin monotherapy patients. Patient demographics are presented in Table 3. 

All baseline laboratory data were collected within 90 days of therapy initiation (Table 4). Statin monotherapy patients had lower triglyceride levels compared with those of the fish oil and statin combination patients. However, both high-density lipoprotein (HDL) and low-density lipoprotein (LDL) levels were higher in the statin monotherapy patients. As seen in Table 5, diagnosis of heart failure, hypertension, hypothyroidism, and dyslipidemia were higher in the statin monotherapy cohort, while tobacco use and pancreatitis were more prevalent in the fish oil and statin combination cohort. 

Kaplan Meier curves of the primary outcome, time to aggregate adverse cardiovascular event, and the secondary outcome, time to all-cause mortality are shown in Figure 1 and Figure 2, respectively.  This shows adverse cardiovascular events and all-cause mortality for approximately 4,500 days for the fish oil and statin cohort and approximately 6,000 days for the statin monotherapy cohort.

Discussion

Analysis of this study failed to detect a statistically significant difference for time to aggregate adverse cardiovascular events or all-cause mortality. This may be due to fewer adverse cardiovascular events and mortality in the 2 cohorts than was anticipated.

There are no studies examining fish oil and statin therapy in the veteran population and only limited studies comparing statin and fish oil combination therapy vs statin monotherapy for adverse cardiovascular outcomes and all-cause mortality. One of the few comparison studies was by Macchia and colleagues and consisted of 7,924 post MI patients in Italy. Over a 4-year period, researchers found a slight improvement in the adjusted paired-matched population for all-cause mortality in the fish oil and statin therapy cohort vs statin monotherapy (8.6% vs 13.6% P < .001).³ A benefit also was seen in the fish oil and statin cohort vs statin monotherapy in the adjusted paired-matched population for death or stroke (16.7% vs 11.5% P < .001).³

In contrast, this study did not address postmyocardial infarction patients exclusively. Rather, patients in this study had lower morbidity, which resulted in fewer adverse cardiovascular outcomes and a greater difficulty to detect a difference in this healthier population. These healthier patients may derive less benefit from primary or secondary prevention with statin and fish oil combination therapy.

In this study, there were extensive inclusion and exclusion criteria to assess the relationship between the cohorts for adverse cardiovascular events caused by atherosclerotic disease. Veterans were required to take fish oil and statin therapy or statin monotherapy for at least 1 year. Other literature has only examined clinical impact on adverse cardiovascular event outcomes if therapy was a year or longer.⁷ Therefore, to prevent confounders from other medications, veterans who used any hyperlipidemia agent other than fish oil and statin therapy for longer than 1 year were excluded. Extensive exclusion criteria eliminated many veterans. However, the robust exclusion of clotting disorders, arrhythmias, chronic anticoagulation other than aspirin, hormonal medication use, or illegal substance abuse prevented the potential confounder of nonatherosclerotic adverse cardiovascular events, for example, a stroke due to poorly controlled atrial fibrillation.

Comparison of demographic data showed both cohorts were of similar age, sex, and race. Of note the Fargo veteran population was primarily white (> 80% in both cohorts). This is slightly higher than the percentage of whites for all US veterans. The slight difference most likely had a minimal clinical impact. Laboratory values recorded within 90 days of initiation of therapy were largely clinically similar except for triglycerides being significantly higher in the fish oil and statin combination cohort (Table 4). This may reflect selection bias, where providers may be more likely to add fish oil therapy for the potential to further control triglycerides.

Diagnoses of hypertension, heart failure, and dyslipidemia were higher in the statin monotherapy cohort. However, body mass index, tobacco use, and pancreatitis were statistically higher in fish oil and statin combination cohort. Even though there was a statistically significant difference in disease diagnoses, this likely created a minor clinical difference between the groups. This is further illustrated by the similarity of the Charlson Comorbidity Index of 1.6 for fish oil and statin cohort and 1.4 in statin monotherapy cohort.

Strengths

A strength of this study was its adherence rates. Adherence rates were high in both cohorts (Table 6). Fish oil and statin cohort did have slightly lower adherence compared with that of statin monotherapy. This may demonstrate extra pill burden influencing adherence. Overall demographics, laboratory values, disease, and adherence rates were clinically similar in both cohorts, thus reducing the potential for confounders.

Limitations

Limitations of this study include its retrospective chart review design. This design is susceptible to incorrect recording of events. The primary outcome, aggregate adverse cardiovascular events, may have been incorrectly recorded in the medical record as other diseases, such as coronary artery disease or heart disease, and therefore not captured by ICD 9 code retrieval. Also, important information, such as laboratory data, disease, and medication adherence, may not have been documented for all patients. Of note 1 patient in the fish oil and statin combination cohort did not have any recorded laboratory data, disease, or adherence data.

Another limitation is lack of access to medical notes from non-VA providers, which can result in missed data collection. To reduce this limitation, the study excluded veterans that received non-VA fish oil, statins, or other hyperlipidemia medications for > 1 year. Veterans were included only if they used VA-provided fish oil or statins. This inclusion and exclusion criteria reduced the chance of missing data from other facilities because it favored inclusion of only subjects that received care exclusively through the VA.

Last, on study initiation it was not realized that fish oil was not provided by the health care system until about the year 2004. This resulted in less risk days for the fish oil and statin cohort. However, Kaplan Meier analysis lessens this issue from being a confounder. Time to event rates for both the primary and secondary outcomes were similar and most likely would have continued to trend together with the same therapy duration.

Conclusion

Fish oil and statin combination therapy when compared with statin monotherapy failed to show that a statistically significant difference exists in the rates of MI, stroke, transient ischemic attack, coronary artery bypass graft, and percutaneous intervention. The clinical difference of fish oil and statin combination therapy vs statin monotherapy is most likely small or nonexistent. From our literature search, this is the only study concerning the use of fish oil and statin combination therapy in the veteran population. It is most likely that fish oil and statin combination therapy and statin monotherapy are similar for the reduction of time to aggregate adverse cardiovascular events and all-cause mortality in the veteran population.

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Fargo VA Healthcare System.

The Centers for Disease Control and Prevention lists cardiovascular-related diseases as a leading cause of mortality.¹ The medication class of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors, more commonly known as statins, is first-line therapy to prevent negative cardiovascular outcomes and reduce premature death.² Additional hyperlipidemia medications, such as fish oil, can be added for potential cardiovascular benefit.

Yokoyama and colleagues demonstrated that fish oil is a promising treatment for the prevention of major coronary events in patients with hypercholesterolemia.³ Furthermore, Macchia and colleagues found reductions in cardiovascular outcomes and all-cause mortality in postmyocardial infarction patients treated with fish oil and statin combination therapy.⁴ In contrast, the Outcomes Reduction with an Initial Glargine Intervention (ORIGIN) trial found glucose intolerant and patients with diabetes mellitus did not have improved cardiovascular outcomes with fish oil therapy.⁵ Likewise The Risk and Prevention Study Collaborative Group found fish oil supplementation provided no benefit for primary prevention in patients with multiple cardiovascular risk factors.⁶ These studies demonstrate fish oil therapy can cause diverse cardiovascular outcomes in different patient populations.

Currently, there are no studies examining the impact of fish oil and statin combination therapy on the US veteran population. The research of Yokoyama, Macchia, and The Risk and Prevention Study Collaborative Group took place in Japanese and Italian populations, which impacts their external validity.^3,4,6 Furthermore, these studies had higher rates of female subjects when compared with the US veteran population. For example, 68% of female subjects in the Yokoyama study received fish oil therapy.³ Also, the ORIGIN trial subjects were restricted to patients with diabetes mellitus or who were glucose intolerant, which is not reflective of the entire veteran population.⁵ These differences can make it difficult to define the role of fish oil and statin combination therapy in treating cardiovascular disease and reducing mortality in the veteran population.

This study aims to help the US Department of Veterans Affairs (VA) primary care providers and clinical pharmacists address the role of fish oil and statin combination therapy in the prevention of cardiovascular disease and all-cause mortality in the veteran population. The addition of fish oil to statin therapy was compared with an established standard of care, statin monotherapy, in veterans at the Fargo Veterans Affairs Health Care System (FVAHCS).

Methods

A retrospective chart review was conducted using the FVAHCS Computerized Patient Record System (CPRS). The institution’s review board and VA medical center approved the study. Eligible veterans with prescriptions for fish oil or statin therapy between January 1, 2000 and September 30, 2015 were randomly selected, reviewed, and sorted based on inclusion and exclusion criteria. 

The primary outcome was time to aggregate cardiovascular events, specifically myocardial infarction (MI), stroke, transient ischemic attack, coronary artery bypass graft, and percutaneous intervention. Adverse cardiovascular event data were obtained from the veterans’ International Classification of Disease (ICD) 9 codes. Furthermore, the secondary outcome—time to all-cause mortality—was gathered by death records in CPRS. Time to these events was compared in veterans on fish oil and statin combination therapy or statin monotherapy. The date of the cardiovascular event or death was recorded for each outcome and was obtained by reviewing provider notes that documented the incidence. If a specific day or month of incidence was not documented, July 1 was selected as the default date for the adverse cardiovascular event.

Demographics, medication adherence, diagnoses, lab values within 90 days of initiation of therapy, and primary and secondary outcomes were collected. Demographics that included age, race, and sex all were obtained via chart review. Diagnoses were gathered using ICD 9 codes. Refill history was retrieved to assess adherence. Adherence was calculated by the total days of medication therapy divided by the total days within the study. Total days in the study was calculated by the duration of therapy days between therapy initiation date and a terminating factor. Terminating factors included an adverse cardiovascular event, death, or the study termination date.

Statistics

Demographic and other cohort characterization variables were compared either by a t test, rank sum test, or Fisher exact test given the character of the variable. Kaplan Meier analysis was used to evaluate time to aggregate cardiovascular events and all-cause mortality. VA Informatics and Computing Infrastructure (VINCI) R 3.4.3 was used for data analysis. One of the few combination studies by Macchia and colleagues gave an estimate of unadjusted incident rates for patients treated with statin monotherapy vs fish oil and statin combination therapy in patients having a recent MI.⁴ Based on this information, a power analysis determined that a 2% difference in incidence rate of adverse cardiovascular events could be detected between the 2 cohorts with 1,000 veterans in the statin cohort, and 500 veterans in the fish oil and statin cohort assuming a time to event interval of about 7.5 years. An α level of 0.05 was set to determine statistical significance.

Results

A total of 3,940 veterans with prescriptions for fish oil or statin therapy were randomly reviewed and sorted based on inclusion and exclusion criteria. This inclusion criteria resulted in 2,575 fish oil and statin combination patients and 1,365 statin monotherapy patients. Exclusion criteria produced a final total of 437 fish oil and statin combination patients and 559 statin monotherapy patients. Patient demographics are presented in Table 3. 

All baseline laboratory data were collected within 90 days of therapy initiation (Table 4). Statin monotherapy patients had lower triglyceride levels compared with those of the fish oil and statin combination patients. However, both high-density lipoprotein (HDL) and low-density lipoprotein (LDL) levels were higher in the statin monotherapy patients. As seen in Table 5, diagnosis of heart failure, hypertension, hypothyroidism, and dyslipidemia were higher in the statin monotherapy cohort, while tobacco use and pancreatitis were more prevalent in the fish oil and statin combination cohort. 

Kaplan Meier curves of the primary outcome, time to aggregate adverse cardiovascular event, and the secondary outcome, time to all-cause mortality are shown in Figure 1 and Figure 2, respectively.  This shows adverse cardiovascular events and all-cause mortality for approximately 4,500 days for the fish oil and statin cohort and approximately 6,000 days for the statin monotherapy cohort.

Discussion

Analysis of this study failed to detect a statistically significant difference for time to aggregate adverse cardiovascular events or all-cause mortality. This may be due to fewer adverse cardiovascular events and mortality in the 2 cohorts than was anticipated.

There are no studies examining fish oil and statin therapy in the veteran population and only limited studies comparing statin and fish oil combination therapy vs statin monotherapy for adverse cardiovascular outcomes and all-cause mortality. One of the few comparison studies was by Macchia and colleagues and consisted of 7,924 post MI patients in Italy. Over a 4-year period, researchers found a slight improvement in the adjusted paired-matched population for all-cause mortality in the fish oil and statin therapy cohort vs statin monotherapy (8.6% vs 13.6% P < .001).³ A benefit also was seen in the fish oil and statin cohort vs statin monotherapy in the adjusted paired-matched population for death or stroke (16.7% vs 11.5% P < .001).³

In contrast, this study did not address postmyocardial infarction patients exclusively. Rather, patients in this study had lower morbidity, which resulted in fewer adverse cardiovascular outcomes and a greater difficulty to detect a difference in this healthier population. These healthier patients may derive less benefit from primary or secondary prevention with statin and fish oil combination therapy.

In this study, there were extensive inclusion and exclusion criteria to assess the relationship between the cohorts for adverse cardiovascular events caused by atherosclerotic disease. Veterans were required to take fish oil and statin therapy or statin monotherapy for at least 1 year. Other literature has only examined clinical impact on adverse cardiovascular event outcomes if therapy was a year or longer.⁷ Therefore, to prevent confounders from other medications, veterans who used any hyperlipidemia agent other than fish oil and statin therapy for longer than 1 year were excluded. Extensive exclusion criteria eliminated many veterans. However, the robust exclusion of clotting disorders, arrhythmias, chronic anticoagulation other than aspirin, hormonal medication use, or illegal substance abuse prevented the potential confounder of nonatherosclerotic adverse cardiovascular events, for example, a stroke due to poorly controlled atrial fibrillation.

Comparison of demographic data showed both cohorts were of similar age, sex, and race. Of note the Fargo veteran population was primarily white (> 80% in both cohorts). This is slightly higher than the percentage of whites for all US veterans. The slight difference most likely had a minimal clinical impact. Laboratory values recorded within 90 days of initiation of therapy were largely clinically similar except for triglycerides being significantly higher in the fish oil and statin combination cohort (Table 4). This may reflect selection bias, where providers may be more likely to add fish oil therapy for the potential to further control triglycerides.

Diagnoses of hypertension, heart failure, and dyslipidemia were higher in the statin monotherapy cohort. However, body mass index, tobacco use, and pancreatitis were statistically higher in fish oil and statin combination cohort. Even though there was a statistically significant difference in disease diagnoses, this likely created a minor clinical difference between the groups. This is further illustrated by the similarity of the Charlson Comorbidity Index of 1.6 for fish oil and statin cohort and 1.4 in statin monotherapy cohort.

Strengths

A strength of this study was its adherence rates. Adherence rates were high in both cohorts (Table 6). Fish oil and statin cohort did have slightly lower adherence compared with that of statin monotherapy. This may demonstrate extra pill burden influencing adherence. Overall demographics, laboratory values, disease, and adherence rates were clinically similar in both cohorts, thus reducing the potential for confounders.

Limitations

Limitations of this study include its retrospective chart review design. This design is susceptible to incorrect recording of events. The primary outcome, aggregate adverse cardiovascular events, may have been incorrectly recorded in the medical record as other diseases, such as coronary artery disease or heart disease, and therefore not captured by ICD 9 code retrieval. Also, important information, such as laboratory data, disease, and medication adherence, may not have been documented for all patients. Of note 1 patient in the fish oil and statin combination cohort did not have any recorded laboratory data, disease, or adherence data.

Another limitation is lack of access to medical notes from non-VA providers, which can result in missed data collection. To reduce this limitation, the study excluded veterans that received non-VA fish oil, statins, or other hyperlipidemia medications for > 1 year. Veterans were included only if they used VA-provided fish oil or statins. This inclusion and exclusion criteria reduced the chance of missing data from other facilities because it favored inclusion of only subjects that received care exclusively through the VA.

Last, on study initiation it was not realized that fish oil was not provided by the health care system until about the year 2004. This resulted in less risk days for the fish oil and statin cohort. However, Kaplan Meier analysis lessens this issue from being a confounder. Time to event rates for both the primary and secondary outcomes were similar and most likely would have continued to trend together with the same therapy duration.

Conclusion

Fish oil and statin combination therapy when compared with statin monotherapy failed to show that a statistically significant difference exists in the rates of MI, stroke, transient ischemic attack, coronary artery bypass graft, and percutaneous intervention. The clinical difference of fish oil and statin combination therapy vs statin monotherapy is most likely small or nonexistent. From our literature search, this is the only study concerning the use of fish oil and statin combination therapy in the veteran population. It is most likely that fish oil and statin combination therapy and statin monotherapy are similar for the reduction of time to aggregate adverse cardiovascular events and all-cause mortality in the veteran population.

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the Fargo VA Healthcare System.

ABSTRACT

In this study, we determine the in vivo effects of injecting sub-populations of leukocytes into normal rat Achilles tendons via a controlled laboratory study. Allogenic monocytes, granulocytes, or plasma were injected into 24 healthy rat Achilles tendons. Treated and contralateral un-treated control tendons then assessed for cellularity, histologic morphology, and vascularity after 7 and 14 days. Significant increases of 221% and 249% in cellularity (P = 0.014) were seen on day 14 within Achilles tendons injected with granulocytes as compared to plasma and monocytes, respectively. Also, significant improvement in morphology (P = 0.029) between days 7 and 14 was seen for the granulocyte injected Achilles tendons. Significant increases in cellularity after an injection of granulocytes, compared to monocytes and plasma, corresponds to a significant increase in inflammation within the tissue, suggesting that leukocyte-rich platelet-rich plasma (PRP) preparations are proinflammatory and potentially catabolic when injected into tendon tissue. The concentration and composition of white blood cells within PRP preparations is variable and needs to be better understood in order to optimize clinical utility of PRP injections.

Continue to: Tendinopathies are debilitating conditions...

Tendinopathies are debilitating conditions affecting patients worldwide every day. They arise most frequently from tendon overuse resulting in pathology.¹ There are 2 major subtypes of tendinopathy: tendinosis and tendinitis. Tendinosis, the more common condition, is characterized by long-term, chronic degradation of tendon tissue resulting in fibrosis from infiltrating fibroblasts.² Tendinitis, the less common condition, is characterized by an acute inflammatory response and inflammatory cell infiltrate.² Both conditions are common, with Achilles tendinopathy affecting 11% of runners and lateral epicondylitis affecting 1% to 3% of the general population.^3,4 Many sports-related overuse injuries, such as tendinopathies, go undiagnosed for extended periods of time because medical attention is avoided in order to prevent time loss from training or competing.⁵ These delays could be eliminated if a non-surgical option for treating tendon pathology was available.

Tendinopathies are believed to result from tendon overuse that causes micro-damage to collagen, as well as from significant changes in protein and enzyme composition within the tendon.⁶ The damage accumulates over time and eventually leads to chronic inflammation or fibrotic change within tendons, in both cases weakening the tendon and causing pain. Currently, accepted treatments for tendinopathies include: nonsteroidal anti-inflammatory drugs, physical therapy, ultrasound, laser-therapy, corticosteroids, glyceryl trinitrate patches, extracorporeal shock wave therapy, sclerotherapy, and surgery.⁷ Recently, platelet-rich plasma (PRP) therapy has emerged as a promising treatment for tendinopathies, as well as a variety of other orthopedic indications.

PRP consists of autologous blood from the patient, centrifuged to increase the amount of platelets in the sample above baseline, and subsequently injected around an affected tendon or joint.⁸ PRP is used to treat tendinopathy because it can supply injured tendons with blood components that aid in healing, which tendons do not receive due to poor vascularity.⁹ These components include growth factors, such as platelet derived growth factor (PDGF), transforming growth factor-β (TGF-β), vascular endothelial growth factor (VEGF), endothelial growth factor, and leukocytes that can stimulate an inflammatory response within the injured tissue.¹⁰ The inflammatory response from the PRP induces a more robust reconstruction and revascularization of the injured tissue, stimulating proliferation, and remodeling.^11,12However, significant variability exists within the platelets, leukocytes, and growth factors that comprise PRP. This is attributed to 3 major causes. First, current commercial preparations of PRP result in differing platelet concentrations, as well as leukocyte-rich and leukocyte-poor compositions.^13,14 Variability in platelet concentrations results in unreliable amounts of growth factors, including cytokines, TGF-β, PDGF, VEGF and basic fibroblast growth factor in each preparation, while leukocyte levels affect inflammation, all leading to variable effects for each preparation.^15,16Second, despite sex and age of the PRP donor not being significant factors influencing variation in growth factor concentrations, the existence of an unexplained variation in concentrations of growth factors between different donors has been observed.¹⁷ Third, the selection of activating agents, bovine thrombin or calcium chloride, and their application, whether to the elbow, shoulder, or knee, produces variability.¹⁸

While the effects of platelets and growth factors in PRP have been well studied, less is known about the effects of differing cell types. Recently it was reported that the concentrations of leukocytes directly affect the outcomes of PRP injections. McCarrel and colleagues^19,20 found that as the number of leukocytes increased, there was a concomitant increase in the expression of inflammatory cytokines and catabolic activity. This effect may result in inferior healing of injured tissues and is attributed to the release of pro-inflammatory cytokines such as interleukin-1β from the leukocytes.²¹ There is also evidence that minimizing the catabolic effect of leukocytes may be just as important to tissue healing as the maximizing anabolic effect of platelets and growth factors.²²

The use of PRP has been highly disputed in recent years due to conflicting reports of its success in treating orthopedic conditions. Numerous favorable studies have shown benefit for treating chronic and acute orthopedic injuries including; rotator cuff tear repair, chronic refractory patellar tendinopathy, and chronic lateral tendinosis/epicondylitis.^23-26 Concurrently, articles demonstrating no significant effects from PRP have also been published. One study claiming that PRP injections did not improve outcomes of chronic Achilles tendinopathy did not differentiate whether patients had tendinosis or tendinitis, and did not consider leukocyte concentration in their PRP preparations²⁷ Another study that determined PRP is not beneficial to the healing of ruptured Achilles tendons after surgical repair also failed to consider the concentration of leukocytes in their PRP preparations.²⁸ One of the difficulties in comparing these studies is their heterogeneous nature. This arises from the use of different conditions in each study that makes the studies incomparable. Variations in PRP preparations lead to different concentrations of growth factors, platelets, and leukocyte concentrations. Additionally, tendinopathy models were not specified as tendinosis and tendonitis, and models or patients were not controlled for age, sex, or comorbidities. Given that leukocyte-rich and leukocyte-poor PRP preparations are currently widely used in clinical practice, the discovery of which type of preparation is indicated in which setting is paramount to evidence-based use of this treatment modality. Due to reports suggesting that leukocytes may be detrimental to tendon healing, determining which types of leukocytes are responsible for these effects is vital. As such, the purpose of this study is to determine the in vivo effects of sub-populations of leukocytes on normal rat tendons. This study design allowed us to isolate the effects of the injections to induce a response and remove confounding effects of normal healing response to a damaged tendon and effects from the injection itself. Our hypothesis was that the injection of leukocytes would cause an inflammatory response in rat tendons, leading to catabolic outcomes.

Continue to: METHODS...

 

 

METHODS

This was a prospective, in vivo, placebo controlled, randomized animal study. The University’s Institutional Animal Care and Use Committee approved all procedures prior to initiation. Twenty-four male Sprague-Dawley rats were randomized to 3 treatment groups (n = 8): monocytes; granulocytes, and; plasma, as a negative control.

Allogenic blood from 6 additional rats was collected into K2EDTA tubes via cardiac puncture. Allogenic, as opposed to autogenic, blood is commonly used in rat models because of low immunogenic response to blood from rats of the same strain and litter.^29,30 The blood was then pooled and the red cells lysed by incubation with Red Blood Cell Lysis Buffer (Roche). The samples were then sorted into fractions containing monocytes and granulocytes using fluorescence activated cell sorting (FACS) using a FACSAria (BD Biosciences). Cells were sorted using Purified PE Mouse Anti-Rat CD11b/c antibodies (BD Pharmingen) specific to monocytes, granulocytes, macrophages, dendritic cells, and microglia, APC-Cy7 Mouse Anti-Rat CD45 antibodies (BD Pharmingen) specific to all hematopoietic cells except erythrocytes, and FITC Mouse Anti-Rat CD42d antibodies (BD Pharmingen) specific to megakaryocytes and platelets. 20 μL of 0.2 mg/mL CD11b/c, 20 μL of 0.2 mg/mL CD 45, and 10 μL of 0.5 mg/mL CD42d antibodies were added to 1 mL of condensed non-red cells collected from the 6 rats and incubated at room temperature in the dark for 15 minutes. A fraction containing only platelet-poor plasma was also collected. For all treatments the injection volume was 75 μL. Rats in the monocyte group were injected with 200,000 cells in platelet-poor plasma, those in the granulocyte group were injected with 900,000 cells in platelet-poor plasma, and rats in the plasma control group received only platelet-poor plasma. The cell concentrations were based on previous studies that documented these concentrations that are found in typical leukocyte-rich PRP preparations.¹³

The animals were anesthetized with isoflurane gas and then injected aseptically once into their right Achilles tendon. The left Achilles tendon was used as an un-injected control, giving a total of 48 total Achilles tendons studied. At days 7 and 14 post-injection, 4 rats from each group were euthanized and the Achilles tendons were harvested.

The tendons were fixed in neutral buffered formalin for 24 hours and then embedded in paraffin and sectioned sagittally at 12 μm. The tendons were then stained with hematoxylin and eosin (H&E) using standard histological protocols and examined by 3 individuals trained to assess cellularity and morphology. All samples were assigned unrecognizable numbers and randomized prior to examination by individuals. Cell counts were based on the number of nuclei present in 3 mid-tendon high-power fields (400x) per sample. Morphology was graded on a scale of 1 to 3, with 1 being a normal tendon and 3 having severe pathology with total loss of alignment and crimping on 3 low-power fields (100x) per sample (Figures 1A-1G).

Vascularity was assessed by immunohistochemical staining using Rabbit Polyclonal Anti-CD31 antibodies (Abcam), a marker for vascular endothelial cells, using a Vectastain ABC Kit (Vector Laboratories) system and the ImmPACT AEC Peroxidase (HRP) Substrate (Vector Laboratories). Following staining, automated image analysis was performed (Bioquant). Briefly, all areas that did not contain tendon were masked. CD31 positive areas were then quantified using global thresholding. Vascularity was then calculated as ratio of CD31 positive area to total tendon area. Analyses were performed on 3 mid-tendon medium-power (200x) fields per sample.

For cellularity and morphology, the results for the injected tendons were normalized to those of their contralateral untreated controls and reported as a percentage. Results for vascularity were compared directly between treated tendons. Differences were assessed between groups at each time-point using Independent Samples Median Tests. When significant differences were identified, pairwise comparisons were performed to identify the source of the differences. All analyses were conducted using SPSS (V22, SAS Institute) with significant differences determined for values of P < 0.05.

RESULTS

No significant differences in cellularity between groups were seen at day 7 (P = 0.368) (Figures 1A-1G). However, a significant difference in cellularity between groups was seen at day 14 (P = 0.014). Pairwise tests showed there to be a significant increase in the number of cells in the tendons treated with granulocytes from 221% and 249% in cellularity (P = 0.014) on day 14, as compared to both monocytes and plasma, respectively. Morphologically, no significant differences were seen between groups at either time-point (P = 0.091 for day 7 and P = 1.000 for day 14) (Figures 2A-2G). However, a significant improvement in morphology was observed from day 7 to day 14 in the granulocyte group from 60% to 165% (P = 0.029). Finally, no differences were seen in vascularity between treatment groups at either time-point (P = 0.368 for day 7 and P = 0.535 for day 14) (Figures 3A-3G).

Continue to: DISCUSSION...

 

 

DISCUSSION

Our hypothesis that the injection of leukocytes would cause an inflammatory response in rat tendons leading to catabolic outcomes was confirmed in the granulocyte group. It should be noted that prior to the catabolic outcome, there was a transient anabolic effect in the granulocyte group during the second week. Deterioration in morphology was observed in the tendons injected with granulocytes on day 7, which subsequently recovered in the following week. We found that injecting granulocytes into normal tendons resulted in an increase in inflammatory cellularity, when compared to monocytes and plasma injections. 

Limitations inherent in this study are those similar to other in vivo studies. To begin with, the results of injections into rat tendons may not be translatable to human tendons. Despite this limitation, the rat is a common model for tendon research.³¹ Another limitation is that this study injected healthy Achilles tendons, rather than tendons with preexisting tendinopathy. In a naturally occurring tendinopathy, there may be other factors present that interact with PRP, and this model negates the contribution of these factors. Finally, while the immunohistochemistry (IHC) and morphological data are clear, the cellularity data are not clear in identifying the type of cells that were increased by granulocyte injection.  However, the cells appeared rounded, resembling inflammatory infiltrate; a common cell type seen in tendons.² While fibroblasts are also a common infiltrate during chronic tendinopathy, they are generally flat and appear on H&E as long spindle shaped cells. Thus, we believe the increased cellularity of the tendons after granulocyte injections is representative of an increase in inflammation. The increased cellularity could be due to the increased number of cells injected into the tendon; however, our conclusions are consistent with the increased inflammation previously reported linking leukocytes to tendon inflammation.^20,22,32

In terms of morphology, we hypothesized that degenerative changes would be seen in the tendons that were injected with granulocytes due to the inflammatory action of these cells. As part of the granulocyte response, neutrophils release proteases and macrophages can stimulate collagen synthesis via fibroblasts, both causing change within the extracellular matrix.^33,34 Indeed, we observed a significant change in tissue morphology in the granulocyte group over the course of 14 days. As the degenerative and regenerative effects of granulocytes take time to present, this is likely what we observed to occur between day 7 and 14 after treatment. These observations are also consistent with prior observations that leukocyte-rich PRP injections can be detrimental to tendon healing, but beneficial to tissue degeneration in the setting of chronic tendonitis.²⁰

We hypothesized that the vascularity of the tendons would be similar in all preparations. This was based on previous studies demonstrating that the lack of platelets in the platelet-poor plasma fraction is sufficient to deplete VEGF, the angiogenic agent in PRP.³⁵ In this study, there were no observable differences in vascularity of platelet-poor plasma, monocyte, and granulocyte injections. We attribute this to the lack of VEGF in any of these preparations. The aforementioned study also showed that the lack of platelets in injection was enough to prevent the angiogenic effect of this treatment.³⁵

Continue to: The goal of this study was...

 

 

The goal of this study was to assess the morphology, cellularity, and vascularity of normal tendons after injections of different leukocyte populations. This is clinically important because of the potential to tailor future PRP injections on a patient-by-patient basis. In patients requiring an anabolic response, leukocyte-poor PRP may be the best option. In contrast, when patient pathology requires an inflammatory response to improve healing³⁶ or breakdown fibrotic tissue, as seen in tendinosis, leukocyte-rich PRP may be warranted. Further, properly controlled clinical studies are needed to validate these recommendations.

Limitations inherent in this study are those similar to other in vivo studies. First, the results of injections into rat tendons may not be translatable to human tendons. Despite this limitation, the rat is a common model for tendon research.³¹ A second limitation is that this study injected healthy Achilles tendons, rather than tendons with preexisting tendinopathy. In a naturally occurring tendinopathy, there may be other factors present that interact with PRP, and this model negates the contribution of these factors. Finally, while the IHC and morphological data show clear changes, the cellularity data are not clear in identifying the type of cells that were increased by granulocyte injection. However, the cells appeared rounded, resembling inflammatory infiltrate; a common cell type seen in tendons.² While fibroblasts are also a common infiltrate during chronic tendinopathy, they are generally flat and appear on H&E as long spindle shaped cells. The last limitation of this study is the lack of functional mechanical testing since, clinically, healing of the tendon is also related to the strength of the tendon. Thus, we believe the increased cellularity of the tendons after granulocyte injections is representative of an increase in inflammation. Moreover, our results are consistent with the increased inflammation previously reported linking leukocytes to tendon inflammation.^20,22,32 It is interesting to note that the increase in inflammation does not lead to an increase in vascularity as could be expected.

CONCLUSION

We found that the injection of leukocytes into healthy rat Achilles tendons increases inflammation, as evidenced by increased cellularity and disrupted morphology, which suggests that leukocyte-rich PRP preparations may be contraindicated in settings of acute tendonitis. However, these preparations may be useful for a specific subset of tendinopathies, including chronic tendinosis.

ABSTRACT

In this study, we determine the in vivo effects of injecting sub-populations of leukocytes into normal rat Achilles tendons via a controlled laboratory study. Allogenic monocytes, granulocytes, or plasma were injected into 24 healthy rat Achilles tendons. Treated and contralateral un-treated control tendons then assessed for cellularity, histologic morphology, and vascularity after 7 and 14 days. Significant increases of 221% and 249% in cellularity (P = 0.014) were seen on day 14 within Achilles tendons injected with granulocytes as compared to plasma and monocytes, respectively. Also, significant improvement in morphology (P = 0.029) between days 7 and 14 was seen for the granulocyte injected Achilles tendons. Significant increases in cellularity after an injection of granulocytes, compared to monocytes and plasma, corresponds to a significant increase in inflammation within the tissue, suggesting that leukocyte-rich platelet-rich plasma (PRP) preparations are proinflammatory and potentially catabolic when injected into tendon tissue. The concentration and composition of white blood cells within PRP preparations is variable and needs to be better understood in order to optimize clinical utility of PRP injections.

Continue to: Tendinopathies are debilitating conditions...

Tendinopathies are debilitating conditions affecting patients worldwide every day. They arise most frequently from tendon overuse resulting in pathology.¹ There are 2 major subtypes of tendinopathy: tendinosis and tendinitis. Tendinosis, the more common condition, is characterized by long-term, chronic degradation of tendon tissue resulting in fibrosis from infiltrating fibroblasts.² Tendinitis, the less common condition, is characterized by an acute inflammatory response and inflammatory cell infiltrate.² Both conditions are common, with Achilles tendinopathy affecting 11% of runners and lateral epicondylitis affecting 1% to 3% of the general population.^3,4 Many sports-related overuse injuries, such as tendinopathies, go undiagnosed for extended periods of time because medical attention is avoided in order to prevent time loss from training or competing.⁵ These delays could be eliminated if a non-surgical option for treating tendon pathology was available.

Tendinopathies are believed to result from tendon overuse that causes micro-damage to collagen, as well as from significant changes in protein and enzyme composition within the tendon.⁶ The damage accumulates over time and eventually leads to chronic inflammation or fibrotic change within tendons, in both cases weakening the tendon and causing pain. Currently, accepted treatments for tendinopathies include: nonsteroidal anti-inflammatory drugs, physical therapy, ultrasound, laser-therapy, corticosteroids, glyceryl trinitrate patches, extracorporeal shock wave therapy, sclerotherapy, and surgery.⁷ Recently, platelet-rich plasma (PRP) therapy has emerged as a promising treatment for tendinopathies, as well as a variety of other orthopedic indications.

PRP consists of autologous blood from the patient, centrifuged to increase the amount of platelets in the sample above baseline, and subsequently injected around an affected tendon or joint.⁸ PRP is used to treat tendinopathy because it can supply injured tendons with blood components that aid in healing, which tendons do not receive due to poor vascularity.⁹ These components include growth factors, such as platelet derived growth factor (PDGF), transforming growth factor-β (TGF-β), vascular endothelial growth factor (VEGF), endothelial growth factor, and leukocytes that can stimulate an inflammatory response within the injured tissue.¹⁰ The inflammatory response from the PRP induces a more robust reconstruction and revascularization of the injured tissue, stimulating proliferation, and remodeling.^11,12However, significant variability exists within the platelets, leukocytes, and growth factors that comprise PRP. This is attributed to 3 major causes. First, current commercial preparations of PRP result in differing platelet concentrations, as well as leukocyte-rich and leukocyte-poor compositions.^13,14 Variability in platelet concentrations results in unreliable amounts of growth factors, including cytokines, TGF-β, PDGF, VEGF and basic fibroblast growth factor in each preparation, while leukocyte levels affect inflammation, all leading to variable effects for each preparation.^15,16Second, despite sex and age of the PRP donor not being significant factors influencing variation in growth factor concentrations, the existence of an unexplained variation in concentrations of growth factors between different donors has been observed.¹⁷ Third, the selection of activating agents, bovine thrombin or calcium chloride, and their application, whether to the elbow, shoulder, or knee, produces variability.¹⁸

While the effects of platelets and growth factors in PRP have been well studied, less is known about the effects of differing cell types. Recently it was reported that the concentrations of leukocytes directly affect the outcomes of PRP injections. McCarrel and colleagues^19,20 found that as the number of leukocytes increased, there was a concomitant increase in the expression of inflammatory cytokines and catabolic activity. This effect may result in inferior healing of injured tissues and is attributed to the release of pro-inflammatory cytokines such as interleukin-1β from the leukocytes.²¹ There is also evidence that minimizing the catabolic effect of leukocytes may be just as important to tissue healing as the maximizing anabolic effect of platelets and growth factors.²²

The use of PRP has been highly disputed in recent years due to conflicting reports of its success in treating orthopedic conditions. Numerous favorable studies have shown benefit for treating chronic and acute orthopedic injuries including; rotator cuff tear repair, chronic refractory patellar tendinopathy, and chronic lateral tendinosis/epicondylitis.^23-26 Concurrently, articles demonstrating no significant effects from PRP have also been published. One study claiming that PRP injections did not improve outcomes of chronic Achilles tendinopathy did not differentiate whether patients had tendinosis or tendinitis, and did not consider leukocyte concentration in their PRP preparations²⁷ Another study that determined PRP is not beneficial to the healing of ruptured Achilles tendons after surgical repair also failed to consider the concentration of leukocytes in their PRP preparations.²⁸ One of the difficulties in comparing these studies is their heterogeneous nature. This arises from the use of different conditions in each study that makes the studies incomparable. Variations in PRP preparations lead to different concentrations of growth factors, platelets, and leukocyte concentrations. Additionally, tendinopathy models were not specified as tendinosis and tendonitis, and models or patients were not controlled for age, sex, or comorbidities. Given that leukocyte-rich and leukocyte-poor PRP preparations are currently widely used in clinical practice, the discovery of which type of preparation is indicated in which setting is paramount to evidence-based use of this treatment modality. Due to reports suggesting that leukocytes may be detrimental to tendon healing, determining which types of leukocytes are responsible for these effects is vital. As such, the purpose of this study is to determine the in vivo effects of sub-populations of leukocytes on normal rat tendons. This study design allowed us to isolate the effects of the injections to induce a response and remove confounding effects of normal healing response to a damaged tendon and effects from the injection itself. Our hypothesis was that the injection of leukocytes would cause an inflammatory response in rat tendons, leading to catabolic outcomes.

Continue to: METHODS...

 

 

METHODS

This was a prospective, in vivo, placebo controlled, randomized animal study. The University’s Institutional Animal Care and Use Committee approved all procedures prior to initiation. Twenty-four male Sprague-Dawley rats were randomized to 3 treatment groups (n = 8): monocytes; granulocytes, and; plasma, as a negative control.

Allogenic blood from 6 additional rats was collected into K2EDTA tubes via cardiac puncture. Allogenic, as opposed to autogenic, blood is commonly used in rat models because of low immunogenic response to blood from rats of the same strain and litter.^29,30 The blood was then pooled and the red cells lysed by incubation with Red Blood Cell Lysis Buffer (Roche). The samples were then sorted into fractions containing monocytes and granulocytes using fluorescence activated cell sorting (FACS) using a FACSAria (BD Biosciences). Cells were sorted using Purified PE Mouse Anti-Rat CD11b/c antibodies (BD Pharmingen) specific to monocytes, granulocytes, macrophages, dendritic cells, and microglia, APC-Cy7 Mouse Anti-Rat CD45 antibodies (BD Pharmingen) specific to all hematopoietic cells except erythrocytes, and FITC Mouse Anti-Rat CD42d antibodies (BD Pharmingen) specific to megakaryocytes and platelets. 20 μL of 0.2 mg/mL CD11b/c, 20 μL of 0.2 mg/mL CD 45, and 10 μL of 0.5 mg/mL CD42d antibodies were added to 1 mL of condensed non-red cells collected from the 6 rats and incubated at room temperature in the dark for 15 minutes. A fraction containing only platelet-poor plasma was also collected. For all treatments the injection volume was 75 μL. Rats in the monocyte group were injected with 200,000 cells in platelet-poor plasma, those in the granulocyte group were injected with 900,000 cells in platelet-poor plasma, and rats in the plasma control group received only platelet-poor plasma. The cell concentrations were based on previous studies that documented these concentrations that are found in typical leukocyte-rich PRP preparations.¹³

The animals were anesthetized with isoflurane gas and then injected aseptically once into their right Achilles tendon. The left Achilles tendon was used as an un-injected control, giving a total of 48 total Achilles tendons studied. At days 7 and 14 post-injection, 4 rats from each group were euthanized and the Achilles tendons were harvested.

The tendons were fixed in neutral buffered formalin for 24 hours and then embedded in paraffin and sectioned sagittally at 12 μm. The tendons were then stained with hematoxylin and eosin (H&E) using standard histological protocols and examined by 3 individuals trained to assess cellularity and morphology. All samples were assigned unrecognizable numbers and randomized prior to examination by individuals. Cell counts were based on the number of nuclei present in 3 mid-tendon high-power fields (400x) per sample. Morphology was graded on a scale of 1 to 3, with 1 being a normal tendon and 3 having severe pathology with total loss of alignment and crimping on 3 low-power fields (100x) per sample (Figures 1A-1G).

Vascularity was assessed by immunohistochemical staining using Rabbit Polyclonal Anti-CD31 antibodies (Abcam), a marker for vascular endothelial cells, using a Vectastain ABC Kit (Vector Laboratories) system and the ImmPACT AEC Peroxidase (HRP) Substrate (Vector Laboratories). Following staining, automated image analysis was performed (Bioquant). Briefly, all areas that did not contain tendon were masked. CD31 positive areas were then quantified using global thresholding. Vascularity was then calculated as ratio of CD31 positive area to total tendon area. Analyses were performed on 3 mid-tendon medium-power (200x) fields per sample.

For cellularity and morphology, the results for the injected tendons were normalized to those of their contralateral untreated controls and reported as a percentage. Results for vascularity were compared directly between treated tendons. Differences were assessed between groups at each time-point using Independent Samples Median Tests. When significant differences were identified, pairwise comparisons were performed to identify the source of the differences. All analyses were conducted using SPSS (V22, SAS Institute) with significant differences determined for values of P < 0.05.

RESULTS

No significant differences in cellularity between groups were seen at day 7 (P = 0.368) (Figures 1A-1G). However, a significant difference in cellularity between groups was seen at day 14 (P = 0.014). Pairwise tests showed there to be a significant increase in the number of cells in the tendons treated with granulocytes from 221% and 249% in cellularity (P = 0.014) on day 14, as compared to both monocytes and plasma, respectively. Morphologically, no significant differences were seen between groups at either time-point (P = 0.091 for day 7 and P = 1.000 for day 14) (Figures 2A-2G). However, a significant improvement in morphology was observed from day 7 to day 14 in the granulocyte group from 60% to 165% (P = 0.029). Finally, no differences were seen in vascularity between treatment groups at either time-point (P = 0.368 for day 7 and P = 0.535 for day 14) (Figures 3A-3G).

Continue to: DISCUSSION...

 

 

DISCUSSION

Our hypothesis that the injection of leukocytes would cause an inflammatory response in rat tendons leading to catabolic outcomes was confirmed in the granulocyte group. It should be noted that prior to the catabolic outcome, there was a transient anabolic effect in the granulocyte group during the second week. Deterioration in morphology was observed in the tendons injected with granulocytes on day 7, which subsequently recovered in the following week. We found that injecting granulocytes into normal tendons resulted in an increase in inflammatory cellularity, when compared to monocytes and plasma injections. 

Limitations inherent in this study are those similar to other in vivo studies. To begin with, the results of injections into rat tendons may not be translatable to human tendons. Despite this limitation, the rat is a common model for tendon research.³¹ Another limitation is that this study injected healthy Achilles tendons, rather than tendons with preexisting tendinopathy. In a naturally occurring tendinopathy, there may be other factors present that interact with PRP, and this model negates the contribution of these factors. Finally, while the immunohistochemistry (IHC) and morphological data are clear, the cellularity data are not clear in identifying the type of cells that were increased by granulocyte injection.  However, the cells appeared rounded, resembling inflammatory infiltrate; a common cell type seen in tendons.² While fibroblasts are also a common infiltrate during chronic tendinopathy, they are generally flat and appear on H&E as long spindle shaped cells. Thus, we believe the increased cellularity of the tendons after granulocyte injections is representative of an increase in inflammation. The increased cellularity could be due to the increased number of cells injected into the tendon; however, our conclusions are consistent with the increased inflammation previously reported linking leukocytes to tendon inflammation.^20,22,32

In terms of morphology, we hypothesized that degenerative changes would be seen in the tendons that were injected with granulocytes due to the inflammatory action of these cells. As part of the granulocyte response, neutrophils release proteases and macrophages can stimulate collagen synthesis via fibroblasts, both causing change within the extracellular matrix.^33,34 Indeed, we observed a significant change in tissue morphology in the granulocyte group over the course of 14 days. As the degenerative and regenerative effects of granulocytes take time to present, this is likely what we observed to occur between day 7 and 14 after treatment. These observations are also consistent with prior observations that leukocyte-rich PRP injections can be detrimental to tendon healing, but beneficial to tissue degeneration in the setting of chronic tendonitis.²⁰

We hypothesized that the vascularity of the tendons would be similar in all preparations. This was based on previous studies demonstrating that the lack of platelets in the platelet-poor plasma fraction is sufficient to deplete VEGF, the angiogenic agent in PRP.³⁵ In this study, there were no observable differences in vascularity of platelet-poor plasma, monocyte, and granulocyte injections. We attribute this to the lack of VEGF in any of these preparations. The aforementioned study also showed that the lack of platelets in injection was enough to prevent the angiogenic effect of this treatment.³⁵

Continue to: The goal of this study was...

 

 

The goal of this study was to assess the morphology, cellularity, and vascularity of normal tendons after injections of different leukocyte populations. This is clinically important because of the potential to tailor future PRP injections on a patient-by-patient basis. In patients requiring an anabolic response, leukocyte-poor PRP may be the best option. In contrast, when patient pathology requires an inflammatory response to improve healing³⁶ or breakdown fibrotic tissue, as seen in tendinosis, leukocyte-rich PRP may be warranted. Further, properly controlled clinical studies are needed to validate these recommendations.

Limitations inherent in this study are those similar to other in vivo studies. First, the results of injections into rat tendons may not be translatable to human tendons. Despite this limitation, the rat is a common model for tendon research.³¹ A second limitation is that this study injected healthy Achilles tendons, rather than tendons with preexisting tendinopathy. In a naturally occurring tendinopathy, there may be other factors present that interact with PRP, and this model negates the contribution of these factors. Finally, while the IHC and morphological data show clear changes, the cellularity data are not clear in identifying the type of cells that were increased by granulocyte injection. However, the cells appeared rounded, resembling inflammatory infiltrate; a common cell type seen in tendons.² While fibroblasts are also a common infiltrate during chronic tendinopathy, they are generally flat and appear on H&E as long spindle shaped cells. The last limitation of this study is the lack of functional mechanical testing since, clinically, healing of the tendon is also related to the strength of the tendon. Thus, we believe the increased cellularity of the tendons after granulocyte injections is representative of an increase in inflammation. Moreover, our results are consistent with the increased inflammation previously reported linking leukocytes to tendon inflammation.^20,22,32 It is interesting to note that the increase in inflammation does not lead to an increase in vascularity as could be expected.

CONCLUSION

We found that the injection of leukocytes into healthy rat Achilles tendons increases inflammation, as evidenced by increased cellularity and disrupted morphology, which suggests that leukocyte-rich PRP preparations may be contraindicated in settings of acute tendonitis. However, these preparations may be useful for a specific subset of tendinopathies, including chronic tendinosis.

ABSTRACT

In this study, we determine the in vivo effects of injecting sub-populations of leukocytes into normal rat Achilles tendons via a controlled laboratory study. Allogenic monocytes, granulocytes, or plasma were injected into 24 healthy rat Achilles tendons. Treated and contralateral un-treated control tendons then assessed for cellularity, histologic morphology, and vascularity after 7 and 14 days. Significant increases of 221% and 249% in cellularity (P = 0.014) were seen on day 14 within Achilles tendons injected with granulocytes as compared to plasma and monocytes, respectively. Also, significant improvement in morphology (P = 0.029) between days 7 and 14 was seen for the granulocyte injected Achilles tendons. Significant increases in cellularity after an injection of granulocytes, compared to monocytes and plasma, corresponds to a significant increase in inflammation within the tissue, suggesting that leukocyte-rich platelet-rich plasma (PRP) preparations are proinflammatory and potentially catabolic when injected into tendon tissue. The concentration and composition of white blood cells within PRP preparations is variable and needs to be better understood in order to optimize clinical utility of PRP injections.

Continue to: Tendinopathies are debilitating conditions...

Tendinopathies are debilitating conditions affecting patients worldwide every day. They arise most frequently from tendon overuse resulting in pathology.¹ There are 2 major subtypes of tendinopathy: tendinosis and tendinitis. Tendinosis, the more common condition, is characterized by long-term, chronic degradation of tendon tissue resulting in fibrosis from infiltrating fibroblasts.² Tendinitis, the less common condition, is characterized by an acute inflammatory response and inflammatory cell infiltrate.² Both conditions are common, with Achilles tendinopathy affecting 11% of runners and lateral epicondylitis affecting 1% to 3% of the general population.^3,4 Many sports-related overuse injuries, such as tendinopathies, go undiagnosed for extended periods of time because medical attention is avoided in order to prevent time loss from training or competing.⁵ These delays could be eliminated if a non-surgical option for treating tendon pathology was available.

Tendinopathies are believed to result from tendon overuse that causes micro-damage to collagen, as well as from significant changes in protein and enzyme composition within the tendon.⁶ The damage accumulates over time and eventually leads to chronic inflammation or fibrotic change within tendons, in both cases weakening the tendon and causing pain. Currently, accepted treatments for tendinopathies include: nonsteroidal anti-inflammatory drugs, physical therapy, ultrasound, laser-therapy, corticosteroids, glyceryl trinitrate patches, extracorporeal shock wave therapy, sclerotherapy, and surgery.⁷ Recently, platelet-rich plasma (PRP) therapy has emerged as a promising treatment for tendinopathies, as well as a variety of other orthopedic indications.

PRP consists of autologous blood from the patient, centrifuged to increase the amount of platelets in the sample above baseline, and subsequently injected around an affected tendon or joint.⁸ PRP is used to treat tendinopathy because it can supply injured tendons with blood components that aid in healing, which tendons do not receive due to poor vascularity.⁹ These components include growth factors, such as platelet derived growth factor (PDGF), transforming growth factor-β (TGF-β), vascular endothelial growth factor (VEGF), endothelial growth factor, and leukocytes that can stimulate an inflammatory response within the injured tissue.¹⁰ The inflammatory response from the PRP induces a more robust reconstruction and revascularization of the injured tissue, stimulating proliferation, and remodeling.^11,12However, significant variability exists within the platelets, leukocytes, and growth factors that comprise PRP. This is attributed to 3 major causes. First, current commercial preparations of PRP result in differing platelet concentrations, as well as leukocyte-rich and leukocyte-poor compositions.^13,14 Variability in platelet concentrations results in unreliable amounts of growth factors, including cytokines, TGF-β, PDGF, VEGF and basic fibroblast growth factor in each preparation, while leukocyte levels affect inflammation, all leading to variable effects for each preparation.^15,16Second, despite sex and age of the PRP donor not being significant factors influencing variation in growth factor concentrations, the existence of an unexplained variation in concentrations of growth factors between different donors has been observed.¹⁷ Third, the selection of activating agents, bovine thrombin or calcium chloride, and their application, whether to the elbow, shoulder, or knee, produces variability.¹⁸

While the effects of platelets and growth factors in PRP have been well studied, less is known about the effects of differing cell types. Recently it was reported that the concentrations of leukocytes directly affect the outcomes of PRP injections. McCarrel and colleagues^19,20 found that as the number of leukocytes increased, there was a concomitant increase in the expression of inflammatory cytokines and catabolic activity. This effect may result in inferior healing of injured tissues and is attributed to the release of pro-inflammatory cytokines such as interleukin-1β from the leukocytes.²¹ There is also evidence that minimizing the catabolic effect of leukocytes may be just as important to tissue healing as the maximizing anabolic effect of platelets and growth factors.²²

The use of PRP has been highly disputed in recent years due to conflicting reports of its success in treating orthopedic conditions. Numerous favorable studies have shown benefit for treating chronic and acute orthopedic injuries including; rotator cuff tear repair, chronic refractory patellar tendinopathy, and chronic lateral tendinosis/epicondylitis.^23-26 Concurrently, articles demonstrating no significant effects from PRP have also been published. One study claiming that PRP injections did not improve outcomes of chronic Achilles tendinopathy did not differentiate whether patients had tendinosis or tendinitis, and did not consider leukocyte concentration in their PRP preparations²⁷ Another study that determined PRP is not beneficial to the healing of ruptured Achilles tendons after surgical repair also failed to consider the concentration of leukocytes in their PRP preparations.²⁸ One of the difficulties in comparing these studies is their heterogeneous nature. This arises from the use of different conditions in each study that makes the studies incomparable. Variations in PRP preparations lead to different concentrations of growth factors, platelets, and leukocyte concentrations. Additionally, tendinopathy models were not specified as tendinosis and tendonitis, and models or patients were not controlled for age, sex, or comorbidities. Given that leukocyte-rich and leukocyte-poor PRP preparations are currently widely used in clinical practice, the discovery of which type of preparation is indicated in which setting is paramount to evidence-based use of this treatment modality. Due to reports suggesting that leukocytes may be detrimental to tendon healing, determining which types of leukocytes are responsible for these effects is vital. As such, the purpose of this study is to determine the in vivo effects of sub-populations of leukocytes on normal rat tendons. This study design allowed us to isolate the effects of the injections to induce a response and remove confounding effects of normal healing response to a damaged tendon and effects from the injection itself. Our hypothesis was that the injection of leukocytes would cause an inflammatory response in rat tendons, leading to catabolic outcomes.

Continue to: METHODS...

 

 

METHODS

This was a prospective, in vivo, placebo controlled, randomized animal study. The University’s Institutional Animal Care and Use Committee approved all procedures prior to initiation. Twenty-four male Sprague-Dawley rats were randomized to 3 treatment groups (n = 8): monocytes; granulocytes, and; plasma, as a negative control.

Allogenic blood from 6 additional rats was collected into K2EDTA tubes via cardiac puncture. Allogenic, as opposed to autogenic, blood is commonly used in rat models because of low immunogenic response to blood from rats of the same strain and litter.^29,30 The blood was then pooled and the red cells lysed by incubation with Red Blood Cell Lysis Buffer (Roche). The samples were then sorted into fractions containing monocytes and granulocytes using fluorescence activated cell sorting (FACS) using a FACSAria (BD Biosciences). Cells were sorted using Purified PE Mouse Anti-Rat CD11b/c antibodies (BD Pharmingen) specific to monocytes, granulocytes, macrophages, dendritic cells, and microglia, APC-Cy7 Mouse Anti-Rat CD45 antibodies (BD Pharmingen) specific to all hematopoietic cells except erythrocytes, and FITC Mouse Anti-Rat CD42d antibodies (BD Pharmingen) specific to megakaryocytes and platelets. 20 μL of 0.2 mg/mL CD11b/c, 20 μL of 0.2 mg/mL CD 45, and 10 μL of 0.5 mg/mL CD42d antibodies were added to 1 mL of condensed non-red cells collected from the 6 rats and incubated at room temperature in the dark for 15 minutes. A fraction containing only platelet-poor plasma was also collected. For all treatments the injection volume was 75 μL. Rats in the monocyte group were injected with 200,000 cells in platelet-poor plasma, those in the granulocyte group were injected with 900,000 cells in platelet-poor plasma, and rats in the plasma control group received only platelet-poor plasma. The cell concentrations were based on previous studies that documented these concentrations that are found in typical leukocyte-rich PRP preparations.¹³

The animals were anesthetized with isoflurane gas and then injected aseptically once into their right Achilles tendon. The left Achilles tendon was used as an un-injected control, giving a total of 48 total Achilles tendons studied. At days 7 and 14 post-injection, 4 rats from each group were euthanized and the Achilles tendons were harvested.

The tendons were fixed in neutral buffered formalin for 24 hours and then embedded in paraffin and sectioned sagittally at 12 μm. The tendons were then stained with hematoxylin and eosin (H&E) using standard histological protocols and examined by 3 individuals trained to assess cellularity and morphology. All samples were assigned unrecognizable numbers and randomized prior to examination by individuals. Cell counts were based on the number of nuclei present in 3 mid-tendon high-power fields (400x) per sample. Morphology was graded on a scale of 1 to 3, with 1 being a normal tendon and 3 having severe pathology with total loss of alignment and crimping on 3 low-power fields (100x) per sample (Figures 1A-1G).

Vascularity was assessed by immunohistochemical staining using Rabbit Polyclonal Anti-CD31 antibodies (Abcam), a marker for vascular endothelial cells, using a Vectastain ABC Kit (Vector Laboratories) system and the ImmPACT AEC Peroxidase (HRP) Substrate (Vector Laboratories). Following staining, automated image analysis was performed (Bioquant). Briefly, all areas that did not contain tendon were masked. CD31 positive areas were then quantified using global thresholding. Vascularity was then calculated as ratio of CD31 positive area to total tendon area. Analyses were performed on 3 mid-tendon medium-power (200x) fields per sample.

For cellularity and morphology, the results for the injected tendons were normalized to those of their contralateral untreated controls and reported as a percentage. Results for vascularity were compared directly between treated tendons. Differences were assessed between groups at each time-point using Independent Samples Median Tests. When significant differences were identified, pairwise comparisons were performed to identify the source of the differences. All analyses were conducted using SPSS (V22, SAS Institute) with significant differences determined for values of P < 0.05.

RESULTS

No significant differences in cellularity between groups were seen at day 7 (P = 0.368) (Figures 1A-1G). However, a significant difference in cellularity between groups was seen at day 14 (P = 0.014). Pairwise tests showed there to be a significant increase in the number of cells in the tendons treated with granulocytes from 221% and 249% in cellularity (P = 0.014) on day 14, as compared to both monocytes and plasma, respectively. Morphologically, no significant differences were seen between groups at either time-point (P = 0.091 for day 7 and P = 1.000 for day 14) (Figures 2A-2G). However, a significant improvement in morphology was observed from day 7 to day 14 in the granulocyte group from 60% to 165% (P = 0.029). Finally, no differences were seen in vascularity between treatment groups at either time-point (P = 0.368 for day 7 and P = 0.535 for day 14) (Figures 3A-3G).

Continue to: DISCUSSION...

 

 

DISCUSSION

Our hypothesis that the injection of leukocytes would cause an inflammatory response in rat tendons leading to catabolic outcomes was confirmed in the granulocyte group. It should be noted that prior to the catabolic outcome, there was a transient anabolic effect in the granulocyte group during the second week. Deterioration in morphology was observed in the tendons injected with granulocytes on day 7, which subsequently recovered in the following week. We found that injecting granulocytes into normal tendons resulted in an increase in inflammatory cellularity, when compared to monocytes and plasma injections. 

Limitations inherent in this study are those similar to other in vivo studies. To begin with, the results of injections into rat tendons may not be translatable to human tendons. Despite this limitation, the rat is a common model for tendon research.³¹ Another limitation is that this study injected healthy Achilles tendons, rather than tendons with preexisting tendinopathy. In a naturally occurring tendinopathy, there may be other factors present that interact with PRP, and this model negates the contribution of these factors. Finally, while the immunohistochemistry (IHC) and morphological data are clear, the cellularity data are not clear in identifying the type of cells that were increased by granulocyte injection.  However, the cells appeared rounded, resembling inflammatory infiltrate; a common cell type seen in tendons.² While fibroblasts are also a common infiltrate during chronic tendinopathy, they are generally flat and appear on H&E as long spindle shaped cells. Thus, we believe the increased cellularity of the tendons after granulocyte injections is representative of an increase in inflammation. The increased cellularity could be due to the increased number of cells injected into the tendon; however, our conclusions are consistent with the increased inflammation previously reported linking leukocytes to tendon inflammation.^20,22,32

In terms of morphology, we hypothesized that degenerative changes would be seen in the tendons that were injected with granulocytes due to the inflammatory action of these cells. As part of the granulocyte response, neutrophils release proteases and macrophages can stimulate collagen synthesis via fibroblasts, both causing change within the extracellular matrix.^33,34 Indeed, we observed a significant change in tissue morphology in the granulocyte group over the course of 14 days. As the degenerative and regenerative effects of granulocytes take time to present, this is likely what we observed to occur between day 7 and 14 after treatment. These observations are also consistent with prior observations that leukocyte-rich PRP injections can be detrimental to tendon healing, but beneficial to tissue degeneration in the setting of chronic tendonitis.²⁰

We hypothesized that the vascularity of the tendons would be similar in all preparations. This was based on previous studies demonstrating that the lack of platelets in the platelet-poor plasma fraction is sufficient to deplete VEGF, the angiogenic agent in PRP.³⁵ In this study, there were no observable differences in vascularity of platelet-poor plasma, monocyte, and granulocyte injections. We attribute this to the lack of VEGF in any of these preparations. The aforementioned study also showed that the lack of platelets in injection was enough to prevent the angiogenic effect of this treatment.³⁵

Continue to: The goal of this study was...

 

 

The goal of this study was to assess the morphology, cellularity, and vascularity of normal tendons after injections of different leukocyte populations. This is clinically important because of the potential to tailor future PRP injections on a patient-by-patient basis. In patients requiring an anabolic response, leukocyte-poor PRP may be the best option. In contrast, when patient pathology requires an inflammatory response to improve healing³⁶ or breakdown fibrotic tissue, as seen in tendinosis, leukocyte-rich PRP may be warranted. Further, properly controlled clinical studies are needed to validate these recommendations.

Limitations inherent in this study are those similar to other in vivo studies. First, the results of injections into rat tendons may not be translatable to human tendons. Despite this limitation, the rat is a common model for tendon research.³¹ A second limitation is that this study injected healthy Achilles tendons, rather than tendons with preexisting tendinopathy. In a naturally occurring tendinopathy, there may be other factors present that interact with PRP, and this model negates the contribution of these factors. Finally, while the IHC and morphological data show clear changes, the cellularity data are not clear in identifying the type of cells that were increased by granulocyte injection. However, the cells appeared rounded, resembling inflammatory infiltrate; a common cell type seen in tendons.² While fibroblasts are also a common infiltrate during chronic tendinopathy, they are generally flat and appear on H&E as long spindle shaped cells. The last limitation of this study is the lack of functional mechanical testing since, clinically, healing of the tendon is also related to the strength of the tendon. Thus, we believe the increased cellularity of the tendons after granulocyte injections is representative of an increase in inflammation. Moreover, our results are consistent with the increased inflammation previously reported linking leukocytes to tendon inflammation.^20,22,32 It is interesting to note that the increase in inflammation does not lead to an increase in vascularity as could be expected.

CONCLUSION

We found that the injection of leukocytes into healthy rat Achilles tendons increases inflammation, as evidenced by increased cellularity and disrupted morphology, which suggests that leukocyte-rich PRP preparations may be contraindicated in settings of acute tendonitis. However, these preparations may be useful for a specific subset of tendinopathies, including chronic tendinosis.

Patient experience and 30-day readmission are important measures of quality of care for hospitalized patients. Performance on both of these measures impact hospitals financially. Performance on the Hospital Consumer Assessment of Healthcare Systems and Providers (HCAHPS) survey is linked to 25% of the incentive payment under Value Based Purchasing (VBP) Program.¹ Starting in 2012, the Centers for Medicare and Medicaid Services (CMS) introduced the Readmission Reduction Program, penalizing hospitals financially for excessive readmissions.²

A relationship between patient experience and readmissions has been explored at the hospital level. Studies have mostly found that higher patient experience scores are associated with lower 30-day readmission rates. In a study of the relationship between 30-day risk-standardized readmission rates for three medical conditions (acute myocardial infarction, heart failure, and pneumonia) and patient experience, the authors noted that higher experience scores for overall care and discharge planning were associated with lower readmission rates for these conditions. They also concluded that patient experience scores were more predictive of 30-day readmission than clinical performance measures. Additionally, the authors predicted that if a hospital increased its total experience scores from the 25th percentile to the 75th percentile, there would be an associated decrease in readmissions by at least 2.3% for each of these conditions.³ Practice management companies and the media have cited this finding to conclude that higher patient experience drives clinical outcomes such as 30-day readmission and that patients are often the best judges of the quality of care delivered.^4,5

Other hospital-level studies have found that high 30-day readmission rates are associated with lower overall experience scores in a mixed surgical patient population; worse reports of pain control and overall care in the colorectal surgery population; lower experience scores with discharge preparedness in vascular surgery patients; and lower experience scores with physician communication, nurse communication, and discharge preparedness.^6-9 A patient-level study noted higher readmissions are associated with worse experience with physician and nursing communication along with a paradoxically better experience with discharge information.¹⁰

Because these studies used an observational design, they demonstrated associations rather than causality. An alternative hypothesis is that readmitted patients complete their patient experience survey after readmission and the low experience is the result, rather than the cause, of their readmission. For patients who are readmitted, it is unclear whether there is an opportunity to complete the survey prior to readmission and whether being readmitted may impact patient perception of quality of care. Using patient-level data, we sought to assess HCAHPS patient-experience responses linked to the index admission of the patients who were readmitted in 30 days and compare it with those patients who were not readmitted during this time period. We paid particular attention to when the surveys were returned.

Study Design

We conducted a retrospective analysis of prospectively collected 10-year HCAHPS and Press Ganey patient survey data for a single tertiary care academic hospital.

Participants

All adult patients discharged from the hospital and who responded to the routinely sent patient-experience survey were included. Surveys were sent to a random sample of 50% of the discharged patients.

The exposure group was comprised of patients who responded to the survey and were readmitted within 30 days of discharge. After subtracting 5 days from the survey receipt date for expected delays related to mail delivery time and processing time, survey response date was calculated. The exposure group was further divided into patients who responded to the survey prior to their 30-day readmission (“Pre-readmission responders”) and those that responded to the survey after their readmission (“Postreadmission responders”). A sensitivity analysis was performed by changing the number of days subtracted from the survey receipt date by 2 days in either direction. This approach did not result in any significant changes in the results.

The control group comprised patients who were not readmitted to the hospital within 30 days of discharge and who did not have an admission in the previous 30 days as well (“Not readmitted” group). An additional comparison group for exploratory analysis included patients who had experienced an admission in the prior 30 days but were not readmitted after the admission linked to the survey. These patients responded to the patient-experience surveys that were linked to their second admission in 30 days (“2^nd-admission responders” group; Figure).

Time Periods

All survey responders from the third quarter of 2006 to the first quarter of 2016 were included in the study. Additionally, administrative data on non-responders were available from 7/2006 to 8/2012. These data were used to estimate response rates. Patient level experience and administrative data were obtained in a linked fashion for these time periods.

Instruments

Press Ganey and HCAHPS surveys were sent via mail in the same envelope. Fifty percent of the discharged patients were randomized to receive the surveys. The Press Ganey survey contained 33 items encompassing several subdomains, including room, meal, nursing, physician, ancillary staff, visitor, discharge, and overall experience.

The HCAHPS survey contained 29 CMS-mandated items, of which 21 are related to patient experience. The development, testing, and methods for administration and reporting of the HCAHPS survey have been previously described and studies using this instrument have been reported in the literature.¹¹ Press Ganey patient satisfaction survey results have also been reported in the literature.¹²

Outcome Variables and Covariates

HCAHPS and Press Ganey experience survey individual item responses were the primary outcome variables of this study. Age, self-reported health status, education, primary language spoken, service line, and time taken to respond to the surveys served as the covariates. These variables are used by CMS for patient-mix adjustment and are collected on the HCAHPS survey. Additionally, the number of days to respond to the survey were included in all regression analysis to adjust for early responder effect.^13-15

Statistical Analysis

“Percent top-box” scores were calculated for each survey item for patients in each group. The percent top-box scores were calculated as the percent of patients who responded “very good” for a given item on Press Ganey survey items and “always” or “definitely yes” or “yes” or “9” or “10” on HCAHPS survey items. CMS utilizes “percent top-box scores” to calculate payments under the VBP program and to report the results publicly. Numerous studies have also reported percent top-box scores for HCAHPS survey results.¹²

We hypothesized that whether patients complete the HCAHPS survey before or after the readmission influences their reporting of experience. To test this hypothesis, HCAHPS and Press Ganey item top-box scores of “Pre-readmission responders” and “Postreadmission responders” were compared with those of the control group using multivariate logistic regression. “Pre-readmission responders” were also compared with “Postreadmission responders”.

“2nd-admission responders” were similarly compared with the control group for an exploratory analysis. Finally, “Postreadmission responders” and “2nd-admission responders” were compared in another exploratory analysis since both these groups responded to the survey after being exposed to the readmission, even though the “Postreadmission responders” group is administratively linked to the index admission.

The Johns Hopkins Institutional Review Board approved this study.

There were 43,737 survey responders, among whom 4,707 were subsequently readmitted within 30 days of discharge. Among the readmitted patients who responded to the surveys linked to their index admission, only 15.8% returned the survey before readmission (pre-readmission responders’) and 84.2% returned the survey after readmission (postreadmission responders). Additionally, 1,663 patients responded to experience surveys linked to their readmission. There were 37,365 patients in the control arm (ie, patients who responded to the survey and were not readmitted within 30 days of discharge or in the prior 30 days; Figure 1). The readmission rate among survey responders was 10.6%. Among the readmitted patients, the median number of days to readmission was 10 days while the median number of days to respond to the survey for this group was 33 days. Among the nonreadmitted patients, the median number of days to return the survey was 29 days.

In this retrospective analysis of prospectively collected Press Ganey and HCAHPS patient-experience survey data, we found that the overwhelming majority of patients readmitted within 30 days of discharge respond to HCAHPS surveys after readmission even though the survey is sent linked to the first admission. This is not unexpected since the median time to survey response is 33 days for this group, while median time to readmission is 10 days. The dissatisfaction pattern of Postreadmission responders was similar to those who responded to the survey linked to the readmission. When a patient is readmitted prior to completing the survey, their responses appear to reflect the cumulative experience of the index admission and the readmission. The lower scores of those who respond to the survey after their readmission appear to be a driver for lower patient-experience scores related to readmissions. Overall, readmission was associated with lower scores on items in five of the nine domains used to calculate patient experience related payments under VBP.¹⁶

These findings have important implications in inferring the direction of potential causal relationship between readmissions and patient experience at the hospital level. Additionally, these patients show broad dissatisfaction with areas beyond physician communication and discharge planning. These include staff responsiveness, phlebotomy, meals, hospital cleanliness, and noise level. This pattern of dissatisfaction may represent impatience and frustration with spending additional time in the hospital environment.

Our results are consistent with findings of many of the earlier studies, but our study goes a step further by using patient-level data and incorporating survey response time in our analysis.^3,7,9,10 By separating out the readmitted patients who responded to the survey prior to admission, we attempted to address the ability of patients’ perception of care to predict future readmissions. Our results do not support this idea, since pre-readmission responders had similar experience scores to non-readmitted patients. However, because of the low numbers of pre-readmission responders, the comparison lacks precision. Current HCAHPS and Press Ganey questions may lack the ability to predict future readmissions because of the timing of the survey (postdischarge) or the questions themselves.

Overall, postreadmission responders are dissatisfied with multiple domains of hospital care. Many of these survey responses may simply be related to general frustration. Alternatively, they may represent a patient population with a high degree of needs that are not as easily met by a hospital’s routine processes of care. Even though the readmission rates were 10.6% among survey responders, 14.6% of the survey responses were associated with readmissions after accounting for those who respond to surveys linked to readmission. These patients could have significant impact on cumulative experience scores.

Our study has a few limitations. First, it involves a single tertiary care academic center study, and our results may not be generalizable. Second, we did not adjust for some of the patient characteristics associated with readmissions. Patients who were admitted within 30 days are different than those not readmitted based on payor, race, length of stay, and severity of illness, and we did not adjust for these factors in our analysis. This was intentional, however. Our goal was to better understand the relationship between 30-day readmission and patient experience scores as they are used for hospital-level studies, VBP, and public reporting. For these purposes, the scores are not adjusted for factors, such as payor and length of stay. We did adjust for patient-mix adjustment factors used by CMS. Third, the response rates to the HCAHPS were low and may have biased the scores. However, HCAHPS is widely used for comparisons between hospitals has been validated, and our study results have implications with regard to comparing hospital-level performance. HCAHPS results are relevant to policy and have financial consequences.¹⁷ Fourth, our study did not directly compare whether the relationship between patient experience for the postreadmission group and nonreadmitted group was different from the relationship between the pre-readmission group and postreadmission group. It is possible that there is no difference in relationship between the groups. However, despite the small number of pre-readmission responders, these patients tended to have more favorable experience responses than those who responded after being readmitted, even after adjusting for response time. Although the P values are nonsignificant for many comparisons, the directionality of the effect is relatively consistent. Also, the vast majority of the patients fall in the postreadmission group, and these patients appear to drive the overall experience related to readmissions. Finally, since relatively few patients turned in surveys prior to readmission, we had limited power to detect a significant difference between these pre-readmission responders and nonreadmitted patients.

Our study has implications for policy makers, researchers, and providers. The HCAHPS scores of patients who are readmitted and completed the survey after being readmitted reflects their experience of both the index admission and the readmission. We did not find evidence to support that HCAHPS survey responses predict future readmissions at the patient level. Our findings do support the concept that lower readmissions rates (whether due to the patient population or processes of care that decrease readmission rates) may improve HCAHPS scores. We suggest caution in assuming that improving patient experience is likely to reduce readmission rates.

Disclosures

The authors declare no conflicts of interest.

Patient experience and 30-day readmission are important measures of quality of care for hospitalized patients. Performance on both of these measures impact hospitals financially. Performance on the Hospital Consumer Assessment of Healthcare Systems and Providers (HCAHPS) survey is linked to 25% of the incentive payment under Value Based Purchasing (VBP) Program.¹ Starting in 2012, the Centers for Medicare and Medicaid Services (CMS) introduced the Readmission Reduction Program, penalizing hospitals financially for excessive readmissions.²

A relationship between patient experience and readmissions has been explored at the hospital level. Studies have mostly found that higher patient experience scores are associated with lower 30-day readmission rates. In a study of the relationship between 30-day risk-standardized readmission rates for three medical conditions (acute myocardial infarction, heart failure, and pneumonia) and patient experience, the authors noted that higher experience scores for overall care and discharge planning were associated with lower readmission rates for these conditions. They also concluded that patient experience scores were more predictive of 30-day readmission than clinical performance measures. Additionally, the authors predicted that if a hospital increased its total experience scores from the 25th percentile to the 75th percentile, there would be an associated decrease in readmissions by at least 2.3% for each of these conditions.³ Practice management companies and the media have cited this finding to conclude that higher patient experience drives clinical outcomes such as 30-day readmission and that patients are often the best judges of the quality of care delivered.^4,5

Other hospital-level studies have found that high 30-day readmission rates are associated with lower overall experience scores in a mixed surgical patient population; worse reports of pain control and overall care in the colorectal surgery population; lower experience scores with discharge preparedness in vascular surgery patients; and lower experience scores with physician communication, nurse communication, and discharge preparedness.^6-9 A patient-level study noted higher readmissions are associated with worse experience with physician and nursing communication along with a paradoxically better experience with discharge information.¹⁰

Because these studies used an observational design, they demonstrated associations rather than causality. An alternative hypothesis is that readmitted patients complete their patient experience survey after readmission and the low experience is the result, rather than the cause, of their readmission. For patients who are readmitted, it is unclear whether there is an opportunity to complete the survey prior to readmission and whether being readmitted may impact patient perception of quality of care. Using patient-level data, we sought to assess HCAHPS patient-experience responses linked to the index admission of the patients who were readmitted in 30 days and compare it with those patients who were not readmitted during this time period. We paid particular attention to when the surveys were returned.

Study Design

We conducted a retrospective analysis of prospectively collected 10-year HCAHPS and Press Ganey patient survey data for a single tertiary care academic hospital.

Participants

All adult patients discharged from the hospital and who responded to the routinely sent patient-experience survey were included. Surveys were sent to a random sample of 50% of the discharged patients.

The exposure group was comprised of patients who responded to the survey and were readmitted within 30 days of discharge. After subtracting 5 days from the survey receipt date for expected delays related to mail delivery time and processing time, survey response date was calculated. The exposure group was further divided into patients who responded to the survey prior to their 30-day readmission (“Pre-readmission responders”) and those that responded to the survey after their readmission (“Postreadmission responders”). A sensitivity analysis was performed by changing the number of days subtracted from the survey receipt date by 2 days in either direction. This approach did not result in any significant changes in the results.

The control group comprised patients who were not readmitted to the hospital within 30 days of discharge and who did not have an admission in the previous 30 days as well (“Not readmitted” group). An additional comparison group for exploratory analysis included patients who had experienced an admission in the prior 30 days but were not readmitted after the admission linked to the survey. These patients responded to the patient-experience surveys that were linked to their second admission in 30 days (“2^nd-admission responders” group; Figure).

Time Periods

All survey responders from the third quarter of 2006 to the first quarter of 2016 were included in the study. Additionally, administrative data on non-responders were available from 7/2006 to 8/2012. These data were used to estimate response rates. Patient level experience and administrative data were obtained in a linked fashion for these time periods.

Instruments

Press Ganey and HCAHPS surveys were sent via mail in the same envelope. Fifty percent of the discharged patients were randomized to receive the surveys. The Press Ganey survey contained 33 items encompassing several subdomains, including room, meal, nursing, physician, ancillary staff, visitor, discharge, and overall experience.

The HCAHPS survey contained 29 CMS-mandated items, of which 21 are related to patient experience. The development, testing, and methods for administration and reporting of the HCAHPS survey have been previously described and studies using this instrument have been reported in the literature.¹¹ Press Ganey patient satisfaction survey results have also been reported in the literature.¹²

Outcome Variables and Covariates

HCAHPS and Press Ganey experience survey individual item responses were the primary outcome variables of this study. Age, self-reported health status, education, primary language spoken, service line, and time taken to respond to the surveys served as the covariates. These variables are used by CMS for patient-mix adjustment and are collected on the HCAHPS survey. Additionally, the number of days to respond to the survey were included in all regression analysis to adjust for early responder effect.^13-15

Statistical Analysis

“Percent top-box” scores were calculated for each survey item for patients in each group. The percent top-box scores were calculated as the percent of patients who responded “very good” for a given item on Press Ganey survey items and “always” or “definitely yes” or “yes” or “9” or “10” on HCAHPS survey items. CMS utilizes “percent top-box scores” to calculate payments under the VBP program and to report the results publicly. Numerous studies have also reported percent top-box scores for HCAHPS survey results.¹²

We hypothesized that whether patients complete the HCAHPS survey before or after the readmission influences their reporting of experience. To test this hypothesis, HCAHPS and Press Ganey item top-box scores of “Pre-readmission responders” and “Postreadmission responders” were compared with those of the control group using multivariate logistic regression. “Pre-readmission responders” were also compared with “Postreadmission responders”.

“2nd-admission responders” were similarly compared with the control group for an exploratory analysis. Finally, “Postreadmission responders” and “2nd-admission responders” were compared in another exploratory analysis since both these groups responded to the survey after being exposed to the readmission, even though the “Postreadmission responders” group is administratively linked to the index admission.

The Johns Hopkins Institutional Review Board approved this study.

There were 43,737 survey responders, among whom 4,707 were subsequently readmitted within 30 days of discharge. Among the readmitted patients who responded to the surveys linked to their index admission, only 15.8% returned the survey before readmission (pre-readmission responders’) and 84.2% returned the survey after readmission (postreadmission responders). Additionally, 1,663 patients responded to experience surveys linked to their readmission. There were 37,365 patients in the control arm (ie, patients who responded to the survey and were not readmitted within 30 days of discharge or in the prior 30 days; Figure 1). The readmission rate among survey responders was 10.6%. Among the readmitted patients, the median number of days to readmission was 10 days while the median number of days to respond to the survey for this group was 33 days. Among the nonreadmitted patients, the median number of days to return the survey was 29 days.

In this retrospective analysis of prospectively collected Press Ganey and HCAHPS patient-experience survey data, we found that the overwhelming majority of patients readmitted within 30 days of discharge respond to HCAHPS surveys after readmission even though the survey is sent linked to the first admission. This is not unexpected since the median time to survey response is 33 days for this group, while median time to readmission is 10 days. The dissatisfaction pattern of Postreadmission responders was similar to those who responded to the survey linked to the readmission. When a patient is readmitted prior to completing the survey, their responses appear to reflect the cumulative experience of the index admission and the readmission. The lower scores of those who respond to the survey after their readmission appear to be a driver for lower patient-experience scores related to readmissions. Overall, readmission was associated with lower scores on items in five of the nine domains used to calculate patient experience related payments under VBP.¹⁶

These findings have important implications in inferring the direction of potential causal relationship between readmissions and patient experience at the hospital level. Additionally, these patients show broad dissatisfaction with areas beyond physician communication and discharge planning. These include staff responsiveness, phlebotomy, meals, hospital cleanliness, and noise level. This pattern of dissatisfaction may represent impatience and frustration with spending additional time in the hospital environment.

Our results are consistent with findings of many of the earlier studies, but our study goes a step further by using patient-level data and incorporating survey response time in our analysis.^3,7,9,10 By separating out the readmitted patients who responded to the survey prior to admission, we attempted to address the ability of patients’ perception of care to predict future readmissions. Our results do not support this idea, since pre-readmission responders had similar experience scores to non-readmitted patients. However, because of the low numbers of pre-readmission responders, the comparison lacks precision. Current HCAHPS and Press Ganey questions may lack the ability to predict future readmissions because of the timing of the survey (postdischarge) or the questions themselves.

Overall, postreadmission responders are dissatisfied with multiple domains of hospital care. Many of these survey responses may simply be related to general frustration. Alternatively, they may represent a patient population with a high degree of needs that are not as easily met by a hospital’s routine processes of care. Even though the readmission rates were 10.6% among survey responders, 14.6% of the survey responses were associated with readmissions after accounting for those who respond to surveys linked to readmission. These patients could have significant impact on cumulative experience scores.

Our study has a few limitations. First, it involves a single tertiary care academic center study, and our results may not be generalizable. Second, we did not adjust for some of the patient characteristics associated with readmissions. Patients who were admitted within 30 days are different than those not readmitted based on payor, race, length of stay, and severity of illness, and we did not adjust for these factors in our analysis. This was intentional, however. Our goal was to better understand the relationship between 30-day readmission and patient experience scores as they are used for hospital-level studies, VBP, and public reporting. For these purposes, the scores are not adjusted for factors, such as payor and length of stay. We did adjust for patient-mix adjustment factors used by CMS. Third, the response rates to the HCAHPS were low and may have biased the scores. However, HCAHPS is widely used for comparisons between hospitals has been validated, and our study results have implications with regard to comparing hospital-level performance. HCAHPS results are relevant to policy and have financial consequences.¹⁷ Fourth, our study did not directly compare whether the relationship between patient experience for the postreadmission group and nonreadmitted group was different from the relationship between the pre-readmission group and postreadmission group. It is possible that there is no difference in relationship between the groups. However, despite the small number of pre-readmission responders, these patients tended to have more favorable experience responses than those who responded after being readmitted, even after adjusting for response time. Although the P values are nonsignificant for many comparisons, the directionality of the effect is relatively consistent. Also, the vast majority of the patients fall in the postreadmission group, and these patients appear to drive the overall experience related to readmissions. Finally, since relatively few patients turned in surveys prior to readmission, we had limited power to detect a significant difference between these pre-readmission responders and nonreadmitted patients.

Our study has implications for policy makers, researchers, and providers. The HCAHPS scores of patients who are readmitted and completed the survey after being readmitted reflects their experience of both the index admission and the readmission. We did not find evidence to support that HCAHPS survey responses predict future readmissions at the patient level. Our findings do support the concept that lower readmissions rates (whether due to the patient population or processes of care that decrease readmission rates) may improve HCAHPS scores. We suggest caution in assuming that improving patient experience is likely to reduce readmission rates.

Disclosures

The authors declare no conflicts of interest.

Patient experience and 30-day readmission are important measures of quality of care for hospitalized patients. Performance on both of these measures impact hospitals financially. Performance on the Hospital Consumer Assessment of Healthcare Systems and Providers (HCAHPS) survey is linked to 25% of the incentive payment under Value Based Purchasing (VBP) Program.¹ Starting in 2012, the Centers for Medicare and Medicaid Services (CMS) introduced the Readmission Reduction Program, penalizing hospitals financially for excessive readmissions.²

A relationship between patient experience and readmissions has been explored at the hospital level. Studies have mostly found that higher patient experience scores are associated with lower 30-day readmission rates. In a study of the relationship between 30-day risk-standardized readmission rates for three medical conditions (acute myocardial infarction, heart failure, and pneumonia) and patient experience, the authors noted that higher experience scores for overall care and discharge planning were associated with lower readmission rates for these conditions. They also concluded that patient experience scores were more predictive of 30-day readmission than clinical performance measures. Additionally, the authors predicted that if a hospital increased its total experience scores from the 25th percentile to the 75th percentile, there would be an associated decrease in readmissions by at least 2.3% for each of these conditions.³ Practice management companies and the media have cited this finding to conclude that higher patient experience drives clinical outcomes such as 30-day readmission and that patients are often the best judges of the quality of care delivered.^4,5

Other hospital-level studies have found that high 30-day readmission rates are associated with lower overall experience scores in a mixed surgical patient population; worse reports of pain control and overall care in the colorectal surgery population; lower experience scores with discharge preparedness in vascular surgery patients; and lower experience scores with physician communication, nurse communication, and discharge preparedness.^6-9 A patient-level study noted higher readmissions are associated with worse experience with physician and nursing communication along with a paradoxically better experience with discharge information.¹⁰

Because these studies used an observational design, they demonstrated associations rather than causality. An alternative hypothesis is that readmitted patients complete their patient experience survey after readmission and the low experience is the result, rather than the cause, of their readmission. For patients who are readmitted, it is unclear whether there is an opportunity to complete the survey prior to readmission and whether being readmitted may impact patient perception of quality of care. Using patient-level data, we sought to assess HCAHPS patient-experience responses linked to the index admission of the patients who were readmitted in 30 days and compare it with those patients who were not readmitted during this time period. We paid particular attention to when the surveys were returned.

Study Design

We conducted a retrospective analysis of prospectively collected 10-year HCAHPS and Press Ganey patient survey data for a single tertiary care academic hospital.

Participants

All adult patients discharged from the hospital and who responded to the routinely sent patient-experience survey were included. Surveys were sent to a random sample of 50% of the discharged patients.

The exposure group was comprised of patients who responded to the survey and were readmitted within 30 days of discharge. After subtracting 5 days from the survey receipt date for expected delays related to mail delivery time and processing time, survey response date was calculated. The exposure group was further divided into patients who responded to the survey prior to their 30-day readmission (“Pre-readmission responders”) and those that responded to the survey after their readmission (“Postreadmission responders”). A sensitivity analysis was performed by changing the number of days subtracted from the survey receipt date by 2 days in either direction. This approach did not result in any significant changes in the results.

The control group comprised patients who were not readmitted to the hospital within 30 days of discharge and who did not have an admission in the previous 30 days as well (“Not readmitted” group). An additional comparison group for exploratory analysis included patients who had experienced an admission in the prior 30 days but were not readmitted after the admission linked to the survey. These patients responded to the patient-experience surveys that were linked to their second admission in 30 days (“2^nd-admission responders” group; Figure).

Time Periods

All survey responders from the third quarter of 2006 to the first quarter of 2016 were included in the study. Additionally, administrative data on non-responders were available from 7/2006 to 8/2012. These data were used to estimate response rates. Patient level experience and administrative data were obtained in a linked fashion for these time periods.

Instruments

Press Ganey and HCAHPS surveys were sent via mail in the same envelope. Fifty percent of the discharged patients were randomized to receive the surveys. The Press Ganey survey contained 33 items encompassing several subdomains, including room, meal, nursing, physician, ancillary staff, visitor, discharge, and overall experience.

The HCAHPS survey contained 29 CMS-mandated items, of which 21 are related to patient experience. The development, testing, and methods for administration and reporting of the HCAHPS survey have been previously described and studies using this instrument have been reported in the literature.¹¹ Press Ganey patient satisfaction survey results have also been reported in the literature.¹²

Outcome Variables and Covariates

HCAHPS and Press Ganey experience survey individual item responses were the primary outcome variables of this study. Age, self-reported health status, education, primary language spoken, service line, and time taken to respond to the surveys served as the covariates. These variables are used by CMS for patient-mix adjustment and are collected on the HCAHPS survey. Additionally, the number of days to respond to the survey were included in all regression analysis to adjust for early responder effect.^13-15

Statistical Analysis

“Percent top-box” scores were calculated for each survey item for patients in each group. The percent top-box scores were calculated as the percent of patients who responded “very good” for a given item on Press Ganey survey items and “always” or “definitely yes” or “yes” or “9” or “10” on HCAHPS survey items. CMS utilizes “percent top-box scores” to calculate payments under the VBP program and to report the results publicly. Numerous studies have also reported percent top-box scores for HCAHPS survey results.¹²

We hypothesized that whether patients complete the HCAHPS survey before or after the readmission influences their reporting of experience. To test this hypothesis, HCAHPS and Press Ganey item top-box scores of “Pre-readmission responders” and “Postreadmission responders” were compared with those of the control group using multivariate logistic regression. “Pre-readmission responders” were also compared with “Postreadmission responders”.

“2nd-admission responders” were similarly compared with the control group for an exploratory analysis. Finally, “Postreadmission responders” and “2nd-admission responders” were compared in another exploratory analysis since both these groups responded to the survey after being exposed to the readmission, even though the “Postreadmission responders” group is administratively linked to the index admission.

The Johns Hopkins Institutional Review Board approved this study.

There were 43,737 survey responders, among whom 4,707 were subsequently readmitted within 30 days of discharge. Among the readmitted patients who responded to the surveys linked to their index admission, only 15.8% returned the survey before readmission (pre-readmission responders’) and 84.2% returned the survey after readmission (postreadmission responders). Additionally, 1,663 patients responded to experience surveys linked to their readmission. There were 37,365 patients in the control arm (ie, patients who responded to the survey and were not readmitted within 30 days of discharge or in the prior 30 days; Figure 1). The readmission rate among survey responders was 10.6%. Among the readmitted patients, the median number of days to readmission was 10 days while the median number of days to respond to the survey for this group was 33 days. Among the nonreadmitted patients, the median number of days to return the survey was 29 days.

In this retrospective analysis of prospectively collected Press Ganey and HCAHPS patient-experience survey data, we found that the overwhelming majority of patients readmitted within 30 days of discharge respond to HCAHPS surveys after readmission even though the survey is sent linked to the first admission. This is not unexpected since the median time to survey response is 33 days for this group, while median time to readmission is 10 days. The dissatisfaction pattern of Postreadmission responders was similar to those who responded to the survey linked to the readmission. When a patient is readmitted prior to completing the survey, their responses appear to reflect the cumulative experience of the index admission and the readmission. The lower scores of those who respond to the survey after their readmission appear to be a driver for lower patient-experience scores related to readmissions. Overall, readmission was associated with lower scores on items in five of the nine domains used to calculate patient experience related payments under VBP.¹⁶

These findings have important implications in inferring the direction of potential causal relationship between readmissions and patient experience at the hospital level. Additionally, these patients show broad dissatisfaction with areas beyond physician communication and discharge planning. These include staff responsiveness, phlebotomy, meals, hospital cleanliness, and noise level. This pattern of dissatisfaction may represent impatience and frustration with spending additional time in the hospital environment.

Our results are consistent with findings of many of the earlier studies, but our study goes a step further by using patient-level data and incorporating survey response time in our analysis.^3,7,9,10 By separating out the readmitted patients who responded to the survey prior to admission, we attempted to address the ability of patients’ perception of care to predict future readmissions. Our results do not support this idea, since pre-readmission responders had similar experience scores to non-readmitted patients. However, because of the low numbers of pre-readmission responders, the comparison lacks precision. Current HCAHPS and Press Ganey questions may lack the ability to predict future readmissions because of the timing of the survey (postdischarge) or the questions themselves.

Overall, postreadmission responders are dissatisfied with multiple domains of hospital care. Many of these survey responses may simply be related to general frustration. Alternatively, they may represent a patient population with a high degree of needs that are not as easily met by a hospital’s routine processes of care. Even though the readmission rates were 10.6% among survey responders, 14.6% of the survey responses were associated with readmissions after accounting for those who respond to surveys linked to readmission. These patients could have significant impact on cumulative experience scores.

Our study has a few limitations. First, it involves a single tertiary care academic center study, and our results may not be generalizable. Second, we did not adjust for some of the patient characteristics associated with readmissions. Patients who were admitted within 30 days are different than those not readmitted based on payor, race, length of stay, and severity of illness, and we did not adjust for these factors in our analysis. This was intentional, however. Our goal was to better understand the relationship between 30-day readmission and patient experience scores as they are used for hospital-level studies, VBP, and public reporting. For these purposes, the scores are not adjusted for factors, such as payor and length of stay. We did adjust for patient-mix adjustment factors used by CMS. Third, the response rates to the HCAHPS were low and may have biased the scores. However, HCAHPS is widely used for comparisons between hospitals has been validated, and our study results have implications with regard to comparing hospital-level performance. HCAHPS results are relevant to policy and have financial consequences.¹⁷ Fourth, our study did not directly compare whether the relationship between patient experience for the postreadmission group and nonreadmitted group was different from the relationship between the pre-readmission group and postreadmission group. It is possible that there is no difference in relationship between the groups. However, despite the small number of pre-readmission responders, these patients tended to have more favorable experience responses than those who responded after being readmitted, even after adjusting for response time. Although the P values are nonsignificant for many comparisons, the directionality of the effect is relatively consistent. Also, the vast majority of the patients fall in the postreadmission group, and these patients appear to drive the overall experience related to readmissions. Finally, since relatively few patients turned in surveys prior to readmission, we had limited power to detect a significant difference between these pre-readmission responders and nonreadmitted patients.

Our study has implications for policy makers, researchers, and providers. The HCAHPS scores of patients who are readmitted and completed the survey after being readmitted reflects their experience of both the index admission and the readmission. We did not find evidence to support that HCAHPS survey responses predict future readmissions at the patient level. Our findings do support the concept that lower readmissions rates (whether due to the patient population or processes of care that decrease readmission rates) may improve HCAHPS scores. We suggest caution in assuming that improving patient experience is likely to reduce readmission rates.

Disclosures

The authors declare no conflicts of interest.

Over 7 million surgeries are performed in United States hospitals each year. Among these surgeries, approximately 85% are noncardiac, nonvascular (NCNV) procedures.^1,2 Although the preoperative use of an angiotensin-converting enzyme inhibitor (ACEI) can be expected in as many as 13% of these surgeries,³ the optimal preoperative ACEI management strategy for patients undergoing NCNV surgeries is poorly understood.

High-quality evidence suggests that renin–angiotensin–aldosterone system (RAAS) inhibitors are associated with intraoperative hypotension among patients undergoing cardiac or vascular surgeries.^4-6 Intraoperative hypotension increases the risk of 30-day mortality,⁷ and the duration of intraoperative hypotension increases the risk of end organ damage.^8,9 This body of evidence suggests that withholding ACEIs prior to cardiac and vascular surgeries is safer than continuing ACEIs without interruption.

The evidence concerning perioperative management of ACEIs is inconclusive for patients undergoing NCNV procedures. Some studies comparing patients taking or not taking a RAAS inhibitor preoperatively describe negligible differences in the frequency of intraoperative hypotensive episodes or complications.^3,10 Others have found an increased risk of intraoperative hypotension and associated postoperative adverse events in patients continuing RAAS inhibitors preoperatively.^11,12 Current guideline discrepancies reflect the uncertainty of the evidence. The guidelines set by the American College of Cardiology and American Heart Association (ACC/AHA) suggest the uninterrupted perioperative continuation of RAAS inhibitors.¹³ The guidelines provided by the European Society of Cardiology and European Society of Anaesthesiology also suggest the continuation of RAAS inhibitors throughout the perioperative period for patients with systolic heart failure but recommend transient discontinuation for patients with hypertension.¹⁴

This randomized study aimed to compare the effect of two practical strategies for preoperative ACEI management on the perioperative blood pressure of patients undergoing NCNV surgery. The two strategies studied were the omission of the final preoperative ACEI dose and the uninterrupted continuation of ACEI therapy. We hypothesized that patients randomized to ACEI omission would experience intraoperative hypotensive episodes less frequently than those randomized to ACEI continuation.

Study Design and Setting

We performed a prospective randomized controlled trial (ClinicalTrials.gov: NCT01669434). The study was carried out in a preoperative evaluation clinic and its affiliated 489-bed academic medical center. Anesthesiologists and internal medicine physicians work collaboratively in the clinic to assess more than 5,000 patients annually (one-third of the institution’s elective surgeries). Patients were randomized 1:1 in block sizes of 5 and 10 and stratified by age < 65 and ≥ 65 years to the omission or continuation of the final preoperative ACEI dose (whether that dose was scheduled for the morning of surgery or the night prior). Preoperative clinicians enrolled patients and subsequently assigned them to intervention groups on the basis of a sequentially numbered list. Patients and healthcare providers were not blinded to allocation status. Intraoperative and postoperative management was provided in accordance with usual care as decided by treatment team.

Participants

Patients who presented to the preoperative evaluation clinic between May 2015 and November 2016 and who had been taking an ACEI for at least 6 weeks were eligible for inclusion. Patients taking angiotensin receptor blockers were excluded. Enrollment was limited to patients planning NCNV surgery. Patients planning intrathoracic, major vascular, organ transplant, and oncologic surgery were excluded. Patients undergoing outpatient procedures not requiring an overnight stay in the hospital were also excluded. Patients with preoperative clinic systolic blood pressure (SBP) <90 or ≥160 or diastolic blood pressure (DBP) <60 or ≥ 95 were excluded. Patients with moderate to severe or clinically decompensated heart failure (left ventricular ejection fraction < 40% or New York Heart Association class III or IV) and those with end-stage renal disease requiring dialysis were also excluded. Patients presenting more than once during the accrual period were eligible for the initial surgery only. All participating patients provided written informed consent. This project was approved by the University of Nebraska Medical Center Institutional Review Board.

Data Collection

Baseline characteristics were recorded by study personnel at the time of enrollment. We measured serum creatinine level at the preoperative visit and on postoperative day 1. An automated anesthesia information management system was used to measure intraoperative blood pressures every three minutes. Postoperative blood pressures through discharge were measured by hospital staff per usual care. During postoperative hospitalization, we queried patients about preoperative adherence to allocation. The digital abstraction of data from the electronic medical record was supplemented by chart review when necessary.

Outcomes

The primary outcome was intraoperative hypotension defined as any SBP < 80 mm Hg occurring from the administration of the first induction agent through transfer to the postanesthesia care unit (PACU). We also examined hypotension during anesthesia induction, which we defined as the 20-minute period following the administration of the first anesthesia induction agent. Episodes of SBP < 80 were defined as being associated with vasopressor administration when any vasopressor was administered during or within 10 min of the episode.

Secondary analyses included postoperative acute kidney injury (AKI), postoperative hypotensive and hypertensive episodes, cardiac events, and mortality. When comparing postoperative day 1 creatinine levels to preoperative creatinine levels, we used the Acute Kidney Injury Network definition of AKI as an increase in creatinine of 0.3 mg/dl or 50%.¹⁵ Postoperative hypotension was defined as any SBP < 90 mm Hg and postoperative hypertension as any SBP > 180 mm Hg occurring after arrival in the PACU. Major adverse cardiac events (MACE) were defined as a composite of acute coronary syndrome, acute heart failure, or new-onset arrhythmia. Discharge from the hospital served as the study endpoint for each patient.

Analysis

Fisher’s exact test was used to compare categorical outcomes between groups. The independent sample t-test or Wilcoxon rank–sum test, as appropriate, was used to compare continuous measures. We selected Fisher’s exact test over χ²-test to produce conservative estimates. Patients were maintained in their allocated group as randomized for analytical purposes regardless of adherence to allocation. We performed all analyses using SAS version 9.4 for Windows (SAS institute, Cary, North Carolina).

We estimated that a sample size of 300 patients would achieve 80% power to detect a difference of 0.17 between the group proportions of 0.33 and 0.50 at a significance level (ɑ) of 0.05 by using a two-sided z-test with continuity correction, assuming 15% loss to follow-up. This estimate allowed for 1 interim analysis using the O’Brien-Fleming spending function truncated at three standard deviations to determine the test boundaries. The monitoring boundary P values associated with the interim analysis were .003, and the threshold P value for the final analysis was .049.

Study Flow

A total of 453 patients were screened for eligibility. Among these patients, 162 were excluded, and the remaining 291 patients were randomized (Figure 1). Surgery was cancelled in six patients allocated to omission and in four patients allocated to continuation arms, respectively. Moreover, three patients in the omission arm were excluded from the analysis following randomization. Specifically, one was excluded because of early discharge without overnight stay, one was excluded because of withdrawal of consent, and one was excluded because of missing primary outcome data. In addition, three cases in the continuation arm were excluded following randomization because of the preoperative (permanent) discontinuation of ACEI therapy in two cases and discharge without an overnight stay in one case. Finally, 275 patients were included in the analysis: 137 in the ACEI omission group and 138 in the ACEI continuation group. Adherence to allocation was 88% and 92% in the omission and continuation groups, respectively.

Baseline Characteristics

The demographic data of patients allocated to ACEI omission and those allocated to ACEI continuation were similar (Table 1). A large majority of patients in both groups took the ACEI lisinopril. Overall, 187 of 275 (68%) patients were taking at least 1 antihypertensive agent, most commonly a diuretic, in addition to an ACEI. SBP measured during the preoperative clinic visit averaged 136.5 mm Hg and did not differ significantly between groups (P = .84).

Surgical Variables

General anesthesia was the most commonly utilized technique, although spinal and regional anesthesia were also represented (Table 1). The majority of cases in both groups were planning for orthopedic and spinal surgery. The method of anesthesia or type of surgery between patients allocated to ACEI omission and those allocated to continuation did not differ (P = .61 and P = .45 respectively).

Episodes of Intraoperative Hypotension

Intraoperative SBPs are displayed in Figure 2, and hemodynamic outcomes are summarized in Table 2. Episodes of SBP < 80 mm Hg during anesthesia induction were numerically less frequent in the omission group than in the continuation group; the difference between groups, however, was not statistically significant (24 of 137 [18%] vs 38 of 138 [28%], RR: 0.64, 95% CI: 0.40 to 1.00, P = .06).

Duration of Intraoperative Hypotension

The median cumulative duration of intraoperative SBP < 80 was two minutes (range 0-41) in patients allocated to the ACEI omission group compared with seven minutes (range 0-214) in those allocated to the continuation group (P < .01). The median cumulative duration of mean arterial pressure < 55 mm Hg was also shorter in the omission group (median 0 min [range 0-39] vs 3 min [range 0-122], P < .01) than in the continuation group. The duration of surgery did not differ between groups (median 141 min [range 77-554] vs 142 min [range 57-665], P = .97).

Postoperative Outcomes

RAAS inhibitor therapy was resumed within 48 h after surgery in 122 of 137 (89%) patients allocated to the omission group and in 128 of 138 (93%) patients allocated to the continuation group (RR: 0.96, 95% CI: 0.89-1.03, P = .30).

Patients allocated to the omission group were significantly less likely to experience postoperative hypotension (15 of 137 [11%] vs 31 of 138 [22%], RR: 0.49, 95% CI: 0.28 to 0.86, P = .02) and significantly more likely to experience severe postoperative hypertension (33 of 137 [24%] vs 17 of 138 [12%], RR: 1.95, 95% CI: 1.14 to 3.34, P = .01) than those allocated to the continuation group. The occurrences of postoperative AKI (RR: 0.60, 95% CI: 0.23 to 1.60, P = .44) or MACE (RR: 4.03, 95% CI: 0.46 to 35.59, P = .21) in the omission group did not differ from the continuation group. The two groups exhibited similar PACU recovery time (mean 97.2 min) and overall hospital length of stay (mean 3.0 days) (P = .49 and P = .56 ). No episodes of inpatient mortality in either group were observed.

The omission of the final preoperative ACEI dose was associated with a significant reduction in the risk of intraoperative hypotension in patients undergoing NCNV surgery. This result confirmed our hypothesis. Coupled with the knowledge that intraoperative hypotension is associated with an increased risk of complications and mortality,^7-9,16 this study favors the omission of the final preoperative ACEI dose prior to NCNV surgeries.

Our findings are in agreement with those of previous randomized studies that explored this question^4,5 and help extend results from cardiac and vascular surgeries to NCNV surgeries. Previous studies on the use of RAAS inhibitors in NCNV surgeries did not employ randomization and yielded mixed results.^3,10-12,17 A large single-institution study (n = 18,056) noted no difference in intraoperative blood pressure between patients taking ACEIs and a matched group of non-ACEI users.³ More recently, a subgroup analysis of the international VISION study showed that omitting RAAS inhibitors on the day of surgery reduced the risk of intraoperative hypotension.¹¹ In that analysis, however, only a small amount of the variability in preoperative RAAS inhibitor management was explainable by modeling known factors, thus allowing for the possibility of unmeasured confounding. Our study, which minimized confounding through randomization, is the first to prospectively compare protocols for patients undergoing NCNV surgery. In contrast to previous studies, the present study was able to report the lack of difference in postoperative RAAS inhibitor administration between study groups. Postoperative RAAS inhibitor management affects complications and mortality.^18,19

Our present finding that preoperative ACEI management affects postoperative hypotensive and hypertensive events conflicts with some previous findings.^11,20 However, recent evidence has revealed that postoperative hypotensive episodes are associated with vascular events and mortality.^11,21 In the context of that evidence, our study lends further support to the omission of the final preoperative ACEI dose. However, we did not detect any decrease in AKI, MACE, or mortality in the ACEI omission group.

This study should be considered in light of its limitations. The pragmatic nature of the study allowed for certain potential biases. Although adherence to allocation was high, the specific ACEI agent taken and the exact timing of the final dose in relation to surgery were not controlled. Anesthetic and postoperative management decisions were made by the treatment team and may have systematically varied given that the treatment team was not blinded to allocation. Furthermore, all outcome data were collected as part of routine care and may not have captured events with great fidelity. Generalizability is limited by the execution of the study at a single academic institution, the preponderance of orthopedic and spine surgeries, and by the negligible representation of ethnicities other than Caucasian. Additionally, recruitment from the preoperative evaluation clinic likely resulted in a patient group with greater comorbidity than the overall population of patients undergoing NCNV surgery. This study was powered for intraoperative hypotension and not postoperative outcomes. Our primary outcome, intraoperative hypotension, is an intermediate measure but one that has well-established associations with adverse outcomes, including mortality. One study showed that sustaining an intraoperative SBP below 70 mm Hg for longer than 5 min increased the risk of mortality from less than 1% to nearly 6%.¹⁶ A large study detected an increase in mortality associated with SBP sustained below 80 mm Hg for 10 min or longer.⁷ Intraoperative hypotension has also been associated with postoperative AKI and myocardial injury.^8,9,12

Many of the limitations of the current study could be addressed by a large randomized controlled trial of ACEI management prior to NCNV surgeries that examines clinically important endpoints beyond intraoperative hypotension. Several specific aspects of perioperative RAAS inhibitor management also deserve further investigation. Our findings may not be generalizable to patients taking ARBs or to patients with congestive heart failure. The preoperative management of ARBs and the preoperative management of RAAS inhibitors in those with congestive heart failure are important areas of focus for future research. Lastly, our finding that preoperative ACEI management decisions can affect postoperative hypotensive and hypertensive events should be substantiated by future research, and any negative consequences of those events should be further explored.

Nonetheless, our study is the largest randomized study of preoperative RAAS inhibition published to date. More than twice as many patients were randomized in this study than in all previous randomized studies combined.^4-6 To the best of our knowledge, this is also the first randomized study evaluating NCNV surgeries. Finally, our use of a practical ACEI omission protocol based on known pharmacokinetics allows for direct application to clinical practice.

Hypertension is among the most common chronic conditions encountered in patients planning surgery, and ACEIs are among the most frequently prescribed antihypertensive medications. This study showed that ACEI continuation is associated with an increased frequency and cumulative duration of intraoperative hypotension. These findings, while at odds with current ACC/AHA guidelines, align with the findings of a meta-analysis on this subject and with recent literature.^3,11-13,22

Acknowledgments

The authors wish to thank Miranda M Fricke, MS, PA-C; Tiffany K Hillyard, APRN-FNP; and Barbara Sink, MPAS, PA-C who assisted in the design and conduct of patient enrollment and randomization procedures.

Disclosures

The authors have no relevant financial conflicts of interest to report.

Funding

This study was subsidized by a grant from the University of Nebraska Medical Center Research Support Fund. The funding source had no role in the design, conduct, analysis, or reporting of the study.

Over 7 million surgeries are performed in United States hospitals each year. Among these surgeries, approximately 85% are noncardiac, nonvascular (NCNV) procedures.^1,2 Although the preoperative use of an angiotensin-converting enzyme inhibitor (ACEI) can be expected in as many as 13% of these surgeries,³ the optimal preoperative ACEI management strategy for patients undergoing NCNV surgeries is poorly understood.

High-quality evidence suggests that renin–angiotensin–aldosterone system (RAAS) inhibitors are associated with intraoperative hypotension among patients undergoing cardiac or vascular surgeries.^4-6 Intraoperative hypotension increases the risk of 30-day mortality,⁷ and the duration of intraoperative hypotension increases the risk of end organ damage.^8,9 This body of evidence suggests that withholding ACEIs prior to cardiac and vascular surgeries is safer than continuing ACEIs without interruption.

The evidence concerning perioperative management of ACEIs is inconclusive for patients undergoing NCNV procedures. Some studies comparing patients taking or not taking a RAAS inhibitor preoperatively describe negligible differences in the frequency of intraoperative hypotensive episodes or complications.^3,10 Others have found an increased risk of intraoperative hypotension and associated postoperative adverse events in patients continuing RAAS inhibitors preoperatively.^11,12 Current guideline discrepancies reflect the uncertainty of the evidence. The guidelines set by the American College of Cardiology and American Heart Association (ACC/AHA) suggest the uninterrupted perioperative continuation of RAAS inhibitors.¹³ The guidelines provided by the European Society of Cardiology and European Society of Anaesthesiology also suggest the continuation of RAAS inhibitors throughout the perioperative period for patients with systolic heart failure but recommend transient discontinuation for patients with hypertension.¹⁴

This randomized study aimed to compare the effect of two practical strategies for preoperative ACEI management on the perioperative blood pressure of patients undergoing NCNV surgery. The two strategies studied were the omission of the final preoperative ACEI dose and the uninterrupted continuation of ACEI therapy. We hypothesized that patients randomized to ACEI omission would experience intraoperative hypotensive episodes less frequently than those randomized to ACEI continuation.

Study Design and Setting

We performed a prospective randomized controlled trial (ClinicalTrials.gov: NCT01669434). The study was carried out in a preoperative evaluation clinic and its affiliated 489-bed academic medical center. Anesthesiologists and internal medicine physicians work collaboratively in the clinic to assess more than 5,000 patients annually (one-third of the institution’s elective surgeries). Patients were randomized 1:1 in block sizes of 5 and 10 and stratified by age < 65 and ≥ 65 years to the omission or continuation of the final preoperative ACEI dose (whether that dose was scheduled for the morning of surgery or the night prior). Preoperative clinicians enrolled patients and subsequently assigned them to intervention groups on the basis of a sequentially numbered list. Patients and healthcare providers were not blinded to allocation status. Intraoperative and postoperative management was provided in accordance with usual care as decided by treatment team.

Participants

Patients who presented to the preoperative evaluation clinic between May 2015 and November 2016 and who had been taking an ACEI for at least 6 weeks were eligible for inclusion. Patients taking angiotensin receptor blockers were excluded. Enrollment was limited to patients planning NCNV surgery. Patients planning intrathoracic, major vascular, organ transplant, and oncologic surgery were excluded. Patients undergoing outpatient procedures not requiring an overnight stay in the hospital were also excluded. Patients with preoperative clinic systolic blood pressure (SBP) <90 or ≥160 or diastolic blood pressure (DBP) <60 or ≥ 95 were excluded. Patients with moderate to severe or clinically decompensated heart failure (left ventricular ejection fraction < 40% or New York Heart Association class III or IV) and those with end-stage renal disease requiring dialysis were also excluded. Patients presenting more than once during the accrual period were eligible for the initial surgery only. All participating patients provided written informed consent. This project was approved by the University of Nebraska Medical Center Institutional Review Board.

Data Collection

Baseline characteristics were recorded by study personnel at the time of enrollment. We measured serum creatinine level at the preoperative visit and on postoperative day 1. An automated anesthesia information management system was used to measure intraoperative blood pressures every three minutes. Postoperative blood pressures through discharge were measured by hospital staff per usual care. During postoperative hospitalization, we queried patients about preoperative adherence to allocation. The digital abstraction of data from the electronic medical record was supplemented by chart review when necessary.

Outcomes

The primary outcome was intraoperative hypotension defined as any SBP < 80 mm Hg occurring from the administration of the first induction agent through transfer to the postanesthesia care unit (PACU). We also examined hypotension during anesthesia induction, which we defined as the 20-minute period following the administration of the first anesthesia induction agent. Episodes of SBP < 80 were defined as being associated with vasopressor administration when any vasopressor was administered during or within 10 min of the episode.

Secondary analyses included postoperative acute kidney injury (AKI), postoperative hypotensive and hypertensive episodes, cardiac events, and mortality. When comparing postoperative day 1 creatinine levels to preoperative creatinine levels, we used the Acute Kidney Injury Network definition of AKI as an increase in creatinine of 0.3 mg/dl or 50%.¹⁵ Postoperative hypotension was defined as any SBP < 90 mm Hg and postoperative hypertension as any SBP > 180 mm Hg occurring after arrival in the PACU. Major adverse cardiac events (MACE) were defined as a composite of acute coronary syndrome, acute heart failure, or new-onset arrhythmia. Discharge from the hospital served as the study endpoint for each patient.

Analysis

Fisher’s exact test was used to compare categorical outcomes between groups. The independent sample t-test or Wilcoxon rank–sum test, as appropriate, was used to compare continuous measures. We selected Fisher’s exact test over χ²-test to produce conservative estimates. Patients were maintained in their allocated group as randomized for analytical purposes regardless of adherence to allocation. We performed all analyses using SAS version 9.4 for Windows (SAS institute, Cary, North Carolina).

We estimated that a sample size of 300 patients would achieve 80% power to detect a difference of 0.17 between the group proportions of 0.33 and 0.50 at a significance level (ɑ) of 0.05 by using a two-sided z-test with continuity correction, assuming 15% loss to follow-up. This estimate allowed for 1 interim analysis using the O’Brien-Fleming spending function truncated at three standard deviations to determine the test boundaries. The monitoring boundary P values associated with the interim analysis were .003, and the threshold P value for the final analysis was .049.

Study Flow

A total of 453 patients were screened for eligibility. Among these patients, 162 were excluded, and the remaining 291 patients were randomized (Figure 1). Surgery was cancelled in six patients allocated to omission and in four patients allocated to continuation arms, respectively. Moreover, three patients in the omission arm were excluded from the analysis following randomization. Specifically, one was excluded because of early discharge without overnight stay, one was excluded because of withdrawal of consent, and one was excluded because of missing primary outcome data. In addition, three cases in the continuation arm were excluded following randomization because of the preoperative (permanent) discontinuation of ACEI therapy in two cases and discharge without an overnight stay in one case. Finally, 275 patients were included in the analysis: 137 in the ACEI omission group and 138 in the ACEI continuation group. Adherence to allocation was 88% and 92% in the omission and continuation groups, respectively.

Baseline Characteristics

The demographic data of patients allocated to ACEI omission and those allocated to ACEI continuation were similar (Table 1). A large majority of patients in both groups took the ACEI lisinopril. Overall, 187 of 275 (68%) patients were taking at least 1 antihypertensive agent, most commonly a diuretic, in addition to an ACEI. SBP measured during the preoperative clinic visit averaged 136.5 mm Hg and did not differ significantly between groups (P = .84).

Surgical Variables

General anesthesia was the most commonly utilized technique, although spinal and regional anesthesia were also represented (Table 1). The majority of cases in both groups were planning for orthopedic and spinal surgery. The method of anesthesia or type of surgery between patients allocated to ACEI omission and those allocated to continuation did not differ (P = .61 and P = .45 respectively).

Episodes of Intraoperative Hypotension

Intraoperative SBPs are displayed in Figure 2, and hemodynamic outcomes are summarized in Table 2. Episodes of SBP < 80 mm Hg during anesthesia induction were numerically less frequent in the omission group than in the continuation group; the difference between groups, however, was not statistically significant (24 of 137 [18%] vs 38 of 138 [28%], RR: 0.64, 95% CI: 0.40 to 1.00, P = .06).

Duration of Intraoperative Hypotension

The median cumulative duration of intraoperative SBP < 80 was two minutes (range 0-41) in patients allocated to the ACEI omission group compared with seven minutes (range 0-214) in those allocated to the continuation group (P < .01). The median cumulative duration of mean arterial pressure < 55 mm Hg was also shorter in the omission group (median 0 min [range 0-39] vs 3 min [range 0-122], P < .01) than in the continuation group. The duration of surgery did not differ between groups (median 141 min [range 77-554] vs 142 min [range 57-665], P = .97).

Postoperative Outcomes

RAAS inhibitor therapy was resumed within 48 h after surgery in 122 of 137 (89%) patients allocated to the omission group and in 128 of 138 (93%) patients allocated to the continuation group (RR: 0.96, 95% CI: 0.89-1.03, P = .30).

Patients allocated to the omission group were significantly less likely to experience postoperative hypotension (15 of 137 [11%] vs 31 of 138 [22%], RR: 0.49, 95% CI: 0.28 to 0.86, P = .02) and significantly more likely to experience severe postoperative hypertension (33 of 137 [24%] vs 17 of 138 [12%], RR: 1.95, 95% CI: 1.14 to 3.34, P = .01) than those allocated to the continuation group. The occurrences of postoperative AKI (RR: 0.60, 95% CI: 0.23 to 1.60, P = .44) or MACE (RR: 4.03, 95% CI: 0.46 to 35.59, P = .21) in the omission group did not differ from the continuation group. The two groups exhibited similar PACU recovery time (mean 97.2 min) and overall hospital length of stay (mean 3.0 days) (P = .49 and P = .56 ). No episodes of inpatient mortality in either group were observed.

The omission of the final preoperative ACEI dose was associated with a significant reduction in the risk of intraoperative hypotension in patients undergoing NCNV surgery. This result confirmed our hypothesis. Coupled with the knowledge that intraoperative hypotension is associated with an increased risk of complications and mortality,^7-9,16 this study favors the omission of the final preoperative ACEI dose prior to NCNV surgeries.

Our findings are in agreement with those of previous randomized studies that explored this question^4,5 and help extend results from cardiac and vascular surgeries to NCNV surgeries. Previous studies on the use of RAAS inhibitors in NCNV surgeries did not employ randomization and yielded mixed results.^3,10-12,17 A large single-institution study (n = 18,056) noted no difference in intraoperative blood pressure between patients taking ACEIs and a matched group of non-ACEI users.³ More recently, a subgroup analysis of the international VISION study showed that omitting RAAS inhibitors on the day of surgery reduced the risk of intraoperative hypotension.¹¹ In that analysis, however, only a small amount of the variability in preoperative RAAS inhibitor management was explainable by modeling known factors, thus allowing for the possibility of unmeasured confounding. Our study, which minimized confounding through randomization, is the first to prospectively compare protocols for patients undergoing NCNV surgery. In contrast to previous studies, the present study was able to report the lack of difference in postoperative RAAS inhibitor administration between study groups. Postoperative RAAS inhibitor management affects complications and mortality.^18,19

Our present finding that preoperative ACEI management affects postoperative hypotensive and hypertensive events conflicts with some previous findings.^11,20 However, recent evidence has revealed that postoperative hypotensive episodes are associated with vascular events and mortality.^11,21 In the context of that evidence, our study lends further support to the omission of the final preoperative ACEI dose. However, we did not detect any decrease in AKI, MACE, or mortality in the ACEI omission group.

This study should be considered in light of its limitations. The pragmatic nature of the study allowed for certain potential biases. Although adherence to allocation was high, the specific ACEI agent taken and the exact timing of the final dose in relation to surgery were not controlled. Anesthetic and postoperative management decisions were made by the treatment team and may have systematically varied given that the treatment team was not blinded to allocation. Furthermore, all outcome data were collected as part of routine care and may not have captured events with great fidelity. Generalizability is limited by the execution of the study at a single academic institution, the preponderance of orthopedic and spine surgeries, and by the negligible representation of ethnicities other than Caucasian. Additionally, recruitment from the preoperative evaluation clinic likely resulted in a patient group with greater comorbidity than the overall population of patients undergoing NCNV surgery. This study was powered for intraoperative hypotension and not postoperative outcomes. Our primary outcome, intraoperative hypotension, is an intermediate measure but one that has well-established associations with adverse outcomes, including mortality. One study showed that sustaining an intraoperative SBP below 70 mm Hg for longer than 5 min increased the risk of mortality from less than 1% to nearly 6%.¹⁶ A large study detected an increase in mortality associated with SBP sustained below 80 mm Hg for 10 min or longer.⁷ Intraoperative hypotension has also been associated with postoperative AKI and myocardial injury.^8,9,12

Many of the limitations of the current study could be addressed by a large randomized controlled trial of ACEI management prior to NCNV surgeries that examines clinically important endpoints beyond intraoperative hypotension. Several specific aspects of perioperative RAAS inhibitor management also deserve further investigation. Our findings may not be generalizable to patients taking ARBs or to patients with congestive heart failure. The preoperative management of ARBs and the preoperative management of RAAS inhibitors in those with congestive heart failure are important areas of focus for future research. Lastly, our finding that preoperative ACEI management decisions can affect postoperative hypotensive and hypertensive events should be substantiated by future research, and any negative consequences of those events should be further explored.

Nonetheless, our study is the largest randomized study of preoperative RAAS inhibition published to date. More than twice as many patients were randomized in this study than in all previous randomized studies combined.^4-6 To the best of our knowledge, this is also the first randomized study evaluating NCNV surgeries. Finally, our use of a practical ACEI omission protocol based on known pharmacokinetics allows for direct application to clinical practice.

Hypertension is among the most common chronic conditions encountered in patients planning surgery, and ACEIs are among the most frequently prescribed antihypertensive medications. This study showed that ACEI continuation is associated with an increased frequency and cumulative duration of intraoperative hypotension. These findings, while at odds with current ACC/AHA guidelines, align with the findings of a meta-analysis on this subject and with recent literature.^3,11-13,22

Acknowledgments

The authors wish to thank Miranda M Fricke, MS, PA-C; Tiffany K Hillyard, APRN-FNP; and Barbara Sink, MPAS, PA-C who assisted in the design and conduct of patient enrollment and randomization procedures.

Disclosures

The authors have no relevant financial conflicts of interest to report.

Funding

This study was subsidized by a grant from the University of Nebraska Medical Center Research Support Fund. The funding source had no role in the design, conduct, analysis, or reporting of the study.

Over 7 million surgeries are performed in United States hospitals each year. Among these surgeries, approximately 85% are noncardiac, nonvascular (NCNV) procedures.^1,2 Although the preoperative use of an angiotensin-converting enzyme inhibitor (ACEI) can be expected in as many as 13% of these surgeries,³ the optimal preoperative ACEI management strategy for patients undergoing NCNV surgeries is poorly understood.

High-quality evidence suggests that renin–angiotensin–aldosterone system (RAAS) inhibitors are associated with intraoperative hypotension among patients undergoing cardiac or vascular surgeries.^4-6 Intraoperative hypotension increases the risk of 30-day mortality,⁷ and the duration of intraoperative hypotension increases the risk of end organ damage.^8,9 This body of evidence suggests that withholding ACEIs prior to cardiac and vascular surgeries is safer than continuing ACEIs without interruption.

The evidence concerning perioperative management of ACEIs is inconclusive for patients undergoing NCNV procedures. Some studies comparing patients taking or not taking a RAAS inhibitor preoperatively describe negligible differences in the frequency of intraoperative hypotensive episodes or complications.^3,10 Others have found an increased risk of intraoperative hypotension and associated postoperative adverse events in patients continuing RAAS inhibitors preoperatively.^11,12 Current guideline discrepancies reflect the uncertainty of the evidence. The guidelines set by the American College of Cardiology and American Heart Association (ACC/AHA) suggest the uninterrupted perioperative continuation of RAAS inhibitors.¹³ The guidelines provided by the European Society of Cardiology and European Society of Anaesthesiology also suggest the continuation of RAAS inhibitors throughout the perioperative period for patients with systolic heart failure but recommend transient discontinuation for patients with hypertension.¹⁴

This randomized study aimed to compare the effect of two practical strategies for preoperative ACEI management on the perioperative blood pressure of patients undergoing NCNV surgery. The two strategies studied were the omission of the final preoperative ACEI dose and the uninterrupted continuation of ACEI therapy. We hypothesized that patients randomized to ACEI omission would experience intraoperative hypotensive episodes less frequently than those randomized to ACEI continuation.

Study Design and Setting

We performed a prospective randomized controlled trial (ClinicalTrials.gov: NCT01669434). The study was carried out in a preoperative evaluation clinic and its affiliated 489-bed academic medical center. Anesthesiologists and internal medicine physicians work collaboratively in the clinic to assess more than 5,000 patients annually (one-third of the institution’s elective surgeries). Patients were randomized 1:1 in block sizes of 5 and 10 and stratified by age < 65 and ≥ 65 years to the omission or continuation of the final preoperative ACEI dose (whether that dose was scheduled for the morning of surgery or the night prior). Preoperative clinicians enrolled patients and subsequently assigned them to intervention groups on the basis of a sequentially numbered list. Patients and healthcare providers were not blinded to allocation status. Intraoperative and postoperative management was provided in accordance with usual care as decided by treatment team.

Participants

Patients who presented to the preoperative evaluation clinic between May 2015 and November 2016 and who had been taking an ACEI for at least 6 weeks were eligible for inclusion. Patients taking angiotensin receptor blockers were excluded. Enrollment was limited to patients planning NCNV surgery. Patients planning intrathoracic, major vascular, organ transplant, and oncologic surgery were excluded. Patients undergoing outpatient procedures not requiring an overnight stay in the hospital were also excluded. Patients with preoperative clinic systolic blood pressure (SBP) <90 or ≥160 or diastolic blood pressure (DBP) <60 or ≥ 95 were excluded. Patients with moderate to severe or clinically decompensated heart failure (left ventricular ejection fraction < 40% or New York Heart Association class III or IV) and those with end-stage renal disease requiring dialysis were also excluded. Patients presenting more than once during the accrual period were eligible for the initial surgery only. All participating patients provided written informed consent. This project was approved by the University of Nebraska Medical Center Institutional Review Board.

Data Collection

Baseline characteristics were recorded by study personnel at the time of enrollment. We measured serum creatinine level at the preoperative visit and on postoperative day 1. An automated anesthesia information management system was used to measure intraoperative blood pressures every three minutes. Postoperative blood pressures through discharge were measured by hospital staff per usual care. During postoperative hospitalization, we queried patients about preoperative adherence to allocation. The digital abstraction of data from the electronic medical record was supplemented by chart review when necessary.

Outcomes

The primary outcome was intraoperative hypotension defined as any SBP < 80 mm Hg occurring from the administration of the first induction agent through transfer to the postanesthesia care unit (PACU). We also examined hypotension during anesthesia induction, which we defined as the 20-minute period following the administration of the first anesthesia induction agent. Episodes of SBP < 80 were defined as being associated with vasopressor administration when any vasopressor was administered during or within 10 min of the episode.

Secondary analyses included postoperative acute kidney injury (AKI), postoperative hypotensive and hypertensive episodes, cardiac events, and mortality. When comparing postoperative day 1 creatinine levels to preoperative creatinine levels, we used the Acute Kidney Injury Network definition of AKI as an increase in creatinine of 0.3 mg/dl or 50%.¹⁵ Postoperative hypotension was defined as any SBP < 90 mm Hg and postoperative hypertension as any SBP > 180 mm Hg occurring after arrival in the PACU. Major adverse cardiac events (MACE) were defined as a composite of acute coronary syndrome, acute heart failure, or new-onset arrhythmia. Discharge from the hospital served as the study endpoint for each patient.

Analysis

Fisher’s exact test was used to compare categorical outcomes between groups. The independent sample t-test or Wilcoxon rank–sum test, as appropriate, was used to compare continuous measures. We selected Fisher’s exact test over χ²-test to produce conservative estimates. Patients were maintained in their allocated group as randomized for analytical purposes regardless of adherence to allocation. We performed all analyses using SAS version 9.4 for Windows (SAS institute, Cary, North Carolina).

We estimated that a sample size of 300 patients would achieve 80% power to detect a difference of 0.17 between the group proportions of 0.33 and 0.50 at a significance level (ɑ) of 0.05 by using a two-sided z-test with continuity correction, assuming 15% loss to follow-up. This estimate allowed for 1 interim analysis using the O’Brien-Fleming spending function truncated at three standard deviations to determine the test boundaries. The monitoring boundary P values associated with the interim analysis were .003, and the threshold P value for the final analysis was .049.

Study Flow

A total of 453 patients were screened for eligibility. Among these patients, 162 were excluded, and the remaining 291 patients were randomized (Figure 1). Surgery was cancelled in six patients allocated to omission and in four patients allocated to continuation arms, respectively. Moreover, three patients in the omission arm were excluded from the analysis following randomization. Specifically, one was excluded because of early discharge without overnight stay, one was excluded because of withdrawal of consent, and one was excluded because of missing primary outcome data. In addition, three cases in the continuation arm were excluded following randomization because of the preoperative (permanent) discontinuation of ACEI therapy in two cases and discharge without an overnight stay in one case. Finally, 275 patients were included in the analysis: 137 in the ACEI omission group and 138 in the ACEI continuation group. Adherence to allocation was 88% and 92% in the omission and continuation groups, respectively.

Baseline Characteristics

The demographic data of patients allocated to ACEI omission and those allocated to ACEI continuation were similar (Table 1). A large majority of patients in both groups took the ACEI lisinopril. Overall, 187 of 275 (68%) patients were taking at least 1 antihypertensive agent, most commonly a diuretic, in addition to an ACEI. SBP measured during the preoperative clinic visit averaged 136.5 mm Hg and did not differ significantly between groups (P = .84).

Surgical Variables

General anesthesia was the most commonly utilized technique, although spinal and regional anesthesia were also represented (Table 1). The majority of cases in both groups were planning for orthopedic and spinal surgery. The method of anesthesia or type of surgery between patients allocated to ACEI omission and those allocated to continuation did not differ (P = .61 and P = .45 respectively).

Episodes of Intraoperative Hypotension

Intraoperative SBPs are displayed in Figure 2, and hemodynamic outcomes are summarized in Table 2. Episodes of SBP < 80 mm Hg during anesthesia induction were numerically less frequent in the omission group than in the continuation group; the difference between groups, however, was not statistically significant (24 of 137 [18%] vs 38 of 138 [28%], RR: 0.64, 95% CI: 0.40 to 1.00, P = .06).

Duration of Intraoperative Hypotension

The median cumulative duration of intraoperative SBP < 80 was two minutes (range 0-41) in patients allocated to the ACEI omission group compared with seven minutes (range 0-214) in those allocated to the continuation group (P < .01). The median cumulative duration of mean arterial pressure < 55 mm Hg was also shorter in the omission group (median 0 min [range 0-39] vs 3 min [range 0-122], P < .01) than in the continuation group. The duration of surgery did not differ between groups (median 141 min [range 77-554] vs 142 min [range 57-665], P = .97).

Postoperative Outcomes

RAAS inhibitor therapy was resumed within 48 h after surgery in 122 of 137 (89%) patients allocated to the omission group and in 128 of 138 (93%) patients allocated to the continuation group (RR: 0.96, 95% CI: 0.89-1.03, P = .30).

Patients allocated to the omission group were significantly less likely to experience postoperative hypotension (15 of 137 [11%] vs 31 of 138 [22%], RR: 0.49, 95% CI: 0.28 to 0.86, P = .02) and significantly more likely to experience severe postoperative hypertension (33 of 137 [24%] vs 17 of 138 [12%], RR: 1.95, 95% CI: 1.14 to 3.34, P = .01) than those allocated to the continuation group. The occurrences of postoperative AKI (RR: 0.60, 95% CI: 0.23 to 1.60, P = .44) or MACE (RR: 4.03, 95% CI: 0.46 to 35.59, P = .21) in the omission group did not differ from the continuation group. The two groups exhibited similar PACU recovery time (mean 97.2 min) and overall hospital length of stay (mean 3.0 days) (P = .49 and P = .56 ). No episodes of inpatient mortality in either group were observed.

The omission of the final preoperative ACEI dose was associated with a significant reduction in the risk of intraoperative hypotension in patients undergoing NCNV surgery. This result confirmed our hypothesis. Coupled with the knowledge that intraoperative hypotension is associated with an increased risk of complications and mortality,^7-9,16 this study favors the omission of the final preoperative ACEI dose prior to NCNV surgeries.

Our findings are in agreement with those of previous randomized studies that explored this question^4,5 and help extend results from cardiac and vascular surgeries to NCNV surgeries. Previous studies on the use of RAAS inhibitors in NCNV surgeries did not employ randomization and yielded mixed results.^3,10-12,17 A large single-institution study (n = 18,056) noted no difference in intraoperative blood pressure between patients taking ACEIs and a matched group of non-ACEI users.³ More recently, a subgroup analysis of the international VISION study showed that omitting RAAS inhibitors on the day of surgery reduced the risk of intraoperative hypotension.¹¹ In that analysis, however, only a small amount of the variability in preoperative RAAS inhibitor management was explainable by modeling known factors, thus allowing for the possibility of unmeasured confounding. Our study, which minimized confounding through randomization, is the first to prospectively compare protocols for patients undergoing NCNV surgery. In contrast to previous studies, the present study was able to report the lack of difference in postoperative RAAS inhibitor administration between study groups. Postoperative RAAS inhibitor management affects complications and mortality.^18,19

Our present finding that preoperative ACEI management affects postoperative hypotensive and hypertensive events conflicts with some previous findings.^11,20 However, recent evidence has revealed that postoperative hypotensive episodes are associated with vascular events and mortality.^11,21 In the context of that evidence, our study lends further support to the omission of the final preoperative ACEI dose. However, we did not detect any decrease in AKI, MACE, or mortality in the ACEI omission group.

This study should be considered in light of its limitations. The pragmatic nature of the study allowed for certain potential biases. Although adherence to allocation was high, the specific ACEI agent taken and the exact timing of the final dose in relation to surgery were not controlled. Anesthetic and postoperative management decisions were made by the treatment team and may have systematically varied given that the treatment team was not blinded to allocation. Furthermore, all outcome data were collected as part of routine care and may not have captured events with great fidelity. Generalizability is limited by the execution of the study at a single academic institution, the preponderance of orthopedic and spine surgeries, and by the negligible representation of ethnicities other than Caucasian. Additionally, recruitment from the preoperative evaluation clinic likely resulted in a patient group with greater comorbidity than the overall population of patients undergoing NCNV surgery. This study was powered for intraoperative hypotension and not postoperative outcomes. Our primary outcome, intraoperative hypotension, is an intermediate measure but one that has well-established associations with adverse outcomes, including mortality. One study showed that sustaining an intraoperative SBP below 70 mm Hg for longer than 5 min increased the risk of mortality from less than 1% to nearly 6%.¹⁶ A large study detected an increase in mortality associated with SBP sustained below 80 mm Hg for 10 min or longer.⁷ Intraoperative hypotension has also been associated with postoperative AKI and myocardial injury.^8,9,12

Many of the limitations of the current study could be addressed by a large randomized controlled trial of ACEI management prior to NCNV surgeries that examines clinically important endpoints beyond intraoperative hypotension. Several specific aspects of perioperative RAAS inhibitor management also deserve further investigation. Our findings may not be generalizable to patients taking ARBs or to patients with congestive heart failure. The preoperative management of ARBs and the preoperative management of RAAS inhibitors in those with congestive heart failure are important areas of focus for future research. Lastly, our finding that preoperative ACEI management decisions can affect postoperative hypotensive and hypertensive events should be substantiated by future research, and any negative consequences of those events should be further explored.

Nonetheless, our study is the largest randomized study of preoperative RAAS inhibition published to date. More than twice as many patients were randomized in this study than in all previous randomized studies combined.^4-6 To the best of our knowledge, this is also the first randomized study evaluating NCNV surgeries. Finally, our use of a practical ACEI omission protocol based on known pharmacokinetics allows for direct application to clinical practice.

Hypertension is among the most common chronic conditions encountered in patients planning surgery, and ACEIs are among the most frequently prescribed antihypertensive medications. This study showed that ACEI continuation is associated with an increased frequency and cumulative duration of intraoperative hypotension. These findings, while at odds with current ACC/AHA guidelines, align with the findings of a meta-analysis on this subject and with recent literature.^3,11-13,22

Acknowledgments

The authors wish to thank Miranda M Fricke, MS, PA-C; Tiffany K Hillyard, APRN-FNP; and Barbara Sink, MPAS, PA-C who assisted in the design and conduct of patient enrollment and randomization procedures.

Disclosures

The authors have no relevant financial conflicts of interest to report.

Funding

This study was subsidized by a grant from the University of Nebraska Medical Center Research Support Fund. The funding source had no role in the design, conduct, analysis, or reporting of the study.

Approximately 14% of children who sustain a concussion are admitted to the hospital,¹ although admission rates reportedly vary substantially among pediatric hospitals.² Children hospitalized for concussion may be at a higher risk for persistent postconcussive symptoms,^3,4 yet little is known about this subset of children and how they are managed while in the hospital. Characterizing children hospitalized for concussion and describing the inpatient care they received will promote hypothesis generation for further inquiry into indications for admission, as well as the relationship between inpatient management and concussion recovery.

We described a cohort of children admitted to 40 pediatric hospitals primarily for concussion and detailed care delivered during hospitalization. We explored individual-level factors and their association with prolonged length of stay (LOS) and emergency department (ED) readmission. Finally, we evaluated if there had been changes in inpatient care over the 8-year study period.

Study Design

The Institutional Review Board determined that this retrospective cohort study was exempt from review.

Data Source

The Children’s Hospital Association’s Pediatric Health Information System (PHIS) is an administrative database from pediatric hospitals located within 17 major metropolitan areas in the United States. Data include: service dates, patient demographics, payer type, diagnosis codes, resource utilization information (eg, medications), and hospital characteristics.^1,5 De-identified data undergo reliability and validity checks prior to inclusion.^1,5 We analyzed data from 40 of 43 hospitals that contributed inpatient data during our study period. 2 hospitals were excluded due to inconsistent data submission, and 1 removed their data.

Study Population

Data were extracted for children 0 to 17 years old who were admitted to an inpatient or observational unit between January 1, 2007 and December 31, 2014 for traumatic brain injury (TBI). Children were identified using International Classification of Diseases, Clinical Modification, Ninth Revision (ICD-9-CM) diagnosis codes that denote TBI per the Centers for Disease Control (CDC): 800.0–801.9, 803.0–804.9, 850–854.1, and 959.01.^6–8 To examine inpatient care for concussion, we only retained children with a primary (ie, first) concussion-related diagnosis code (850.0–850.9⁹) for analyses. For patients with multiple visits during our study period, only the index admission was analyzed. We refined our cohort using 2 injury scores calculated from ICD-9-CM diagnosis codes using validated ICDMAP-90 injury coding software.^6,10–12 The Abbreviated Injury Scale (AIS) ranges from 1 (minor injury) to 6 (not survivable). The total Injury Severity Score (ISS) is based on 6 body regions (head/neck, face, chest, abdomen, extremity, and external) and calculated by summing the squares of the 3 worst AIS scores.¹³ A concussion receives a head AIS score of 2 if there is an associated loss of consciousness or a score of 1 if there is not; therefore, children were excluded if the head AIS score was >2. We also excluded children with the following features, as they may be indicative of more severe injuries that were likely the cause of admission: ISS > 6, secondary diagnosis code of skull fracture or intracranial injury, intensive care unit (ICU) or operating room (OR) charges, or a LOS > 7 days. Because some children are hospitalized for potentially abusive minor head trauma pending a safe discharge plan, we excluded children 0 to 4 years of age with child abuse, which was determined using a specific set of diagnosis codes (E960-E96820, 995.54, and 995.55) similar to previous research.¹⁴

Data Elements and Outcomes

Outcomes

Based on previous reports,^1,15 a LOS ≥ 2 days distinguished a typical hospitalization from a prolonged one. ED revisit was identified when a child had a visit with a TBI-related primary diagnosis code at a PHIS hospital within 30 days of initial admission and was discharged home. We limited analyses to children discharged, as children readmitted may have had an initially missed intracranial injury.

Patient Characteristics

We examined the following patient variables: age, race, sex, presence of chronic medical condition, payer type, household income, area of residence (eg, rural versus urban), and mechanism of injury. Age was categorized to represent early childhood (0 to 4 years), school age (5 to 12 years), and adolescence (12 to 17 years). Race was grouped as white, black, or other (Asian, Pacific Islander, American Indian, and “other” per PHIS). Ethnicity was described as Hispanic/Latino or not Hispanic/Latino. Children with medical conditions lasting at least 12 months and comorbidities that may impact TBI recovery were identified using a subgrouping of ICD-9-CM codes for children with “complex chronic conditions”.¹⁶ Payer type was categorized as government, private, and self-pay. We extracted a PHIS variable representing the 2010 median household income for the child’s home zip code and categorized it into quartiles based on the Federal Poverty Level for a family of 4.^17,18 Area of residence was defined using a Rural–Urban Commuting Area (RUCA) classification system¹⁹ and grouped into large urban core, suburban area, large rural town, or small rural town/isolated rural area.¹⁷ Mechanism of injury was determined using E-codes and categorized using the CDC injury framework,²⁰ with sports-related injuries identified using a previously described set of E-codes.¹ Mechanisms of injury included fall, motor vehicle collision, other motorized transport (eg, all-terrain vehicles), sports-related, struck by or against (ie, objects), and all others (eg, cyclists).

Hospital Characteristics

Hospitals were characterized by region (Northeast, Central, South, and West) and size (small <200, medium 200–400, and large >400 beds). The trauma-level accreditation was identified with Level 1 reflecting the highest possible trauma resources.

Medical Care Variables

Care variables included medications, neuroimaging, and cost of stay. Medication classes included oral non-narcotic analgesics [acetaminophen, ibuprofen, and others (aspirin, tramadol, and naproxen)], oral narcotics (codeine, oxycodone, and narcotic–non-narcotic combinations), intravenous (IV) non-narcotics (ketorolac), IV narcotics (morphine, fentanyl, and hydromorphone), antiemetics [ondansetron, metoclopramide, and phenothiazines (prochlorperazine, chlorpromazine, and promethazine)], maintenance IV fluids (dextrose with electrolytes or 0.45% sodium chloride), and resuscitation IV fluids (0.9% sodium chloride or lactated Ringer’s solution). Receipt of neuroimaging was determined if head computed tomography (CT) had been conducted at the admitting hospital. Adjusted cost of stay was calculated using a hospital-specific cost-to-charge ratio with additional adjustments using the Center for Medicare & Medicaid’s Wage Index.

Statistical Analyses

Descriptive statistics were calculated for individual, injury, and hospital, and care data elements, LOS, and ED readmissions. The number of children admitted with TBI was used as the denominator to assess the proportion of pediatric TBI admissions that were due to concussions. To identify factors associated with prolonged LOS (ie, ≥2 days) and ED readmission, we employed a mixed models approach that accounted for clustering of observations within hospitals. Independent variables included age, sex, race, ethnicity, payer type, household income, RUCA code, chronic medical condition, and injury mechanism. Models were adjusted for hospital location, size, and trauma-level accreditation. The binary distribution was specified along with a logit link function. A 2-phase process determined factors associated with each outcome. First, bivariable models were developed, followed by multivariable models that included independent variables with P values < .25 in the bivariable analysis. Backward step-wise elimination was performed, deleting variables with the highest P value one at a time. After each deletion, the percentage change in odds ratios was examined; if variable removal resulted in >10% change, the variable was retained as a potential confounder. This process was repeated until all remaining variables were significant (P < .05) with the exception of potential confounders. Finally, we examined the proportion of children receiving selected care practices annually. Descriptive and trend analyses were used to analyze adjusted median cost of stay. Analyses were performed using SAS software (Version 9.3, SAS Institute Inc., Cary, North Carolina).

Over 8 years, 88,526 children were admitted to 40 PHIS hospitals with a TBI-related diagnosis, among whom 13,708 had a primary diagnosis of concussion. We excluded 2,973 children with 1 or more of the following characteristics: a secondary diagnosis of intracranial injury (n = 58), head AIS score > 2 (n = 218), LOS > 7 days (n = 50), OR charges (n = 132), ICU charges (n = 1947), and ISS > 6 (n = 568). Six additional children aging 0 to 4 years were excluded due to child abuse. The remaining 10,729 children, averaging 1300 hospitalizations annually, were identified as being hospitalized primarily for concussion.

Table 1 summarizes the individual characteristics for this cohort. The average (standard deviation) age was 9.5 (5.1) years. Ethnicity was missing for 25.3% and therefore excluded from the multivariable models. Almost all children had a head AIS score of 2 (99.2%), and the majority had a total ISS ≤ 4 (73.4%). The majority of admissions were admitted to Level 1 trauma-accredited hospitals (78.7%) and medium-sized hospitals (63.9%).

From 2007 to 2014, approximately 15% of children admitted to PHIS hospitals for TBI were admitted primarily for concussion. Since almost all children had a head AIS score of 2 and an ISS ≤ 4, our data suggest that most children had an associated loss of consciousness and that concussion was the only injury sustained, respectively. This study identified important subgroups that necessitated inpatient care but are rarely the focus of concussion research (eg, toddlers and those injured due to a motor vehicle collision). Most children (83.3%) received medications to treat common postconcussive symptoms (eg, pain and nausea), with almost half receiving 3 or more medication classes. Factors associated with the development of postconcussive syndrome (eg, female sex and adolescent age)^4,21 were significantly associated with hospitalization of 2 or more days and ED revisit within 30 days of admission. In the absence of evidenced-based guidelines for inpatient concussion management, we identified significant trends in care, including increased use of specific pain [ie, oral and IV nonsteroidal anti-inflammatory drugs (NSAIDs)] and antiemetic (ie, ondansetron) medications and decreased use of head CT. Given the number of children admitted and receiving treatment for concussion symptomatology, influences on the decision to deliver specific care practices, as well as the impact and benefit of hospitalization, require closer examination.

Our study extends previous reports from the PHIS database by characterizing children admitted for concussion.¹ We found that children admitted for concussion had similar characteristics to the broader population of children who sustain concussion (eg, school-aged children, male, and injured due to a fall or during sports).^1,3,22 However, approximately 20% of the cohort were less than 5 years old, and less is known regarding appropriate treatment and outcomes of concussion in this age group.²³ Uncertainty regarding optimal management and a young child’s inability to articulate symptoms may contribute to a physician’s decision to admit for close observation. Similar to Blinman et al., we found that a substantial proportion of children admitted with concussion were injured due to a motor vehicle collision,³ suggesting that although sports-related injuries are responsible for a significant proportion of pediatric concussions, children injured by other preventable mechanisms may also be incurring significant concussive injuries. Finally, the majority of our cohort was from an urban core, relative to a rural area, which is likely a reflection of the regionalization of trauma care, as well as variations in access to health care.

Although most children recover fully from concussion without specific interventions, 20%-30% may remain symptomatic at 1 month,^3,4,21,24 and children who are hospitalized with concussion may be at higher risk for protracted symptoms. While specific individual or injury-related factors (eg, female sex, adolescent age, and injury due to motor vehicle collision) may contribute to more significant postconcussive symptoms, it is unclear how inpatient management affects recovery trajectory. Frequent sleep disruptions associated with inpatient care²⁵ contradict current acute concussion management recommendations for physical and cognitive rest²⁶ and could potentially impair symptom recovery. Additionally, we found widespread use of NSAIDs, although there is evidence suggesting that NSAIDs may potentially worsen concussive symptoms.²⁶ We identified an increase in medication usage over time despite limited evidence of their effectiveness for pediatric concussion.^27–29 This change may reflect improved symptom screening^4,30 and/or increased awareness of specific medication safety profiles in pediatric trauma patients, especially for NSAIDs and ondansetron. Although we saw an increase in NSAID use, we did not see a proportional decrease in narcotic use. Similarly, while two-thirds of our cohort received IV medications, there is controversy about the need for IV fluids and medications for other pediatric illnesses, with research demonstrating that IV treatment may not reduce recovery time and may contribute to prolonged hospitalization and phlebitis.^31,32 Thus, there is a need to understand the therapeutic effectiveness and benefits of medications and fluids on postconcussion recovery.

Neuroimaging rates for children receiving ED evaluation for concussion have been reported to be up to 60%-70%,^1,22 although a more recent study spanning 2006 to 2011 found a 35-%–40% head CT rate in pediatric patients by hospital-based EDs in the United States.³³ Our results appear to support decreasing head CT use over time in pediatric hospitals. Hospitalization for observation is costly¹ but could decrease a child’s risk of malignancy from radiation exposure. Further work on balancing cost, risk, and shared decision-making with parents could guide decisions regarding emergent neuroimaging versus admission.

This study has limitations inherent to the use of an administrative dataset, including lack of information regarding why the child was admitted. Since the focus was to describe inpatient care of children with concussion, those discharged home from the ED were not included in this dataset. Consequently, we could not contrast the ED care of those discharged home with those who were admitted or assess trends in admission rates for concussion. Although the overall number of concussion admissions has continued to remain stable over time,¹ due to a lack of prospectively collected clinical information, we are unable to determine whether observed trends in care are secondary to changes in practice or changes in concussion severity. However, there has been no research to date supporting the latter. Ethnicity was excluded due to high levels of missing data. Cost of stay was not extensively analyzed given hospital variation in designation of observational or inpatient status, which subsequently affects billing.³⁴ Rates of neuroimaging and ED revisit may have been underestimated since children could have received care at a non-PHIS hospital. Similarly, the decrease in the proportion of children receiving neuroimaging over time may have been associated with an increase in children being transferred from a non-PHIS hospital for admission, although with increased regionalization in trauma care, we would not expect transfers of children with only concussion to have significantly increased. Finally, data were limited to the pediatric tertiary care centers participating in PHIS, thereby reducing generalizability and introducing selection bias by only including children who were able to access care at PHIS hospitals. Although the care practices we evaluated (eg, NSAIDs and head CT) are available at all hospitals, our analyses only reflect care delivered within the PHIS.

Concussion accounted for 15% of all pediatric TBI admissions during our study period. Further investigation of potential factors associated with admission and protracted recovery (eg, adolescent females needing treatment for severe symptomatology) could facilitate better understanding of how hospitalization affects recovery. Additionally, research on acute pharmacotherapies (eg, IV therapies and/or inpatient treatment until symptoms resolve) is needed to fully elucidate the acute and long-term benefits of interventions delivered to children.

Colleen Mangeot: Biostatistician with extensive PHIS knowledge who contributed to database creation and statistical analysis. Yanhong (Amy) Liu: Research database programmer who developed the database, ran quality assurance measures, and cleaned all study data.

Disclosures

The authors have nothing to disclose.

Funding

This study was supported by grant R40 MC 268060102 from the Maternal and Child Health Research Program, Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Services Administration, Department of Health and Human Services. The funding source was not involved in development of the study design; in the collection, analysis and interpretation of data; or in the writing of this report.

Approximately 14% of children who sustain a concussion are admitted to the hospital,¹ although admission rates reportedly vary substantially among pediatric hospitals.² Children hospitalized for concussion may be at a higher risk for persistent postconcussive symptoms,^3,4 yet little is known about this subset of children and how they are managed while in the hospital. Characterizing children hospitalized for concussion and describing the inpatient care they received will promote hypothesis generation for further inquiry into indications for admission, as well as the relationship between inpatient management and concussion recovery.

We described a cohort of children admitted to 40 pediatric hospitals primarily for concussion and detailed care delivered during hospitalization. We explored individual-level factors and their association with prolonged length of stay (LOS) and emergency department (ED) readmission. Finally, we evaluated if there had been changes in inpatient care over the 8-year study period.

Study Design

The Institutional Review Board determined that this retrospective cohort study was exempt from review.

Data Source

The Children’s Hospital Association’s Pediatric Health Information System (PHIS) is an administrative database from pediatric hospitals located within 17 major metropolitan areas in the United States. Data include: service dates, patient demographics, payer type, diagnosis codes, resource utilization information (eg, medications), and hospital characteristics.^1,5 De-identified data undergo reliability and validity checks prior to inclusion.^1,5 We analyzed data from 40 of 43 hospitals that contributed inpatient data during our study period. 2 hospitals were excluded due to inconsistent data submission, and 1 removed their data.

Study Population

Data were extracted for children 0 to 17 years old who were admitted to an inpatient or observational unit between January 1, 2007 and December 31, 2014 for traumatic brain injury (TBI). Children were identified using International Classification of Diseases, Clinical Modification, Ninth Revision (ICD-9-CM) diagnosis codes that denote TBI per the Centers for Disease Control (CDC): 800.0–801.9, 803.0–804.9, 850–854.1, and 959.01.^6–8 To examine inpatient care for concussion, we only retained children with a primary (ie, first) concussion-related diagnosis code (850.0–850.9⁹) for analyses. For patients with multiple visits during our study period, only the index admission was analyzed. We refined our cohort using 2 injury scores calculated from ICD-9-CM diagnosis codes using validated ICDMAP-90 injury coding software.^6,10–12 The Abbreviated Injury Scale (AIS) ranges from 1 (minor injury) to 6 (not survivable). The total Injury Severity Score (ISS) is based on 6 body regions (head/neck, face, chest, abdomen, extremity, and external) and calculated by summing the squares of the 3 worst AIS scores.¹³ A concussion receives a head AIS score of 2 if there is an associated loss of consciousness or a score of 1 if there is not; therefore, children were excluded if the head AIS score was >2. We also excluded children with the following features, as they may be indicative of more severe injuries that were likely the cause of admission: ISS > 6, secondary diagnosis code of skull fracture or intracranial injury, intensive care unit (ICU) or operating room (OR) charges, or a LOS > 7 days. Because some children are hospitalized for potentially abusive minor head trauma pending a safe discharge plan, we excluded children 0 to 4 years of age with child abuse, which was determined using a specific set of diagnosis codes (E960-E96820, 995.54, and 995.55) similar to previous research.¹⁴

Data Elements and Outcomes

Outcomes

Based on previous reports,^1,15 a LOS ≥ 2 days distinguished a typical hospitalization from a prolonged one. ED revisit was identified when a child had a visit with a TBI-related primary diagnosis code at a PHIS hospital within 30 days of initial admission and was discharged home. We limited analyses to children discharged, as children readmitted may have had an initially missed intracranial injury.

Patient Characteristics

We examined the following patient variables: age, race, sex, presence of chronic medical condition, payer type, household income, area of residence (eg, rural versus urban), and mechanism of injury. Age was categorized to represent early childhood (0 to 4 years), school age (5 to 12 years), and adolescence (12 to 17 years). Race was grouped as white, black, or other (Asian, Pacific Islander, American Indian, and “other” per PHIS). Ethnicity was described as Hispanic/Latino or not Hispanic/Latino. Children with medical conditions lasting at least 12 months and comorbidities that may impact TBI recovery were identified using a subgrouping of ICD-9-CM codes for children with “complex chronic conditions”.¹⁶ Payer type was categorized as government, private, and self-pay. We extracted a PHIS variable representing the 2010 median household income for the child’s home zip code and categorized it into quartiles based on the Federal Poverty Level for a family of 4.^17,18 Area of residence was defined using a Rural–Urban Commuting Area (RUCA) classification system¹⁹ and grouped into large urban core, suburban area, large rural town, or small rural town/isolated rural area.¹⁷ Mechanism of injury was determined using E-codes and categorized using the CDC injury framework,²⁰ with sports-related injuries identified using a previously described set of E-codes.¹ Mechanisms of injury included fall, motor vehicle collision, other motorized transport (eg, all-terrain vehicles), sports-related, struck by or against (ie, objects), and all others (eg, cyclists).

Hospital Characteristics

Hospitals were characterized by region (Northeast, Central, South, and West) and size (small <200, medium 200–400, and large >400 beds). The trauma-level accreditation was identified with Level 1 reflecting the highest possible trauma resources.

Medical Care Variables

Care variables included medications, neuroimaging, and cost of stay. Medication classes included oral non-narcotic analgesics [acetaminophen, ibuprofen, and others (aspirin, tramadol, and naproxen)], oral narcotics (codeine, oxycodone, and narcotic–non-narcotic combinations), intravenous (IV) non-narcotics (ketorolac), IV narcotics (morphine, fentanyl, and hydromorphone), antiemetics [ondansetron, metoclopramide, and phenothiazines (prochlorperazine, chlorpromazine, and promethazine)], maintenance IV fluids (dextrose with electrolytes or 0.45% sodium chloride), and resuscitation IV fluids (0.9% sodium chloride or lactated Ringer’s solution). Receipt of neuroimaging was determined if head computed tomography (CT) had been conducted at the admitting hospital. Adjusted cost of stay was calculated using a hospital-specific cost-to-charge ratio with additional adjustments using the Center for Medicare & Medicaid’s Wage Index.

Statistical Analyses

Descriptive statistics were calculated for individual, injury, and hospital, and care data elements, LOS, and ED readmissions. The number of children admitted with TBI was used as the denominator to assess the proportion of pediatric TBI admissions that were due to concussions. To identify factors associated with prolonged LOS (ie, ≥2 days) and ED readmission, we employed a mixed models approach that accounted for clustering of observations within hospitals. Independent variables included age, sex, race, ethnicity, payer type, household income, RUCA code, chronic medical condition, and injury mechanism. Models were adjusted for hospital location, size, and trauma-level accreditation. The binary distribution was specified along with a logit link function. A 2-phase process determined factors associated with each outcome. First, bivariable models were developed, followed by multivariable models that included independent variables with P values < .25 in the bivariable analysis. Backward step-wise elimination was performed, deleting variables with the highest P value one at a time. After each deletion, the percentage change in odds ratios was examined; if variable removal resulted in >10% change, the variable was retained as a potential confounder. This process was repeated until all remaining variables were significant (P < .05) with the exception of potential confounders. Finally, we examined the proportion of children receiving selected care practices annually. Descriptive and trend analyses were used to analyze adjusted median cost of stay. Analyses were performed using SAS software (Version 9.3, SAS Institute Inc., Cary, North Carolina).

Over 8 years, 88,526 children were admitted to 40 PHIS hospitals with a TBI-related diagnosis, among whom 13,708 had a primary diagnosis of concussion. We excluded 2,973 children with 1 or more of the following characteristics: a secondary diagnosis of intracranial injury (n = 58), head AIS score > 2 (n = 218), LOS > 7 days (n = 50), OR charges (n = 132), ICU charges (n = 1947), and ISS > 6 (n = 568). Six additional children aging 0 to 4 years were excluded due to child abuse. The remaining 10,729 children, averaging 1300 hospitalizations annually, were identified as being hospitalized primarily for concussion.

Table 1 summarizes the individual characteristics for this cohort. The average (standard deviation) age was 9.5 (5.1) years. Ethnicity was missing for 25.3% and therefore excluded from the multivariable models. Almost all children had a head AIS score of 2 (99.2%), and the majority had a total ISS ≤ 4 (73.4%). The majority of admissions were admitted to Level 1 trauma-accredited hospitals (78.7%) and medium-sized hospitals (63.9%).

From 2007 to 2014, approximately 15% of children admitted to PHIS hospitals for TBI were admitted primarily for concussion. Since almost all children had a head AIS score of 2 and an ISS ≤ 4, our data suggest that most children had an associated loss of consciousness and that concussion was the only injury sustained, respectively. This study identified important subgroups that necessitated inpatient care but are rarely the focus of concussion research (eg, toddlers and those injured due to a motor vehicle collision). Most children (83.3%) received medications to treat common postconcussive symptoms (eg, pain and nausea), with almost half receiving 3 or more medication classes. Factors associated with the development of postconcussive syndrome (eg, female sex and adolescent age)^4,21 were significantly associated with hospitalization of 2 or more days and ED revisit within 30 days of admission. In the absence of evidenced-based guidelines for inpatient concussion management, we identified significant trends in care, including increased use of specific pain [ie, oral and IV nonsteroidal anti-inflammatory drugs (NSAIDs)] and antiemetic (ie, ondansetron) medications and decreased use of head CT. Given the number of children admitted and receiving treatment for concussion symptomatology, influences on the decision to deliver specific care practices, as well as the impact and benefit of hospitalization, require closer examination.

Our study extends previous reports from the PHIS database by characterizing children admitted for concussion.¹ We found that children admitted for concussion had similar characteristics to the broader population of children who sustain concussion (eg, school-aged children, male, and injured due to a fall or during sports).^1,3,22 However, approximately 20% of the cohort were less than 5 years old, and less is known regarding appropriate treatment and outcomes of concussion in this age group.²³ Uncertainty regarding optimal management and a young child’s inability to articulate symptoms may contribute to a physician’s decision to admit for close observation. Similar to Blinman et al., we found that a substantial proportion of children admitted with concussion were injured due to a motor vehicle collision,³ suggesting that although sports-related injuries are responsible for a significant proportion of pediatric concussions, children injured by other preventable mechanisms may also be incurring significant concussive injuries. Finally, the majority of our cohort was from an urban core, relative to a rural area, which is likely a reflection of the regionalization of trauma care, as well as variations in access to health care.

Although most children recover fully from concussion without specific interventions, 20%-30% may remain symptomatic at 1 month,^3,4,21,24 and children who are hospitalized with concussion may be at higher risk for protracted symptoms. While specific individual or injury-related factors (eg, female sex, adolescent age, and injury due to motor vehicle collision) may contribute to more significant postconcussive symptoms, it is unclear how inpatient management affects recovery trajectory. Frequent sleep disruptions associated with inpatient care²⁵ contradict current acute concussion management recommendations for physical and cognitive rest²⁶ and could potentially impair symptom recovery. Additionally, we found widespread use of NSAIDs, although there is evidence suggesting that NSAIDs may potentially worsen concussive symptoms.²⁶ We identified an increase in medication usage over time despite limited evidence of their effectiveness for pediatric concussion.^27–29 This change may reflect improved symptom screening^4,30 and/or increased awareness of specific medication safety profiles in pediatric trauma patients, especially for NSAIDs and ondansetron. Although we saw an increase in NSAID use, we did not see a proportional decrease in narcotic use. Similarly, while two-thirds of our cohort received IV medications, there is controversy about the need for IV fluids and medications for other pediatric illnesses, with research demonstrating that IV treatment may not reduce recovery time and may contribute to prolonged hospitalization and phlebitis.^31,32 Thus, there is a need to understand the therapeutic effectiveness and benefits of medications and fluids on postconcussion recovery.

Neuroimaging rates for children receiving ED evaluation for concussion have been reported to be up to 60%-70%,^1,22 although a more recent study spanning 2006 to 2011 found a 35-%–40% head CT rate in pediatric patients by hospital-based EDs in the United States.³³ Our results appear to support decreasing head CT use over time in pediatric hospitals. Hospitalization for observation is costly¹ but could decrease a child’s risk of malignancy from radiation exposure. Further work on balancing cost, risk, and shared decision-making with parents could guide decisions regarding emergent neuroimaging versus admission.

This study has limitations inherent to the use of an administrative dataset, including lack of information regarding why the child was admitted. Since the focus was to describe inpatient care of children with concussion, those discharged home from the ED were not included in this dataset. Consequently, we could not contrast the ED care of those discharged home with those who were admitted or assess trends in admission rates for concussion. Although the overall number of concussion admissions has continued to remain stable over time,¹ due to a lack of prospectively collected clinical information, we are unable to determine whether observed trends in care are secondary to changes in practice or changes in concussion severity. However, there has been no research to date supporting the latter. Ethnicity was excluded due to high levels of missing data. Cost of stay was not extensively analyzed given hospital variation in designation of observational or inpatient status, which subsequently affects billing.³⁴ Rates of neuroimaging and ED revisit may have been underestimated since children could have received care at a non-PHIS hospital. Similarly, the decrease in the proportion of children receiving neuroimaging over time may have been associated with an increase in children being transferred from a non-PHIS hospital for admission, although with increased regionalization in trauma care, we would not expect transfers of children with only concussion to have significantly increased. Finally, data were limited to the pediatric tertiary care centers participating in PHIS, thereby reducing generalizability and introducing selection bias by only including children who were able to access care at PHIS hospitals. Although the care practices we evaluated (eg, NSAIDs and head CT) are available at all hospitals, our analyses only reflect care delivered within the PHIS.

Concussion accounted for 15% of all pediatric TBI admissions during our study period. Further investigation of potential factors associated with admission and protracted recovery (eg, adolescent females needing treatment for severe symptomatology) could facilitate better understanding of how hospitalization affects recovery. Additionally, research on acute pharmacotherapies (eg, IV therapies and/or inpatient treatment until symptoms resolve) is needed to fully elucidate the acute and long-term benefits of interventions delivered to children.

Colleen Mangeot: Biostatistician with extensive PHIS knowledge who contributed to database creation and statistical analysis. Yanhong (Amy) Liu: Research database programmer who developed the database, ran quality assurance measures, and cleaned all study data.

Disclosures

The authors have nothing to disclose.

Funding

This study was supported by grant R40 MC 268060102 from the Maternal and Child Health Research Program, Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Services Administration, Department of Health and Human Services. The funding source was not involved in development of the study design; in the collection, analysis and interpretation of data; or in the writing of this report.

Approximately 14% of children who sustain a concussion are admitted to the hospital,¹ although admission rates reportedly vary substantially among pediatric hospitals.² Children hospitalized for concussion may be at a higher risk for persistent postconcussive symptoms,^3,4 yet little is known about this subset of children and how they are managed while in the hospital. Characterizing children hospitalized for concussion and describing the inpatient care they received will promote hypothesis generation for further inquiry into indications for admission, as well as the relationship between inpatient management and concussion recovery.

We described a cohort of children admitted to 40 pediatric hospitals primarily for concussion and detailed care delivered during hospitalization. We explored individual-level factors and their association with prolonged length of stay (LOS) and emergency department (ED) readmission. Finally, we evaluated if there had been changes in inpatient care over the 8-year study period.

Study Design

The Institutional Review Board determined that this retrospective cohort study was exempt from review.

Data Source

The Children’s Hospital Association’s Pediatric Health Information System (PHIS) is an administrative database from pediatric hospitals located within 17 major metropolitan areas in the United States. Data include: service dates, patient demographics, payer type, diagnosis codes, resource utilization information (eg, medications), and hospital characteristics.^1,5 De-identified data undergo reliability and validity checks prior to inclusion.^1,5 We analyzed data from 40 of 43 hospitals that contributed inpatient data during our study period. 2 hospitals were excluded due to inconsistent data submission, and 1 removed their data.

Study Population

Data were extracted for children 0 to 17 years old who were admitted to an inpatient or observational unit between January 1, 2007 and December 31, 2014 for traumatic brain injury (TBI). Children were identified using International Classification of Diseases, Clinical Modification, Ninth Revision (ICD-9-CM) diagnosis codes that denote TBI per the Centers for Disease Control (CDC): 800.0–801.9, 803.0–804.9, 850–854.1, and 959.01.^6–8 To examine inpatient care for concussion, we only retained children with a primary (ie, first) concussion-related diagnosis code (850.0–850.9⁹) for analyses. For patients with multiple visits during our study period, only the index admission was analyzed. We refined our cohort using 2 injury scores calculated from ICD-9-CM diagnosis codes using validated ICDMAP-90 injury coding software.^6,10–12 The Abbreviated Injury Scale (AIS) ranges from 1 (minor injury) to 6 (not survivable). The total Injury Severity Score (ISS) is based on 6 body regions (head/neck, face, chest, abdomen, extremity, and external) and calculated by summing the squares of the 3 worst AIS scores.¹³ A concussion receives a head AIS score of 2 if there is an associated loss of consciousness or a score of 1 if there is not; therefore, children were excluded if the head AIS score was >2. We also excluded children with the following features, as they may be indicative of more severe injuries that were likely the cause of admission: ISS > 6, secondary diagnosis code of skull fracture or intracranial injury, intensive care unit (ICU) or operating room (OR) charges, or a LOS > 7 days. Because some children are hospitalized for potentially abusive minor head trauma pending a safe discharge plan, we excluded children 0 to 4 years of age with child abuse, which was determined using a specific set of diagnosis codes (E960-E96820, 995.54, and 995.55) similar to previous research.¹⁴

Data Elements and Outcomes

Outcomes

Based on previous reports,^1,15 a LOS ≥ 2 days distinguished a typical hospitalization from a prolonged one. ED revisit was identified when a child had a visit with a TBI-related primary diagnosis code at a PHIS hospital within 30 days of initial admission and was discharged home. We limited analyses to children discharged, as children readmitted may have had an initially missed intracranial injury.

Patient Characteristics

We examined the following patient variables: age, race, sex, presence of chronic medical condition, payer type, household income, area of residence (eg, rural versus urban), and mechanism of injury. Age was categorized to represent early childhood (0 to 4 years), school age (5 to 12 years), and adolescence (12 to 17 years). Race was grouped as white, black, or other (Asian, Pacific Islander, American Indian, and “other” per PHIS). Ethnicity was described as Hispanic/Latino or not Hispanic/Latino. Children with medical conditions lasting at least 12 months and comorbidities that may impact TBI recovery were identified using a subgrouping of ICD-9-CM codes for children with “complex chronic conditions”.¹⁶ Payer type was categorized as government, private, and self-pay. We extracted a PHIS variable representing the 2010 median household income for the child’s home zip code and categorized it into quartiles based on the Federal Poverty Level for a family of 4.^17,18 Area of residence was defined using a Rural–Urban Commuting Area (RUCA) classification system¹⁹ and grouped into large urban core, suburban area, large rural town, or small rural town/isolated rural area.¹⁷ Mechanism of injury was determined using E-codes and categorized using the CDC injury framework,²⁰ with sports-related injuries identified using a previously described set of E-codes.¹ Mechanisms of injury included fall, motor vehicle collision, other motorized transport (eg, all-terrain vehicles), sports-related, struck by or against (ie, objects), and all others (eg, cyclists).

Hospital Characteristics

Hospitals were characterized by region (Northeast, Central, South, and West) and size (small <200, medium 200–400, and large >400 beds). The trauma-level accreditation was identified with Level 1 reflecting the highest possible trauma resources.

Medical Care Variables

Care variables included medications, neuroimaging, and cost of stay. Medication classes included oral non-narcotic analgesics [acetaminophen, ibuprofen, and others (aspirin, tramadol, and naproxen)], oral narcotics (codeine, oxycodone, and narcotic–non-narcotic combinations), intravenous (IV) non-narcotics (ketorolac), IV narcotics (morphine, fentanyl, and hydromorphone), antiemetics [ondansetron, metoclopramide, and phenothiazines (prochlorperazine, chlorpromazine, and promethazine)], maintenance IV fluids (dextrose with electrolytes or 0.45% sodium chloride), and resuscitation IV fluids (0.9% sodium chloride or lactated Ringer’s solution). Receipt of neuroimaging was determined if head computed tomography (CT) had been conducted at the admitting hospital. Adjusted cost of stay was calculated using a hospital-specific cost-to-charge ratio with additional adjustments using the Center for Medicare & Medicaid’s Wage Index.

Statistical Analyses

Descriptive statistics were calculated for individual, injury, and hospital, and care data elements, LOS, and ED readmissions. The number of children admitted with TBI was used as the denominator to assess the proportion of pediatric TBI admissions that were due to concussions. To identify factors associated with prolonged LOS (ie, ≥2 days) and ED readmission, we employed a mixed models approach that accounted for clustering of observations within hospitals. Independent variables included age, sex, race, ethnicity, payer type, household income, RUCA code, chronic medical condition, and injury mechanism. Models were adjusted for hospital location, size, and trauma-level accreditation. The binary distribution was specified along with a logit link function. A 2-phase process determined factors associated with each outcome. First, bivariable models were developed, followed by multivariable models that included independent variables with P values < .25 in the bivariable analysis. Backward step-wise elimination was performed, deleting variables with the highest P value one at a time. After each deletion, the percentage change in odds ratios was examined; if variable removal resulted in >10% change, the variable was retained as a potential confounder. This process was repeated until all remaining variables were significant (P < .05) with the exception of potential confounders. Finally, we examined the proportion of children receiving selected care practices annually. Descriptive and trend analyses were used to analyze adjusted median cost of stay. Analyses were performed using SAS software (Version 9.3, SAS Institute Inc., Cary, North Carolina).

Over 8 years, 88,526 children were admitted to 40 PHIS hospitals with a TBI-related diagnosis, among whom 13,708 had a primary diagnosis of concussion. We excluded 2,973 children with 1 or more of the following characteristics: a secondary diagnosis of intracranial injury (n = 58), head AIS score > 2 (n = 218), LOS > 7 days (n = 50), OR charges (n = 132), ICU charges (n = 1947), and ISS > 6 (n = 568). Six additional children aging 0 to 4 years were excluded due to child abuse. The remaining 10,729 children, averaging 1300 hospitalizations annually, were identified as being hospitalized primarily for concussion.

Table 1 summarizes the individual characteristics for this cohort. The average (standard deviation) age was 9.5 (5.1) years. Ethnicity was missing for 25.3% and therefore excluded from the multivariable models. Almost all children had a head AIS score of 2 (99.2%), and the majority had a total ISS ≤ 4 (73.4%). The majority of admissions were admitted to Level 1 trauma-accredited hospitals (78.7%) and medium-sized hospitals (63.9%).

From 2007 to 2014, approximately 15% of children admitted to PHIS hospitals for TBI were admitted primarily for concussion. Since almost all children had a head AIS score of 2 and an ISS ≤ 4, our data suggest that most children had an associated loss of consciousness and that concussion was the only injury sustained, respectively. This study identified important subgroups that necessitated inpatient care but are rarely the focus of concussion research (eg, toddlers and those injured due to a motor vehicle collision). Most children (83.3%) received medications to treat common postconcussive symptoms (eg, pain and nausea), with almost half receiving 3 or more medication classes. Factors associated with the development of postconcussive syndrome (eg, female sex and adolescent age)^4,21 were significantly associated with hospitalization of 2 or more days and ED revisit within 30 days of admission. In the absence of evidenced-based guidelines for inpatient concussion management, we identified significant trends in care, including increased use of specific pain [ie, oral and IV nonsteroidal anti-inflammatory drugs (NSAIDs)] and antiemetic (ie, ondansetron) medications and decreased use of head CT. Given the number of children admitted and receiving treatment for concussion symptomatology, influences on the decision to deliver specific care practices, as well as the impact and benefit of hospitalization, require closer examination.

Our study extends previous reports from the PHIS database by characterizing children admitted for concussion.¹ We found that children admitted for concussion had similar characteristics to the broader population of children who sustain concussion (eg, school-aged children, male, and injured due to a fall or during sports).^1,3,22 However, approximately 20% of the cohort were less than 5 years old, and less is known regarding appropriate treatment and outcomes of concussion in this age group.²³ Uncertainty regarding optimal management and a young child’s inability to articulate symptoms may contribute to a physician’s decision to admit for close observation. Similar to Blinman et al., we found that a substantial proportion of children admitted with concussion were injured due to a motor vehicle collision,³ suggesting that although sports-related injuries are responsible for a significant proportion of pediatric concussions, children injured by other preventable mechanisms may also be incurring significant concussive injuries. Finally, the majority of our cohort was from an urban core, relative to a rural area, which is likely a reflection of the regionalization of trauma care, as well as variations in access to health care.

Although most children recover fully from concussion without specific interventions, 20%-30% may remain symptomatic at 1 month,^3,4,21,24 and children who are hospitalized with concussion may be at higher risk for protracted symptoms. While specific individual or injury-related factors (eg, female sex, adolescent age, and injury due to motor vehicle collision) may contribute to more significant postconcussive symptoms, it is unclear how inpatient management affects recovery trajectory. Frequent sleep disruptions associated with inpatient care²⁵ contradict current acute concussion management recommendations for physical and cognitive rest²⁶ and could potentially impair symptom recovery. Additionally, we found widespread use of NSAIDs, although there is evidence suggesting that NSAIDs may potentially worsen concussive symptoms.²⁶ We identified an increase in medication usage over time despite limited evidence of their effectiveness for pediatric concussion.^27–29 This change may reflect improved symptom screening^4,30 and/or increased awareness of specific medication safety profiles in pediatric trauma patients, especially for NSAIDs and ondansetron. Although we saw an increase in NSAID use, we did not see a proportional decrease in narcotic use. Similarly, while two-thirds of our cohort received IV medications, there is controversy about the need for IV fluids and medications for other pediatric illnesses, with research demonstrating that IV treatment may not reduce recovery time and may contribute to prolonged hospitalization and phlebitis.^31,32 Thus, there is a need to understand the therapeutic effectiveness and benefits of medications and fluids on postconcussion recovery.

Neuroimaging rates for children receiving ED evaluation for concussion have been reported to be up to 60%-70%,^1,22 although a more recent study spanning 2006 to 2011 found a 35-%–40% head CT rate in pediatric patients by hospital-based EDs in the United States.³³ Our results appear to support decreasing head CT use over time in pediatric hospitals. Hospitalization for observation is costly¹ but could decrease a child’s risk of malignancy from radiation exposure. Further work on balancing cost, risk, and shared decision-making with parents could guide decisions regarding emergent neuroimaging versus admission.

This study has limitations inherent to the use of an administrative dataset, including lack of information regarding why the child was admitted. Since the focus was to describe inpatient care of children with concussion, those discharged home from the ED were not included in this dataset. Consequently, we could not contrast the ED care of those discharged home with those who were admitted or assess trends in admission rates for concussion. Although the overall number of concussion admissions has continued to remain stable over time,¹ due to a lack of prospectively collected clinical information, we are unable to determine whether observed trends in care are secondary to changes in practice or changes in concussion severity. However, there has been no research to date supporting the latter. Ethnicity was excluded due to high levels of missing data. Cost of stay was not extensively analyzed given hospital variation in designation of observational or inpatient status, which subsequently affects billing.³⁴ Rates of neuroimaging and ED revisit may have been underestimated since children could have received care at a non-PHIS hospital. Similarly, the decrease in the proportion of children receiving neuroimaging over time may have been associated with an increase in children being transferred from a non-PHIS hospital for admission, although with increased regionalization in trauma care, we would not expect transfers of children with only concussion to have significantly increased. Finally, data were limited to the pediatric tertiary care centers participating in PHIS, thereby reducing generalizability and introducing selection bias by only including children who were able to access care at PHIS hospitals. Although the care practices we evaluated (eg, NSAIDs and head CT) are available at all hospitals, our analyses only reflect care delivered within the PHIS.

Concussion accounted for 15% of all pediatric TBI admissions during our study period. Further investigation of potential factors associated with admission and protracted recovery (eg, adolescent females needing treatment for severe symptomatology) could facilitate better understanding of how hospitalization affects recovery. Additionally, research on acute pharmacotherapies (eg, IV therapies and/or inpatient treatment until symptoms resolve) is needed to fully elucidate the acute and long-term benefits of interventions delivered to children.

Colleen Mangeot: Biostatistician with extensive PHIS knowledge who contributed to database creation and statistical analysis. Yanhong (Amy) Liu: Research database programmer who developed the database, ran quality assurance measures, and cleaned all study data.

Disclosures

The authors have nothing to disclose.

Funding

This study was supported by grant R40 MC 268060102 from the Maternal and Child Health Research Program, Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Services Administration, Department of Health and Human Services. The funding source was not involved in development of the study design; in the collection, analysis and interpretation of data; or in the writing of this report.

Diagnostic error—defined as a failure to establish an accurate and timely explanation of the patient’s health problem—is an important source of patient harm.¹ Data suggest that all patients will experience at least 1 diagnostic error in their lifetime.^2-4 Not surprisingly, diagnostic errors are among the leading categories of paid malpractice claims in the United States.⁵

Despite diagnostic errors being morbid and sometimes deadly in the hospital,^6,7 little is known about how residents and learners approach diagnostic decision making. Errors in diagnosis are believed to stem from cognitive or system failures,⁸ with errors in cognition believed to occur due to rapid, reflexive thinking operating in the absence of a more analytical, deliberate process. System-based problems (eg, lack of expert availability, technology barriers, and access to data) have also been cited as contributors.⁹ However, whether and how these apply to trainees is not known.

Therefore, we conducted a focused ethnography of inpatient medicine teams (ie, attendings, residents, interns, and medical students) in 2 affiliated teaching hospitals, aiming to (a) observe the process of diagnosis by trainees and (b) identify methods to improve the diagnostic process and prevent errors.

We designed a multimethod, focused ethnographic study to examine diagnostic decision making in hospital settings.^10,11 In contrast to anthropologic ethnographies that study entire fields using open-ended questions, our study was designed to examine the process of diagnosis from the perspective of clinicians engaged in this activity.¹¹ This approach allowed us to capture diagnostic decisions and cognitive and system-based factors in a manner currently lacking in the literature.¹²

Setting and Participants

Between January 2016 and May 2016, we observed the members of four inpatient internal medicine teaching teams at 2 affiliated teaching hospitals. We purposefully selected teaching teams for observation because they are the primary model of care in academic settings and we have expertise in carrying out similar studies.^13,14 Teaching teams typically consisted of a medical attending (senior-level physician), 1 senior resident (a second- or third-year postgraduate trainee), two interns (a trainee in their first postgraduate year), and two to four  medical students. Teams were selected at random using existing schedules and followed Monday to Friday so as to permit observation of work on call and noncall days. Owing to manpower limitations, weekend and night shifts were not observed. However, overnight events were captured during morning rounds.

Most of the teams began rounds at 8:30 AM. Typically, rounds lasted for 90–120 min and concluded with a recap (ie, “running the list”) with a review of explicit plans for patients after they had been evaluated by the attending. This discussion often occurred in the team rooms, with the attending leading the discussion with the trainees.

Data Collection

A multidisciplinary team, including clinicians (eg, physicians, nurses), nonclinicians (eg, qualitative researchers, social scientists), and healthcare engineers, conducted the observations. We observed preround activities of interns and residents before arrival of the attending (7:00 AM - 8:30 AM), followed by morning rounds with the entire team, and afternoon work that included senior residents, interns, and students.

To capture multiple aspects of the diagnostic process, we collected data using field notes modeled on components of the National Academy of Science model for diagnosis (Appendix).^1,15 This model encompasses phases of the diagnostic process (eg, data gathering, integration, formulation of a working diagnosis, treatment delivery, and outcomes) and the work system (team members, organization, technology and tools, physical environment, tasks).

Focus Groups and Interviews

At the end of weekly observations, we conducted focus groups with the residents and one-on- one interviews with the attendings. Focus groups with the residents were conducted to encourage a group discussion about the diagnostic process. Separate interviews with the attendings were performed to ensure that power differentials did not influence discussions. During focus groups, we specifically asked about challenges and possible solutions to improve diagnosis. Experienced qualitative methodologists (J.F., M.H., M.Q.) used semistructured interview guides for discussions (Appendix).

Data Analysis

After aggregating and reading the data, three reviewers (V.C., S.K., S.S.) began inductive analysis by handwriting notes and initial reflective thoughts to create preliminary codes. Multiple team members then reread the original field notes and the focus group/interview data to refine the preliminary codes and develop additional codes. Next, relationships between codes were identified and used to develop key themes. Triangulation of data collected from observations and interview/focus group sessions was carried out to compare data that we surmised with data that were verbalized by the team. The developed themes were discussed as a group to ensure consistency of major findings.

Ethical and Regulatory Oversight

This study was reviewed and approved by the Institutional Review Boards at the University of Michigan Health System (HUM-00106657) and the VA Ann Arbor Healthcare System (1-2016-010040).

Four teaching teams (4 attendings, 4 senior residents, 9 interns, and 14 medical students) were observed over 33 distinct shifts and 168 hours. Observations included morning rounds (96 h), postround call days (52 h), and postround non-call days (20 h). Morning rounds lasted an average of 127 min (range: 48-232 min) and included an average of 9 patients (range: 4-16 patients).

Themes Regarding the Diagnostic Process

We identified the following 4 primary themes related to the diagnostic process in teaching hospitals: (1) diagnosis is a social phenomenon; (2) data necessary to make diagnoses are fragmented; (3) distractions undermine the diagnostic process; and (4) time pressures interfere with diagnostic decision making (Appendix Table 1).

(1) Diagnosis is a Social Phenomenon.

Team members viewed the process of diagnosis as a social exchange of facts, findings, and strategies within a defined structure. The opportunity to discuss impressions with others was valued as a means to share, test, and process assumptions.

“Rounds are the most important part of the process. That is where we make most decisions in a collective, collaborative way with the attending present. We bounce ideas off each other.” (Intern)

Typical of social processes, variations based on time of day and schedule were observed. For instance, during call days, learners gathered data and formed working diagnosis and treatment plans with minimal attending interaction. This separation of roles and responsibilities introduced a hierarchy within diagnosis as follows:

“The interns would not call me first; they would talk to the senior resident and then if the senior thought he should chat with me, then they would call. But for the most part, they gather information and come up with the plan.” (Attending).

The work system was suited to facilitate social interactions. For instance, designated rooms (with team members informally assigned to a computer) provided physical proximity of the resident to interns and medical students. In this space, numerous informal discussions between team members (eg, “What do you think about this test?” “I’m not sure what to do about this finding.” “Should I call a [consult] on this patient?”) were observed. Although proximity to each other was viewed as beneficial, dangers to the social nature of diagnosis in the form of anchoring (ie, a cognitive bias where emphasis is placed on the first piece of data)¹⁶ were also mentioned. Similarly, the paradox associated with social proof (ie, the pressure to assume conformity within a group) was also observed as disagreement between team members and attendings rarely occurred during observations.

“I mean, they’re the attending, right? It’s hard to argue with them when they want a test or something done. When I do push back, it’s rare that others will support me–so it’s usually me and the attending.” (Resident)

“I would push back if I think it’s really bad for the patient or could cause harm–but the truth is, it doesn’t happen much.” (Intern)

(2) Data Necessary to Make Diagnoses are Fragmented

Team members universally cited fragmentation in data delivery, retrieval, and processing as a barrier to diagnosis. Team members indicated that test results might not be looked at or acted upon in a timely manner, and participants pointed to the electronic medical record as a source of this challenge.

“Before I knew about [the app for Epic], I would literally sit on the computer to get all the information we would need on rounds. Its key to making decisions. We often say we will do something, only to find the test result doesn’t support it–and then we’re back to square 1.” (Intern)

Information used by teams came from myriad sources (eg, patients, family members, electronic records) and from various settings (eg, emergency department, patient rooms, discussions with consultants). Additionally, test results often appeared without warning. Thus, availability of information was poorly aligned with clinical duties.

“They (the lab) will call us when a blood culture is positive or something is off. That is very helpful but it often comes later in the day, when we’re done with rounds.” (Resident)

The work system was highlighted as a key contributor to data fragmentation. Peculiarities of our electronic medical record (EMR) and how data were collected, stored, or presented were described as “frustrating,” and “unsafe,” by team members. Correspondingly, we frequently observed interns asking for assistance for tasks such as ordering tests or finding information despite being “trained” to use the EMR.

“People have to learn how to filter, how to recognize the most important points and link data streams together in terms of causality. But we assume they know where to find that information. It’s actually a very hard thing to do, for both the house staff and me.” (Attending)

(3) Distractions Undermine the Diagnostic Process

Distractions often created cognitive difficulties. For example, ambient noise and interruptions from neighbors working on other teams were cited as barriers to diagnosis. In addition, we observed several team members using headphones to drown out ambient noise while working on the computer.

“I know I shouldn’t do it (wear headphones), but I have no other way of turning down the noise so I can concentrate.” (Intern)

Similarly, the unpredictable nature and the volume of pages often interrupted thinking about diagnosis.

“Sometimes the pager just goes off all the time and (after making sure its not an urgent issue), I will just ignore it for a bit, especially if I am in the middle of something. It would be great if I could finish my thought process knowing I would not be interrupted.” (Resident)

To mitigate this problem, 1 attending described how he would proactively seek out nurses caring for his patients to “head off” questions (eg, “I will renew the restraints and medications this morning,” and “Is there anything you need in terms of orders for this patient that I can take care of now?”) that might lead to pages. Another resident described his approach as follows:

“I make it a point to tell the nurses where I will be hanging out and where they can find me if they have any questions. I tell them to come talk to me rather than page me since that will be less distracting.” (Resident).

Most of the interns described documentation work such as writing admission and progress notes in negative terms (“an academic exercise,” “part of the billing activity”). However, in the context of interruptions, some described this as helpful.

“The most valuable part of the thinking process was writing the assessment and plan because that’s actually my schema for all problems. It literally is the only time where I can sit and collect my thoughts to formulate a diagnosis and plan.” (Intern)

(4) Time Pressures Interfere With Diagnostic Decision Making

All team members spoke about the challenge of finding time for diagnosis during the workday. Often, they had to skip learning sessions for this purpose.

“They tell us we should go to morning report or noon conference but when I’m running around trying to get things done. I hate having to choose between my education and doing what’s best for the patient–but that’s often what it comes down to.” (Intern)

When specifically asked whether setting aside dedicated time to specifically review and formulate diagnoses would be valuable, respondents were uniformly enthusiastic. Team members described attentional conflicts as being the worst when “cross covering” other teams on call days, as their patient load effectively doubled during this time. Of note, cross-covering occurred when teams were also on call—and thus took them away from important diagnostic activities such as data gathering or synthesis for patients they were admitting.

“If you were to ever design a system where errors were likely–this is how you would design it: take a team with little supervision, double their patient load, keep them busy with new challenging cases and then ask questions about patients they know little about.” (Resident)

Although diagnostic errors have been called “the next frontier for patient safety,”¹⁷ little is known about the process, barriers, and facilitators to diagnosis in teaching hospitals. In this focused ethnography conducted at 2 academic medical centers, we identified multiple cognitive and system-level challenges and potential strategies to improve diagnosis from trainees engaged in this activity. Key themes identified by those we observed included the social nature of diagnosis, fragmented information delivery, constant distractions and interruptions, and time pressures. In turn, these insights allow us to generate strategies that can be applied to improve the diagnostic process in teaching hospitals.

Our study underscores the importance of social interactions in diagnosis. In contrast, most of the interventions to prevent diagnostic errors target individual providers through practices such as metacognition and “thinking about thinking.”^18-20 These interventions are based on Daniel Kahnemann’s work on dual thought process. Type 1 thought processes are fast, subconscious, reflexive, largely intuitive, and more vulnerable to error. In contrast, Type 2 processes are slower, deliberate, analytic, and less prone to error.²¹ Although an individual’s Type 2 thought capacity is limited, a major goal of cognitive interventions is to encourage Type 2 over Type 1 thinking, an approach termed “de-biasing.”^22-24 Unfortunately, cognitive interventions testing such approaches have suffered mixed results–perhaps because of lack of focus on collective wisdom or group thinking, which may be key to diagnosis from our findings.^9,25 In this sense, morning rounds were a social gathering used to strategize and develop care plans, but with limited time to think about diagnosis.²⁶ Introduction of defined periods for individuals to engage in diagnostic activities such as de-biasing (ie, asking “what else could this be)²⁷ before or after rounds may provide an opportunity for reflection and improving diagnosis. In addition, embedding tools such as diagnosis expanders and checklists within these defined time slots^28,29 may prove to be useful in reflecting on diagnosis and preventing diagnostic errors.

An unexpected yet important finding from this study were the challenges posed by distractions and the physical environment. Potentially maladaptive workarounds to these interruptions included use of headphones; more productive strategies included updating nurses with plans to avert pages and creating a list of activities to ensure that key tasks were not forgotten.^30,31 Applying lessons from aviation, a focused effort to limit distractions during key portions of the day, might be worth considering for diagnostic safety.³² Similarly, improving the environment in which diagnosis occurs—including creating spaces that are quiet, orderly, and optimized for thinking—may be valuable.³³Our study has limitations. First, our findings are limited to direct observations; we are thus unable to comment on how unobserved aspects of care (eg, cognitive processes) might have influenced our findings. Our observations of clinical care might also have introduced a Hawthorne effect. However, because we were closely integrated with teams and conducted focus groups to corroborate our assessments, we believe that this was not the case. Second, we did not identify diagnostic errors or link processes we observed to errors. Third, our approach is limited to 2 teaching centers, thereby limiting the generalizability of findings. Relatedly, we were only able to conduct observations during weekdays; differences in weekend and night resources might affect our insights.

The cognitive and system-based barriers faced by clinicians in teaching hospitals suggest that new methods to improve diagnosis are needed. Future interventions such as defined “time-outs” for diagnosis, strategies focused on limiting distractions, and methods to improve communication between team members are novel and have parallels in other industries. As challenges to quantify diagnostic errors abound,³⁴ improving cognitive- and system-based factors via reflection through communication, concentration, and organization is necessary to improve medical decision making in academic medical centers.

Disclosures

None declared for all coauthors.

Funding

This project was supported by grant number P30HS024385 from the Agency for Healthcare Research and Quality. The funding source played no role in study design, data acquisition, analysis or decision to report these data. Dr. Chopra is supported by a career development award from the Agency of Healthcare Research and Quality (1-K08-HS022835-01). Dr. Krein is supported by a VA Health Services Research and Development Research Career Scientist Award (RCS 11-222). Dr. Singh is partially supported by Houston VA HSR&D Center for Innovations in Quality, Effectiveness and Safety (CIN 13-413). The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality or the Department of Veterans Affairs.

Diagnostic error—defined as a failure to establish an accurate and timely explanation of the patient’s health problem—is an important source of patient harm.¹ Data suggest that all patients will experience at least 1 diagnostic error in their lifetime.^2-4 Not surprisingly, diagnostic errors are among the leading categories of paid malpractice claims in the United States.⁵

Despite diagnostic errors being morbid and sometimes deadly in the hospital,^6,7 little is known about how residents and learners approach diagnostic decision making. Errors in diagnosis are believed to stem from cognitive or system failures,⁸ with errors in cognition believed to occur due to rapid, reflexive thinking operating in the absence of a more analytical, deliberate process. System-based problems (eg, lack of expert availability, technology barriers, and access to data) have also been cited as contributors.⁹ However, whether and how these apply to trainees is not known.

Therefore, we conducted a focused ethnography of inpatient medicine teams (ie, attendings, residents, interns, and medical students) in 2 affiliated teaching hospitals, aiming to (a) observe the process of diagnosis by trainees and (b) identify methods to improve the diagnostic process and prevent errors.

We designed a multimethod, focused ethnographic study to examine diagnostic decision making in hospital settings.^10,11 In contrast to anthropologic ethnographies that study entire fields using open-ended questions, our study was designed to examine the process of diagnosis from the perspective of clinicians engaged in this activity.¹¹ This approach allowed us to capture diagnostic decisions and cognitive and system-based factors in a manner currently lacking in the literature.¹²

Setting and Participants

Between January 2016 and May 2016, we observed the members of four inpatient internal medicine teaching teams at 2 affiliated teaching hospitals. We purposefully selected teaching teams for observation because they are the primary model of care in academic settings and we have expertise in carrying out similar studies.^13,14 Teaching teams typically consisted of a medical attending (senior-level physician), 1 senior resident (a second- or third-year postgraduate trainee), two interns (a trainee in their first postgraduate year), and two to four  medical students. Teams were selected at random using existing schedules and followed Monday to Friday so as to permit observation of work on call and noncall days. Owing to manpower limitations, weekend and night shifts were not observed. However, overnight events were captured during morning rounds.

Most of the teams began rounds at 8:30 AM. Typically, rounds lasted for 90–120 min and concluded with a recap (ie, “running the list”) with a review of explicit plans for patients after they had been evaluated by the attending. This discussion often occurred in the team rooms, with the attending leading the discussion with the trainees.

Data Collection

A multidisciplinary team, including clinicians (eg, physicians, nurses), nonclinicians (eg, qualitative researchers, social scientists), and healthcare engineers, conducted the observations. We observed preround activities of interns and residents before arrival of the attending (7:00 AM - 8:30 AM), followed by morning rounds with the entire team, and afternoon work that included senior residents, interns, and students.

To capture multiple aspects of the diagnostic process, we collected data using field notes modeled on components of the National Academy of Science model for diagnosis (Appendix).^1,15 This model encompasses phases of the diagnostic process (eg, data gathering, integration, formulation of a working diagnosis, treatment delivery, and outcomes) and the work system (team members, organization, technology and tools, physical environment, tasks).

Focus Groups and Interviews

At the end of weekly observations, we conducted focus groups with the residents and one-on- one interviews with the attendings. Focus groups with the residents were conducted to encourage a group discussion about the diagnostic process. Separate interviews with the attendings were performed to ensure that power differentials did not influence discussions. During focus groups, we specifically asked about challenges and possible solutions to improve diagnosis. Experienced qualitative methodologists (J.F., M.H., M.Q.) used semistructured interview guides for discussions (Appendix).

Data Analysis

After aggregating and reading the data, three reviewers (V.C., S.K., S.S.) began inductive analysis by handwriting notes and initial reflective thoughts to create preliminary codes. Multiple team members then reread the original field notes and the focus group/interview data to refine the preliminary codes and develop additional codes. Next, relationships between codes were identified and used to develop key themes. Triangulation of data collected from observations and interview/focus group sessions was carried out to compare data that we surmised with data that were verbalized by the team. The developed themes were discussed as a group to ensure consistency of major findings.

Ethical and Regulatory Oversight

This study was reviewed and approved by the Institutional Review Boards at the University of Michigan Health System (HUM-00106657) and the VA Ann Arbor Healthcare System (1-2016-010040).

Four teaching teams (4 attendings, 4 senior residents, 9 interns, and 14 medical students) were observed over 33 distinct shifts and 168 hours. Observations included morning rounds (96 h), postround call days (52 h), and postround non-call days (20 h). Morning rounds lasted an average of 127 min (range: 48-232 min) and included an average of 9 patients (range: 4-16 patients).

Themes Regarding the Diagnostic Process

We identified the following 4 primary themes related to the diagnostic process in teaching hospitals: (1) diagnosis is a social phenomenon; (2) data necessary to make diagnoses are fragmented; (3) distractions undermine the diagnostic process; and (4) time pressures interfere with diagnostic decision making (Appendix Table 1).

(1) Diagnosis is a Social Phenomenon.

Team members viewed the process of diagnosis as a social exchange of facts, findings, and strategies within a defined structure. The opportunity to discuss impressions with others was valued as a means to share, test, and process assumptions.

“Rounds are the most important part of the process. That is where we make most decisions in a collective, collaborative way with the attending present. We bounce ideas off each other.” (Intern)

Typical of social processes, variations based on time of day and schedule were observed. For instance, during call days, learners gathered data and formed working diagnosis and treatment plans with minimal attending interaction. This separation of roles and responsibilities introduced a hierarchy within diagnosis as follows:

“The interns would not call me first; they would talk to the senior resident and then if the senior thought he should chat with me, then they would call. But for the most part, they gather information and come up with the plan.” (Attending).

The work system was suited to facilitate social interactions. For instance, designated rooms (with team members informally assigned to a computer) provided physical proximity of the resident to interns and medical students. In this space, numerous informal discussions between team members (eg, “What do you think about this test?” “I’m not sure what to do about this finding.” “Should I call a [consult] on this patient?”) were observed. Although proximity to each other was viewed as beneficial, dangers to the social nature of diagnosis in the form of anchoring (ie, a cognitive bias where emphasis is placed on the first piece of data)¹⁶ were also mentioned. Similarly, the paradox associated with social proof (ie, the pressure to assume conformity within a group) was also observed as disagreement between team members and attendings rarely occurred during observations.

“I mean, they’re the attending, right? It’s hard to argue with them when they want a test or something done. When I do push back, it’s rare that others will support me–so it’s usually me and the attending.” (Resident)

“I would push back if I think it’s really bad for the patient or could cause harm–but the truth is, it doesn’t happen much.” (Intern)

(2) Data Necessary to Make Diagnoses are Fragmented

Team members universally cited fragmentation in data delivery, retrieval, and processing as a barrier to diagnosis. Team members indicated that test results might not be looked at or acted upon in a timely manner, and participants pointed to the electronic medical record as a source of this challenge.

“Before I knew about [the app for Epic], I would literally sit on the computer to get all the information we would need on rounds. Its key to making decisions. We often say we will do something, only to find the test result doesn’t support it–and then we’re back to square 1.” (Intern)

Information used by teams came from myriad sources (eg, patients, family members, electronic records) and from various settings (eg, emergency department, patient rooms, discussions with consultants). Additionally, test results often appeared without warning. Thus, availability of information was poorly aligned with clinical duties.

“They (the lab) will call us when a blood culture is positive or something is off. That is very helpful but it often comes later in the day, when we’re done with rounds.” (Resident)

The work system was highlighted as a key contributor to data fragmentation. Peculiarities of our electronic medical record (EMR) and how data were collected, stored, or presented were described as “frustrating,” and “unsafe,” by team members. Correspondingly, we frequently observed interns asking for assistance for tasks such as ordering tests or finding information despite being “trained” to use the EMR.

“People have to learn how to filter, how to recognize the most important points and link data streams together in terms of causality. But we assume they know where to find that information. It’s actually a very hard thing to do, for both the house staff and me.” (Attending)

(3) Distractions Undermine the Diagnostic Process

Distractions often created cognitive difficulties. For example, ambient noise and interruptions from neighbors working on other teams were cited as barriers to diagnosis. In addition, we observed several team members using headphones to drown out ambient noise while working on the computer.

“I know I shouldn’t do it (wear headphones), but I have no other way of turning down the noise so I can concentrate.” (Intern)

Similarly, the unpredictable nature and the volume of pages often interrupted thinking about diagnosis.

“Sometimes the pager just goes off all the time and (after making sure its not an urgent issue), I will just ignore it for a bit, especially if I am in the middle of something. It would be great if I could finish my thought process knowing I would not be interrupted.” (Resident)

To mitigate this problem, 1 attending described how he would proactively seek out nurses caring for his patients to “head off” questions (eg, “I will renew the restraints and medications this morning,” and “Is there anything you need in terms of orders for this patient that I can take care of now?”) that might lead to pages. Another resident described his approach as follows:

“I make it a point to tell the nurses where I will be hanging out and where they can find me if they have any questions. I tell them to come talk to me rather than page me since that will be less distracting.” (Resident).

Most of the interns described documentation work such as writing admission and progress notes in negative terms (“an academic exercise,” “part of the billing activity”). However, in the context of interruptions, some described this as helpful.

“The most valuable part of the thinking process was writing the assessment and plan because that’s actually my schema for all problems. It literally is the only time where I can sit and collect my thoughts to formulate a diagnosis and plan.” (Intern)

(4) Time Pressures Interfere With Diagnostic Decision Making

All team members spoke about the challenge of finding time for diagnosis during the workday. Often, they had to skip learning sessions for this purpose.

“They tell us we should go to morning report or noon conference but when I’m running around trying to get things done. I hate having to choose between my education and doing what’s best for the patient–but that’s often what it comes down to.” (Intern)

When specifically asked whether setting aside dedicated time to specifically review and formulate diagnoses would be valuable, respondents were uniformly enthusiastic. Team members described attentional conflicts as being the worst when “cross covering” other teams on call days, as their patient load effectively doubled during this time. Of note, cross-covering occurred when teams were also on call—and thus took them away from important diagnostic activities such as data gathering or synthesis for patients they were admitting.

“If you were to ever design a system where errors were likely–this is how you would design it: take a team with little supervision, double their patient load, keep them busy with new challenging cases and then ask questions about patients they know little about.” (Resident)

Although diagnostic errors have been called “the next frontier for patient safety,”¹⁷ little is known about the process, barriers, and facilitators to diagnosis in teaching hospitals. In this focused ethnography conducted at 2 academic medical centers, we identified multiple cognitive and system-level challenges and potential strategies to improve diagnosis from trainees engaged in this activity. Key themes identified by those we observed included the social nature of diagnosis, fragmented information delivery, constant distractions and interruptions, and time pressures. In turn, these insights allow us to generate strategies that can be applied to improve the diagnostic process in teaching hospitals.

Our study underscores the importance of social interactions in diagnosis. In contrast, most of the interventions to prevent diagnostic errors target individual providers through practices such as metacognition and “thinking about thinking.”^18-20 These interventions are based on Daniel Kahnemann’s work on dual thought process. Type 1 thought processes are fast, subconscious, reflexive, largely intuitive, and more vulnerable to error. In contrast, Type 2 processes are slower, deliberate, analytic, and less prone to error.²¹ Although an individual’s Type 2 thought capacity is limited, a major goal of cognitive interventions is to encourage Type 2 over Type 1 thinking, an approach termed “de-biasing.”^22-24 Unfortunately, cognitive interventions testing such approaches have suffered mixed results–perhaps because of lack of focus on collective wisdom or group thinking, which may be key to diagnosis from our findings.^9,25 In this sense, morning rounds were a social gathering used to strategize and develop care plans, but with limited time to think about diagnosis.²⁶ Introduction of defined periods for individuals to engage in diagnostic activities such as de-biasing (ie, asking “what else could this be)²⁷ before or after rounds may provide an opportunity for reflection and improving diagnosis. In addition, embedding tools such as diagnosis expanders and checklists within these defined time slots^28,29 may prove to be useful in reflecting on diagnosis and preventing diagnostic errors.

An unexpected yet important finding from this study were the challenges posed by distractions and the physical environment. Potentially maladaptive workarounds to these interruptions included use of headphones; more productive strategies included updating nurses with plans to avert pages and creating a list of activities to ensure that key tasks were not forgotten.^30,31 Applying lessons from aviation, a focused effort to limit distractions during key portions of the day, might be worth considering for diagnostic safety.³² Similarly, improving the environment in which diagnosis occurs—including creating spaces that are quiet, orderly, and optimized for thinking—may be valuable.³³Our study has limitations. First, our findings are limited to direct observations; we are thus unable to comment on how unobserved aspects of care (eg, cognitive processes) might have influenced our findings. Our observations of clinical care might also have introduced a Hawthorne effect. However, because we were closely integrated with teams and conducted focus groups to corroborate our assessments, we believe that this was not the case. Second, we did not identify diagnostic errors or link processes we observed to errors. Third, our approach is limited to 2 teaching centers, thereby limiting the generalizability of findings. Relatedly, we were only able to conduct observations during weekdays; differences in weekend and night resources might affect our insights.

The cognitive and system-based barriers faced by clinicians in teaching hospitals suggest that new methods to improve diagnosis are needed. Future interventions such as defined “time-outs” for diagnosis, strategies focused on limiting distractions, and methods to improve communication between team members are novel and have parallels in other industries. As challenges to quantify diagnostic errors abound,³⁴ improving cognitive- and system-based factors via reflection through communication, concentration, and organization is necessary to improve medical decision making in academic medical centers.

Disclosures

None declared for all coauthors.

Funding

This project was supported by grant number P30HS024385 from the Agency for Healthcare Research and Quality. The funding source played no role in study design, data acquisition, analysis or decision to report these data. Dr. Chopra is supported by a career development award from the Agency of Healthcare Research and Quality (1-K08-HS022835-01). Dr. Krein is supported by a VA Health Services Research and Development Research Career Scientist Award (RCS 11-222). Dr. Singh is partially supported by Houston VA HSR&D Center for Innovations in Quality, Effectiveness and Safety (CIN 13-413). The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality or the Department of Veterans Affairs.

Diagnostic error—defined as a failure to establish an accurate and timely explanation of the patient’s health problem—is an important source of patient harm.¹ Data suggest that all patients will experience at least 1 diagnostic error in their lifetime.^2-4 Not surprisingly, diagnostic errors are among the leading categories of paid malpractice claims in the United States.⁵

Despite diagnostic errors being morbid and sometimes deadly in the hospital,^6,7 little is known about how residents and learners approach diagnostic decision making. Errors in diagnosis are believed to stem from cognitive or system failures,⁸ with errors in cognition believed to occur due to rapid, reflexive thinking operating in the absence of a more analytical, deliberate process. System-based problems (eg, lack of expert availability, technology barriers, and access to data) have also been cited as contributors.⁹ However, whether and how these apply to trainees is not known.

Therefore, we conducted a focused ethnography of inpatient medicine teams (ie, attendings, residents, interns, and medical students) in 2 affiliated teaching hospitals, aiming to (a) observe the process of diagnosis by trainees and (b) identify methods to improve the diagnostic process and prevent errors.

We designed a multimethod, focused ethnographic study to examine diagnostic decision making in hospital settings.^10,11 In contrast to anthropologic ethnographies that study entire fields using open-ended questions, our study was designed to examine the process of diagnosis from the perspective of clinicians engaged in this activity.¹¹ This approach allowed us to capture diagnostic decisions and cognitive and system-based factors in a manner currently lacking in the literature.¹²

Setting and Participants

Between January 2016 and May 2016, we observed the members of four inpatient internal medicine teaching teams at 2 affiliated teaching hospitals. We purposefully selected teaching teams for observation because they are the primary model of care in academic settings and we have expertise in carrying out similar studies.^13,14 Teaching teams typically consisted of a medical attending (senior-level physician), 1 senior resident (a second- or third-year postgraduate trainee), two interns (a trainee in their first postgraduate year), and two to four  medical students. Teams were selected at random using existing schedules and followed Monday to Friday so as to permit observation of work on call and noncall days. Owing to manpower limitations, weekend and night shifts were not observed. However, overnight events were captured during morning rounds.

Most of the teams began rounds at 8:30 AM. Typically, rounds lasted for 90–120 min and concluded with a recap (ie, “running the list”) with a review of explicit plans for patients after they had been evaluated by the attending. This discussion often occurred in the team rooms, with the attending leading the discussion with the trainees.

Data Collection

A multidisciplinary team, including clinicians (eg, physicians, nurses), nonclinicians (eg, qualitative researchers, social scientists), and healthcare engineers, conducted the observations. We observed preround activities of interns and residents before arrival of the attending (7:00 AM - 8:30 AM), followed by morning rounds with the entire team, and afternoon work that included senior residents, interns, and students.

To capture multiple aspects of the diagnostic process, we collected data using field notes modeled on components of the National Academy of Science model for diagnosis (Appendix).^1,15 This model encompasses phases of the diagnostic process (eg, data gathering, integration, formulation of a working diagnosis, treatment delivery, and outcomes) and the work system (team members, organization, technology and tools, physical environment, tasks).

Focus Groups and Interviews

At the end of weekly observations, we conducted focus groups with the residents and one-on- one interviews with the attendings. Focus groups with the residents were conducted to encourage a group discussion about the diagnostic process. Separate interviews with the attendings were performed to ensure that power differentials did not influence discussions. During focus groups, we specifically asked about challenges and possible solutions to improve diagnosis. Experienced qualitative methodologists (J.F., M.H., M.Q.) used semistructured interview guides for discussions (Appendix).

Data Analysis

After aggregating and reading the data, three reviewers (V.C., S.K., S.S.) began inductive analysis by handwriting notes and initial reflective thoughts to create preliminary codes. Multiple team members then reread the original field notes and the focus group/interview data to refine the preliminary codes and develop additional codes. Next, relationships between codes were identified and used to develop key themes. Triangulation of data collected from observations and interview/focus group sessions was carried out to compare data that we surmised with data that were verbalized by the team. The developed themes were discussed as a group to ensure consistency of major findings.

Ethical and Regulatory Oversight

This study was reviewed and approved by the Institutional Review Boards at the University of Michigan Health System (HUM-00106657) and the VA Ann Arbor Healthcare System (1-2016-010040).

Four teaching teams (4 attendings, 4 senior residents, 9 interns, and 14 medical students) were observed over 33 distinct shifts and 168 hours. Observations included morning rounds (96 h), postround call days (52 h), and postround non-call days (20 h). Morning rounds lasted an average of 127 min (range: 48-232 min) and included an average of 9 patients (range: 4-16 patients).

Themes Regarding the Diagnostic Process

We identified the following 4 primary themes related to the diagnostic process in teaching hospitals: (1) diagnosis is a social phenomenon; (2) data necessary to make diagnoses are fragmented; (3) distractions undermine the diagnostic process; and (4) time pressures interfere with diagnostic decision making (Appendix Table 1).

(1) Diagnosis is a Social Phenomenon.

Team members viewed the process of diagnosis as a social exchange of facts, findings, and strategies within a defined structure. The opportunity to discuss impressions with others was valued as a means to share, test, and process assumptions.

“Rounds are the most important part of the process. That is where we make most decisions in a collective, collaborative way with the attending present. We bounce ideas off each other.” (Intern)

Typical of social processes, variations based on time of day and schedule were observed. For instance, during call days, learners gathered data and formed working diagnosis and treatment plans with minimal attending interaction. This separation of roles and responsibilities introduced a hierarchy within diagnosis as follows:

“The interns would not call me first; they would talk to the senior resident and then if the senior thought he should chat with me, then they would call. But for the most part, they gather information and come up with the plan.” (Attending).

The work system was suited to facilitate social interactions. For instance, designated rooms (with team members informally assigned to a computer) provided physical proximity of the resident to interns and medical students. In this space, numerous informal discussions between team members (eg, “What do you think about this test?” “I’m not sure what to do about this finding.” “Should I call a [consult] on this patient?”) were observed. Although proximity to each other was viewed as beneficial, dangers to the social nature of diagnosis in the form of anchoring (ie, a cognitive bias where emphasis is placed on the first piece of data)¹⁶ were also mentioned. Similarly, the paradox associated with social proof (ie, the pressure to assume conformity within a group) was also observed as disagreement between team members and attendings rarely occurred during observations.

“I mean, they’re the attending, right? It’s hard to argue with them when they want a test or something done. When I do push back, it’s rare that others will support me–so it’s usually me and the attending.” (Resident)

“I would push back if I think it’s really bad for the patient or could cause harm–but the truth is, it doesn’t happen much.” (Intern)

(2) Data Necessary to Make Diagnoses are Fragmented

Team members universally cited fragmentation in data delivery, retrieval, and processing as a barrier to diagnosis. Team members indicated that test results might not be looked at or acted upon in a timely manner, and participants pointed to the electronic medical record as a source of this challenge.

“Before I knew about [the app for Epic], I would literally sit on the computer to get all the information we would need on rounds. Its key to making decisions. We often say we will do something, only to find the test result doesn’t support it–and then we’re back to square 1.” (Intern)

Information used by teams came from myriad sources (eg, patients, family members, electronic records) and from various settings (eg, emergency department, patient rooms, discussions with consultants). Additionally, test results often appeared without warning. Thus, availability of information was poorly aligned with clinical duties.

“They (the lab) will call us when a blood culture is positive or something is off. That is very helpful but it often comes later in the day, when we’re done with rounds.” (Resident)

The work system was highlighted as a key contributor to data fragmentation. Peculiarities of our electronic medical record (EMR) and how data were collected, stored, or presented were described as “frustrating,” and “unsafe,” by team members. Correspondingly, we frequently observed interns asking for assistance for tasks such as ordering tests or finding information despite being “trained” to use the EMR.

“People have to learn how to filter, how to recognize the most important points and link data streams together in terms of causality. But we assume they know where to find that information. It’s actually a very hard thing to do, for both the house staff and me.” (Attending)

(3) Distractions Undermine the Diagnostic Process

Distractions often created cognitive difficulties. For example, ambient noise and interruptions from neighbors working on other teams were cited as barriers to diagnosis. In addition, we observed several team members using headphones to drown out ambient noise while working on the computer.

“I know I shouldn’t do it (wear headphones), but I have no other way of turning down the noise so I can concentrate.” (Intern)

Similarly, the unpredictable nature and the volume of pages often interrupted thinking about diagnosis.

“Sometimes the pager just goes off all the time and (after making sure its not an urgent issue), I will just ignore it for a bit, especially if I am in the middle of something. It would be great if I could finish my thought process knowing I would not be interrupted.” (Resident)

To mitigate this problem, 1 attending described how he would proactively seek out nurses caring for his patients to “head off” questions (eg, “I will renew the restraints and medications this morning,” and “Is there anything you need in terms of orders for this patient that I can take care of now?”) that might lead to pages. Another resident described his approach as follows:

“I make it a point to tell the nurses where I will be hanging out and where they can find me if they have any questions. I tell them to come talk to me rather than page me since that will be less distracting.” (Resident).

Most of the interns described documentation work such as writing admission and progress notes in negative terms (“an academic exercise,” “part of the billing activity”). However, in the context of interruptions, some described this as helpful.

“The most valuable part of the thinking process was writing the assessment and plan because that’s actually my schema for all problems. It literally is the only time where I can sit and collect my thoughts to formulate a diagnosis and plan.” (Intern)

(4) Time Pressures Interfere With Diagnostic Decision Making

All team members spoke about the challenge of finding time for diagnosis during the workday. Often, they had to skip learning sessions for this purpose.

“They tell us we should go to morning report or noon conference but when I’m running around trying to get things done. I hate having to choose between my education and doing what’s best for the patient–but that’s often what it comes down to.” (Intern)

When specifically asked whether setting aside dedicated time to specifically review and formulate diagnoses would be valuable, respondents were uniformly enthusiastic. Team members described attentional conflicts as being the worst when “cross covering” other teams on call days, as their patient load effectively doubled during this time. Of note, cross-covering occurred when teams were also on call—and thus took them away from important diagnostic activities such as data gathering or synthesis for patients they were admitting.

“If you were to ever design a system where errors were likely–this is how you would design it: take a team with little supervision, double their patient load, keep them busy with new challenging cases and then ask questions about patients they know little about.” (Resident)

Although diagnostic errors have been called “the next frontier for patient safety,”¹⁷ little is known about the process, barriers, and facilitators to diagnosis in teaching hospitals. In this focused ethnography conducted at 2 academic medical centers, we identified multiple cognitive and system-level challenges and potential strategies to improve diagnosis from trainees engaged in this activity. Key themes identified by those we observed included the social nature of diagnosis, fragmented information delivery, constant distractions and interruptions, and time pressures. In turn, these insights allow us to generate strategies that can be applied to improve the diagnostic process in teaching hospitals.

Our study underscores the importance of social interactions in diagnosis. In contrast, most of the interventions to prevent diagnostic errors target individual providers through practices such as metacognition and “thinking about thinking.”^18-20 These interventions are based on Daniel Kahnemann’s work on dual thought process. Type 1 thought processes are fast, subconscious, reflexive, largely intuitive, and more vulnerable to error. In contrast, Type 2 processes are slower, deliberate, analytic, and less prone to error.²¹ Although an individual’s Type 2 thought capacity is limited, a major goal of cognitive interventions is to encourage Type 2 over Type 1 thinking, an approach termed “de-biasing.”^22-24 Unfortunately, cognitive interventions testing such approaches have suffered mixed results–perhaps because of lack of focus on collective wisdom or group thinking, which may be key to diagnosis from our findings.^9,25 In this sense, morning rounds were a social gathering used to strategize and develop care plans, but with limited time to think about diagnosis.²⁶ Introduction of defined periods for individuals to engage in diagnostic activities such as de-biasing (ie, asking “what else could this be)²⁷ before or after rounds may provide an opportunity for reflection and improving diagnosis. In addition, embedding tools such as diagnosis expanders and checklists within these defined time slots^28,29 may prove to be useful in reflecting on diagnosis and preventing diagnostic errors.

An unexpected yet important finding from this study were the challenges posed by distractions and the physical environment. Potentially maladaptive workarounds to these interruptions included use of headphones; more productive strategies included updating nurses with plans to avert pages and creating a list of activities to ensure that key tasks were not forgotten.^30,31 Applying lessons from aviation, a focused effort to limit distractions during key portions of the day, might be worth considering for diagnostic safety.³² Similarly, improving the environment in which diagnosis occurs—including creating spaces that are quiet, orderly, and optimized for thinking—may be valuable.³³Our study has limitations. First, our findings are limited to direct observations; we are thus unable to comment on how unobserved aspects of care (eg, cognitive processes) might have influenced our findings. Our observations of clinical care might also have introduced a Hawthorne effect. However, because we were closely integrated with teams and conducted focus groups to corroborate our assessments, we believe that this was not the case. Second, we did not identify diagnostic errors or link processes we observed to errors. Third, our approach is limited to 2 teaching centers, thereby limiting the generalizability of findings. Relatedly, we were only able to conduct observations during weekdays; differences in weekend and night resources might affect our insights.

The cognitive and system-based barriers faced by clinicians in teaching hospitals suggest that new methods to improve diagnosis are needed. Future interventions such as defined “time-outs” for diagnosis, strategies focused on limiting distractions, and methods to improve communication between team members are novel and have parallels in other industries. As challenges to quantify diagnostic errors abound,³⁴ improving cognitive- and system-based factors via reflection through communication, concentration, and organization is necessary to improve medical decision making in academic medical centers.

Disclosures

None declared for all coauthors.

Funding

This project was supported by grant number P30HS024385 from the Agency for Healthcare Research and Quality. The funding source played no role in study design, data acquisition, analysis or decision to report these data. Dr. Chopra is supported by a career development award from the Agency of Healthcare Research and Quality (1-K08-HS022835-01). Dr. Krein is supported by a VA Health Services Research and Development Research Career Scientist Award (RCS 11-222). Dr. Singh is partially supported by Houston VA HSR&D Center for Innovations in Quality, Effectiveness and Safety (CIN 13-413). The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality or the Department of Veterans Affairs.