Original Research

Delirium is an acute neurocognitive disorder¹ that affects up to 25% of older emergency department (ED) and hospitalized patients.^2-4 The relationship between delirium and adverse outcomes is well documented.^5-7 Delirium is a strong predictor of increased length of mechanical ventilation, longer intensive care unit and hospital stays, increased risk of falls, long-term cognitive impairment, and mortality.^8-13 Delirium is frequently missed by healthcare professionals^2,14-16 and goes undetected in up to 3 out of 4 patients by bedside nurses and medical practitioners in many hospital settings.^14,17-22 A significant barrier to recognizing delirium is the absence of brief delirium assessments.

In an effort to improve delirium recognition in the acute care setting, there has been a concerted effort to develop and validate brief delirium assessments. To address this unmet need, 4 ‘A’s Test (4AT), the Brief Confusion Assessment Method (bCAM), and the 3-minute diagnostic assessment for CAM-defined delirium (3D-CAM) are 1- to 3-minute delirium assessments that were validated in acutely ill older patients.²³ However, 1 to 3 minutes may still be too long in busy clinical environments, and briefer (<30 seconds) delirium assessments may be needed.

One potential more-rapid method to screen for delirium is to specifically test for the presence of inattention, which is a cardinal feature of delirium.^24,25 Inattention can be ascertained by having the patient recite the months backwards, recite the days of the week backwards, or spell a word backwards.²⁶ Recent studies have evaluated the diagnostic accuracy of reciting the months of the year backwards for delirium. O’Regan et al.²⁷ evaluated the diagnostic accuracy of the month of the year backwards from December to July (MOTYB-6) and observed that this task was 84% sensitive and 90% specific for delirium in older patients. However, they performed the reference standard delirium assessments in patients who had a positive MOTYB-6, which can overestimate sensitivity and underestimate specificity (verification bias).²⁸ Fick et al.²⁹ examined the diagnostic accuracy of 20 individual elements of the 3D-CAM and observed that reciting the months of the year backwards from December to January (MOTYB-12) was 83% sensitive and 69% specific for delirium. However, this was an exploratory study that was designed to identify an element of the 3D-CAM that had the best diagnostic accuracy.

To address these limitations, we sought to evaluate the diagnostic performance of the MOTYB-6 and MOTYB-12 for delirium as diagnosed by a reference standard. We also explored other brief tests of inattention such as spelling a word (“LUNCH”) backwards, reciting the days of the week backwards, 10-letter vigilance “A” task, and 5 picture recognition task.

METHODS

Study Design and Setting

This was a preplanned secondary analysis of a prospective observational study that validated 3 delirium assessments.^30,31 This study was conducted at a tertiary care, academic ED. The local institutional review board (IRB) reviewed and approved this study. Informed consent from the patient or an authorized surrogate was obtained whenever possible. Because this was an observational study and posed minimal risk to the patient, the IRB granted a waiver of consent for patients who were both unable to provide consent and were without an authorized surrogate available in the ED or by phone.

Selection of Participants

We enrolled a convenience sample of patients between June 2010 and February 2012 Monday through Friday from 8 am to 4 pm. This enrollment window was based upon the psychiatrist’s availability. Because of the extensiveness of the psychiatric evaluations, we limited enrollment to 1 patient per day. Patients who were 65 years or older, not in a hallway bed, and in the ED for less than 12 hours at the time of enrollment were included. We used a 12-hour cutoff so that patients who presented in the evening and early morning hours could be included. Patients were excluded if they were previously enrolled, non-English speaking, deaf or blind, comatose, suffered from end-stage dementia, or were unable to complete all the study assessments. The rationale for excluding patients with end-stage dementia was that diagnosing delirium in this patient population is challenging.

 

Research assistants approached patients who met inclusion criteria and determined if any exclusion criteria were present. If none of the exclusion criteria were present, then the research assistant reviewed the informed consent document with the patient or authorized surrogate if the patient was not capable of providing consent. If a patient was not capable of providing consent and no authorized surrogate was available, then the patient was enrolled (under the waiver of consent) as long as the patient assented to be a part of the study. Once the patient was enrolled, the research assistant contacted the physician rater and reference standard psychiatrists to approach the patient.

Measures of Inattention

An emergency physician (JHH) who had no formal training in the mental status assessment of elders administered a cognitive battery to the patient, including tests of inattention. The following inattention tasks were administered:

• Spell the word “LUNCH” backwards.³⁰ Patients were initially allowed to spell the word “LUNCH” forwards. Patients who were unable to perform the task were assigned 5 errors.
• Recite the months of the year backwards from December to July.^{23,26,27,30,32} Patients who were unable to perform the task were assigned 6 errors.
• Recite the days of the week backwards.^23,26,33 Patients who were unable to perform the task were assigned 7 errors.
• Ten-letter vigilance “A” task.³⁴ The patient was given a series of 10 letters (“S-A-V-E-A-H-A-A-R-T”) every 3 seconds and was asked to squeeze the rater’s hand every time the patient heard the letter “A.” Patients who were unable to perform the task were assigned 10 errors.
• Five picture recognition task.³⁴ Patients were shown 5 objects on picture cards. Afterwards, patients were shown 10 pictures with the previously shown objects intermingled. The patient had to identify which objects were seen previously in the first 5 pictures. Patients who were unable to perform the task were assigned 10 errors.
• Recite the months of the year backwards from December to January.²⁹ Patients who were unable to perform the task were assigned 12 errors.

Reference Standard for Delirium

A comprehensive consultation-liaison psychiatrist assessment was the reference standard for delirium; the diagnosis of delirium was based on Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) criteria.³⁵ Three psychiatrists who each had an average of 11 years of clinical experience and regularly diagnosed delirium as part of their daily clinical practice were available to perform these assessments. To arrive at the diagnosis of delirium, they interviewed those who best understood the patient’s mental status (eg, the patient’s family members or caregivers, physician, and nurses). They also reviewed the patient’s medical record and radiology and laboratory test results. They performed bedside cognitive testing that included, but was not limited to, the Mini-Mental State Examination, Clock Drawing Test, Luria hand sequencing task, and tests for verbal fluency. A focused neurological examination was also performed (ie, screening for paraphasic errors, tremors, tone, asterixis, frontal release signs, etc.), and they also evaluated the patient for affective lability, hallucinations, and level of alertness. If the presence of delirium was still questionable, then confrontational naming, proverb interpretation or similarities, and assessments for apraxias were performed at the discretion of the psychiatrist. The psychiatrists were blinded to the physician’s assessments, and the assessments were conducted within 3 hours of each other.

Additional Variables Collected

Using medical record review, comorbidity burden, severity of illness, and premorbid cognition were ascertained. The Charlson Comorbidity Index, a weighted index that takes into account the number and seriousness of 19 preexisting comorbid conditions, was used to quantify comorbidity burden; higher scores indicate higher comorbid burden.^36,37 The Acute Physiology Score of the Acute Physiology and Chronic Health Evaluation II was used to quantify severity of illness.³⁸ This score is based upon the initial values of 12 routine physiologic measurements such as vital sign and laboratory abnormalities; higher scores represent higher severities of illness.³⁸ The medical record was reviewed to ascertain the presence of premorbid cognitive impairment; any documentation of dementia in the patient’s clinical problem list or physician history and physical examination from the outpatient or inpatient settings was considered positive. The medical record review was performed by a research assistant and was double-checked for accuracy by one of the investigators (JHH).

Data Analyses

Measures of central tendency and dispersion for continuous variables were reported as medians and interquartile ranges. Categorical variables were reported as proportions. Receiver operating characteristic curves were constructed for each inattention task. Area under the receiver operating characteristic curves (AUC) was reported to provide a global measure of diagnostic accuracy. Sensitivities, specificities, positive likelihood ratios (PLRs), and negative likelihood ratios (NLRs) with their 95% CIs were calculated using the psychiatrist’s assessment as the reference standard.³⁹ Cut-points with PLRs greater than 10 (strongly increased the likelihood of delirium) or NLRs less than 0.1 (strongly decreased the likelihood of delirium) were preferentially reported whenever possible.

 

All statistical analyses were performed with open source R statistical software version 3.0.1 (http://www.r-project.org/), SAS 9.4 (SAS Institute, Cary, NC), and Microsoft Excel 2010 (Microsoft Inc., Redmond, WA).

RESULTS

A total of 542 patients were screened; 214 patients refused to participate, and 93 were excluded, leaving 235 patients. The patient characteristics can be seen in Table 1. Compared with all patients (N = 15,359) who presented to the ED during the study period, enrolled patients were similar in age but more likely to be male, have cardiovascular chief complaints, and be admitted to the hospital. Of those enrolled, 25 (10.6%) were delirious. Delirious patients were older, more likely to be nonwhite, have a past history of dementia, have a graduate school degree, and have a chief complaint of altered mental status.

Making any error on the MOTYB-6 task had a sensitivity of 80.0% (95% CI, 60.9%-91.1%), specificity of 57.1% (95% CI, 50.4%-63.7%), PLR of 1.87 (95% CI, 1.45-2.40) and NLR of 0.35 (95% CI, 0.16-0.77) for delirium as diagnosed by a psychiatrist. Making any error on the MOTYB-12 task had a sensitivity of 84.0% (95% CI, 65.4%-93.6%), specificity of 51.9% (95% CI, 45.2%-58.5%), PLR of 1.75 (95% CI, 1.40-2.18), and NLR of 0.31 (95% CI, 0.12-0.76) for delirium. The AUCs for the MOTYB-6 and MOTYB-12 tasks were 0.77 and 0.78, respectively, indicating very good diagnostic performance.

The diagnostic performances of the other inattention tasks and additional cutoff values for the MOTYB-6 and MOTYB-12 tasks can be seen in Table 2. Increasing the MOTYB-6 cut-off to 2 or more errors and MOTYB-12 cut-off to 3 or more errors increased the specificity to 70.0% and 70.5%, respectively, without decreasing their sensitivity. The best combination of sensitivity and specificity was reciting the days of the week backwards task; if the patient made any error, this was 84.0% (95% CI, 65.4%-93.6%) sensitive and 81.9% (95% CI, 76.1%-86.5%) specific for delirium. The inattention tasks that strongly increased the likelihood of delirium (PLR > 10) were the vigilance “A” and picture recognition tasks. If the patient made 2 or more errors on the vigilance task or 3 or more errors on the picture recognition task, then the likelihood of delirium strongly increased, as evidenced by a PLR of 16.80 (95% CI, 8.01-35.23) and 23.10 (95% CI, 7.95-67.12), respectively. No other inattention tasks were able to achieve a PLR of greater than 10, regardless of what cutoff was used. No inattention task was able to achieve a NLR of less than 0.10, which would have strongly decreased the likelihood of delirium. The best NLRs were if the patient made no errors spelling the word “LUNCH” backwards (NLR, 0.16; 95% CI, 0.04-0.60), no errors on the vigilance “A” task (NLR, 0.18; 95% CI, 0.07-0.43), and no errors on the days of the week backwards task (NLR, 0.20; 95% CI, 0.08-0.48).

DISCUSSION

Delirium is frequently missed by healthcare providers because it is not routinely screened for in the acute care setting. To help address this deficiency of care, we evaluated several brief measures of inattention that take less than 30 seconds to complete. We observed that any errors made on the MOTYB-6 and MOTYB-12 tasks had very good sensitivities (80% and 84%) but were limited by their modest specificities (approximately 50%) for delirium. As a result, these assessments have limited clinical utility as standalone delirium screens. We also explored other commonly used brief measures of inattention and at a variety of error cutoffs. Reciting the days of the week backwards appeared to best balance sensitivity and specificity. None of the inattention measures could convincingly rule out delirium (NLR < 0.10), but the vigilance “A” and picture recognition tasks may have clinical utility in ruling in delirium (PLR > 10). Overall, all the inattention tasks, including MOTYB-6 and MOTYB-12, had very good diagnostic performances based upon their AUC. However, achieving a high sensitivity often had to be sacrificed for specificity or, alternatively, achieving a high specificity had to be sacrificed for sensitivity.

Inattention has been shown to be the cardinal feature for delirium,⁴⁰ and its assessment using cognitive testing has been recommended to help identify the presence of delirium according to an expert consensus panel.²⁶ The diagnostic performance of the MOTYB-12 observed in our study is similar to a study by Fick et al., who reported that MOTYB-12 had very good sensitivity (83%) but had modest specificity (69%) with a cutoff of 1 or more errors. Hendry et al. observed that the MOTYB-12 was 91% sensitive and 50% specific using a cutoff of 4 or more errors. With regard to the MOTYB-6, our reported specificity was different from what was observed by O’Regan et al.²⁷ Using 1 or more errors as a cutoff, they observed a much higher specificity for delirium than we did (90% vs 57%). Discordant observations regarding the diagnostic accuracy for other inattention tasks also exist. We observed that making any error on the days of the week backwards task was 84% sensitive and 82% specific for delirium, whereas Fick et al. observed a sensitivity and specificity of 50% and 94%, respectively. For the vigilance “A” task, we observed that making 2 or more errors over a series of 10 letters was 64.0% sensitive and 91.4% specific for delirium, whereas Pompei et al.⁴¹ observed that making 2 or more errors over a series of 60 letters was 51% sensitive and 77% specific for delirium.

The abovementioned discordant findings may be driven by spectrum bias, wherein the sensitivities and specificities for each inattention task may differ in different subgroups. As a result, differences in the age distribution, proportion of college graduates, history of dementia, and susceptibility to delirium can influence overall sensitivity and specificity. Objective measures of delirium, including the inattention screens studied, are particularly prone to spectrum bias.^31,34 However, the strength of this approach is that the assessment of inattention becomes less reliant upon clinical judgment and allows it to be used by raters from a wide range of clinical backgrounds. On the other hand, a subjective interpretation of these inattention tasks may allow the rater to capture the subtleties of inattention (ie, decreased speed of performance in a highly intelligent and well-educated patient without dementia). The disadvantage of this approach, however, is that it is more dependent on clinical judgment and may have decreased diagnostic accuracy in those with less clinical experience or with limited training.^14,42,43 These factors must be carefully considered when determining which delirium assessment to use.

Additional research is required to determine the clinical utility of these brief inattention assessments. These findings need to be further validated in larger studies, and the optimal cutoff of each task for different subgroup of patients (eg, demented vs nondemented) needs to be further clarified. It is not completely clear whether these inattention tests can serve as standalone assessments. Depending on the cutoff used, some of these assessments may have unacceptable false negative or false positive rates that may lead to increased adverse patient outcomes or increased resource utilization, respectively. Additional components or assessments may be needed to improve the diagnostic accuracy of these assessments. In addition to understanding these inattention assessments’ diagnostic accuracies, their ability to predict adverse outcomes also needs to be investigated. While a previous study observed that making any error on the MOTYB-12 task was associated with increased physical restraint use and prolonged hospital length of stay,⁴⁴ these assessments’ ability to prognosticate long-term outcomes such as mortality or long-term cognition or function need to be studied. Lastly, studies should also evaluate how easily implementable these assessments are and whether improved delirium recognition leads to improved patient outcomes.

This study has several notable limitations. Though planned a priori, this was a secondary analysis of a larger investigation designed to validate 3 delirium assessments. Our sample size was also relatively small, causing our 95% CIs to overlap in most cases and limiting the statistical power to truly determine whether one measure is better than the other. We also asked the patient to recite the months backwards from December to July as well as recite the months backwards from December to January. It is possible that the patient may have performed better at going from December to January because of learning effect. Our reference standard for delirium was based upon DSM-IV-TR criteria. The new DSM-V criteria may be more restrictive and may slightly change the sensitivities and specificities of the inattention tasks. We enrolled a convenience sample and enrolled patients who were more likely to be male, have cardiovascular chief complaints, and be admitted to the hospital; as a result, selection bias may have been introduced. Lastly, this study was conducted in a single center and enrolled patients who were 65 years and older. Our findings may not be generalizable to other settings and in those who are less than 65 years of age.

 

CONCLUSIONS

The MOTYB-6 and MOTYB-12 tasks had very good sensitivities but modest specificities (approximately 50%) using any error made as a cutoff; increasing cutoff to 2 errors and 3 errors, respectively, improved their specificities (approximately 70%) with minimal impact to their sensitivities. Reciting the days of the week backwards, spelling the word “LUNCH” backwards, and the 10-letter vigilance “A” task appeared to perform the best in ruling out delirium but only moderately decreased the likelihood of delirium. The 10-letter Vigilance “A” and picture recognition task appeared to perform the best in ruling in delirium. Days of the week backwards appeared to have the best combination of sensitivity and specificity.

Disclosure

This study was funded by the Emergency Medicine Foundation Career Development Award, National Institutes of Health K23AG032355, and National Center for Research Resources, Grant UL1 RR024975-01. The authors report no financial conflicts of interest.

Delirium is an acute neurocognitive disorder¹ that affects up to 25% of older emergency department (ED) and hospitalized patients.^2-4 The relationship between delirium and adverse outcomes is well documented.^5-7 Delirium is a strong predictor of increased length of mechanical ventilation, longer intensive care unit and hospital stays, increased risk of falls, long-term cognitive impairment, and mortality.^8-13 Delirium is frequently missed by healthcare professionals^2,14-16 and goes undetected in up to 3 out of 4 patients by bedside nurses and medical practitioners in many hospital settings.^14,17-22 A significant barrier to recognizing delirium is the absence of brief delirium assessments.

In an effort to improve delirium recognition in the acute care setting, there has been a concerted effort to develop and validate brief delirium assessments. To address this unmet need, 4 ‘A’s Test (4AT), the Brief Confusion Assessment Method (bCAM), and the 3-minute diagnostic assessment for CAM-defined delirium (3D-CAM) are 1- to 3-minute delirium assessments that were validated in acutely ill older patients.²³ However, 1 to 3 minutes may still be too long in busy clinical environments, and briefer (<30 seconds) delirium assessments may be needed.

One potential more-rapid method to screen for delirium is to specifically test for the presence of inattention, which is a cardinal feature of delirium.^24,25 Inattention can be ascertained by having the patient recite the months backwards, recite the days of the week backwards, or spell a word backwards.²⁶ Recent studies have evaluated the diagnostic accuracy of reciting the months of the year backwards for delirium. O’Regan et al.²⁷ evaluated the diagnostic accuracy of the month of the year backwards from December to July (MOTYB-6) and observed that this task was 84% sensitive and 90% specific for delirium in older patients. However, they performed the reference standard delirium assessments in patients who had a positive MOTYB-6, which can overestimate sensitivity and underestimate specificity (verification bias).²⁸ Fick et al.²⁹ examined the diagnostic accuracy of 20 individual elements of the 3D-CAM and observed that reciting the months of the year backwards from December to January (MOTYB-12) was 83% sensitive and 69% specific for delirium. However, this was an exploratory study that was designed to identify an element of the 3D-CAM that had the best diagnostic accuracy.

To address these limitations, we sought to evaluate the diagnostic performance of the MOTYB-6 and MOTYB-12 for delirium as diagnosed by a reference standard. We also explored other brief tests of inattention such as spelling a word (“LUNCH”) backwards, reciting the days of the week backwards, 10-letter vigilance “A” task, and 5 picture recognition task.

METHODS

Study Design and Setting

This was a preplanned secondary analysis of a prospective observational study that validated 3 delirium assessments.^30,31 This study was conducted at a tertiary care, academic ED. The local institutional review board (IRB) reviewed and approved this study. Informed consent from the patient or an authorized surrogate was obtained whenever possible. Because this was an observational study and posed minimal risk to the patient, the IRB granted a waiver of consent for patients who were both unable to provide consent and were without an authorized surrogate available in the ED or by phone.

Selection of Participants

We enrolled a convenience sample of patients between June 2010 and February 2012 Monday through Friday from 8 am to 4 pm. This enrollment window was based upon the psychiatrist’s availability. Because of the extensiveness of the psychiatric evaluations, we limited enrollment to 1 patient per day. Patients who were 65 years or older, not in a hallway bed, and in the ED for less than 12 hours at the time of enrollment were included. We used a 12-hour cutoff so that patients who presented in the evening and early morning hours could be included. Patients were excluded if they were previously enrolled, non-English speaking, deaf or blind, comatose, suffered from end-stage dementia, or were unable to complete all the study assessments. The rationale for excluding patients with end-stage dementia was that diagnosing delirium in this patient population is challenging.

 

Research assistants approached patients who met inclusion criteria and determined if any exclusion criteria were present. If none of the exclusion criteria were present, then the research assistant reviewed the informed consent document with the patient or authorized surrogate if the patient was not capable of providing consent. If a patient was not capable of providing consent and no authorized surrogate was available, then the patient was enrolled (under the waiver of consent) as long as the patient assented to be a part of the study. Once the patient was enrolled, the research assistant contacted the physician rater and reference standard psychiatrists to approach the patient.

Measures of Inattention

An emergency physician (JHH) who had no formal training in the mental status assessment of elders administered a cognitive battery to the patient, including tests of inattention. The following inattention tasks were administered:

• Spell the word “LUNCH” backwards.³⁰ Patients were initially allowed to spell the word “LUNCH” forwards. Patients who were unable to perform the task were assigned 5 errors.
• Recite the months of the year backwards from December to July.^{23,26,27,30,32} Patients who were unable to perform the task were assigned 6 errors.
• Recite the days of the week backwards.^23,26,33 Patients who were unable to perform the task were assigned 7 errors.
• Ten-letter vigilance “A” task.³⁴ The patient was given a series of 10 letters (“S-A-V-E-A-H-A-A-R-T”) every 3 seconds and was asked to squeeze the rater’s hand every time the patient heard the letter “A.” Patients who were unable to perform the task were assigned 10 errors.
• Five picture recognition task.³⁴ Patients were shown 5 objects on picture cards. Afterwards, patients were shown 10 pictures with the previously shown objects intermingled. The patient had to identify which objects were seen previously in the first 5 pictures. Patients who were unable to perform the task were assigned 10 errors.
• Recite the months of the year backwards from December to January.²⁹ Patients who were unable to perform the task were assigned 12 errors.

Reference Standard for Delirium

A comprehensive consultation-liaison psychiatrist assessment was the reference standard for delirium; the diagnosis of delirium was based on Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) criteria.³⁵ Three psychiatrists who each had an average of 11 years of clinical experience and regularly diagnosed delirium as part of their daily clinical practice were available to perform these assessments. To arrive at the diagnosis of delirium, they interviewed those who best understood the patient’s mental status (eg, the patient’s family members or caregivers, physician, and nurses). They also reviewed the patient’s medical record and radiology and laboratory test results. They performed bedside cognitive testing that included, but was not limited to, the Mini-Mental State Examination, Clock Drawing Test, Luria hand sequencing task, and tests for verbal fluency. A focused neurological examination was also performed (ie, screening for paraphasic errors, tremors, tone, asterixis, frontal release signs, etc.), and they also evaluated the patient for affective lability, hallucinations, and level of alertness. If the presence of delirium was still questionable, then confrontational naming, proverb interpretation or similarities, and assessments for apraxias were performed at the discretion of the psychiatrist. The psychiatrists were blinded to the physician’s assessments, and the assessments were conducted within 3 hours of each other.

Additional Variables Collected

Using medical record review, comorbidity burden, severity of illness, and premorbid cognition were ascertained. The Charlson Comorbidity Index, a weighted index that takes into account the number and seriousness of 19 preexisting comorbid conditions, was used to quantify comorbidity burden; higher scores indicate higher comorbid burden.^36,37 The Acute Physiology Score of the Acute Physiology and Chronic Health Evaluation II was used to quantify severity of illness.³⁸ This score is based upon the initial values of 12 routine physiologic measurements such as vital sign and laboratory abnormalities; higher scores represent higher severities of illness.³⁸ The medical record was reviewed to ascertain the presence of premorbid cognitive impairment; any documentation of dementia in the patient’s clinical problem list or physician history and physical examination from the outpatient or inpatient settings was considered positive. The medical record review was performed by a research assistant and was double-checked for accuracy by one of the investigators (JHH).

Data Analyses

Measures of central tendency and dispersion for continuous variables were reported as medians and interquartile ranges. Categorical variables were reported as proportions. Receiver operating characteristic curves were constructed for each inattention task. Area under the receiver operating characteristic curves (AUC) was reported to provide a global measure of diagnostic accuracy. Sensitivities, specificities, positive likelihood ratios (PLRs), and negative likelihood ratios (NLRs) with their 95% CIs were calculated using the psychiatrist’s assessment as the reference standard.³⁹ Cut-points with PLRs greater than 10 (strongly increased the likelihood of delirium) or NLRs less than 0.1 (strongly decreased the likelihood of delirium) were preferentially reported whenever possible.

 

All statistical analyses were performed with open source R statistical software version 3.0.1 (http://www.r-project.org/), SAS 9.4 (SAS Institute, Cary, NC), and Microsoft Excel 2010 (Microsoft Inc., Redmond, WA).

RESULTS

A total of 542 patients were screened; 214 patients refused to participate, and 93 were excluded, leaving 235 patients. The patient characteristics can be seen in Table 1. Compared with all patients (N = 15,359) who presented to the ED during the study period, enrolled patients were similar in age but more likely to be male, have cardiovascular chief complaints, and be admitted to the hospital. Of those enrolled, 25 (10.6%) were delirious. Delirious patients were older, more likely to be nonwhite, have a past history of dementia, have a graduate school degree, and have a chief complaint of altered mental status.

Making any error on the MOTYB-6 task had a sensitivity of 80.0% (95% CI, 60.9%-91.1%), specificity of 57.1% (95% CI, 50.4%-63.7%), PLR of 1.87 (95% CI, 1.45-2.40) and NLR of 0.35 (95% CI, 0.16-0.77) for delirium as diagnosed by a psychiatrist. Making any error on the MOTYB-12 task had a sensitivity of 84.0% (95% CI, 65.4%-93.6%), specificity of 51.9% (95% CI, 45.2%-58.5%), PLR of 1.75 (95% CI, 1.40-2.18), and NLR of 0.31 (95% CI, 0.12-0.76) for delirium. The AUCs for the MOTYB-6 and MOTYB-12 tasks were 0.77 and 0.78, respectively, indicating very good diagnostic performance.

The diagnostic performances of the other inattention tasks and additional cutoff values for the MOTYB-6 and MOTYB-12 tasks can be seen in Table 2. Increasing the MOTYB-6 cut-off to 2 or more errors and MOTYB-12 cut-off to 3 or more errors increased the specificity to 70.0% and 70.5%, respectively, without decreasing their sensitivity. The best combination of sensitivity and specificity was reciting the days of the week backwards task; if the patient made any error, this was 84.0% (95% CI, 65.4%-93.6%) sensitive and 81.9% (95% CI, 76.1%-86.5%) specific for delirium. The inattention tasks that strongly increased the likelihood of delirium (PLR > 10) were the vigilance “A” and picture recognition tasks. If the patient made 2 or more errors on the vigilance task or 3 or more errors on the picture recognition task, then the likelihood of delirium strongly increased, as evidenced by a PLR of 16.80 (95% CI, 8.01-35.23) and 23.10 (95% CI, 7.95-67.12), respectively. No other inattention tasks were able to achieve a PLR of greater than 10, regardless of what cutoff was used. No inattention task was able to achieve a NLR of less than 0.10, which would have strongly decreased the likelihood of delirium. The best NLRs were if the patient made no errors spelling the word “LUNCH” backwards (NLR, 0.16; 95% CI, 0.04-0.60), no errors on the vigilance “A” task (NLR, 0.18; 95% CI, 0.07-0.43), and no errors on the days of the week backwards task (NLR, 0.20; 95% CI, 0.08-0.48).

DISCUSSION

Delirium is frequently missed by healthcare providers because it is not routinely screened for in the acute care setting. To help address this deficiency of care, we evaluated several brief measures of inattention that take less than 30 seconds to complete. We observed that any errors made on the MOTYB-6 and MOTYB-12 tasks had very good sensitivities (80% and 84%) but were limited by their modest specificities (approximately 50%) for delirium. As a result, these assessments have limited clinical utility as standalone delirium screens. We also explored other commonly used brief measures of inattention and at a variety of error cutoffs. Reciting the days of the week backwards appeared to best balance sensitivity and specificity. None of the inattention measures could convincingly rule out delirium (NLR < 0.10), but the vigilance “A” and picture recognition tasks may have clinical utility in ruling in delirium (PLR > 10). Overall, all the inattention tasks, including MOTYB-6 and MOTYB-12, had very good diagnostic performances based upon their AUC. However, achieving a high sensitivity often had to be sacrificed for specificity or, alternatively, achieving a high specificity had to be sacrificed for sensitivity.

Inattention has been shown to be the cardinal feature for delirium,⁴⁰ and its assessment using cognitive testing has been recommended to help identify the presence of delirium according to an expert consensus panel.²⁶ The diagnostic performance of the MOTYB-12 observed in our study is similar to a study by Fick et al., who reported that MOTYB-12 had very good sensitivity (83%) but had modest specificity (69%) with a cutoff of 1 or more errors. Hendry et al. observed that the MOTYB-12 was 91% sensitive and 50% specific using a cutoff of 4 or more errors. With regard to the MOTYB-6, our reported specificity was different from what was observed by O’Regan et al.²⁷ Using 1 or more errors as a cutoff, they observed a much higher specificity for delirium than we did (90% vs 57%). Discordant observations regarding the diagnostic accuracy for other inattention tasks also exist. We observed that making any error on the days of the week backwards task was 84% sensitive and 82% specific for delirium, whereas Fick et al. observed a sensitivity and specificity of 50% and 94%, respectively. For the vigilance “A” task, we observed that making 2 or more errors over a series of 10 letters was 64.0% sensitive and 91.4% specific for delirium, whereas Pompei et al.⁴¹ observed that making 2 or more errors over a series of 60 letters was 51% sensitive and 77% specific for delirium.

The abovementioned discordant findings may be driven by spectrum bias, wherein the sensitivities and specificities for each inattention task may differ in different subgroups. As a result, differences in the age distribution, proportion of college graduates, history of dementia, and susceptibility to delirium can influence overall sensitivity and specificity. Objective measures of delirium, including the inattention screens studied, are particularly prone to spectrum bias.^31,34 However, the strength of this approach is that the assessment of inattention becomes less reliant upon clinical judgment and allows it to be used by raters from a wide range of clinical backgrounds. On the other hand, a subjective interpretation of these inattention tasks may allow the rater to capture the subtleties of inattention (ie, decreased speed of performance in a highly intelligent and well-educated patient without dementia). The disadvantage of this approach, however, is that it is more dependent on clinical judgment and may have decreased diagnostic accuracy in those with less clinical experience or with limited training.^14,42,43 These factors must be carefully considered when determining which delirium assessment to use.

Additional research is required to determine the clinical utility of these brief inattention assessments. These findings need to be further validated in larger studies, and the optimal cutoff of each task for different subgroup of patients (eg, demented vs nondemented) needs to be further clarified. It is not completely clear whether these inattention tests can serve as standalone assessments. Depending on the cutoff used, some of these assessments may have unacceptable false negative or false positive rates that may lead to increased adverse patient outcomes or increased resource utilization, respectively. Additional components or assessments may be needed to improve the diagnostic accuracy of these assessments. In addition to understanding these inattention assessments’ diagnostic accuracies, their ability to predict adverse outcomes also needs to be investigated. While a previous study observed that making any error on the MOTYB-12 task was associated with increased physical restraint use and prolonged hospital length of stay,⁴⁴ these assessments’ ability to prognosticate long-term outcomes such as mortality or long-term cognition or function need to be studied. Lastly, studies should also evaluate how easily implementable these assessments are and whether improved delirium recognition leads to improved patient outcomes.

This study has several notable limitations. Though planned a priori, this was a secondary analysis of a larger investigation designed to validate 3 delirium assessments. Our sample size was also relatively small, causing our 95% CIs to overlap in most cases and limiting the statistical power to truly determine whether one measure is better than the other. We also asked the patient to recite the months backwards from December to July as well as recite the months backwards from December to January. It is possible that the patient may have performed better at going from December to January because of learning effect. Our reference standard for delirium was based upon DSM-IV-TR criteria. The new DSM-V criteria may be more restrictive and may slightly change the sensitivities and specificities of the inattention tasks. We enrolled a convenience sample and enrolled patients who were more likely to be male, have cardiovascular chief complaints, and be admitted to the hospital; as a result, selection bias may have been introduced. Lastly, this study was conducted in a single center and enrolled patients who were 65 years and older. Our findings may not be generalizable to other settings and in those who are less than 65 years of age.

 

CONCLUSIONS

The MOTYB-6 and MOTYB-12 tasks had very good sensitivities but modest specificities (approximately 50%) using any error made as a cutoff; increasing cutoff to 2 errors and 3 errors, respectively, improved their specificities (approximately 70%) with minimal impact to their sensitivities. Reciting the days of the week backwards, spelling the word “LUNCH” backwards, and the 10-letter vigilance “A” task appeared to perform the best in ruling out delirium but only moderately decreased the likelihood of delirium. The 10-letter Vigilance “A” and picture recognition task appeared to perform the best in ruling in delirium. Days of the week backwards appeared to have the best combination of sensitivity and specificity.

Disclosure

This study was funded by the Emergency Medicine Foundation Career Development Award, National Institutes of Health K23AG032355, and National Center for Research Resources, Grant UL1 RR024975-01. The authors report no financial conflicts of interest.

Delirium is an acute neurocognitive disorder¹ that affects up to 25% of older emergency department (ED) and hospitalized patients.^2-4 The relationship between delirium and adverse outcomes is well documented.^5-7 Delirium is a strong predictor of increased length of mechanical ventilation, longer intensive care unit and hospital stays, increased risk of falls, long-term cognitive impairment, and mortality.^8-13 Delirium is frequently missed by healthcare professionals^2,14-16 and goes undetected in up to 3 out of 4 patients by bedside nurses and medical practitioners in many hospital settings.^14,17-22 A significant barrier to recognizing delirium is the absence of brief delirium assessments.

In an effort to improve delirium recognition in the acute care setting, there has been a concerted effort to develop and validate brief delirium assessments. To address this unmet need, 4 ‘A’s Test (4AT), the Brief Confusion Assessment Method (bCAM), and the 3-minute diagnostic assessment for CAM-defined delirium (3D-CAM) are 1- to 3-minute delirium assessments that were validated in acutely ill older patients.²³ However, 1 to 3 minutes may still be too long in busy clinical environments, and briefer (<30 seconds) delirium assessments may be needed.

One potential more-rapid method to screen for delirium is to specifically test for the presence of inattention, which is a cardinal feature of delirium.^24,25 Inattention can be ascertained by having the patient recite the months backwards, recite the days of the week backwards, or spell a word backwards.²⁶ Recent studies have evaluated the diagnostic accuracy of reciting the months of the year backwards for delirium. O’Regan et al.²⁷ evaluated the diagnostic accuracy of the month of the year backwards from December to July (MOTYB-6) and observed that this task was 84% sensitive and 90% specific for delirium in older patients. However, they performed the reference standard delirium assessments in patients who had a positive MOTYB-6, which can overestimate sensitivity and underestimate specificity (verification bias).²⁸ Fick et al.²⁹ examined the diagnostic accuracy of 20 individual elements of the 3D-CAM and observed that reciting the months of the year backwards from December to January (MOTYB-12) was 83% sensitive and 69% specific for delirium. However, this was an exploratory study that was designed to identify an element of the 3D-CAM that had the best diagnostic accuracy.

To address these limitations, we sought to evaluate the diagnostic performance of the MOTYB-6 and MOTYB-12 for delirium as diagnosed by a reference standard. We also explored other brief tests of inattention such as spelling a word (“LUNCH”) backwards, reciting the days of the week backwards, 10-letter vigilance “A” task, and 5 picture recognition task.

METHODS

Study Design and Setting

This was a preplanned secondary analysis of a prospective observational study that validated 3 delirium assessments.^30,31 This study was conducted at a tertiary care, academic ED. The local institutional review board (IRB) reviewed and approved this study. Informed consent from the patient or an authorized surrogate was obtained whenever possible. Because this was an observational study and posed minimal risk to the patient, the IRB granted a waiver of consent for patients who were both unable to provide consent and were without an authorized surrogate available in the ED or by phone.

Selection of Participants

We enrolled a convenience sample of patients between June 2010 and February 2012 Monday through Friday from 8 am to 4 pm. This enrollment window was based upon the psychiatrist’s availability. Because of the extensiveness of the psychiatric evaluations, we limited enrollment to 1 patient per day. Patients who were 65 years or older, not in a hallway bed, and in the ED for less than 12 hours at the time of enrollment were included. We used a 12-hour cutoff so that patients who presented in the evening and early morning hours could be included. Patients were excluded if they were previously enrolled, non-English speaking, deaf or blind, comatose, suffered from end-stage dementia, or were unable to complete all the study assessments. The rationale for excluding patients with end-stage dementia was that diagnosing delirium in this patient population is challenging.

 

Research assistants approached patients who met inclusion criteria and determined if any exclusion criteria were present. If none of the exclusion criteria were present, then the research assistant reviewed the informed consent document with the patient or authorized surrogate if the patient was not capable of providing consent. If a patient was not capable of providing consent and no authorized surrogate was available, then the patient was enrolled (under the waiver of consent) as long as the patient assented to be a part of the study. Once the patient was enrolled, the research assistant contacted the physician rater and reference standard psychiatrists to approach the patient.

Measures of Inattention

An emergency physician (JHH) who had no formal training in the mental status assessment of elders administered a cognitive battery to the patient, including tests of inattention. The following inattention tasks were administered:

• Spell the word “LUNCH” backwards.³⁰ Patients were initially allowed to spell the word “LUNCH” forwards. Patients who were unable to perform the task were assigned 5 errors.
• Recite the months of the year backwards from December to July.^{23,26,27,30,32} Patients who were unable to perform the task were assigned 6 errors.
• Recite the days of the week backwards.^23,26,33 Patients who were unable to perform the task were assigned 7 errors.
• Ten-letter vigilance “A” task.³⁴ The patient was given a series of 10 letters (“S-A-V-E-A-H-A-A-R-T”) every 3 seconds and was asked to squeeze the rater’s hand every time the patient heard the letter “A.” Patients who were unable to perform the task were assigned 10 errors.
• Five picture recognition task.³⁴ Patients were shown 5 objects on picture cards. Afterwards, patients were shown 10 pictures with the previously shown objects intermingled. The patient had to identify which objects were seen previously in the first 5 pictures. Patients who were unable to perform the task were assigned 10 errors.
• Recite the months of the year backwards from December to January.²⁹ Patients who were unable to perform the task were assigned 12 errors.

Reference Standard for Delirium

A comprehensive consultation-liaison psychiatrist assessment was the reference standard for delirium; the diagnosis of delirium was based on Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) criteria.³⁵ Three psychiatrists who each had an average of 11 years of clinical experience and regularly diagnosed delirium as part of their daily clinical practice were available to perform these assessments. To arrive at the diagnosis of delirium, they interviewed those who best understood the patient’s mental status (eg, the patient’s family members or caregivers, physician, and nurses). They also reviewed the patient’s medical record and radiology and laboratory test results. They performed bedside cognitive testing that included, but was not limited to, the Mini-Mental State Examination, Clock Drawing Test, Luria hand sequencing task, and tests for verbal fluency. A focused neurological examination was also performed (ie, screening for paraphasic errors, tremors, tone, asterixis, frontal release signs, etc.), and they also evaluated the patient for affective lability, hallucinations, and level of alertness. If the presence of delirium was still questionable, then confrontational naming, proverb interpretation or similarities, and assessments for apraxias were performed at the discretion of the psychiatrist. The psychiatrists were blinded to the physician’s assessments, and the assessments were conducted within 3 hours of each other.

Additional Variables Collected

Using medical record review, comorbidity burden, severity of illness, and premorbid cognition were ascertained. The Charlson Comorbidity Index, a weighted index that takes into account the number and seriousness of 19 preexisting comorbid conditions, was used to quantify comorbidity burden; higher scores indicate higher comorbid burden.^36,37 The Acute Physiology Score of the Acute Physiology and Chronic Health Evaluation II was used to quantify severity of illness.³⁸ This score is based upon the initial values of 12 routine physiologic measurements such as vital sign and laboratory abnormalities; higher scores represent higher severities of illness.³⁸ The medical record was reviewed to ascertain the presence of premorbid cognitive impairment; any documentation of dementia in the patient’s clinical problem list or physician history and physical examination from the outpatient or inpatient settings was considered positive. The medical record review was performed by a research assistant and was double-checked for accuracy by one of the investigators (JHH).

Data Analyses

Measures of central tendency and dispersion for continuous variables were reported as medians and interquartile ranges. Categorical variables were reported as proportions. Receiver operating characteristic curves were constructed for each inattention task. Area under the receiver operating characteristic curves (AUC) was reported to provide a global measure of diagnostic accuracy. Sensitivities, specificities, positive likelihood ratios (PLRs), and negative likelihood ratios (NLRs) with their 95% CIs were calculated using the psychiatrist’s assessment as the reference standard.³⁹ Cut-points with PLRs greater than 10 (strongly increased the likelihood of delirium) or NLRs less than 0.1 (strongly decreased the likelihood of delirium) were preferentially reported whenever possible.

 

All statistical analyses were performed with open source R statistical software version 3.0.1 (http://www.r-project.org/), SAS 9.4 (SAS Institute, Cary, NC), and Microsoft Excel 2010 (Microsoft Inc., Redmond, WA).

RESULTS

A total of 542 patients were screened; 214 patients refused to participate, and 93 were excluded, leaving 235 patients. The patient characteristics can be seen in Table 1. Compared with all patients (N = 15,359) who presented to the ED during the study period, enrolled patients were similar in age but more likely to be male, have cardiovascular chief complaints, and be admitted to the hospital. Of those enrolled, 25 (10.6%) were delirious. Delirious patients were older, more likely to be nonwhite, have a past history of dementia, have a graduate school degree, and have a chief complaint of altered mental status.

Making any error on the MOTYB-6 task had a sensitivity of 80.0% (95% CI, 60.9%-91.1%), specificity of 57.1% (95% CI, 50.4%-63.7%), PLR of 1.87 (95% CI, 1.45-2.40) and NLR of 0.35 (95% CI, 0.16-0.77) for delirium as diagnosed by a psychiatrist. Making any error on the MOTYB-12 task had a sensitivity of 84.0% (95% CI, 65.4%-93.6%), specificity of 51.9% (95% CI, 45.2%-58.5%), PLR of 1.75 (95% CI, 1.40-2.18), and NLR of 0.31 (95% CI, 0.12-0.76) for delirium. The AUCs for the MOTYB-6 and MOTYB-12 tasks were 0.77 and 0.78, respectively, indicating very good diagnostic performance.

The diagnostic performances of the other inattention tasks and additional cutoff values for the MOTYB-6 and MOTYB-12 tasks can be seen in Table 2. Increasing the MOTYB-6 cut-off to 2 or more errors and MOTYB-12 cut-off to 3 or more errors increased the specificity to 70.0% and 70.5%, respectively, without decreasing their sensitivity. The best combination of sensitivity and specificity was reciting the days of the week backwards task; if the patient made any error, this was 84.0% (95% CI, 65.4%-93.6%) sensitive and 81.9% (95% CI, 76.1%-86.5%) specific for delirium. The inattention tasks that strongly increased the likelihood of delirium (PLR > 10) were the vigilance “A” and picture recognition tasks. If the patient made 2 or more errors on the vigilance task or 3 or more errors on the picture recognition task, then the likelihood of delirium strongly increased, as evidenced by a PLR of 16.80 (95% CI, 8.01-35.23) and 23.10 (95% CI, 7.95-67.12), respectively. No other inattention tasks were able to achieve a PLR of greater than 10, regardless of what cutoff was used. No inattention task was able to achieve a NLR of less than 0.10, which would have strongly decreased the likelihood of delirium. The best NLRs were if the patient made no errors spelling the word “LUNCH” backwards (NLR, 0.16; 95% CI, 0.04-0.60), no errors on the vigilance “A” task (NLR, 0.18; 95% CI, 0.07-0.43), and no errors on the days of the week backwards task (NLR, 0.20; 95% CI, 0.08-0.48).

DISCUSSION

Delirium is frequently missed by healthcare providers because it is not routinely screened for in the acute care setting. To help address this deficiency of care, we evaluated several brief measures of inattention that take less than 30 seconds to complete. We observed that any errors made on the MOTYB-6 and MOTYB-12 tasks had very good sensitivities (80% and 84%) but were limited by their modest specificities (approximately 50%) for delirium. As a result, these assessments have limited clinical utility as standalone delirium screens. We also explored other commonly used brief measures of inattention and at a variety of error cutoffs. Reciting the days of the week backwards appeared to best balance sensitivity and specificity. None of the inattention measures could convincingly rule out delirium (NLR < 0.10), but the vigilance “A” and picture recognition tasks may have clinical utility in ruling in delirium (PLR > 10). Overall, all the inattention tasks, including MOTYB-6 and MOTYB-12, had very good diagnostic performances based upon their AUC. However, achieving a high sensitivity often had to be sacrificed for specificity or, alternatively, achieving a high specificity had to be sacrificed for sensitivity.

Inattention has been shown to be the cardinal feature for delirium,⁴⁰ and its assessment using cognitive testing has been recommended to help identify the presence of delirium according to an expert consensus panel.²⁶ The diagnostic performance of the MOTYB-12 observed in our study is similar to a study by Fick et al., who reported that MOTYB-12 had very good sensitivity (83%) but had modest specificity (69%) with a cutoff of 1 or more errors. Hendry et al. observed that the MOTYB-12 was 91% sensitive and 50% specific using a cutoff of 4 or more errors. With regard to the MOTYB-6, our reported specificity was different from what was observed by O’Regan et al.²⁷ Using 1 or more errors as a cutoff, they observed a much higher specificity for delirium than we did (90% vs 57%). Discordant observations regarding the diagnostic accuracy for other inattention tasks also exist. We observed that making any error on the days of the week backwards task was 84% sensitive and 82% specific for delirium, whereas Fick et al. observed a sensitivity and specificity of 50% and 94%, respectively. For the vigilance “A” task, we observed that making 2 or more errors over a series of 10 letters was 64.0% sensitive and 91.4% specific for delirium, whereas Pompei et al.⁴¹ observed that making 2 or more errors over a series of 60 letters was 51% sensitive and 77% specific for delirium.

The abovementioned discordant findings may be driven by spectrum bias, wherein the sensitivities and specificities for each inattention task may differ in different subgroups. As a result, differences in the age distribution, proportion of college graduates, history of dementia, and susceptibility to delirium can influence overall sensitivity and specificity. Objective measures of delirium, including the inattention screens studied, are particularly prone to spectrum bias.^31,34 However, the strength of this approach is that the assessment of inattention becomes less reliant upon clinical judgment and allows it to be used by raters from a wide range of clinical backgrounds. On the other hand, a subjective interpretation of these inattention tasks may allow the rater to capture the subtleties of inattention (ie, decreased speed of performance in a highly intelligent and well-educated patient without dementia). The disadvantage of this approach, however, is that it is more dependent on clinical judgment and may have decreased diagnostic accuracy in those with less clinical experience or with limited training.^14,42,43 These factors must be carefully considered when determining which delirium assessment to use.

Additional research is required to determine the clinical utility of these brief inattention assessments. These findings need to be further validated in larger studies, and the optimal cutoff of each task for different subgroup of patients (eg, demented vs nondemented) needs to be further clarified. It is not completely clear whether these inattention tests can serve as standalone assessments. Depending on the cutoff used, some of these assessments may have unacceptable false negative or false positive rates that may lead to increased adverse patient outcomes or increased resource utilization, respectively. Additional components or assessments may be needed to improve the diagnostic accuracy of these assessments. In addition to understanding these inattention assessments’ diagnostic accuracies, their ability to predict adverse outcomes also needs to be investigated. While a previous study observed that making any error on the MOTYB-12 task was associated with increased physical restraint use and prolonged hospital length of stay,⁴⁴ these assessments’ ability to prognosticate long-term outcomes such as mortality or long-term cognition or function need to be studied. Lastly, studies should also evaluate how easily implementable these assessments are and whether improved delirium recognition leads to improved patient outcomes.

This study has several notable limitations. Though planned a priori, this was a secondary analysis of a larger investigation designed to validate 3 delirium assessments. Our sample size was also relatively small, causing our 95% CIs to overlap in most cases and limiting the statistical power to truly determine whether one measure is better than the other. We also asked the patient to recite the months backwards from December to July as well as recite the months backwards from December to January. It is possible that the patient may have performed better at going from December to January because of learning effect. Our reference standard for delirium was based upon DSM-IV-TR criteria. The new DSM-V criteria may be more restrictive and may slightly change the sensitivities and specificities of the inattention tasks. We enrolled a convenience sample and enrolled patients who were more likely to be male, have cardiovascular chief complaints, and be admitted to the hospital; as a result, selection bias may have been introduced. Lastly, this study was conducted in a single center and enrolled patients who were 65 years and older. Our findings may not be generalizable to other settings and in those who are less than 65 years of age.

 

CONCLUSIONS

The MOTYB-6 and MOTYB-12 tasks had very good sensitivities but modest specificities (approximately 50%) using any error made as a cutoff; increasing cutoff to 2 errors and 3 errors, respectively, improved their specificities (approximately 70%) with minimal impact to their sensitivities. Reciting the days of the week backwards, spelling the word “LUNCH” backwards, and the 10-letter vigilance “A” task appeared to perform the best in ruling out delirium but only moderately decreased the likelihood of delirium. The 10-letter Vigilance “A” and picture recognition task appeared to perform the best in ruling in delirium. Days of the week backwards appeared to have the best combination of sensitivity and specificity.

Disclosure

This study was funded by the Emergency Medicine Foundation Career Development Award, National Institutes of Health K23AG032355, and National Center for Research Resources, Grant UL1 RR024975-01. The authors report no financial conflicts of interest.

The prevalence of alcohol dependence in the U.S. represents a significant public health concern. Alcohol use disorder (AUD) is estimated to affect 6.7% of Americans and is the fourth leading preventable cause of death.¹ Men and women who have served in the military are at an even higher risk of excessive alcohol use. More than 20% of service members report binge drinking every week.² This risk is further exacerbated in veterans who have experienced active combat or who have comorbid health conditions, such as posttraumatic stress disorder.³

Background

Individuals that regularly consume excessive amounts of alcohol can develop acute alcohol withdrawal syndrome (AWS) after abrupt discontinuation or significant reduction of alcohol intake. Patients admitted for acute alcohol withdrawal may experience complicated courses of treatment and extended lengths of hospitalization.^4,5 Cessation from chronic alcohol intake elicits a pathophysiologic response from increased N-methyl-d-aspartate receptor activity and decreased γ-aminobutyric acid (GABA) receptor function.

Autonomic and psychomotor hyperactivity disturbances, such as anxiety, nausea, tremors, diaphoresis, and tachycardia, may occur as early as 6 to 8 hours after cessation of use. Within 48 to 72 hours of alcohol cessation, patients may be at an increased risk of experiencing tonic-clonic seizures, visual and auditory hallucinations, and delirium tremens (DTs), which may be accompanied by signs of extreme autonomic hyperactivity and agitation.⁶ Patients hospitalized within acute settings require frequent medical supervision for acute alcohol withdrawal, especially in patients at high risk for seizure or DTs because morbidity and mortality risk is increased.⁷

Benzodiazepines remain the standard of care for management of moderate-to-severe symptoms of AWS. Strong evidence supports the use of benzodiazepines to reduce withdrawal severity, incidence of delirium, and seizures in AWS by enhancing GABA activity.⁸ However, the adverse effect (AE) burden associated with benzodiazepines can be a major limitation throughout care. Benzodiazepines also may be limited in their use in select patient populations, such as in older adults or patients who present with hepatic dysfunction due to the risk of increased AEs or metabolite accumulation.⁶ A high dosing requirement of benzodiazepine for symptom management can lead to oversedation to the point of requiring intubation, increasing length of stay in the intensive care unit (ICU), and the risk of nosocomial infections.⁹

Anticonvulsants, such as carbamazepine, valproic acid, and gabapentin, have shown to be superior to placebo and equal in efficacy to benzodiazepines for symptom management in mild-to-moderate alcohol withdrawal in both inpatient and outpatient settings.^6-8 However, these agents are not recommended as first-line monotherapy due to the limited number of randomized trials supporting their efficacy over benzodiazepines in preventing severe symptoms of withdrawal, such as seizures or delirium.^10-12 Nonetheless, the mechanism of action of anticonvulsants may help raise seizure threshold in patients and provide a benzodiazepine-sparing effect by enhancing GABAergic activity and lowering neuronal excitability.¹³

Gabapentin makes an attractive agent for clinical use because of its anxiolytic and sedative properties that can be used to potentially target symptoms analogous with AWS when the use of benzodiazepines becomes a safety concern. Although similar in chemical structure, gabapentin is not metabolized to GABA and does not directly interact with the receptor. Gabapentin may increase GABA concentrations by direct synthesis of GABA and indirectly through interaction with voltage-gated calcium channels.¹³ In addition to its overall safety profile, gabapentin may be a viable adjuvant because emerging data may suggest a potential role in the management of acute alcohol withdrawal.^12,14,15

Gabapentin for Alcohol Withdrawal at VAPORHCS

Although not currently included in the alcohol withdrawal protocol at Veterans Affairs Portland Health Care System (VAPORHCS), gabapentin has been added to the standard of care in select patients per the discretion of the attending physician. Anecdotal reports of patients experiencing milder symptoms and less benzodiazepine administration have facilitated use of gabapentin in alcohol withdrawal management at VAPORHCS. However, routine use of gabapentin is not consistent among all patients treated for acute alcohol withdrawal, and dosing schedules of gabapentin seem highly variable. Standard symptom management for acute alcohol withdrawal should be consistent for all affected individuals, using evidence-based medicine in order to achieve optimal outcomes and improve harm reduction.

The objective of this quality assurance/quality improvement (QA/QI) project was to assess the amount of lorazepam required for symptom management in acute alcohol withdrawal when gabapentin is used as an adjunct to treatment and to evaluate the impact on symptom management using the Clinical Institute Withdrawal Assessment for Alcohol scale, revised version (CIWA-Ar) in patients admitted to the ICU and general medicine wards for acute alcohol withdrawal at VAPORHCS.¹⁶ If a possible adjunct for the treatment of alcohol withdrawal has the potential to reduce benzodiazepine requirements and minimize AEs, a thorough evaluation of the treatment should be conducted before its practice is incorporated into the current standard of care.

 

Methods

The following QA/QI project was approved locally by the VAPORHCS associate chief of staff/Office of Research and Development and is considered to be nonresearch VHA operations activity and exempt from an institutional review board committee review. This project was a single-center, retrospective chart review of patients admitted to the ICU and general medicine wards at VAPORHCS with acute alcohol withdrawal. The CIWA-Ar protocol order sets between January 1, 2014 and December 31, 2015, were retrieved through the Computerized Patient Record System (CPRS) at VAPORHCS.

Patients with an alcohol withdrawal protocol order set who received gabapentin with or without lorazepam during hospitalization were identified for chart review. Patients were eligible for review if they were aged ≥ 18 years with a primary or secondary diagnosis of acute alcohol withdrawal and had a CIWA-Ar protocol order set placed during hospitalization. Patients must have been administered gabapentin, lorazepam, or both while the CIWA-Ar protocol was active. Patients with an active outpatient prescription for gabapentin or benzodiazepine filled within the previous 30 days, documented history of psychosis or epileptic seizure disorder, or other concomitant benzodiazepines or antiepileptics administered while on the CIWA-Ar protocol were excluded from the analysis.

Baseline characteristics for patients eligible for review were collected and included age; sex; race, body mass index (BMI); estimated creatinine clearance (CrCl); toxicology screen at admission (if available), history of substance use disorder, AWS, or history of withdrawal seizures; and history of a sedative hypnotics (not including benzodiazepines) prescription within 30 days prior to admission.¹⁷

The primary endpoint was the total amount of lorazepam administered from the time of admission to the time of discontinuation of the alcohol withdrawal protocol. The dose, frequency, and amount of lorazepam and gabapentin administered daily were collected for each patient while on the CIWA-Ar protocol. Secondary endpoints included rate of the CIWA-Ar score reduction, time to protocol discontinuation, as well as incidence and onset of peak delirium scores during hospitalization. Cumulative CIWA-Ar scores over 24 hours were averaged per patient per day while on CIWA-Ar protocol. Peak CIWA-Ar scores per patient per day on the protocol also were collected. Time to protocol termination was determined by date of order for discontinuation or by date when scoring had ceased and protocol order was inadvertently continued. Peak Intensive Care Delirium Screening Checklist (ICDSC) scores were collected for patients admitted to the ICU.¹⁸ Day of peak ICDSC scores also were evaluated.

Statistical Analysis

The sample size for this analysis was determined by the number of patients identified who met the inclusion criteria and did not meet any of the exclusion criteria. Power was not calculated to estimate sample size needed to determine statistical significance. One hundred patients treated for alcohol withdrawal was established as the target sample size for this project. Descriptive statistics were performed to analyze patient baseline characteristics and primary and secondary objective data.

Results

A total of 1,611 CIWA-Ar protocol orders were identified between January 1, 2014 and December 31, 2015. 

Among these orders, 1,507 met exclusion criteria, leaving 55 patients on CIWA-Ar protocol who received lorazepam only (group 1) and 46 patients who received gabapentin with (n = 36) or without (n = 10) lorazepam (group 2) for alcohol withdrawal management (Figure 1). Baseline characteristics were similar across groups and were characteristic of the patient population treated at VAPORHCS (Table 1), 

consisting primarily of white males with a mean age of 58 years (± 10) and BMI 28 kg/m² (± 7). Groups 1 and 2 had similar estimated CrCl at 123 mL/min/1.73 m² (± 57) and 104 mL/min/1.73 m² (± 40), respectively. Blood alcohol levels taken at admission were available in 56% and 43% of patients in group 1 and 2 with a mean concentration of 0.16 g/dL (± 1.2) and 0.20 g/dL (± 1.4), respectively. Excluding alcohol, tobacco, and marijuana use, both groups had similar incidence of substance use disorder (26% vs 27%), and 13% of patients in group 1 had a history of alcohol withdrawal seizures vs 0% patients in group 2. Patients who received gabapentin (group 2) from the time of CIWA-Ar protocol initiation until discontinuation received an average daily dose of 948 mg (median 900; min 100/max 3,600, ± 786).

Primary Endpoint

The average amount of lorazepam administered for the total duration on CIWA-Ar protocol was 7.9 mg (median 6, ± 8.2) among

patients in group 1 and 5.5 mg (median 2, ± 8.7) among patients in group 2 (Table 2). 

The average amount of lorazepam administered per day on the CIWA-Ar protocol between group 1 and group 2 is shown in Figure 2.

 

Secondary Endpoints

On average, the total number of days spent on CIWA-Ar protocol was 3.8 (median 4, ± 1.5) in group 1 compared with 4.1 (median 4, ± 1.6) in group 2. Rate of CIWA-Ar protocol discontinuation for patients in group 1 and group 2 is shown in Figure 3. 

Delirium scores were measured in a subset of patients in group 1 (n = 13) and group 2 (n = 21) admitted to the ICU. On average, the peak delirium score among patients was 3.54 (± 2.8) and 3.48 (± 2.9) and occurred within an average of 1.71 (± 0.8) and 1.78 (± 0.9) days in group 1 and 2, respectively. Presence of delirium (ICDSC score > 4) occurred in 53% (n = 13) of patients in group 1 compared with 47% (n = 10) of patients in group 2.

Discussion

The purpose of this project was to evaluate gabapentin use at VAPORHCS for alcohol withdrawal and evaluate the impact on symptom management. Patients who were started on gabapentin on the initiation of the alcohol withdrawal protocol received less lorazepam dosing compared with patients who received only lorazepam for symptom management for alcohol withdrawal. Except for day 3, average lorazepam dosage per day on the alcohol withdrawal protocol was lower in patients who were also taking gabapentin.

This trend also can be seen in the recorded peak CIWA-Ar scores per day as illustrated in Figures 4 and 5. 

The wide standard deviation seen in the primary endpoint is likely to be attributed to the nonparametric distributionof data points that were found in both groups. The extensive range of total lorazepam administered between patients may result from not distinguishing patients who may have had very mild withdrawal and those with more severe symptoms of alcohol withdrawal throughout hospitalization. 

The average number of days spent on CIWA-Ar protocol did not differ between groups. From a cost perspective, it could be argued that the need for additional medications for adjunct treatment of acute alcohol withdrawal may not be fiscally reasonable if the number of days a patient is hospitalized is ultimately the same without its use. However, costs also could be potentially offset by lower benzodiazepine dosages and fewer hospital staff resources required to monitor and manage patient’s symptoms.

Limitations

Prior to evaluation, power analysis was not calculated to estimate an appropriate sample size necessary to determine statistical significance. Results from this evaluation are not definitive and are meant to be hypothesis generating for future analysis.

There were several limitations that were identified throughout this project. For this review, history and extent of patient’s prior alcohol use was not assessed. Therefore, the degree of symptom severity in which patients may have experienced during withdrawal may not have been adequately matched between groups. The inherent subjectivity of CIWA-Ar scoring was considered a limitation because scores were determined by clinical interpretation among various nursing staff. As this was a retrospective review, exact timing of medications administered as well as additional supportive care measures, such as ancillary medications for symptom management, were not accounted for and controlled between groups.

Patients presenting to the emergency department or from a facility outside of VAPORHCS for acute AWS may have had incomplete documentation of the onset of symptoms on presentation or of the medications administered prior to being admitted, which may have confounded initial CIWA-Ar scoring and total duration required to be on a withdrawal protocol. Some patients may have received benzodiazepines at initial presentation prior to gabapentin initiation and may have inaccurately reflected its efficacy potential to manage symptoms without the need for lorazepam.

There were 10 patients that were identified who received gabapentin on the alcohol withdrawal protocol and did not receive any lorazepam. This retrospective review could not be determined whether these patients did not require lorazepam because initiating gabapentin reduced severity or simply because their withdrawal symptoms were not severe enough to warrant the need for lorazepam, regardless of gabapentin use.

Gabapentin dosing was not standardized among patients, averaging from 100 mg to 3,600 mg per day. This wide variation in dose may have influenced the requirement of lorazepam needed for symptom management in patients receiving minimal doses or AEs experienced in patients who received large doses. Initiation and/or select dosing of gabapentin may have been dependent on the experience of the provider and familiarity with its use in alcohol withdrawal management. Interestingly, patients with a history of withdrawal seizures (13%) were identified only within the lorazepam-only group. This could suggest that patients with prior symptoms of severe alcohol withdrawal were selected to receive lorazepam-only at the discretion of the provider.

Existing literature investigating gabapentin utilization in alcohol withdrawal has demonstrated benefit for patients with mild-to-moderate symptoms in both inpatient and outpatient studies. However, supporting evidence is limited by the differences in design, methods, and comparators within each trial. Leung and colleagues identified 5 studies that utilized gabapentin as monotherapy or in combination with other agents in alcohol withdrawal.¹³ Three of these studies were performed within an inpatient setting, each differing in trial design, inclusion/exclusion criteria, intervention, and outcomes. Gabapentin dosing strategies were highly variable among studies. Collectively, the differences noted make it difficult to generalize that similar outcomes would result in other patient populations. The purpose of this project was to evaluate gabapentin use at VAPORHCS for alcohol withdrawal and evaluate the impact on symptom management. Future projects could be designed to draw more specific conclusions.

 

Conclusion

On average, the required benzodiazepine dosage was lower with concomitant use of gabapentin in acute AWS management. The duration for patients on alcohol withdrawal protocol was not reduced with use of gabapentin. Between group (ie, history of withdrawal seizures, blood alcohol level) and among group (ie, gabapentin administration) differences prevent direct correlations to be drawn from this evaluation. Future reviews should include power analysis to establish an appropriate sample size to determine statistical significance among identified covariates. Further evaluation of the use of gabapentin for withdrawal management is warranted prior to incorporating its routine use in the current standard of care for patients experiencing acute AWS.

Acknowledgments
The authors thank Ryan Bickel, PharmD, BCCCP, Critical Care Clinical Pharmacist; Stephen M. Smith, PhD, Director of Medical Critical Care; Gordon Wong, PharmD, Clinical Applications Coordinator; and Eileen Wilbur, RPh, Research Pharmacy Supervisor.

The prevalence of alcohol dependence in the U.S. represents a significant public health concern. Alcohol use disorder (AUD) is estimated to affect 6.7% of Americans and is the fourth leading preventable cause of death.¹ Men and women who have served in the military are at an even higher risk of excessive alcohol use. More than 20% of service members report binge drinking every week.² This risk is further exacerbated in veterans who have experienced active combat or who have comorbid health conditions, such as posttraumatic stress disorder.³

Background

Individuals that regularly consume excessive amounts of alcohol can develop acute alcohol withdrawal syndrome (AWS) after abrupt discontinuation or significant reduction of alcohol intake. Patients admitted for acute alcohol withdrawal may experience complicated courses of treatment and extended lengths of hospitalization.^4,5 Cessation from chronic alcohol intake elicits a pathophysiologic response from increased N-methyl-d-aspartate receptor activity and decreased γ-aminobutyric acid (GABA) receptor function.

Autonomic and psychomotor hyperactivity disturbances, such as anxiety, nausea, tremors, diaphoresis, and tachycardia, may occur as early as 6 to 8 hours after cessation of use. Within 48 to 72 hours of alcohol cessation, patients may be at an increased risk of experiencing tonic-clonic seizures, visual and auditory hallucinations, and delirium tremens (DTs), which may be accompanied by signs of extreme autonomic hyperactivity and agitation.⁶ Patients hospitalized within acute settings require frequent medical supervision for acute alcohol withdrawal, especially in patients at high risk for seizure or DTs because morbidity and mortality risk is increased.⁷

Benzodiazepines remain the standard of care for management of moderate-to-severe symptoms of AWS. Strong evidence supports the use of benzodiazepines to reduce withdrawal severity, incidence of delirium, and seizures in AWS by enhancing GABA activity.⁸ However, the adverse effect (AE) burden associated with benzodiazepines can be a major limitation throughout care. Benzodiazepines also may be limited in their use in select patient populations, such as in older adults or patients who present with hepatic dysfunction due to the risk of increased AEs or metabolite accumulation.⁶ A high dosing requirement of benzodiazepine for symptom management can lead to oversedation to the point of requiring intubation, increasing length of stay in the intensive care unit (ICU), and the risk of nosocomial infections.⁹

Anticonvulsants, such as carbamazepine, valproic acid, and gabapentin, have shown to be superior to placebo and equal in efficacy to benzodiazepines for symptom management in mild-to-moderate alcohol withdrawal in both inpatient and outpatient settings.^6-8 However, these agents are not recommended as first-line monotherapy due to the limited number of randomized trials supporting their efficacy over benzodiazepines in preventing severe symptoms of withdrawal, such as seizures or delirium.^10-12 Nonetheless, the mechanism of action of anticonvulsants may help raise seizure threshold in patients and provide a benzodiazepine-sparing effect by enhancing GABAergic activity and lowering neuronal excitability.¹³

Gabapentin makes an attractive agent for clinical use because of its anxiolytic and sedative properties that can be used to potentially target symptoms analogous with AWS when the use of benzodiazepines becomes a safety concern. Although similar in chemical structure, gabapentin is not metabolized to GABA and does not directly interact with the receptor. Gabapentin may increase GABA concentrations by direct synthesis of GABA and indirectly through interaction with voltage-gated calcium channels.¹³ In addition to its overall safety profile, gabapentin may be a viable adjuvant because emerging data may suggest a potential role in the management of acute alcohol withdrawal.^12,14,15

Gabapentin for Alcohol Withdrawal at VAPORHCS

Although not currently included in the alcohol withdrawal protocol at Veterans Affairs Portland Health Care System (VAPORHCS), gabapentin has been added to the standard of care in select patients per the discretion of the attending physician. Anecdotal reports of patients experiencing milder symptoms and less benzodiazepine administration have facilitated use of gabapentin in alcohol withdrawal management at VAPORHCS. However, routine use of gabapentin is not consistent among all patients treated for acute alcohol withdrawal, and dosing schedules of gabapentin seem highly variable. Standard symptom management for acute alcohol withdrawal should be consistent for all affected individuals, using evidence-based medicine in order to achieve optimal outcomes and improve harm reduction.

The objective of this quality assurance/quality improvement (QA/QI) project was to assess the amount of lorazepam required for symptom management in acute alcohol withdrawal when gabapentin is used as an adjunct to treatment and to evaluate the impact on symptom management using the Clinical Institute Withdrawal Assessment for Alcohol scale, revised version (CIWA-Ar) in patients admitted to the ICU and general medicine wards for acute alcohol withdrawal at VAPORHCS.¹⁶ If a possible adjunct for the treatment of alcohol withdrawal has the potential to reduce benzodiazepine requirements and minimize AEs, a thorough evaluation of the treatment should be conducted before its practice is incorporated into the current standard of care.

 

Methods

The following QA/QI project was approved locally by the VAPORHCS associate chief of staff/Office of Research and Development and is considered to be nonresearch VHA operations activity and exempt from an institutional review board committee review. This project was a single-center, retrospective chart review of patients admitted to the ICU and general medicine wards at VAPORHCS with acute alcohol withdrawal. The CIWA-Ar protocol order sets between January 1, 2014 and December 31, 2015, were retrieved through the Computerized Patient Record System (CPRS) at VAPORHCS.

Patients with an alcohol withdrawal protocol order set who received gabapentin with or without lorazepam during hospitalization were identified for chart review. Patients were eligible for review if they were aged ≥ 18 years with a primary or secondary diagnosis of acute alcohol withdrawal and had a CIWA-Ar protocol order set placed during hospitalization. Patients must have been administered gabapentin, lorazepam, or both while the CIWA-Ar protocol was active. Patients with an active outpatient prescription for gabapentin or benzodiazepine filled within the previous 30 days, documented history of psychosis or epileptic seizure disorder, or other concomitant benzodiazepines or antiepileptics administered while on the CIWA-Ar protocol were excluded from the analysis.

Baseline characteristics for patients eligible for review were collected and included age; sex; race, body mass index (BMI); estimated creatinine clearance (CrCl); toxicology screen at admission (if available), history of substance use disorder, AWS, or history of withdrawal seizures; and history of a sedative hypnotics (not including benzodiazepines) prescription within 30 days prior to admission.¹⁷

The primary endpoint was the total amount of lorazepam administered from the time of admission to the time of discontinuation of the alcohol withdrawal protocol. The dose, frequency, and amount of lorazepam and gabapentin administered daily were collected for each patient while on the CIWA-Ar protocol. Secondary endpoints included rate of the CIWA-Ar score reduction, time to protocol discontinuation, as well as incidence and onset of peak delirium scores during hospitalization. Cumulative CIWA-Ar scores over 24 hours were averaged per patient per day while on CIWA-Ar protocol. Peak CIWA-Ar scores per patient per day on the protocol also were collected. Time to protocol termination was determined by date of order for discontinuation or by date when scoring had ceased and protocol order was inadvertently continued. Peak Intensive Care Delirium Screening Checklist (ICDSC) scores were collected for patients admitted to the ICU.¹⁸ Day of peak ICDSC scores also were evaluated.

Statistical Analysis

The sample size for this analysis was determined by the number of patients identified who met the inclusion criteria and did not meet any of the exclusion criteria. Power was not calculated to estimate sample size needed to determine statistical significance. One hundred patients treated for alcohol withdrawal was established as the target sample size for this project. Descriptive statistics were performed to analyze patient baseline characteristics and primary and secondary objective data.

Results

A total of 1,611 CIWA-Ar protocol orders were identified between January 1, 2014 and December 31, 2015. 

Among these orders, 1,507 met exclusion criteria, leaving 55 patients on CIWA-Ar protocol who received lorazepam only (group 1) and 46 patients who received gabapentin with (n = 36) or without (n = 10) lorazepam (group 2) for alcohol withdrawal management (Figure 1). Baseline characteristics were similar across groups and were characteristic of the patient population treated at VAPORHCS (Table 1), 

consisting primarily of white males with a mean age of 58 years (± 10) and BMI 28 kg/m² (± 7). Groups 1 and 2 had similar estimated CrCl at 123 mL/min/1.73 m² (± 57) and 104 mL/min/1.73 m² (± 40), respectively. Blood alcohol levels taken at admission were available in 56% and 43% of patients in group 1 and 2 with a mean concentration of 0.16 g/dL (± 1.2) and 0.20 g/dL (± 1.4), respectively. Excluding alcohol, tobacco, and marijuana use, both groups had similar incidence of substance use disorder (26% vs 27%), and 13% of patients in group 1 had a history of alcohol withdrawal seizures vs 0% patients in group 2. Patients who received gabapentin (group 2) from the time of CIWA-Ar protocol initiation until discontinuation received an average daily dose of 948 mg (median 900; min 100/max 3,600, ± 786).

Primary Endpoint

The average amount of lorazepam administered for the total duration on CIWA-Ar protocol was 7.9 mg (median 6, ± 8.2) among

patients in group 1 and 5.5 mg (median 2, ± 8.7) among patients in group 2 (Table 2). 

The average amount of lorazepam administered per day on the CIWA-Ar protocol between group 1 and group 2 is shown in Figure 2.

 

Secondary Endpoints

On average, the total number of days spent on CIWA-Ar protocol was 3.8 (median 4, ± 1.5) in group 1 compared with 4.1 (median 4, ± 1.6) in group 2. Rate of CIWA-Ar protocol discontinuation for patients in group 1 and group 2 is shown in Figure 3. 

Delirium scores were measured in a subset of patients in group 1 (n = 13) and group 2 (n = 21) admitted to the ICU. On average, the peak delirium score among patients was 3.54 (± 2.8) and 3.48 (± 2.9) and occurred within an average of 1.71 (± 0.8) and 1.78 (± 0.9) days in group 1 and 2, respectively. Presence of delirium (ICDSC score > 4) occurred in 53% (n = 13) of patients in group 1 compared with 47% (n = 10) of patients in group 2.

Discussion

The purpose of this project was to evaluate gabapentin use at VAPORHCS for alcohol withdrawal and evaluate the impact on symptom management. Patients who were started on gabapentin on the initiation of the alcohol withdrawal protocol received less lorazepam dosing compared with patients who received only lorazepam for symptom management for alcohol withdrawal. Except for day 3, average lorazepam dosage per day on the alcohol withdrawal protocol was lower in patients who were also taking gabapentin.

This trend also can be seen in the recorded peak CIWA-Ar scores per day as illustrated in Figures 4 and 5. 

The wide standard deviation seen in the primary endpoint is likely to be attributed to the nonparametric distributionof data points that were found in both groups. The extensive range of total lorazepam administered between patients may result from not distinguishing patients who may have had very mild withdrawal and those with more severe symptoms of alcohol withdrawal throughout hospitalization. 

The average number of days spent on CIWA-Ar protocol did not differ between groups. From a cost perspective, it could be argued that the need for additional medications for adjunct treatment of acute alcohol withdrawal may not be fiscally reasonable if the number of days a patient is hospitalized is ultimately the same without its use. However, costs also could be potentially offset by lower benzodiazepine dosages and fewer hospital staff resources required to monitor and manage patient’s symptoms.

Limitations

Prior to evaluation, power analysis was not calculated to estimate an appropriate sample size necessary to determine statistical significance. Results from this evaluation are not definitive and are meant to be hypothesis generating for future analysis.

There were several limitations that were identified throughout this project. For this review, history and extent of patient’s prior alcohol use was not assessed. Therefore, the degree of symptom severity in which patients may have experienced during withdrawal may not have been adequately matched between groups. The inherent subjectivity of CIWA-Ar scoring was considered a limitation because scores were determined by clinical interpretation among various nursing staff. As this was a retrospective review, exact timing of medications administered as well as additional supportive care measures, such as ancillary medications for symptom management, were not accounted for and controlled between groups.

Patients presenting to the emergency department or from a facility outside of VAPORHCS for acute AWS may have had incomplete documentation of the onset of symptoms on presentation or of the medications administered prior to being admitted, which may have confounded initial CIWA-Ar scoring and total duration required to be on a withdrawal protocol. Some patients may have received benzodiazepines at initial presentation prior to gabapentin initiation and may have inaccurately reflected its efficacy potential to manage symptoms without the need for lorazepam.

There were 10 patients that were identified who received gabapentin on the alcohol withdrawal protocol and did not receive any lorazepam. This retrospective review could not be determined whether these patients did not require lorazepam because initiating gabapentin reduced severity or simply because their withdrawal symptoms were not severe enough to warrant the need for lorazepam, regardless of gabapentin use.

Gabapentin dosing was not standardized among patients, averaging from 100 mg to 3,600 mg per day. This wide variation in dose may have influenced the requirement of lorazepam needed for symptom management in patients receiving minimal doses or AEs experienced in patients who received large doses. Initiation and/or select dosing of gabapentin may have been dependent on the experience of the provider and familiarity with its use in alcohol withdrawal management. Interestingly, patients with a history of withdrawal seizures (13%) were identified only within the lorazepam-only group. This could suggest that patients with prior symptoms of severe alcohol withdrawal were selected to receive lorazepam-only at the discretion of the provider.

Existing literature investigating gabapentin utilization in alcohol withdrawal has demonstrated benefit for patients with mild-to-moderate symptoms in both inpatient and outpatient studies. However, supporting evidence is limited by the differences in design, methods, and comparators within each trial. Leung and colleagues identified 5 studies that utilized gabapentin as monotherapy or in combination with other agents in alcohol withdrawal.¹³ Three of these studies were performed within an inpatient setting, each differing in trial design, inclusion/exclusion criteria, intervention, and outcomes. Gabapentin dosing strategies were highly variable among studies. Collectively, the differences noted make it difficult to generalize that similar outcomes would result in other patient populations. The purpose of this project was to evaluate gabapentin use at VAPORHCS for alcohol withdrawal and evaluate the impact on symptom management. Future projects could be designed to draw more specific conclusions.

 

Conclusion

On average, the required benzodiazepine dosage was lower with concomitant use of gabapentin in acute AWS management. The duration for patients on alcohol withdrawal protocol was not reduced with use of gabapentin. Between group (ie, history of withdrawal seizures, blood alcohol level) and among group (ie, gabapentin administration) differences prevent direct correlations to be drawn from this evaluation. Future reviews should include power analysis to establish an appropriate sample size to determine statistical significance among identified covariates. Further evaluation of the use of gabapentin for withdrawal management is warranted prior to incorporating its routine use in the current standard of care for patients experiencing acute AWS.

Acknowledgments
The authors thank Ryan Bickel, PharmD, BCCCP, Critical Care Clinical Pharmacist; Stephen M. Smith, PhD, Director of Medical Critical Care; Gordon Wong, PharmD, Clinical Applications Coordinator; and Eileen Wilbur, RPh, Research Pharmacy Supervisor.

The prevalence of alcohol dependence in the U.S. represents a significant public health concern. Alcohol use disorder (AUD) is estimated to affect 6.7% of Americans and is the fourth leading preventable cause of death.¹ Men and women who have served in the military are at an even higher risk of excessive alcohol use. More than 20% of service members report binge drinking every week.² This risk is further exacerbated in veterans who have experienced active combat or who have comorbid health conditions, such as posttraumatic stress disorder.³

Background

Individuals that regularly consume excessive amounts of alcohol can develop acute alcohol withdrawal syndrome (AWS) after abrupt discontinuation or significant reduction of alcohol intake. Patients admitted for acute alcohol withdrawal may experience complicated courses of treatment and extended lengths of hospitalization.^4,5 Cessation from chronic alcohol intake elicits a pathophysiologic response from increased N-methyl-d-aspartate receptor activity and decreased γ-aminobutyric acid (GABA) receptor function.

Autonomic and psychomotor hyperactivity disturbances, such as anxiety, nausea, tremors, diaphoresis, and tachycardia, may occur as early as 6 to 8 hours after cessation of use. Within 48 to 72 hours of alcohol cessation, patients may be at an increased risk of experiencing tonic-clonic seizures, visual and auditory hallucinations, and delirium tremens (DTs), which may be accompanied by signs of extreme autonomic hyperactivity and agitation.⁶ Patients hospitalized within acute settings require frequent medical supervision for acute alcohol withdrawal, especially in patients at high risk for seizure or DTs because morbidity and mortality risk is increased.⁷

Benzodiazepines remain the standard of care for management of moderate-to-severe symptoms of AWS. Strong evidence supports the use of benzodiazepines to reduce withdrawal severity, incidence of delirium, and seizures in AWS by enhancing GABA activity.⁸ However, the adverse effect (AE) burden associated with benzodiazepines can be a major limitation throughout care. Benzodiazepines also may be limited in their use in select patient populations, such as in older adults or patients who present with hepatic dysfunction due to the risk of increased AEs or metabolite accumulation.⁶ A high dosing requirement of benzodiazepine for symptom management can lead to oversedation to the point of requiring intubation, increasing length of stay in the intensive care unit (ICU), and the risk of nosocomial infections.⁹

Anticonvulsants, such as carbamazepine, valproic acid, and gabapentin, have shown to be superior to placebo and equal in efficacy to benzodiazepines for symptom management in mild-to-moderate alcohol withdrawal in both inpatient and outpatient settings.^6-8 However, these agents are not recommended as first-line monotherapy due to the limited number of randomized trials supporting their efficacy over benzodiazepines in preventing severe symptoms of withdrawal, such as seizures or delirium.^10-12 Nonetheless, the mechanism of action of anticonvulsants may help raise seizure threshold in patients and provide a benzodiazepine-sparing effect by enhancing GABAergic activity and lowering neuronal excitability.¹³

Gabapentin makes an attractive agent for clinical use because of its anxiolytic and sedative properties that can be used to potentially target symptoms analogous with AWS when the use of benzodiazepines becomes a safety concern. Although similar in chemical structure, gabapentin is not metabolized to GABA and does not directly interact with the receptor. Gabapentin may increase GABA concentrations by direct synthesis of GABA and indirectly through interaction with voltage-gated calcium channels.¹³ In addition to its overall safety profile, gabapentin may be a viable adjuvant because emerging data may suggest a potential role in the management of acute alcohol withdrawal.^12,14,15

Gabapentin for Alcohol Withdrawal at VAPORHCS

Although not currently included in the alcohol withdrawal protocol at Veterans Affairs Portland Health Care System (VAPORHCS), gabapentin has been added to the standard of care in select patients per the discretion of the attending physician. Anecdotal reports of patients experiencing milder symptoms and less benzodiazepine administration have facilitated use of gabapentin in alcohol withdrawal management at VAPORHCS. However, routine use of gabapentin is not consistent among all patients treated for acute alcohol withdrawal, and dosing schedules of gabapentin seem highly variable. Standard symptom management for acute alcohol withdrawal should be consistent for all affected individuals, using evidence-based medicine in order to achieve optimal outcomes and improve harm reduction.

The objective of this quality assurance/quality improvement (QA/QI) project was to assess the amount of lorazepam required for symptom management in acute alcohol withdrawal when gabapentin is used as an adjunct to treatment and to evaluate the impact on symptom management using the Clinical Institute Withdrawal Assessment for Alcohol scale, revised version (CIWA-Ar) in patients admitted to the ICU and general medicine wards for acute alcohol withdrawal at VAPORHCS.¹⁶ If a possible adjunct for the treatment of alcohol withdrawal has the potential to reduce benzodiazepine requirements and minimize AEs, a thorough evaluation of the treatment should be conducted before its practice is incorporated into the current standard of care.

 

Methods

The following QA/QI project was approved locally by the VAPORHCS associate chief of staff/Office of Research and Development and is considered to be nonresearch VHA operations activity and exempt from an institutional review board committee review. This project was a single-center, retrospective chart review of patients admitted to the ICU and general medicine wards at VAPORHCS with acute alcohol withdrawal. The CIWA-Ar protocol order sets between January 1, 2014 and December 31, 2015, were retrieved through the Computerized Patient Record System (CPRS) at VAPORHCS.

Patients with an alcohol withdrawal protocol order set who received gabapentin with or without lorazepam during hospitalization were identified for chart review. Patients were eligible for review if they were aged ≥ 18 years with a primary or secondary diagnosis of acute alcohol withdrawal and had a CIWA-Ar protocol order set placed during hospitalization. Patients must have been administered gabapentin, lorazepam, or both while the CIWA-Ar protocol was active. Patients with an active outpatient prescription for gabapentin or benzodiazepine filled within the previous 30 days, documented history of psychosis or epileptic seizure disorder, or other concomitant benzodiazepines or antiepileptics administered while on the CIWA-Ar protocol were excluded from the analysis.

Baseline characteristics for patients eligible for review were collected and included age; sex; race, body mass index (BMI); estimated creatinine clearance (CrCl); toxicology screen at admission (if available), history of substance use disorder, AWS, or history of withdrawal seizures; and history of a sedative hypnotics (not including benzodiazepines) prescription within 30 days prior to admission.¹⁷

The primary endpoint was the total amount of lorazepam administered from the time of admission to the time of discontinuation of the alcohol withdrawal protocol. The dose, frequency, and amount of lorazepam and gabapentin administered daily were collected for each patient while on the CIWA-Ar protocol. Secondary endpoints included rate of the CIWA-Ar score reduction, time to protocol discontinuation, as well as incidence and onset of peak delirium scores during hospitalization. Cumulative CIWA-Ar scores over 24 hours were averaged per patient per day while on CIWA-Ar protocol. Peak CIWA-Ar scores per patient per day on the protocol also were collected. Time to protocol termination was determined by date of order for discontinuation or by date when scoring had ceased and protocol order was inadvertently continued. Peak Intensive Care Delirium Screening Checklist (ICDSC) scores were collected for patients admitted to the ICU.¹⁸ Day of peak ICDSC scores also were evaluated.

Statistical Analysis

The sample size for this analysis was determined by the number of patients identified who met the inclusion criteria and did not meet any of the exclusion criteria. Power was not calculated to estimate sample size needed to determine statistical significance. One hundred patients treated for alcohol withdrawal was established as the target sample size for this project. Descriptive statistics were performed to analyze patient baseline characteristics and primary and secondary objective data.

Results

A total of 1,611 CIWA-Ar protocol orders were identified between January 1, 2014 and December 31, 2015. 

Among these orders, 1,507 met exclusion criteria, leaving 55 patients on CIWA-Ar protocol who received lorazepam only (group 1) and 46 patients who received gabapentin with (n = 36) or without (n = 10) lorazepam (group 2) for alcohol withdrawal management (Figure 1). Baseline characteristics were similar across groups and were characteristic of the patient population treated at VAPORHCS (Table 1), 

consisting primarily of white males with a mean age of 58 years (± 10) and BMI 28 kg/m² (± 7). Groups 1 and 2 had similar estimated CrCl at 123 mL/min/1.73 m² (± 57) and 104 mL/min/1.73 m² (± 40), respectively. Blood alcohol levels taken at admission were available in 56% and 43% of patients in group 1 and 2 with a mean concentration of 0.16 g/dL (± 1.2) and 0.20 g/dL (± 1.4), respectively. Excluding alcohol, tobacco, and marijuana use, both groups had similar incidence of substance use disorder (26% vs 27%), and 13% of patients in group 1 had a history of alcohol withdrawal seizures vs 0% patients in group 2. Patients who received gabapentin (group 2) from the time of CIWA-Ar protocol initiation until discontinuation received an average daily dose of 948 mg (median 900; min 100/max 3,600, ± 786).

Primary Endpoint

The average amount of lorazepam administered for the total duration on CIWA-Ar protocol was 7.9 mg (median 6, ± 8.2) among

patients in group 1 and 5.5 mg (median 2, ± 8.7) among patients in group 2 (Table 2). 

The average amount of lorazepam administered per day on the CIWA-Ar protocol between group 1 and group 2 is shown in Figure 2.

 

Secondary Endpoints

On average, the total number of days spent on CIWA-Ar protocol was 3.8 (median 4, ± 1.5) in group 1 compared with 4.1 (median 4, ± 1.6) in group 2. Rate of CIWA-Ar protocol discontinuation for patients in group 1 and group 2 is shown in Figure 3. 

Delirium scores were measured in a subset of patients in group 1 (n = 13) and group 2 (n = 21) admitted to the ICU. On average, the peak delirium score among patients was 3.54 (± 2.8) and 3.48 (± 2.9) and occurred within an average of 1.71 (± 0.8) and 1.78 (± 0.9) days in group 1 and 2, respectively. Presence of delirium (ICDSC score > 4) occurred in 53% (n = 13) of patients in group 1 compared with 47% (n = 10) of patients in group 2.

Discussion

The purpose of this project was to evaluate gabapentin use at VAPORHCS for alcohol withdrawal and evaluate the impact on symptom management. Patients who were started on gabapentin on the initiation of the alcohol withdrawal protocol received less lorazepam dosing compared with patients who received only lorazepam for symptom management for alcohol withdrawal. Except for day 3, average lorazepam dosage per day on the alcohol withdrawal protocol was lower in patients who were also taking gabapentin.

This trend also can be seen in the recorded peak CIWA-Ar scores per day as illustrated in Figures 4 and 5. 

The wide standard deviation seen in the primary endpoint is likely to be attributed to the nonparametric distributionof data points that were found in both groups. The extensive range of total lorazepam administered between patients may result from not distinguishing patients who may have had very mild withdrawal and those with more severe symptoms of alcohol withdrawal throughout hospitalization. 

The average number of days spent on CIWA-Ar protocol did not differ between groups. From a cost perspective, it could be argued that the need for additional medications for adjunct treatment of acute alcohol withdrawal may not be fiscally reasonable if the number of days a patient is hospitalized is ultimately the same without its use. However, costs also could be potentially offset by lower benzodiazepine dosages and fewer hospital staff resources required to monitor and manage patient’s symptoms.

Limitations

Prior to evaluation, power analysis was not calculated to estimate an appropriate sample size necessary to determine statistical significance. Results from this evaluation are not definitive and are meant to be hypothesis generating for future analysis.

There were several limitations that were identified throughout this project. For this review, history and extent of patient’s prior alcohol use was not assessed. Therefore, the degree of symptom severity in which patients may have experienced during withdrawal may not have been adequately matched between groups. The inherent subjectivity of CIWA-Ar scoring was considered a limitation because scores were determined by clinical interpretation among various nursing staff. As this was a retrospective review, exact timing of medications administered as well as additional supportive care measures, such as ancillary medications for symptom management, were not accounted for and controlled between groups.

Patients presenting to the emergency department or from a facility outside of VAPORHCS for acute AWS may have had incomplete documentation of the onset of symptoms on presentation or of the medications administered prior to being admitted, which may have confounded initial CIWA-Ar scoring and total duration required to be on a withdrawal protocol. Some patients may have received benzodiazepines at initial presentation prior to gabapentin initiation and may have inaccurately reflected its efficacy potential to manage symptoms without the need for lorazepam.

There were 10 patients that were identified who received gabapentin on the alcohol withdrawal protocol and did not receive any lorazepam. This retrospective review could not be determined whether these patients did not require lorazepam because initiating gabapentin reduced severity or simply because their withdrawal symptoms were not severe enough to warrant the need for lorazepam, regardless of gabapentin use.

Gabapentin dosing was not standardized among patients, averaging from 100 mg to 3,600 mg per day. This wide variation in dose may have influenced the requirement of lorazepam needed for symptom management in patients receiving minimal doses or AEs experienced in patients who received large doses. Initiation and/or select dosing of gabapentin may have been dependent on the experience of the provider and familiarity with its use in alcohol withdrawal management. Interestingly, patients with a history of withdrawal seizures (13%) were identified only within the lorazepam-only group. This could suggest that patients with prior symptoms of severe alcohol withdrawal were selected to receive lorazepam-only at the discretion of the provider.

Existing literature investigating gabapentin utilization in alcohol withdrawal has demonstrated benefit for patients with mild-to-moderate symptoms in both inpatient and outpatient studies. However, supporting evidence is limited by the differences in design, methods, and comparators within each trial. Leung and colleagues identified 5 studies that utilized gabapentin as monotherapy or in combination with other agents in alcohol withdrawal.¹³ Three of these studies were performed within an inpatient setting, each differing in trial design, inclusion/exclusion criteria, intervention, and outcomes. Gabapentin dosing strategies were highly variable among studies. Collectively, the differences noted make it difficult to generalize that similar outcomes would result in other patient populations. The purpose of this project was to evaluate gabapentin use at VAPORHCS for alcohol withdrawal and evaluate the impact on symptom management. Future projects could be designed to draw more specific conclusions.

 

Conclusion

On average, the required benzodiazepine dosage was lower with concomitant use of gabapentin in acute AWS management. The duration for patients on alcohol withdrawal protocol was not reduced with use of gabapentin. Between group (ie, history of withdrawal seizures, blood alcohol level) and among group (ie, gabapentin administration) differences prevent direct correlations to be drawn from this evaluation. Future reviews should include power analysis to establish an appropriate sample size to determine statistical significance among identified covariates. Further evaluation of the use of gabapentin for withdrawal management is warranted prior to incorporating its routine use in the current standard of care for patients experiencing acute AWS.

Acknowledgments
The authors thank Ryan Bickel, PharmD, BCCCP, Critical Care Clinical Pharmacist; Stephen M. Smith, PhD, Director of Medical Critical Care; Gordon Wong, PharmD, Clinical Applications Coordinator; and Eileen Wilbur, RPh, Research Pharmacy Supervisor.

ABSTRACT

The h-index is a commonly utilized metric for academic productivity. Previous studies have proposed that self-citation may limit the utility of the h-index. The purpose of this study is to evaluate the impact of self-citation on the h-index among orthopedic investigators. The study cohort consisted of program directors, chairpersons, and faculty at orthopedic surgery residency programs in the United States. The Scopus database was used to determine the h-index and number of citations ± self-citations. The total number of publications was correlated with the change in the h-index via self-citation. A total of 463 researchers were included (198 National Institutes of Health-funded faculty, 147 chairpersons, 118 program directors). Of these researchers, 83.8% cited previous work at least once (mean, 123.9 ± 277.6). Self-citations accounted for 5.8% of all citations. Including these citations increased the author h-index from 18.5 ± 14.9 to 19.2 ± 15.6 (P < .001). A minority of researchers (36.3%, P < .001) had increased their h-index via self-citation (range, 0-11). The proportional increase in the h-index via self-citation was positively related to the number of publications (P < .001). While the practice of self-citation is prevalent in orthopedics, its impact on the h-index is minimal for most researchers. With more publications, researchers can increase their h-index to a greater degree via self-citation.

Continue to: The competitive nature of academic research...

 

The competitive nature of academic research requires objective metrics to define career end points, such as promotion and funding procurement. While various criteria are used to assess performance in academia, publications and research funding are particularly regarded.¹ Quantifying research dollars is relatively straightforward, but measuring research productivity is more complex. Not all articles are created equal, and disparities exist regarding effort and the ultimate impact of articles. In 2005, a physicist created the h-index to measure both research impact and productivity.² As a bibliometric, the h-index equals the number of publications h that have been cited at least h times. Given its simplicity, the h-index has gained wide popularity in diverse medical specialties, including orthopedic surgery.³ Other recent studies have applied the h-index to hand surgery and spine surgery.^4,5

Importantly, some authors have raised concerns regarding potential limitations of the h-index. One potentially significant limitation is the ability of authors to artificially inflate their h-index via self-citation. The impact of this practice is of particular interest as the h-index becomes widely adopted as a metric for promotion at many academic institutions.^6-7 Furthermore, scholarly productivity has remained a critical component of successful grant funding procurement, and future grant funding applications may evaluate the h-index.^8-10

The purpose of this study is to determine the prevalence and impact of self-citation on the h-index in a large cohort of orthopedic investigators. Given their high level of investment in academic orthopedic surgery, we focused on program directors, chairpersons, and National Institutes of Health (NIH)-funded research faculty at orthopedic surgery residency programs.

METHODS

INCLUSION CRITERIA

This study qualified as non-human and non-animal research and received exemption per the standing policy of the Institutional Review Board. The Fellowship and Residency Electronic Interactive Database (FREIDA) was accessed to generate a list of orthopedic residency program directors.¹¹ This database was also used to generate a list of allopathic orthopedic surgery residency programs. Official program websites were accessed to generate a list of orthopedic chairpersons. Lastly, the NIH RePORTER was used to generate a list of basic science orthopedic investigators who received funding anytime during 2011 to 2014.¹² This methodology was used due to the lack of reporting of basic science investigators on program websites. A list of NIH-funded orthopedic investigators was cross-referenced via an online search to isolate a cohort of PhD investigators.

Orthopedic faculty were defined as chairpersons, program directors, or NIH-funded investigators. In cases of overlap, preference was given in that order. Orthopedic investigators who had not published an article after 1995 were excluded (6 chairpersons, 1 program director).

BIBLIOMETRIC ANALYSIS

While several resources exist to calculate the h-index, the Scopus database (Elsevier) is one of the easiest programs to use.¹³ Author entries are created via institutional affiliations, thereby alleviating the need for manual reconciliations. Investigators were identified on Scopus via “author last name” and “first name, middle initial.” For each author, publications were screened for relevance to the field of orthopedics. Affiliated institutions were cross-referenced with information obtained from individual program websites. The “view h-graph” feature was used to calculate the number of publications, h-index, and number of citations. Then, the “Exclude self-citations” feature was used to calculate the number of corrected citations and the h-index excluding self-citations. Metrics were calculated over a 2-day period.

Continue to: STATISTICAL ANALYSIS

 

 

STATISTICAL ANALYSIS

Bibliometric analyses were presented descriptively with means ± standard deviation. After testing for normality, differences in the h-index between groups were assessed via analysis of variance tests. The proportional increase in the number of citations and the h-index were calculated by dividing the difference between the before and after exclusion of self-citations by the total number of citations before exclusion. The relationship between the number of publications and the proportional change in the h-index was assessed via calculation of the Spearman correlation coefficient. The independent variable was the number of publications, and the proportional increase in the h-index via self-citation was the dependent variable. Statistical tests were performed on STATA 13 (StataCorp) and the results considered significant if P < .05. Figures were created using GraphPad Prism 6.02 Software.

RESULTS

A total of 463 orthopedic investigators were included (147 chairpersons, 118 program directors, and 198 NIH-funded faculty) (Table 1). On average, these researchers produced 72.3 ± 83.0 articles and referenced 2139 ± 3222 articles (mean, 29.6 references per article). The cumulative h-index was 19.2 ± 15.6, and was the highest among NIH-funded researchers (24.3 ± 17.0) (P < .001). In all, 83.8% of orthopedic investigators self-cited their previous work at least once, and the total number of self-citations was highest among NIH-funded investigators (221 ± 355) (P < .001). After these self-citations were excluded, the h-index changed by 0.6 ± 1.1 for all investigators, and this change was greatest among NIH-funded researchers (1.1 ± 1.3) (P < .001).

Table 1. Effect of Self-Citation on NIH-funded Investigators, Chairpersons, and Program Directors in Orthopedics

Investigator	N (%)	Articles, n (mean ± SD)	Total Citations (mean ± SD)	h-index	Self-Citations (mean ± SD)	Corrected h-index	∆ h-index
NIH-funded	198 (42.8)	87.6 ± 84.9	3086 ± 3799	24.3 ± 17.0	221 ± 355	23.2 ± 16.3	1.1 ± 1.3
Chairperson	147 (31.7)	85.3 ± 95.5	2151 ± 3098	19.9 ± 15.0	85.2 ± 221	19.5 ± 14.5	0.4 ± 0.8
Program Director	118 (25.5)	30.5 ± 35.9	536.8 ± 785	9.6 ± 7.2	8.8 ± 19.9	9.5 ± 7.1	0.1 ± 0.3
Total	463 (100)	72.3 ± 83.0	2139 ± 3222	19.2 ± 15.6	123.9 ± 277.6	18.5 ± 14.9	0.6 ± 1.1

Abbreviation: NIH, National Institutes of Health.

Most orthopedic investigators did not increase their h-index via self-citation (63.7%, P < .001). Table 2 categorizes investigators by changes in their h-index after excluding self-citations (range, 0-11). The maximal change in the h-index was seen in the most prolific group of investigators, who produced 261.0 ± 149.3 articles. In this group, the h-index increased by 11.1% ± 5.2%. The Figure investigates the relationship between the number of articles and the proportional increase in the h-index. The number of publications was positively correlated with the change in h-index after self-citations were excluded (r = 0.448, P < .001).

Table 2. Stratification of Orthopedic Researcher Investigators by Change in h-index After Self-Citation

∆ h-index	N (%)	Articles (mean ± SD)	Self-Citations (mean ± SD)	h-index (mean ± SD)	% Increase in h-index
0	295 (63.7)	43.8 ± 51.3	27.6 ± 58.4	13.1 ± 10.7	0
1	101 (21.8)	87.9 ± 68.3	126.0 ± 130.6	24.0 ± 13.3	5.9 ± 4.1
2	42 (9.1)	141.9 ± 111.1	331.6 ± 318.0	32.4 ± 16.6	8.4 ± 5.5
3	14 (3.0)	203.1 ± 92.6	611.6 ± 332.9	45.4 ± 14.9	7.6 ± 3.6
4+	11 (2.4)	261.0 ± 149.3	1277.1 ± 692.4	53.1 ± 18.9	11.1 ± 5.2

DISCUSSION

The practice of self-citation is widely prevalent among experienced orthopedic investigators. However, this practice seems to have minimal impact on the h-index for most investigators. Self-citation had a measurable impact on the h-index only after an investigator had many publications. At a mean of 87.9 ± 68.3 articles, investigators had a ∆h-index of 1. This represented a mean 5.9% increase. Overall, these findings underscore the utility of the h-index in assessing scholarly impact and ameliorate concerns over bibliometric manipulation.

Among a large group of experienced orthopedic investigators, self-citation has minimal effect on the h-index. Importantly, most investigators (63.7%) did not experience a full integer increase in their h-index. At a threshold of ∆ h-index increase of 1, investigators had impressive h-indices (24.0 ± 13.3), which eclipsed those of recent studies of hand surgeons (10.2 ± 9.9) and spine surgeons (13.6 ± 8.7).^4,5 This finding suggests that committees for academic promotion in orthopedic surgery may disregard the impact of self-citation on the h-index. While the thresholds for promotion have not been defined in the orthopedic literature, a study in plastic surgery established an h-index threshold of 14.5 for promotion from associate to full professor.¹⁴ It may be, however, that h-indices are higher among orthopedic surgeons, as a previous study reported an h-index of 20 among full professors.¹⁵ Future research is needed to determine thresholds for promotion within orthopedic surgery, as the h-index varies by specialty according to unique citation patterns.

Continue to: It is worth highlighting...

 

It is worth highlighting the academic performance of NIH-funded PhD researchers in orthopedics. Even including training grant awardees in this group, this cohort exceeded the academic productivity of their orthopedic chairpersons, as measured by the h-index. Previous studies in urology, neurosurgery, and otolaryngology have demonstrated the impact of NIH-funding on academic productivity.^8-10 Ultimately, orthopedic departments could increase academic productivity by recruiting more PhD investigators with NIH funding.

In contrast to academic radiology,¹⁶ this study demonstrated a correlation between the number of publications and the increase in h-index via self-citation. Several reasons may help explain this disparity. The first reason is a technical one, as at the time of this study, the Scopus database had been updated to include citations before 1996. Considering that the h-index increases over time as older publications are cited, the exclusion of older articles is a significant limitation of previous h-index studies. Applying the same logic, the mean h-index for chairpersons of 19.9 quoted in this study contradicts a recent study, which quoted a mean h-index of 15.3.³ This previous study utilized citations that were limited to articles published after 1996.

Previous work on self-citation in the field of orthopedics has been limited to its influence on journal impact factors. Our results build on this literature in several important ways. Firstly, the calculation of a journal’s impact factor is a highly scrutinized process, and authors have criticized the mechanisms employed by editors to inflate impact factors.¹⁷ One study reported that 20% of authors have been encouraged to cite a journal during the revision process.¹⁸ Self-citation practices have been demonstrated in journals of cardiology,¹⁹ diabetes,²⁰ anesthesia,²¹ and medicine.²² A study using a hypothetical model to assess the maximum potential for h-index increase by self-citation demonstrated an h-index inflation of 5 points over 20 years (5/14, 35.7%) by publishing 3 papers per year with 3 self-citations each.²³ This study highlights a potential limitation of the h-index, but our study observed an h-index inflation of ≥4 in only 11 researchers (2.4%). Thus, results from our study ameliorate self-citation concerns in academic orthopedic surgery.

There are several limitations to this study that offer future areas of research. First, the validity of the h-index academic promotion in orthopedic surgery has not been evaluated. This was a motivation for the present study, and the authors have ongoing efforts to characterize the h-index in a larger cohort of orthopedic investigators. Importantly, an appropriate amount of self-citation was not established. It may be necessary for orthopedic researchers to cite their works as they become experts on a specific topic. Lastly, our analyses are prone to limitations inherent in the h-index, which does not account for author contribution or journal impact factors. Despite these limitations, we show that for most orthopedic researchers, the practice of self-citation does not impact the h-index.

In summary, self-citation is a widely prevalent practice among orthopedic investigators, but this practice has minimal impact on an author’s h-index. Approximately one third of orthopedic faculty in our study had a higher h-index due to self-citation. Greater h-index inflation through self-citation correlated with more publications. For the majority of orthopedic faculty, however, self-citation did not inflate the h-index, suggesting that promotional committees may disregard this concern when using the h-index as an adjunct measure for career advancement.

ABSTRACT

The h-index is a commonly utilized metric for academic productivity. Previous studies have proposed that self-citation may limit the utility of the h-index. The purpose of this study is to evaluate the impact of self-citation on the h-index among orthopedic investigators. The study cohort consisted of program directors, chairpersons, and faculty at orthopedic surgery residency programs in the United States. The Scopus database was used to determine the h-index and number of citations ± self-citations. The total number of publications was correlated with the change in the h-index via self-citation. A total of 463 researchers were included (198 National Institutes of Health-funded faculty, 147 chairpersons, 118 program directors). Of these researchers, 83.8% cited previous work at least once (mean, 123.9 ± 277.6). Self-citations accounted for 5.8% of all citations. Including these citations increased the author h-index from 18.5 ± 14.9 to 19.2 ± 15.6 (P < .001). A minority of researchers (36.3%, P < .001) had increased their h-index via self-citation (range, 0-11). The proportional increase in the h-index via self-citation was positively related to the number of publications (P < .001). While the practice of self-citation is prevalent in orthopedics, its impact on the h-index is minimal for most researchers. With more publications, researchers can increase their h-index to a greater degree via self-citation.

Continue to: The competitive nature of academic research...

 

The competitive nature of academic research requires objective metrics to define career end points, such as promotion and funding procurement. While various criteria are used to assess performance in academia, publications and research funding are particularly regarded.¹ Quantifying research dollars is relatively straightforward, but measuring research productivity is more complex. Not all articles are created equal, and disparities exist regarding effort and the ultimate impact of articles. In 2005, a physicist created the h-index to measure both research impact and productivity.² As a bibliometric, the h-index equals the number of publications h that have been cited at least h times. Given its simplicity, the h-index has gained wide popularity in diverse medical specialties, including orthopedic surgery.³ Other recent studies have applied the h-index to hand surgery and spine surgery.^4,5

Importantly, some authors have raised concerns regarding potential limitations of the h-index. One potentially significant limitation is the ability of authors to artificially inflate their h-index via self-citation. The impact of this practice is of particular interest as the h-index becomes widely adopted as a metric for promotion at many academic institutions.^6-7 Furthermore, scholarly productivity has remained a critical component of successful grant funding procurement, and future grant funding applications may evaluate the h-index.^8-10

The purpose of this study is to determine the prevalence and impact of self-citation on the h-index in a large cohort of orthopedic investigators. Given their high level of investment in academic orthopedic surgery, we focused on program directors, chairpersons, and National Institutes of Health (NIH)-funded research faculty at orthopedic surgery residency programs.

METHODS

INCLUSION CRITERIA

This study qualified as non-human and non-animal research and received exemption per the standing policy of the Institutional Review Board. The Fellowship and Residency Electronic Interactive Database (FREIDA) was accessed to generate a list of orthopedic residency program directors.¹¹ This database was also used to generate a list of allopathic orthopedic surgery residency programs. Official program websites were accessed to generate a list of orthopedic chairpersons. Lastly, the NIH RePORTER was used to generate a list of basic science orthopedic investigators who received funding anytime during 2011 to 2014.¹² This methodology was used due to the lack of reporting of basic science investigators on program websites. A list of NIH-funded orthopedic investigators was cross-referenced via an online search to isolate a cohort of PhD investigators.

Orthopedic faculty were defined as chairpersons, program directors, or NIH-funded investigators. In cases of overlap, preference was given in that order. Orthopedic investigators who had not published an article after 1995 were excluded (6 chairpersons, 1 program director).

BIBLIOMETRIC ANALYSIS

While several resources exist to calculate the h-index, the Scopus database (Elsevier) is one of the easiest programs to use.¹³ Author entries are created via institutional affiliations, thereby alleviating the need for manual reconciliations. Investigators were identified on Scopus via “author last name” and “first name, middle initial.” For each author, publications were screened for relevance to the field of orthopedics. Affiliated institutions were cross-referenced with information obtained from individual program websites. The “view h-graph” feature was used to calculate the number of publications, h-index, and number of citations. Then, the “Exclude self-citations” feature was used to calculate the number of corrected citations and the h-index excluding self-citations. Metrics were calculated over a 2-day period.

Continue to: STATISTICAL ANALYSIS

 

 

STATISTICAL ANALYSIS

Bibliometric analyses were presented descriptively with means ± standard deviation. After testing for normality, differences in the h-index between groups were assessed via analysis of variance tests. The proportional increase in the number of citations and the h-index were calculated by dividing the difference between the before and after exclusion of self-citations by the total number of citations before exclusion. The relationship between the number of publications and the proportional change in the h-index was assessed via calculation of the Spearman correlation coefficient. The independent variable was the number of publications, and the proportional increase in the h-index via self-citation was the dependent variable. Statistical tests were performed on STATA 13 (StataCorp) and the results considered significant if P < .05. Figures were created using GraphPad Prism 6.02 Software.

RESULTS

A total of 463 orthopedic investigators were included (147 chairpersons, 118 program directors, and 198 NIH-funded faculty) (Table 1). On average, these researchers produced 72.3 ± 83.0 articles and referenced 2139 ± 3222 articles (mean, 29.6 references per article). The cumulative h-index was 19.2 ± 15.6, and was the highest among NIH-funded researchers (24.3 ± 17.0) (P < .001). In all, 83.8% of orthopedic investigators self-cited their previous work at least once, and the total number of self-citations was highest among NIH-funded investigators (221 ± 355) (P < .001). After these self-citations were excluded, the h-index changed by 0.6 ± 1.1 for all investigators, and this change was greatest among NIH-funded researchers (1.1 ± 1.3) (P < .001).

Table 1. Effect of Self-Citation on NIH-funded Investigators, Chairpersons, and Program Directors in Orthopedics

Investigator	N (%)	Articles, n (mean ± SD)	Total Citations (mean ± SD)	h-index	Self-Citations (mean ± SD)	Corrected h-index	∆ h-index
NIH-funded	198 (42.8)	87.6 ± 84.9	3086 ± 3799	24.3 ± 17.0	221 ± 355	23.2 ± 16.3	1.1 ± 1.3
Chairperson	147 (31.7)	85.3 ± 95.5	2151 ± 3098	19.9 ± 15.0	85.2 ± 221	19.5 ± 14.5	0.4 ± 0.8
Program Director	118 (25.5)	30.5 ± 35.9	536.8 ± 785	9.6 ± 7.2	8.8 ± 19.9	9.5 ± 7.1	0.1 ± 0.3
Total	463 (100)	72.3 ± 83.0	2139 ± 3222	19.2 ± 15.6	123.9 ± 277.6	18.5 ± 14.9	0.6 ± 1.1

Abbreviation: NIH, National Institutes of Health.

Most orthopedic investigators did not increase their h-index via self-citation (63.7%, P < .001). Table 2 categorizes investigators by changes in their h-index after excluding self-citations (range, 0-11). The maximal change in the h-index was seen in the most prolific group of investigators, who produced 261.0 ± 149.3 articles. In this group, the h-index increased by 11.1% ± 5.2%. The Figure investigates the relationship between the number of articles and the proportional increase in the h-index. The number of publications was positively correlated with the change in h-index after self-citations were excluded (r = 0.448, P < .001).

Table 2. Stratification of Orthopedic Researcher Investigators by Change in h-index After Self-Citation

∆ h-index	N (%)	Articles (mean ± SD)	Self-Citations (mean ± SD)	h-index (mean ± SD)	% Increase in h-index
0	295 (63.7)	43.8 ± 51.3	27.6 ± 58.4	13.1 ± 10.7	0
1	101 (21.8)	87.9 ± 68.3	126.0 ± 130.6	24.0 ± 13.3	5.9 ± 4.1
2	42 (9.1)	141.9 ± 111.1	331.6 ± 318.0	32.4 ± 16.6	8.4 ± 5.5
3	14 (3.0)	203.1 ± 92.6	611.6 ± 332.9	45.4 ± 14.9	7.6 ± 3.6
4+	11 (2.4)	261.0 ± 149.3	1277.1 ± 692.4	53.1 ± 18.9	11.1 ± 5.2

DISCUSSION

The practice of self-citation is widely prevalent among experienced orthopedic investigators. However, this practice seems to have minimal impact on the h-index for most investigators. Self-citation had a measurable impact on the h-index only after an investigator had many publications. At a mean of 87.9 ± 68.3 articles, investigators had a ∆h-index of 1. This represented a mean 5.9% increase. Overall, these findings underscore the utility of the h-index in assessing scholarly impact and ameliorate concerns over bibliometric manipulation.

Among a large group of experienced orthopedic investigators, self-citation has minimal effect on the h-index. Importantly, most investigators (63.7%) did not experience a full integer increase in their h-index. At a threshold of ∆ h-index increase of 1, investigators had impressive h-indices (24.0 ± 13.3), which eclipsed those of recent studies of hand surgeons (10.2 ± 9.9) and spine surgeons (13.6 ± 8.7).^4,5 This finding suggests that committees for academic promotion in orthopedic surgery may disregard the impact of self-citation on the h-index. While the thresholds for promotion have not been defined in the orthopedic literature, a study in plastic surgery established an h-index threshold of 14.5 for promotion from associate to full professor.¹⁴ It may be, however, that h-indices are higher among orthopedic surgeons, as a previous study reported an h-index of 20 among full professors.¹⁵ Future research is needed to determine thresholds for promotion within orthopedic surgery, as the h-index varies by specialty according to unique citation patterns.

Continue to: It is worth highlighting...

 

It is worth highlighting the academic performance of NIH-funded PhD researchers in orthopedics. Even including training grant awardees in this group, this cohort exceeded the academic productivity of their orthopedic chairpersons, as measured by the h-index. Previous studies in urology, neurosurgery, and otolaryngology have demonstrated the impact of NIH-funding on academic productivity.^8-10 Ultimately, orthopedic departments could increase academic productivity by recruiting more PhD investigators with NIH funding.

In contrast to academic radiology,¹⁶ this study demonstrated a correlation between the number of publications and the increase in h-index via self-citation. Several reasons may help explain this disparity. The first reason is a technical one, as at the time of this study, the Scopus database had been updated to include citations before 1996. Considering that the h-index increases over time as older publications are cited, the exclusion of older articles is a significant limitation of previous h-index studies. Applying the same logic, the mean h-index for chairpersons of 19.9 quoted in this study contradicts a recent study, which quoted a mean h-index of 15.3.³ This previous study utilized citations that were limited to articles published after 1996.

Previous work on self-citation in the field of orthopedics has been limited to its influence on journal impact factors. Our results build on this literature in several important ways. Firstly, the calculation of a journal’s impact factor is a highly scrutinized process, and authors have criticized the mechanisms employed by editors to inflate impact factors.¹⁷ One study reported that 20% of authors have been encouraged to cite a journal during the revision process.¹⁸ Self-citation practices have been demonstrated in journals of cardiology,¹⁹ diabetes,²⁰ anesthesia,²¹ and medicine.²² A study using a hypothetical model to assess the maximum potential for h-index increase by self-citation demonstrated an h-index inflation of 5 points over 20 years (5/14, 35.7%) by publishing 3 papers per year with 3 self-citations each.²³ This study highlights a potential limitation of the h-index, but our study observed an h-index inflation of ≥4 in only 11 researchers (2.4%). Thus, results from our study ameliorate self-citation concerns in academic orthopedic surgery.

There are several limitations to this study that offer future areas of research. First, the validity of the h-index academic promotion in orthopedic surgery has not been evaluated. This was a motivation for the present study, and the authors have ongoing efforts to characterize the h-index in a larger cohort of orthopedic investigators. Importantly, an appropriate amount of self-citation was not established. It may be necessary for orthopedic researchers to cite their works as they become experts on a specific topic. Lastly, our analyses are prone to limitations inherent in the h-index, which does not account for author contribution or journal impact factors. Despite these limitations, we show that for most orthopedic researchers, the practice of self-citation does not impact the h-index.

In summary, self-citation is a widely prevalent practice among orthopedic investigators, but this practice has minimal impact on an author’s h-index. Approximately one third of orthopedic faculty in our study had a higher h-index due to self-citation. Greater h-index inflation through self-citation correlated with more publications. For the majority of orthopedic faculty, however, self-citation did not inflate the h-index, suggesting that promotional committees may disregard this concern when using the h-index as an adjunct measure for career advancement.

ABSTRACT

The h-index is a commonly utilized metric for academic productivity. Previous studies have proposed that self-citation may limit the utility of the h-index. The purpose of this study is to evaluate the impact of self-citation on the h-index among orthopedic investigators. The study cohort consisted of program directors, chairpersons, and faculty at orthopedic surgery residency programs in the United States. The Scopus database was used to determine the h-index and number of citations ± self-citations. The total number of publications was correlated with the change in the h-index via self-citation. A total of 463 researchers were included (198 National Institutes of Health-funded faculty, 147 chairpersons, 118 program directors). Of these researchers, 83.8% cited previous work at least once (mean, 123.9 ± 277.6). Self-citations accounted for 5.8% of all citations. Including these citations increased the author h-index from 18.5 ± 14.9 to 19.2 ± 15.6 (P < .001). A minority of researchers (36.3%, P < .001) had increased their h-index via self-citation (range, 0-11). The proportional increase in the h-index via self-citation was positively related to the number of publications (P < .001). While the practice of self-citation is prevalent in orthopedics, its impact on the h-index is minimal for most researchers. With more publications, researchers can increase their h-index to a greater degree via self-citation.

Continue to: The competitive nature of academic research...

 

The competitive nature of academic research requires objective metrics to define career end points, such as promotion and funding procurement. While various criteria are used to assess performance in academia, publications and research funding are particularly regarded.¹ Quantifying research dollars is relatively straightforward, but measuring research productivity is more complex. Not all articles are created equal, and disparities exist regarding effort and the ultimate impact of articles. In 2005, a physicist created the h-index to measure both research impact and productivity.² As a bibliometric, the h-index equals the number of publications h that have been cited at least h times. Given its simplicity, the h-index has gained wide popularity in diverse medical specialties, including orthopedic surgery.³ Other recent studies have applied the h-index to hand surgery and spine surgery.^4,5

Importantly, some authors have raised concerns regarding potential limitations of the h-index. One potentially significant limitation is the ability of authors to artificially inflate their h-index via self-citation. The impact of this practice is of particular interest as the h-index becomes widely adopted as a metric for promotion at many academic institutions.^6-7 Furthermore, scholarly productivity has remained a critical component of successful grant funding procurement, and future grant funding applications may evaluate the h-index.^8-10

The purpose of this study is to determine the prevalence and impact of self-citation on the h-index in a large cohort of orthopedic investigators. Given their high level of investment in academic orthopedic surgery, we focused on program directors, chairpersons, and National Institutes of Health (NIH)-funded research faculty at orthopedic surgery residency programs.

METHODS

INCLUSION CRITERIA

This study qualified as non-human and non-animal research and received exemption per the standing policy of the Institutional Review Board. The Fellowship and Residency Electronic Interactive Database (FREIDA) was accessed to generate a list of orthopedic residency program directors.¹¹ This database was also used to generate a list of allopathic orthopedic surgery residency programs. Official program websites were accessed to generate a list of orthopedic chairpersons. Lastly, the NIH RePORTER was used to generate a list of basic science orthopedic investigators who received funding anytime during 2011 to 2014.¹² This methodology was used due to the lack of reporting of basic science investigators on program websites. A list of NIH-funded orthopedic investigators was cross-referenced via an online search to isolate a cohort of PhD investigators.

Orthopedic faculty were defined as chairpersons, program directors, or NIH-funded investigators. In cases of overlap, preference was given in that order. Orthopedic investigators who had not published an article after 1995 were excluded (6 chairpersons, 1 program director).

BIBLIOMETRIC ANALYSIS

While several resources exist to calculate the h-index, the Scopus database (Elsevier) is one of the easiest programs to use.¹³ Author entries are created via institutional affiliations, thereby alleviating the need for manual reconciliations. Investigators were identified on Scopus via “author last name” and “first name, middle initial.” For each author, publications were screened for relevance to the field of orthopedics. Affiliated institutions were cross-referenced with information obtained from individual program websites. The “view h-graph” feature was used to calculate the number of publications, h-index, and number of citations. Then, the “Exclude self-citations” feature was used to calculate the number of corrected citations and the h-index excluding self-citations. Metrics were calculated over a 2-day period.

Continue to: STATISTICAL ANALYSIS

 

 

STATISTICAL ANALYSIS

Bibliometric analyses were presented descriptively with means ± standard deviation. After testing for normality, differences in the h-index between groups were assessed via analysis of variance tests. The proportional increase in the number of citations and the h-index were calculated by dividing the difference between the before and after exclusion of self-citations by the total number of citations before exclusion. The relationship between the number of publications and the proportional change in the h-index was assessed via calculation of the Spearman correlation coefficient. The independent variable was the number of publications, and the proportional increase in the h-index via self-citation was the dependent variable. Statistical tests were performed on STATA 13 (StataCorp) and the results considered significant if P < .05. Figures were created using GraphPad Prism 6.02 Software.

RESULTS

A total of 463 orthopedic investigators were included (147 chairpersons, 118 program directors, and 198 NIH-funded faculty) (Table 1). On average, these researchers produced 72.3 ± 83.0 articles and referenced 2139 ± 3222 articles (mean, 29.6 references per article). The cumulative h-index was 19.2 ± 15.6, and was the highest among NIH-funded researchers (24.3 ± 17.0) (P < .001). In all, 83.8% of orthopedic investigators self-cited their previous work at least once, and the total number of self-citations was highest among NIH-funded investigators (221 ± 355) (P < .001). After these self-citations were excluded, the h-index changed by 0.6 ± 1.1 for all investigators, and this change was greatest among NIH-funded researchers (1.1 ± 1.3) (P < .001).

Table 1. Effect of Self-Citation on NIH-funded Investigators, Chairpersons, and Program Directors in Orthopedics

Investigator	N (%)	Articles, n (mean ± SD)	Total Citations (mean ± SD)	h-index	Self-Citations (mean ± SD)	Corrected h-index	∆ h-index
NIH-funded	198 (42.8)	87.6 ± 84.9	3086 ± 3799	24.3 ± 17.0	221 ± 355	23.2 ± 16.3	1.1 ± 1.3
Chairperson	147 (31.7)	85.3 ± 95.5	2151 ± 3098	19.9 ± 15.0	85.2 ± 221	19.5 ± 14.5	0.4 ± 0.8
Program Director	118 (25.5)	30.5 ± 35.9	536.8 ± 785	9.6 ± 7.2	8.8 ± 19.9	9.5 ± 7.1	0.1 ± 0.3
Total	463 (100)	72.3 ± 83.0	2139 ± 3222	19.2 ± 15.6	123.9 ± 277.6	18.5 ± 14.9	0.6 ± 1.1

Abbreviation: NIH, National Institutes of Health.

Most orthopedic investigators did not increase their h-index via self-citation (63.7%, P < .001). Table 2 categorizes investigators by changes in their h-index after excluding self-citations (range, 0-11). The maximal change in the h-index was seen in the most prolific group of investigators, who produced 261.0 ± 149.3 articles. In this group, the h-index increased by 11.1% ± 5.2%. The Figure investigates the relationship between the number of articles and the proportional increase in the h-index. The number of publications was positively correlated with the change in h-index after self-citations were excluded (r = 0.448, P < .001).

Table 2. Stratification of Orthopedic Researcher Investigators by Change in h-index After Self-Citation

∆ h-index	N (%)	Articles (mean ± SD)	Self-Citations (mean ± SD)	h-index (mean ± SD)	% Increase in h-index
0	295 (63.7)	43.8 ± 51.3	27.6 ± 58.4	13.1 ± 10.7	0
1	101 (21.8)	87.9 ± 68.3	126.0 ± 130.6	24.0 ± 13.3	5.9 ± 4.1
2	42 (9.1)	141.9 ± 111.1	331.6 ± 318.0	32.4 ± 16.6	8.4 ± 5.5
3	14 (3.0)	203.1 ± 92.6	611.6 ± 332.9	45.4 ± 14.9	7.6 ± 3.6
4+	11 (2.4)	261.0 ± 149.3	1277.1 ± 692.4	53.1 ± 18.9	11.1 ± 5.2

DISCUSSION

The practice of self-citation is widely prevalent among experienced orthopedic investigators. However, this practice seems to have minimal impact on the h-index for most investigators. Self-citation had a measurable impact on the h-index only after an investigator had many publications. At a mean of 87.9 ± 68.3 articles, investigators had a ∆h-index of 1. This represented a mean 5.9% increase. Overall, these findings underscore the utility of the h-index in assessing scholarly impact and ameliorate concerns over bibliometric manipulation.

Among a large group of experienced orthopedic investigators, self-citation has minimal effect on the h-index. Importantly, most investigators (63.7%) did not experience a full integer increase in their h-index. At a threshold of ∆ h-index increase of 1, investigators had impressive h-indices (24.0 ± 13.3), which eclipsed those of recent studies of hand surgeons (10.2 ± 9.9) and spine surgeons (13.6 ± 8.7).^4,5 This finding suggests that committees for academic promotion in orthopedic surgery may disregard the impact of self-citation on the h-index. While the thresholds for promotion have not been defined in the orthopedic literature, a study in plastic surgery established an h-index threshold of 14.5 for promotion from associate to full professor.¹⁴ It may be, however, that h-indices are higher among orthopedic surgeons, as a previous study reported an h-index of 20 among full professors.¹⁵ Future research is needed to determine thresholds for promotion within orthopedic surgery, as the h-index varies by specialty according to unique citation patterns.

Continue to: It is worth highlighting...

 

It is worth highlighting the academic performance of NIH-funded PhD researchers in orthopedics. Even including training grant awardees in this group, this cohort exceeded the academic productivity of their orthopedic chairpersons, as measured by the h-index. Previous studies in urology, neurosurgery, and otolaryngology have demonstrated the impact of NIH-funding on academic productivity.^8-10 Ultimately, orthopedic departments could increase academic productivity by recruiting more PhD investigators with NIH funding.

In contrast to academic radiology,¹⁶ this study demonstrated a correlation between the number of publications and the increase in h-index via self-citation. Several reasons may help explain this disparity. The first reason is a technical one, as at the time of this study, the Scopus database had been updated to include citations before 1996. Considering that the h-index increases over time as older publications are cited, the exclusion of older articles is a significant limitation of previous h-index studies. Applying the same logic, the mean h-index for chairpersons of 19.9 quoted in this study contradicts a recent study, which quoted a mean h-index of 15.3.³ This previous study utilized citations that were limited to articles published after 1996.

Previous work on self-citation in the field of orthopedics has been limited to its influence on journal impact factors. Our results build on this literature in several important ways. Firstly, the calculation of a journal’s impact factor is a highly scrutinized process, and authors have criticized the mechanisms employed by editors to inflate impact factors.¹⁷ One study reported that 20% of authors have been encouraged to cite a journal during the revision process.¹⁸ Self-citation practices have been demonstrated in journals of cardiology,¹⁹ diabetes,²⁰ anesthesia,²¹ and medicine.²² A study using a hypothetical model to assess the maximum potential for h-index increase by self-citation demonstrated an h-index inflation of 5 points over 20 years (5/14, 35.7%) by publishing 3 papers per year with 3 self-citations each.²³ This study highlights a potential limitation of the h-index, but our study observed an h-index inflation of ≥4 in only 11 researchers (2.4%). Thus, results from our study ameliorate self-citation concerns in academic orthopedic surgery.

There are several limitations to this study that offer future areas of research. First, the validity of the h-index academic promotion in orthopedic surgery has not been evaluated. This was a motivation for the present study, and the authors have ongoing efforts to characterize the h-index in a larger cohort of orthopedic investigators. Importantly, an appropriate amount of self-citation was not established. It may be necessary for orthopedic researchers to cite their works as they become experts on a specific topic. Lastly, our analyses are prone to limitations inherent in the h-index, which does not account for author contribution or journal impact factors. Despite these limitations, we show that for most orthopedic researchers, the practice of self-citation does not impact the h-index.

In summary, self-citation is a widely prevalent practice among orthopedic investigators, but this practice has minimal impact on an author’s h-index. Approximately one third of orthopedic faculty in our study had a higher h-index due to self-citation. Greater h-index inflation through self-citation correlated with more publications. For the majority of orthopedic faculty, however, self-citation did not inflate the h-index, suggesting that promotional committees may disregard this concern when using the h-index as an adjunct measure for career advancement.

Hypertension affects more than 65 million individuals in the U.S., accounting for nearly 30% of the adult population.¹ Less than 50% of those with hypertension are taking appropriate pharmacotherapy.² Hypertension contributes to cardiovascular events, including cerebrovascular accident, transient ischemic attack, hypertensive retinopathy, renal failure, myocardial infarction, and heart failure.¹ Chronic hypertension mainly is an asymptomatic condition, earning the nickname “the silent killer.”² An acute, symptomatic elevation in blood pressure (BP) often is referred to as hypertensive emergency. Symptoms of end-organ damage can include headache, blurry vision, chest pain, shortness of breath, altered mental status, epistaxis, and oliguria.² Although rare, hypertensive emergencies should be treated immediately. The Seventh Report of the Joint National Committee (JNC 7), and the more recent JNC 8, have published guidelines on managing chronic hypertension.^3,4 However, neither report provides guidance on hypertensive emergency or the appropriate actions in cases of extremely elevated BP in an asymptomatic patient.^3,4

Acute hypertensive episodes—often referred to as hypertensive crises—are responsible for nearly 8 million hospitalizations each year and 20 million visits to the emergency department (ED).^5,6 Most of these visits are same-day “treat-and-release” events.⁵ There is no universally accepted BP value associated with a hypertensive crisis, but most resources state that a BP ≥ 180/110 mm Hg requires attention.^2,7 Without other symptoms, elevated BP is not an emergency, yet ED referral for acute management is common.⁷

Three terms fall under the umbrella of hypertensive crises: hypertensive emergency, hypertensive urgency, and asymptomatic hypertension (AH).² In a 2007 article, the American College of Chest Physicians defined hypertensive emergency as BP ≥ 180/110 mm Hg with evidence of end-organ damage.² Symptoms are almost always present in true hypertensive emergencies, and immediate medical intervention is required to halt further organ damage. In the same article, hypertensive urgency is defined as BP ≥ 180/110 mm Hg without end-organ damage.² The definition of hypertensive urgency could be further refined to include the presence of cardiovascular and renal risk factors, although this additional point is not consistent across the literature. Asymptomatic hypertension is similar to hypertensive urgency; however, there is an absence of signs or symptoms of end-organ damage.² There is ambiguity in the literature concerning managing hypertensive urgency and AH, but both share a basic tenet: Immediate BP reduction is not essential. Gradual dosage adjustment(s) of oral medications, preferably by a primary care provider (PCP), and follow-up within 7 days are recommended.⁷

Limited evidence exists to guide ED providers in managing AH. Long-term outcomes and guidelines intended for the primary care setting should not be extrapolated to acute management in the ED. With limited treatment guidelines, providers might be more likely to refer patients with AH to the ED for evaluation. In 2013, the American College of Emergency Physicians (ACEP) created a clinical policy concerning AH in the ED. The ACEP concluded that screening for target organ injury and medical intervention in the ED does not reduce rates of adverse events (AEs) and could lead to overtreatment and acute hypoperfusion.⁷ More recently, Patel and colleagues published findings on hypertensive urgency in the ambulatory care setting, which similarly found that referral to the ED was associated with increased use of health care resources and no change in short-term major AEs.⁸ The ACEP recommends that patients presenting with AH be referred to primary care clinics where long-term monitoring and medication adjustments can be achieved more cost-effectively.⁷

The objective of this retrospective evaluation was to assess the incidence and management of AH within a VA ED. The authors aimed to provide insight into how these patients are managed and discuss alternatives to ED use.

Methods

This retrospective observational study was conducted within the North Florida/South Georgia Veterans Health System (NFSGVHS), which provides patient care at 2 medical centers in Gainesville and Lake City, Florida, as well as 11 outpatient clinics located throughout North Florida and South Georgia. The NFSGVHS serves rural and urban veteran populations. Study approval was granted by the NFSGVHS Institutional Review Board and Research and Development Committee.

Inclusion/Exclusion Criteria

Adult patients who were ordered at least 1 antihypertensive medication in the ED from July 1, 2011 to July 1, 2014, in addition to being asymptomatic with BP ≥ 180/110 mm Hg at ED triage were included. Based on clinical experience, the authors estimated that 3 years would provide a sample size of more than 100 patients. Patients were excluded if they presented with any acute symptoms or were hospitalized for further management.

Data Collection

Baseline demographics were collected for all participants. During the ED encounter, pre- and postintervention vital signs were recorded and prespecified laboratory data obtained. Interrater reliability was accounted for by performing random reviews of previously collected data to ensure consistency during the chart review process. Renal end-organ damage was defined using Acute Kidney Injury Network criteria, a serum creatinine 50% above baseline, or an absolute increase in baseline serum creatinine by 0.3 mg/dL.⁹ Additional laboratory markers of organ damage included cardiac troponin levels. Urinalysis results also were assessed to determine the presence of hematuria or proteinuria. Patient-reported nonadherence with medications was determined by reviewing ED provider and/or nurse documentation notes for the index ED encounter.

 

Investigators documented the route (IV or oral) and antihypertensive(s) medication selected for each patient. Adverse effects and any changes to patients’ outpatient medication regimens were noted. Investigators also assessed days to next medical contact after ED discharge to determine whether follow-up occurred according to the recommended standard of 7 days.⁹ Days to next medical contact was defined as any contact—in person or by telephone—that was documented in the electronic health record after the index ED visit.

Statistical Analysis

Descriptive statistics, including mean, median, and standard deviation, were used to analyze data.

Results

A total of 1,052 patients presented with BP ≥ 180/110 mm Hg and for whom antihypertensive medication was ordered but not necessarily given in the ED. Of the total, 724 patients were excluded because of hospital admission for other primary diagnoses; however, 6 of these patients were admitted for hypertensive urgency. The final analysis included 132 patients who presented with the primary condition of elevated BP without any accompanying symptoms. Among these patients, 2 had repeat ED visits for AH during the specified time frame.

Each ED visit was treated as a separate occurrence.

Most patients were male with an average age of 63 years and documented history of hypertension. Nearly all patients had established primary care within the NFSGVHS. The most common comorbidity was diabetes mellitus (36%), followed by coronary artery disease (27%) and chronic kidney disease (CKD) (21%) (Table 1). About one-third of patients presented to the ED on their own volition, and slightly more than half were referred to the ED by primary care or specialty clinics.

The average BP and heart rate at ED presentation was 199/112 mm Hg and 76 beats per minute, respectively.

In the ED, 130 patients received BP treatment (Table 2). Medication was ordered for 2 patients who did not receive treatment. In total, 12 different medication classes were used for treating patients with AH in the ED (Figure). 

Most were treated with at least 1 oral antihypertensive; clonidine was the most common (48% of orally administered doses). In this study, 13% of patients received IV-only intervention; most were treated with hydralazine. Among the patients in the study, 22% were treated with a combination of oral and IV antihypertensives. No immediate AEs were noted for medications administered in the ED; however, 1 patient returned to the ED with angioedema after initiating an angiotensin-converting enzyme inhibitor at an ED visit 5 days earlier.

Treatment in the ED resulted in an average BP and heart rate reduction of 27/20 mm Hg and 5 beats per minute, respectively. About 80% of patients had a basic metabolic panel drawn, and there were no instances of acute kidney injury. Of the patients in the study 38% had cardiac enzymes collected, and only 1 patient had a positive result, which was determined to be unrelated to acute coronary syndrome. Forty-one (31%) of patients had a urinalysis; 12 has positive results for hematuria, and 18 revealed proteinuria. Of note, the 6 patients who were hospitalized for hypertensive urgency had neither symptoms at presentation to the ED nor laboratory findings indicating end-organ damage. The reason these patients were admitted is unclear.

At discharge, ED providers made changes to 54% of patients’ outpatient antihypertensive regimens. These changes included adding a new medication (68%), increasing the dosage of an existing medication (24%), or multiple changes (8%). Refills were provided for 18% of prescriptions. Follow-up within 7 days from ED discharge was recorded for 34% of patients. One patient received follow-up outside the NFSGVHS and was not included in this analysis.

Discussion

The aim of this retrospective study was to determine the incidence of AH in a VA ED and describe how these patients were managed. Overall, the rate of patients presenting to the ED with AH during the study period was about 1 patient every 8 days or 45 patients per year. By comparison, more than 30,000 patients are seen at the NFSGVHS ED annually. Although AH seems to be an uncommon occurrence, study findings raise questions about the value of managing the condition in the ED.

This study found several management strategies as well as noteworthy trends. For example, laboratory tests were not ordered routinely for all patients, suggesting that some ED providers question their use for AH. There were no patients with acute elevations in serum creatinine that indicated acute kidney injury, and although hematuria and proteinuria were common findings, neither were specific for acute injury. However, there were findings typical of chronic hypertension, and urinalysis may provide little benefit when testing for acute kidney injury. Only 1 patient showed elevated cardiac enzymes, which was determined to be a result of CKD.

Although not included in the final analysis, the 6 patients who were hospitalized for hypertensive urgency were similar in that they had neither symptoms at presentation to the ED nor laboratory findings indicating end-organ damage. Collectively, these findings support existing literature that questions the utility of laboratory testing of patients with AH in the ED.¹⁰

Patients also were treated with a variety of antihypertensive agents in the ED. One explanation might be outpatient nonadherence with medications. In patients with AH, it is common to provide doses of chronic medications that the patient might have missed and should be taking on a regular basis. Therefore, assessing adherence with current medications before modifying chronic therapy is an important initial step when managing AH.

Although oral agents primarily were used, IV antihypertensives were administered to about one-third of patients. Preference for IV administration in the ED might be related to its ability to lower BP quickly. The practice of obtaining IV access for medication in a patient with AH is costly, unnecessary, and potentially harmful.⁷ The authors theorize that this practice is performed, in many cases, as an attempt to expedite ED discharge after an acceptable BP reading is documented.

Rapid reductions in BP can precipitate hypoperfusion inadvertently and are more likely to occur with IV agents than with oral ones. Therefore, the safety, convenience, and cost savings associated with oral administration make it the preferred route for managing AH. 

Oral antihypertensives with desired therapeutic and pharmacokinetic properties are listed in Table 3. When used appropriately, these agents are well tolerated and effective and could be given in an ambulatory care clinic without the need for intensive monitoring.

 

Best Practices

Primary care clinics are best suited to manage AH because medication adjustments and long-term monitoring are easier to perform and at substantially lower costs when compared with that of the ED. Rather than immediately referring a patient to the ED, clinicians should consider factors that could elevate BP, such as medication nonadherence, anxiety, acute pain, recent tobacco or caffeine use, or white coat syndrome. Staff should be well educated on proper BP measurement and instructed to repeat the reading for confirmation. Before measuring BP, allow the patient to sit quietly for 5 minutes with the feet flat on the floor and arm supported.³ Ideally, the measurement used should be the average of 3 BP readings on an automated device.¹¹ If BP readings are high, staff should ask the patient about medication adherence and missed medication(s) should be administered.

It also is reasonable to have the patient rest quietly for up to 30 minutes because rest has been shown to reduce BP in some patients.¹² The drawback to the prolonged rest strategy is the potential to cause delays in care for other patients. However, it is important to remember that wait times in the ED often are measured in hours, which causes frustration for patients referred to the ED for AH management. Before completing the office visit, the provider should recheck BP using proper technique and confirm that the patient has antihypertensive medication(s) in his/her possession; a follow-up appointment should be scheduled for no later than 1 week.

Primary care providers might be concerned about taking on additional liability and could favor ED referral, but legislation makes it difficult for EDs to defer nonemergent issues to primary care clinics. The Emergency Medical Treatment and Labor Act states that hospitals are prohibited from denying a patient care during an emergency.¹³ Despite evidence that AH is not an emergency, many patients continue to be referred to the ED. One-third of patients presented to the ED on their own volition and more than one-half were referred by health care personnel. This strongly suggests that both patients and health care personnel consider AH an emergency medical condition requiring immediate attention. However, patients with AH rarely are found to have any acute end-organ damage; therefore, acute treatment and extensive laboratory or diagnostic testing in the ED provides little, if any, benefit.¹⁰ The authors believe the ACEP clinical policy should be adopted into mainstream practice to help reduce health care costs and preserve ED resources for patients with true emergencies.

Another pervasive issue that could contribute to inappropriate AH referrals to the ED is the shortage of PCPs and limited same-day appointments for nonemergent conditions. In a 2017 survey, the average wait time for a PCP appointment ranged between 12 and 109 days, depending on the metropolitan area. The national average wait time conducted by this survey was 29.3 days.¹⁴ When primary care appointments are unavailable, triage staff could recommend that patients seek care in the ED. Additionally, patients might choose to seek ED care rather than wait for the next available PCP appointment. Clinic proximity to an ED could influence referral rates. In other words, medical centers or health systems with primary care clinics and ED services under one roof could experience more frequent ED referrals.

A promising strategy to help overcome the challenges of addressing AH and avoiding ED referrals is increasing patient access to and use of qualified, nonphysician providers, such as clinical pharmacists and nurse practitioners. Large health systems such as the VA and Kaiser Permanente have employed clinical pharmacist providers to reduce follow-up times for patients in primary care settings.¹⁵ Furthermore, there is substantial evidence that supports the cost-effectiveness and clinical success of pharmacist-driven hypertension clinics.^16-18 Nurse-driven efforts to improve hypertension control have been successfully implemented in health systems.¹⁹ Both clinical pharmacist and nurse-managed hypertension clinics are effective solutions to manage patients with AH who might otherwise use costly ED services.For example, the average cost of a single ED visit is $740 to $3,437.²⁰ In comparison, a 2010 report from the Agency for Healthcare Research and Quality showed the average annual cost of managing hypertension in ambulatory care clinics was $442 per adult, a cost considerably lower than that of the ED.²¹

Limitations

The retrospective and observational design of this study are inherent limitations. This study was not designed to evaluate cardiovascular outcomes after ED encounters. The sample size could have been larger if patients with BP < 180/110 mm Hg at ED triage were included; however, the 180/110 mm Hg threshold was chosen because it was the most widely agreed on BP value in the literature. This study did not capture patients who presented with AH and did not receive any acute treatment in the ED.Prescribing patterns based on provider training (eg, emergency medicine, family medicine, or internal medicine) were not tracked and might have accounted for differences in selection of diagnostic tests, laboratory ordering, and route of drug administration preference.

 

A small subset of patients reported positive pain scores at triage but did not describe acute pain. Pain scores are highly subjective, and few primary literature sources link chronic pain with increased BP.^22,23 Nevertheless, patients who reported acute pain and elevated BP were excluded in order to identify truly asymptomatic patients. VA hospitals are unique health systems and data obtained from this study might not be applicable to other public or private facilities. Last, the study did not take into account patients’ psychosocial circumstances that might have fostered a disproportionate reliance on the ED for health care.

Conclusion

Asymptomatic patients with elevated BP are treated in the ED despite no evidence supporting improved outcomes after acute BP lowering in this population. Follow-up after ED encounters for AH did not occur consistently within guideline-recommended 7 days, a trend that also occurs in non-VA systems.⁸ Clinics and health care systems could establish policies to prevent or minimize management of AH in the ED. Ideally, AH should be managed in a clinic setting by a PCP, but growing clinician workload might lead to increasing wait times and difficultly obtaining same-day appointments. Nurse-led clinics and clinical pharmacists operating under a scope of practice and working closely with a PCP are a cost-effective solution to ensure timely treatment and appropriate follow-up of patients with uncontrolled hypertension.

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the North Florida South Georgia Veterans Health System in Gainesville, Florida.

Hypertension affects more than 65 million individuals in the U.S., accounting for nearly 30% of the adult population.¹ Less than 50% of those with hypertension are taking appropriate pharmacotherapy.² Hypertension contributes to cardiovascular events, including cerebrovascular accident, transient ischemic attack, hypertensive retinopathy, renal failure, myocardial infarction, and heart failure.¹ Chronic hypertension mainly is an asymptomatic condition, earning the nickname “the silent killer.”² An acute, symptomatic elevation in blood pressure (BP) often is referred to as hypertensive emergency. Symptoms of end-organ damage can include headache, blurry vision, chest pain, shortness of breath, altered mental status, epistaxis, and oliguria.² Although rare, hypertensive emergencies should be treated immediately. The Seventh Report of the Joint National Committee (JNC 7), and the more recent JNC 8, have published guidelines on managing chronic hypertension.^3,4 However, neither report provides guidance on hypertensive emergency or the appropriate actions in cases of extremely elevated BP in an asymptomatic patient.^3,4

Acute hypertensive episodes—often referred to as hypertensive crises—are responsible for nearly 8 million hospitalizations each year and 20 million visits to the emergency department (ED).^5,6 Most of these visits are same-day “treat-and-release” events.⁵ There is no universally accepted BP value associated with a hypertensive crisis, but most resources state that a BP ≥ 180/110 mm Hg requires attention.^2,7 Without other symptoms, elevated BP is not an emergency, yet ED referral for acute management is common.⁷

Three terms fall under the umbrella of hypertensive crises: hypertensive emergency, hypertensive urgency, and asymptomatic hypertension (AH).² In a 2007 article, the American College of Chest Physicians defined hypertensive emergency as BP ≥ 180/110 mm Hg with evidence of end-organ damage.² Symptoms are almost always present in true hypertensive emergencies, and immediate medical intervention is required to halt further organ damage. In the same article, hypertensive urgency is defined as BP ≥ 180/110 mm Hg without end-organ damage.² The definition of hypertensive urgency could be further refined to include the presence of cardiovascular and renal risk factors, although this additional point is not consistent across the literature. Asymptomatic hypertension is similar to hypertensive urgency; however, there is an absence of signs or symptoms of end-organ damage.² There is ambiguity in the literature concerning managing hypertensive urgency and AH, but both share a basic tenet: Immediate BP reduction is not essential. Gradual dosage adjustment(s) of oral medications, preferably by a primary care provider (PCP), and follow-up within 7 days are recommended.⁷

Limited evidence exists to guide ED providers in managing AH. Long-term outcomes and guidelines intended for the primary care setting should not be extrapolated to acute management in the ED. With limited treatment guidelines, providers might be more likely to refer patients with AH to the ED for evaluation. In 2013, the American College of Emergency Physicians (ACEP) created a clinical policy concerning AH in the ED. The ACEP concluded that screening for target organ injury and medical intervention in the ED does not reduce rates of adverse events (AEs) and could lead to overtreatment and acute hypoperfusion.⁷ More recently, Patel and colleagues published findings on hypertensive urgency in the ambulatory care setting, which similarly found that referral to the ED was associated with increased use of health care resources and no change in short-term major AEs.⁸ The ACEP recommends that patients presenting with AH be referred to primary care clinics where long-term monitoring and medication adjustments can be achieved more cost-effectively.⁷

The objective of this retrospective evaluation was to assess the incidence and management of AH within a VA ED. The authors aimed to provide insight into how these patients are managed and discuss alternatives to ED use.

Methods

This retrospective observational study was conducted within the North Florida/South Georgia Veterans Health System (NFSGVHS), which provides patient care at 2 medical centers in Gainesville and Lake City, Florida, as well as 11 outpatient clinics located throughout North Florida and South Georgia. The NFSGVHS serves rural and urban veteran populations. Study approval was granted by the NFSGVHS Institutional Review Board and Research and Development Committee.

Inclusion/Exclusion Criteria

Adult patients who were ordered at least 1 antihypertensive medication in the ED from July 1, 2011 to July 1, 2014, in addition to being asymptomatic with BP ≥ 180/110 mm Hg at ED triage were included. Based on clinical experience, the authors estimated that 3 years would provide a sample size of more than 100 patients. Patients were excluded if they presented with any acute symptoms or were hospitalized for further management.

Data Collection

Baseline demographics were collected for all participants. During the ED encounter, pre- and postintervention vital signs were recorded and prespecified laboratory data obtained. Interrater reliability was accounted for by performing random reviews of previously collected data to ensure consistency during the chart review process. Renal end-organ damage was defined using Acute Kidney Injury Network criteria, a serum creatinine 50% above baseline, or an absolute increase in baseline serum creatinine by 0.3 mg/dL.⁹ Additional laboratory markers of organ damage included cardiac troponin levels. Urinalysis results also were assessed to determine the presence of hematuria or proteinuria. Patient-reported nonadherence with medications was determined by reviewing ED provider and/or nurse documentation notes for the index ED encounter.

 

Investigators documented the route (IV or oral) and antihypertensive(s) medication selected for each patient. Adverse effects and any changes to patients’ outpatient medication regimens were noted. Investigators also assessed days to next medical contact after ED discharge to determine whether follow-up occurred according to the recommended standard of 7 days.⁹ Days to next medical contact was defined as any contact—in person or by telephone—that was documented in the electronic health record after the index ED visit.

Statistical Analysis

Descriptive statistics, including mean, median, and standard deviation, were used to analyze data.

Results

A total of 1,052 patients presented with BP ≥ 180/110 mm Hg and for whom antihypertensive medication was ordered but not necessarily given in the ED. Of the total, 724 patients were excluded because of hospital admission for other primary diagnoses; however, 6 of these patients were admitted for hypertensive urgency. The final analysis included 132 patients who presented with the primary condition of elevated BP without any accompanying symptoms. Among these patients, 2 had repeat ED visits for AH during the specified time frame.

Each ED visit was treated as a separate occurrence.

Most patients were male with an average age of 63 years and documented history of hypertension. Nearly all patients had established primary care within the NFSGVHS. The most common comorbidity was diabetes mellitus (36%), followed by coronary artery disease (27%) and chronic kidney disease (CKD) (21%) (Table 1). About one-third of patients presented to the ED on their own volition, and slightly more than half were referred to the ED by primary care or specialty clinics.

The average BP and heart rate at ED presentation was 199/112 mm Hg and 76 beats per minute, respectively.

In the ED, 130 patients received BP treatment (Table 2). Medication was ordered for 2 patients who did not receive treatment. In total, 12 different medication classes were used for treating patients with AH in the ED (Figure). 

Most were treated with at least 1 oral antihypertensive; clonidine was the most common (48% of orally administered doses). In this study, 13% of patients received IV-only intervention; most were treated with hydralazine. Among the patients in the study, 22% were treated with a combination of oral and IV antihypertensives. No immediate AEs were noted for medications administered in the ED; however, 1 patient returned to the ED with angioedema after initiating an angiotensin-converting enzyme inhibitor at an ED visit 5 days earlier.

Treatment in the ED resulted in an average BP and heart rate reduction of 27/20 mm Hg and 5 beats per minute, respectively. About 80% of patients had a basic metabolic panel drawn, and there were no instances of acute kidney injury. Of the patients in the study 38% had cardiac enzymes collected, and only 1 patient had a positive result, which was determined to be unrelated to acute coronary syndrome. Forty-one (31%) of patients had a urinalysis; 12 has positive results for hematuria, and 18 revealed proteinuria. Of note, the 6 patients who were hospitalized for hypertensive urgency had neither symptoms at presentation to the ED nor laboratory findings indicating end-organ damage. The reason these patients were admitted is unclear.

At discharge, ED providers made changes to 54% of patients’ outpatient antihypertensive regimens. These changes included adding a new medication (68%), increasing the dosage of an existing medication (24%), or multiple changes (8%). Refills were provided for 18% of prescriptions. Follow-up within 7 days from ED discharge was recorded for 34% of patients. One patient received follow-up outside the NFSGVHS and was not included in this analysis.

Discussion

The aim of this retrospective study was to determine the incidence of AH in a VA ED and describe how these patients were managed. Overall, the rate of patients presenting to the ED with AH during the study period was about 1 patient every 8 days or 45 patients per year. By comparison, more than 30,000 patients are seen at the NFSGVHS ED annually. Although AH seems to be an uncommon occurrence, study findings raise questions about the value of managing the condition in the ED.

This study found several management strategies as well as noteworthy trends. For example, laboratory tests were not ordered routinely for all patients, suggesting that some ED providers question their use for AH. There were no patients with acute elevations in serum creatinine that indicated acute kidney injury, and although hematuria and proteinuria were common findings, neither were specific for acute injury. However, there were findings typical of chronic hypertension, and urinalysis may provide little benefit when testing for acute kidney injury. Only 1 patient showed elevated cardiac enzymes, which was determined to be a result of CKD.

Although not included in the final analysis, the 6 patients who were hospitalized for hypertensive urgency were similar in that they had neither symptoms at presentation to the ED nor laboratory findings indicating end-organ damage. Collectively, these findings support existing literature that questions the utility of laboratory testing of patients with AH in the ED.¹⁰

Patients also were treated with a variety of antihypertensive agents in the ED. One explanation might be outpatient nonadherence with medications. In patients with AH, it is common to provide doses of chronic medications that the patient might have missed and should be taking on a regular basis. Therefore, assessing adherence with current medications before modifying chronic therapy is an important initial step when managing AH.

Although oral agents primarily were used, IV antihypertensives were administered to about one-third of patients. Preference for IV administration in the ED might be related to its ability to lower BP quickly. The practice of obtaining IV access for medication in a patient with AH is costly, unnecessary, and potentially harmful.⁷ The authors theorize that this practice is performed, in many cases, as an attempt to expedite ED discharge after an acceptable BP reading is documented.

Rapid reductions in BP can precipitate hypoperfusion inadvertently and are more likely to occur with IV agents than with oral ones. Therefore, the safety, convenience, and cost savings associated with oral administration make it the preferred route for managing AH. 

Oral antihypertensives with desired therapeutic and pharmacokinetic properties are listed in Table 3. When used appropriately, these agents are well tolerated and effective and could be given in an ambulatory care clinic without the need for intensive monitoring.

 

Best Practices

Primary care clinics are best suited to manage AH because medication adjustments and long-term monitoring are easier to perform and at substantially lower costs when compared with that of the ED. Rather than immediately referring a patient to the ED, clinicians should consider factors that could elevate BP, such as medication nonadherence, anxiety, acute pain, recent tobacco or caffeine use, or white coat syndrome. Staff should be well educated on proper BP measurement and instructed to repeat the reading for confirmation. Before measuring BP, allow the patient to sit quietly for 5 minutes with the feet flat on the floor and arm supported.³ Ideally, the measurement used should be the average of 3 BP readings on an automated device.¹¹ If BP readings are high, staff should ask the patient about medication adherence and missed medication(s) should be administered.

It also is reasonable to have the patient rest quietly for up to 30 minutes because rest has been shown to reduce BP in some patients.¹² The drawback to the prolonged rest strategy is the potential to cause delays in care for other patients. However, it is important to remember that wait times in the ED often are measured in hours, which causes frustration for patients referred to the ED for AH management. Before completing the office visit, the provider should recheck BP using proper technique and confirm that the patient has antihypertensive medication(s) in his/her possession; a follow-up appointment should be scheduled for no later than 1 week.

Primary care providers might be concerned about taking on additional liability and could favor ED referral, but legislation makes it difficult for EDs to defer nonemergent issues to primary care clinics. The Emergency Medical Treatment and Labor Act states that hospitals are prohibited from denying a patient care during an emergency.¹³ Despite evidence that AH is not an emergency, many patients continue to be referred to the ED. One-third of patients presented to the ED on their own volition and more than one-half were referred by health care personnel. This strongly suggests that both patients and health care personnel consider AH an emergency medical condition requiring immediate attention. However, patients with AH rarely are found to have any acute end-organ damage; therefore, acute treatment and extensive laboratory or diagnostic testing in the ED provides little, if any, benefit.¹⁰ The authors believe the ACEP clinical policy should be adopted into mainstream practice to help reduce health care costs and preserve ED resources for patients with true emergencies.

Another pervasive issue that could contribute to inappropriate AH referrals to the ED is the shortage of PCPs and limited same-day appointments for nonemergent conditions. In a 2017 survey, the average wait time for a PCP appointment ranged between 12 and 109 days, depending on the metropolitan area. The national average wait time conducted by this survey was 29.3 days.¹⁴ When primary care appointments are unavailable, triage staff could recommend that patients seek care in the ED. Additionally, patients might choose to seek ED care rather than wait for the next available PCP appointment. Clinic proximity to an ED could influence referral rates. In other words, medical centers or health systems with primary care clinics and ED services under one roof could experience more frequent ED referrals.

A promising strategy to help overcome the challenges of addressing AH and avoiding ED referrals is increasing patient access to and use of qualified, nonphysician providers, such as clinical pharmacists and nurse practitioners. Large health systems such as the VA and Kaiser Permanente have employed clinical pharmacist providers to reduce follow-up times for patients in primary care settings.¹⁵ Furthermore, there is substantial evidence that supports the cost-effectiveness and clinical success of pharmacist-driven hypertension clinics.^16-18 Nurse-driven efforts to improve hypertension control have been successfully implemented in health systems.¹⁹ Both clinical pharmacist and nurse-managed hypertension clinics are effective solutions to manage patients with AH who might otherwise use costly ED services.For example, the average cost of a single ED visit is $740 to $3,437.²⁰ In comparison, a 2010 report from the Agency for Healthcare Research and Quality showed the average annual cost of managing hypertension in ambulatory care clinics was $442 per adult, a cost considerably lower than that of the ED.²¹

Limitations

The retrospective and observational design of this study are inherent limitations. This study was not designed to evaluate cardiovascular outcomes after ED encounters. The sample size could have been larger if patients with BP < 180/110 mm Hg at ED triage were included; however, the 180/110 mm Hg threshold was chosen because it was the most widely agreed on BP value in the literature. This study did not capture patients who presented with AH and did not receive any acute treatment in the ED.Prescribing patterns based on provider training (eg, emergency medicine, family medicine, or internal medicine) were not tracked and might have accounted for differences in selection of diagnostic tests, laboratory ordering, and route of drug administration preference.

 

A small subset of patients reported positive pain scores at triage but did not describe acute pain. Pain scores are highly subjective, and few primary literature sources link chronic pain with increased BP.^22,23 Nevertheless, patients who reported acute pain and elevated BP were excluded in order to identify truly asymptomatic patients. VA hospitals are unique health systems and data obtained from this study might not be applicable to other public or private facilities. Last, the study did not take into account patients’ psychosocial circumstances that might have fostered a disproportionate reliance on the ED for health care.

Conclusion

Asymptomatic patients with elevated BP are treated in the ED despite no evidence supporting improved outcomes after acute BP lowering in this population. Follow-up after ED encounters for AH did not occur consistently within guideline-recommended 7 days, a trend that also occurs in non-VA systems.⁸ Clinics and health care systems could establish policies to prevent or minimize management of AH in the ED. Ideally, AH should be managed in a clinic setting by a PCP, but growing clinician workload might lead to increasing wait times and difficultly obtaining same-day appointments. Nurse-led clinics and clinical pharmacists operating under a scope of practice and working closely with a PCP are a cost-effective solution to ensure timely treatment and appropriate follow-up of patients with uncontrolled hypertension.

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the North Florida South Georgia Veterans Health System in Gainesville, Florida.

Hypertension affects more than 65 million individuals in the U.S., accounting for nearly 30% of the adult population.¹ Less than 50% of those with hypertension are taking appropriate pharmacotherapy.² Hypertension contributes to cardiovascular events, including cerebrovascular accident, transient ischemic attack, hypertensive retinopathy, renal failure, myocardial infarction, and heart failure.¹ Chronic hypertension mainly is an asymptomatic condition, earning the nickname “the silent killer.”² An acute, symptomatic elevation in blood pressure (BP) often is referred to as hypertensive emergency. Symptoms of end-organ damage can include headache, blurry vision, chest pain, shortness of breath, altered mental status, epistaxis, and oliguria.² Although rare, hypertensive emergencies should be treated immediately. The Seventh Report of the Joint National Committee (JNC 7), and the more recent JNC 8, have published guidelines on managing chronic hypertension.^3,4 However, neither report provides guidance on hypertensive emergency or the appropriate actions in cases of extremely elevated BP in an asymptomatic patient.^3,4

Acute hypertensive episodes—often referred to as hypertensive crises—are responsible for nearly 8 million hospitalizations each year and 20 million visits to the emergency department (ED).^5,6 Most of these visits are same-day “treat-and-release” events.⁵ There is no universally accepted BP value associated with a hypertensive crisis, but most resources state that a BP ≥ 180/110 mm Hg requires attention.^2,7 Without other symptoms, elevated BP is not an emergency, yet ED referral for acute management is common.⁷

Three terms fall under the umbrella of hypertensive crises: hypertensive emergency, hypertensive urgency, and asymptomatic hypertension (AH).² In a 2007 article, the American College of Chest Physicians defined hypertensive emergency as BP ≥ 180/110 mm Hg with evidence of end-organ damage.² Symptoms are almost always present in true hypertensive emergencies, and immediate medical intervention is required to halt further organ damage. In the same article, hypertensive urgency is defined as BP ≥ 180/110 mm Hg without end-organ damage.² The definition of hypertensive urgency could be further refined to include the presence of cardiovascular and renal risk factors, although this additional point is not consistent across the literature. Asymptomatic hypertension is similar to hypertensive urgency; however, there is an absence of signs or symptoms of end-organ damage.² There is ambiguity in the literature concerning managing hypertensive urgency and AH, but both share a basic tenet: Immediate BP reduction is not essential. Gradual dosage adjustment(s) of oral medications, preferably by a primary care provider (PCP), and follow-up within 7 days are recommended.⁷

Limited evidence exists to guide ED providers in managing AH. Long-term outcomes and guidelines intended for the primary care setting should not be extrapolated to acute management in the ED. With limited treatment guidelines, providers might be more likely to refer patients with AH to the ED for evaluation. In 2013, the American College of Emergency Physicians (ACEP) created a clinical policy concerning AH in the ED. The ACEP concluded that screening for target organ injury and medical intervention in the ED does not reduce rates of adverse events (AEs) and could lead to overtreatment and acute hypoperfusion.⁷ More recently, Patel and colleagues published findings on hypertensive urgency in the ambulatory care setting, which similarly found that referral to the ED was associated with increased use of health care resources and no change in short-term major AEs.⁸ The ACEP recommends that patients presenting with AH be referred to primary care clinics where long-term monitoring and medication adjustments can be achieved more cost-effectively.⁷

The objective of this retrospective evaluation was to assess the incidence and management of AH within a VA ED. The authors aimed to provide insight into how these patients are managed and discuss alternatives to ED use.

Methods

This retrospective observational study was conducted within the North Florida/South Georgia Veterans Health System (NFSGVHS), which provides patient care at 2 medical centers in Gainesville and Lake City, Florida, as well as 11 outpatient clinics located throughout North Florida and South Georgia. The NFSGVHS serves rural and urban veteran populations. Study approval was granted by the NFSGVHS Institutional Review Board and Research and Development Committee.

Inclusion/Exclusion Criteria

Adult patients who were ordered at least 1 antihypertensive medication in the ED from July 1, 2011 to July 1, 2014, in addition to being asymptomatic with BP ≥ 180/110 mm Hg at ED triage were included. Based on clinical experience, the authors estimated that 3 years would provide a sample size of more than 100 patients. Patients were excluded if they presented with any acute symptoms or were hospitalized for further management.

Data Collection

Baseline demographics were collected for all participants. During the ED encounter, pre- and postintervention vital signs were recorded and prespecified laboratory data obtained. Interrater reliability was accounted for by performing random reviews of previously collected data to ensure consistency during the chart review process. Renal end-organ damage was defined using Acute Kidney Injury Network criteria, a serum creatinine 50% above baseline, or an absolute increase in baseline serum creatinine by 0.3 mg/dL.⁹ Additional laboratory markers of organ damage included cardiac troponin levels. Urinalysis results also were assessed to determine the presence of hematuria or proteinuria. Patient-reported nonadherence with medications was determined by reviewing ED provider and/or nurse documentation notes for the index ED encounter.

 

Investigators documented the route (IV or oral) and antihypertensive(s) medication selected for each patient. Adverse effects and any changes to patients’ outpatient medication regimens were noted. Investigators also assessed days to next medical contact after ED discharge to determine whether follow-up occurred according to the recommended standard of 7 days.⁹ Days to next medical contact was defined as any contact—in person or by telephone—that was documented in the electronic health record after the index ED visit.

Statistical Analysis

Descriptive statistics, including mean, median, and standard deviation, were used to analyze data.

Results

A total of 1,052 patients presented with BP ≥ 180/110 mm Hg and for whom antihypertensive medication was ordered but not necessarily given in the ED. Of the total, 724 patients were excluded because of hospital admission for other primary diagnoses; however, 6 of these patients were admitted for hypertensive urgency. The final analysis included 132 patients who presented with the primary condition of elevated BP without any accompanying symptoms. Among these patients, 2 had repeat ED visits for AH during the specified time frame.

Each ED visit was treated as a separate occurrence.

Most patients were male with an average age of 63 years and documented history of hypertension. Nearly all patients had established primary care within the NFSGVHS. The most common comorbidity was diabetes mellitus (36%), followed by coronary artery disease (27%) and chronic kidney disease (CKD) (21%) (Table 1). About one-third of patients presented to the ED on their own volition, and slightly more than half were referred to the ED by primary care or specialty clinics.

The average BP and heart rate at ED presentation was 199/112 mm Hg and 76 beats per minute, respectively.

In the ED, 130 patients received BP treatment (Table 2). Medication was ordered for 2 patients who did not receive treatment. In total, 12 different medication classes were used for treating patients with AH in the ED (Figure). 

Most were treated with at least 1 oral antihypertensive; clonidine was the most common (48% of orally administered doses). In this study, 13% of patients received IV-only intervention; most were treated with hydralazine. Among the patients in the study, 22% were treated with a combination of oral and IV antihypertensives. No immediate AEs were noted for medications administered in the ED; however, 1 patient returned to the ED with angioedema after initiating an angiotensin-converting enzyme inhibitor at an ED visit 5 days earlier.

Treatment in the ED resulted in an average BP and heart rate reduction of 27/20 mm Hg and 5 beats per minute, respectively. About 80% of patients had a basic metabolic panel drawn, and there were no instances of acute kidney injury. Of the patients in the study 38% had cardiac enzymes collected, and only 1 patient had a positive result, which was determined to be unrelated to acute coronary syndrome. Forty-one (31%) of patients had a urinalysis; 12 has positive results for hematuria, and 18 revealed proteinuria. Of note, the 6 patients who were hospitalized for hypertensive urgency had neither symptoms at presentation to the ED nor laboratory findings indicating end-organ damage. The reason these patients were admitted is unclear.

At discharge, ED providers made changes to 54% of patients’ outpatient antihypertensive regimens. These changes included adding a new medication (68%), increasing the dosage of an existing medication (24%), or multiple changes (8%). Refills were provided for 18% of prescriptions. Follow-up within 7 days from ED discharge was recorded for 34% of patients. One patient received follow-up outside the NFSGVHS and was not included in this analysis.

Discussion

The aim of this retrospective study was to determine the incidence of AH in a VA ED and describe how these patients were managed. Overall, the rate of patients presenting to the ED with AH during the study period was about 1 patient every 8 days or 45 patients per year. By comparison, more than 30,000 patients are seen at the NFSGVHS ED annually. Although AH seems to be an uncommon occurrence, study findings raise questions about the value of managing the condition in the ED.

This study found several management strategies as well as noteworthy trends. For example, laboratory tests were not ordered routinely for all patients, suggesting that some ED providers question their use for AH. There were no patients with acute elevations in serum creatinine that indicated acute kidney injury, and although hematuria and proteinuria were common findings, neither were specific for acute injury. However, there were findings typical of chronic hypertension, and urinalysis may provide little benefit when testing for acute kidney injury. Only 1 patient showed elevated cardiac enzymes, which was determined to be a result of CKD.

Although not included in the final analysis, the 6 patients who were hospitalized for hypertensive urgency were similar in that they had neither symptoms at presentation to the ED nor laboratory findings indicating end-organ damage. Collectively, these findings support existing literature that questions the utility of laboratory testing of patients with AH in the ED.¹⁰

Patients also were treated with a variety of antihypertensive agents in the ED. One explanation might be outpatient nonadherence with medications. In patients with AH, it is common to provide doses of chronic medications that the patient might have missed and should be taking on a regular basis. Therefore, assessing adherence with current medications before modifying chronic therapy is an important initial step when managing AH.

Although oral agents primarily were used, IV antihypertensives were administered to about one-third of patients. Preference for IV administration in the ED might be related to its ability to lower BP quickly. The practice of obtaining IV access for medication in a patient with AH is costly, unnecessary, and potentially harmful.⁷ The authors theorize that this practice is performed, in many cases, as an attempt to expedite ED discharge after an acceptable BP reading is documented.

Rapid reductions in BP can precipitate hypoperfusion inadvertently and are more likely to occur with IV agents than with oral ones. Therefore, the safety, convenience, and cost savings associated with oral administration make it the preferred route for managing AH. 

Oral antihypertensives with desired therapeutic and pharmacokinetic properties are listed in Table 3. When used appropriately, these agents are well tolerated and effective and could be given in an ambulatory care clinic without the need for intensive monitoring.

 

Best Practices

Primary care clinics are best suited to manage AH because medication adjustments and long-term monitoring are easier to perform and at substantially lower costs when compared with that of the ED. Rather than immediately referring a patient to the ED, clinicians should consider factors that could elevate BP, such as medication nonadherence, anxiety, acute pain, recent tobacco or caffeine use, or white coat syndrome. Staff should be well educated on proper BP measurement and instructed to repeat the reading for confirmation. Before measuring BP, allow the patient to sit quietly for 5 minutes with the feet flat on the floor and arm supported.³ Ideally, the measurement used should be the average of 3 BP readings on an automated device.¹¹ If BP readings are high, staff should ask the patient about medication adherence and missed medication(s) should be administered.

It also is reasonable to have the patient rest quietly for up to 30 minutes because rest has been shown to reduce BP in some patients.¹² The drawback to the prolonged rest strategy is the potential to cause delays in care for other patients. However, it is important to remember that wait times in the ED often are measured in hours, which causes frustration for patients referred to the ED for AH management. Before completing the office visit, the provider should recheck BP using proper technique and confirm that the patient has antihypertensive medication(s) in his/her possession; a follow-up appointment should be scheduled for no later than 1 week.

Primary care providers might be concerned about taking on additional liability and could favor ED referral, but legislation makes it difficult for EDs to defer nonemergent issues to primary care clinics. The Emergency Medical Treatment and Labor Act states that hospitals are prohibited from denying a patient care during an emergency.¹³ Despite evidence that AH is not an emergency, many patients continue to be referred to the ED. One-third of patients presented to the ED on their own volition and more than one-half were referred by health care personnel. This strongly suggests that both patients and health care personnel consider AH an emergency medical condition requiring immediate attention. However, patients with AH rarely are found to have any acute end-organ damage; therefore, acute treatment and extensive laboratory or diagnostic testing in the ED provides little, if any, benefit.¹⁰ The authors believe the ACEP clinical policy should be adopted into mainstream practice to help reduce health care costs and preserve ED resources for patients with true emergencies.

Another pervasive issue that could contribute to inappropriate AH referrals to the ED is the shortage of PCPs and limited same-day appointments for nonemergent conditions. In a 2017 survey, the average wait time for a PCP appointment ranged between 12 and 109 days, depending on the metropolitan area. The national average wait time conducted by this survey was 29.3 days.¹⁴ When primary care appointments are unavailable, triage staff could recommend that patients seek care in the ED. Additionally, patients might choose to seek ED care rather than wait for the next available PCP appointment. Clinic proximity to an ED could influence referral rates. In other words, medical centers or health systems with primary care clinics and ED services under one roof could experience more frequent ED referrals.

A promising strategy to help overcome the challenges of addressing AH and avoiding ED referrals is increasing patient access to and use of qualified, nonphysician providers, such as clinical pharmacists and nurse practitioners. Large health systems such as the VA and Kaiser Permanente have employed clinical pharmacist providers to reduce follow-up times for patients in primary care settings.¹⁵ Furthermore, there is substantial evidence that supports the cost-effectiveness and clinical success of pharmacist-driven hypertension clinics.^16-18 Nurse-driven efforts to improve hypertension control have been successfully implemented in health systems.¹⁹ Both clinical pharmacist and nurse-managed hypertension clinics are effective solutions to manage patients with AH who might otherwise use costly ED services.For example, the average cost of a single ED visit is $740 to $3,437.²⁰ In comparison, a 2010 report from the Agency for Healthcare Research and Quality showed the average annual cost of managing hypertension in ambulatory care clinics was $442 per adult, a cost considerably lower than that of the ED.²¹

Limitations

The retrospective and observational design of this study are inherent limitations. This study was not designed to evaluate cardiovascular outcomes after ED encounters. The sample size could have been larger if patients with BP < 180/110 mm Hg at ED triage were included; however, the 180/110 mm Hg threshold was chosen because it was the most widely agreed on BP value in the literature. This study did not capture patients who presented with AH and did not receive any acute treatment in the ED.Prescribing patterns based on provider training (eg, emergency medicine, family medicine, or internal medicine) were not tracked and might have accounted for differences in selection of diagnostic tests, laboratory ordering, and route of drug administration preference.

 

A small subset of patients reported positive pain scores at triage but did not describe acute pain. Pain scores are highly subjective, and few primary literature sources link chronic pain with increased BP.^22,23 Nevertheless, patients who reported acute pain and elevated BP were excluded in order to identify truly asymptomatic patients. VA hospitals are unique health systems and data obtained from this study might not be applicable to other public or private facilities. Last, the study did not take into account patients’ psychosocial circumstances that might have fostered a disproportionate reliance on the ED for health care.

Conclusion

Asymptomatic patients with elevated BP are treated in the ED despite no evidence supporting improved outcomes after acute BP lowering in this population. Follow-up after ED encounters for AH did not occur consistently within guideline-recommended 7 days, a trend that also occurs in non-VA systems.⁸ Clinics and health care systems could establish policies to prevent or minimize management of AH in the ED. Ideally, AH should be managed in a clinic setting by a PCP, but growing clinician workload might lead to increasing wait times and difficultly obtaining same-day appointments. Nurse-led clinics and clinical pharmacists operating under a scope of practice and working closely with a PCP are a cost-effective solution to ensure timely treatment and appropriate follow-up of patients with uncontrolled hypertension.

Acknowledgments
This material is the result of work supported with resources and the use of facilities at the North Florida South Georgia Veterans Health System in Gainesville, Florida.

From the Northwestern University, Feinberg School of Medicine, Chicago, IL (Mr. Bavishi, Dr. Lavin), the Johns Hopkins University, Baltimore, MD (Dr. Boss), Children’s National Medical Center, Washington, DC (Dr. Shah), and Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL (Dr. Lavin).

 

Abstract

• Background: Endoscopic management of pediatric subglottic stenosis is common; however, no multiinstitutional studies have assessed its perioperative outcomes. The American College of Surgeon’s National Surgical Quality Improvement Program – Pediatric (ACS-NSQIP-P) represents a source of such data.
• Objective: To investigate 30-day outcomes of endoscopic dilation of the pediatric airway and to compare these outcomes to those seen with open reconstruction techniques.
• Methods: Current procedural terminology (CPT) codes were queried for endoscopic or open airway reconstruction in the 2015 ACS-NSQIP-P Public Use File (PUF). Demo­graphics and 30-day events were abstracted to compare open to endoscopic techniques and to assess for risk factors for varied outcomes after endoscopic dilation. Outcome measures included length of stay (LOS), 30-day rates of reintubation, readmission, and reoperation.
• Results: 171 endoscopic and 116 open procedures were identified. Mean age at endoscopic and open procedures was 4.1 (SEM = 0.37) and 5.4 years (SEM = 0.40). Mean LOS was shorter after endoscopic procedures (5.5 days, SEM = 1.13 vs. 11.3 days SEM = 1.01, P < 0.001). Open procedures had higher rates of reintubation (OR = 7.41, P = 0.026) and reoperation (OR = 3.09, P = 0.009). In patients undergoing endoscopic dilation, children < 1 year were more likely to require readmission (OR = 4.21, P = 0.03) and reoperation (OR = 4.39, P = 0.03) when compared with older children.
• Conclusion: Open airway reconstruction is associated with longer LOS and increased reintubations and reoperations, suggesting a possible opportunity to improve value in health care in the appropriately selected patient. Reoperations and readmissions following endoscopic dilation are more prevalent in children younger than 1 year.

Keywords: airway stenosis; subglottic stenosis; endoscopic dilation; pediatrics; outcomes.

 

Historically, pediatric laryngotracheal stenosis was managed using open reconstruction techniques, including laryngoplasty, tracheal resection, and cervical tracheoplasty. Initial reports of endoscopic dilation were described in the 1980s as a means to salvage re-stenosis after open reconstruction [1]. Currently, primary endoscopic dilation has become commonplace in otolaryngology due to its less invasive nature as well as—in cases of balloon dilation—minimization of tissue damage [2]. The advancements made in endoscopic balloon dilation have reduced the frequency with which open reconstruction is performed.

Systematic reviews and case series investigating endoscopic dilation indicate a 70% to 80% success rate in preventing future open surgery or tracheostomy [2–5]. While increased severity of stenosis has been associated with poorer outcomes in endoscopic procedures, few other risk factors that influence surgical success have been identified [4,5]. In a single study in the adult literature, open surgical management of idiopathic subglottic stenosis was associated with improved outcomes when compared to endoscopic techniques [5]. Such findings suggest a need to identify these factors for the purpose of optimizing clinical decision-making.

As laryngotracheal stenosis is rare, postoperative outcomes and risk factors are best identified on a multi­institutional level. Due to its participation from 80 hospitals and its accurate and reliable reporting of both demographic and risk-stratified 30-day outcomes data, the American College of Surgeon’s National Surgical Quality Improvement Program – Pediatric (ACS NSQIP-P) provides such a platform [6–8]. Thirty-day outcomes and risk factors for open reconstruction utilizing the ACS NSQIP-P database have previously been reported; however, no such outcomes for endoscopic dilation have been described, and no comparison between endoscopic and open procedures has been made [9]. The purpose of this study was to utilize the 2015 ACS-NSQIP-P database to investigate 30-day outcomes of endoscopic dilation of the pediatric airway and to compare these outcomes to open reconstruction techniques. Secondarily, we aimed to determine if any demographic factors or medical comorbidities are associated with varied outcomes in endoscopic reconstruction. While these data reflect safety and quality of this procedure in the United States, findings may potentially be applied across international settings.

Methods

Data Source

Data was obtained from the 2015 ACS-NSQIP-P Public Use File (PUF). Due to the de-identified and public nature of these data, this research was exempt from review by the Ann & Robert H. Lurie Children’s Hospital of Chicago review board. Data collection methods for ACS-NSQIP-P have previously been described [10]. In brief, data was collected from 80 hospitals on approximately 120 preoperative, intraoperative, and postoperative variables. Cases are systematically sampled on an 8-day cycle basis, where the first 35 cases meeting the inclusion criteria in each hospital in each cycle are submitted to ACS-NSQIP-P.

Variables and Outcomes

Airway procedures for endoscopic dilations and open reconstructions were obtained by CPT code. Endoscopic dilations (CPT 31528) were compared to open reconstructions, which included laryngoplasty (31580, 31582), cervical tracheoplasty (31750), cricoid split (31587), and tracheal resection (31780). Demographic variables included age, sex, race, and history of prematurity. Presence of specific comorbid diseases were also collected and tested for significance.

Dependent outcomes of interest were unplanned 30-day postoperative events grouped as reoperation, unplanned readmission, and postoperative reintubation. In the case of endoscopic procedures, the presence of salvage open reconstruction or tracheostomy within 30 days of surgery was also recorded. Length of stay (LOS) after the procedure was collected. Specific postoperative complications and reasons for readmission were recorded within the limitations of data available in the PUF.

Analysis

Analysis was performed using descriptive statistics and frequency analysis where appropriate. Chi-square analysis was used to compare adverse events between open and endoscopic procedures. Logistic regression with calculation of odds ratio (OR) was performed to determine predictive factors for reoperation, readmission, and reintubation in all pediatric airway reconstructive procedures in adjusted and unadjusted models. T-test and linear regression was performed on the continuous outcome of length of stay. For all analyses, a p value of < 0.05 was considered statistically significant. All variable recoding and statistical analyses were performed in SAS/STAT software (Cary, NC).

Results

A total of 84,056 pediatric procedures were extracted from the 2015 NSQIP-P PUFs. Using the above CPT codes, 171 endoscopic dilations and 116 open airway reconstructions were identified, with patient age ranging from 0 days to 17.6 years. Average age of patients undergoing endoscopic dilation and open reconstruction was 4.1 and 5.4 years, respectively (Table 1).

Potential confounders were tested with univariate logistic regression to determine if they had a significant impact on readmission, reintubation, or reoperation rates. These variables (Table 2) 

included age, tracheostomy status, prematurity, sex, race, congenital malformations, prior cardiac surgery, underlying pulmonary disease, immune disease, brain disease, gastrointestinal disease, cardiac disease. Only age was significantly associated with reoperation rates. These variables were also tested against length of stay using univariate linear regression: age, pulmonary disease and tracheostomy were significantly associated with change in length of stay. All models were adjusted for these 3 variables accordingly.

In patients undergoing endoscopic dilation, average length of stay was 5.5 days (SEM = 1.13), with 79 (48.5%) patients having a length of stay of zero days. Of all patients who had endoscopic dilations, 70 (40.1%) had a pre-existing tracheostomy and these accounted for the majority (73%) of patients who had zero days as their LOS. LOS after endoscopic management was significantly shorter than the mean of 11.3 days (SEM = 1.01) reported in those undergoing open reconstruction (P < 0.001).

With respect to 30-day adverse events, 2 patients in the endoscopic group (1.1%) required reintubation. Thirteen endoscopic dilation cases (7.6%) had an unplanned readmission, four (2.3%) of which were associated with reoperation within 30 days of the primary surgical procedure. There were 9 other reoperations unassociated with unplanned readmission. Three of these reoperations were due to failed endoscopic dilations, resulting in 2 tracheostomies and one open airway reconstruction. There was one patient death, in a 0-day old with tetralogy of Fallot, trachea-esophageal fistula, and ventilator dependence who underwent emergent endoscopic dilation and died the same day.

Open procedures were associated with 11 unplanned readmissions (9.5%), 7 re-intubations (6%) and 18 reoperations (15.5%). Of patents undergoing reoperation, one patient undergoing open reconstruction underwent tracheostomy within 30 days of surgery.

When comparing open reconstruction to endoscopic dilation, there was a significant increase in reintubation (OR = 7.41, P = 0.026) and reoperation (OR = 3.09, P = 0.009) for open procedures, even with adjustment for age, tracheostomy status, and pulmonary disease. There was no significant difference between the two for unplanned readmissions (OR = 1.19, P = 0.79) (Figure).

Younger age was also found to be significantly associated with reoperation rates, in an adjusted logistic model that accounted for tracheostomy status, type of surgery, and pulmonary disease. Per year of life, younger children had higher reoperation rates than older children (OR = 1.91, P = 0.017). When endoscopic dilation was individually examined, children younger than 1 year of age were more likely to undergo reoperation after an endoscopic dilation than children older than 1 (OR = 4.39, P = 0.03). Children under age 1 were also more likely to have an unplanned readmission after an endoscopic dilation (OR = 4.21, P = 0.03). The relationship between age and re-intubation was not significant (OR = 0, P = 0.95). For open reconstruction, this age dichotomization was not associated with any increased reoperation (OR = 2.3, P = 0.52), readmission (OR = 0, P = 0.97), or reintubation (OR = 0, P = 0.94).

T-test analysis was performed to determine if children < 1 year old also had significantly longer hospital stays after endoscopic dilation than older children (mean 14.1 days vs 1.9 days, P < 0.001). This relationship held true in a linear regression after adjustment for pulmonary disease and tracheostomy, with length of stay decreasing by 0.48 days per year of life (P = 0.03). For endoscopic dilations, the same relationship held true, where length of stay decreased by 0.75 days per year of life.

 

Discussion

Endoscopic dilation for primary management of pediatric laryngotracheal stenosis has become commonplace. Despite this, outcomes of this procedure have only been described in case series and meta-analyses [2–5]. The relative rarity of pediatric laryngotracheal stenosis suggests the need for large, multi-institutional data for purposes of patient selection and medical decision-making.

This study utilized the ACS-NSQIP-Pediatric database to highlight 30-day outcomes of endoscopic dilation and to compare these outcomes to those of open airway reconstruction procedures. The ACS-NSQIP database has been endorsed by multiple organizations, including the Center for Medicare and Medicaid Services, the Joint Commission, and the American Medical Association. It has been shown to have higher sensitivity and a lower false-positive rate when compared to administrative data, in part due to data collection from trained personnel [11]. Furthermore, ACS-NSQIP use has the additional benefit of reporting an unplanned admission—a feature unavailable in review of claims data [12].

With respect to adverse events, our study demonstrates that endoscopic dilation is associated with an equally high rate of unplanned readmission when compared to open reconstruction. The high prevalence of comorbid disease such as chronic lung disease (32% of endoscopic dilation and 43% of open reconstruction) can account for some of the morbidity associated with any airway procedures.

Despite high rates of unplanned readmission, patients undergoing endoscopic dilation were less likely to have reoperations within 30 days of initial surgery when compared to those undergoing open reconstruction. While differences in disease severity may be partially responsible for this difference in the reoperation rate, this finding is notable given the health care costs associated with multiple operations as well as safety concerns with multiple anesthetics in the very young [13,14].

The ACS-NSQIP platform does not distinguish unplanned from planned reoperations. In the setting of airway surgery, where multiple planned reoperations are commonplace, this metric is a suboptimal stand-alone indicator of adverse outcomes. Other markers available in the database—such as reintubations and performance of tracheostomy or open airway reconstruction within 30-days of surgery—are more indicative of surgical outcome in the setting of airway surgery. We found that both reintubations and salvage open reconstruction within 30-days were rare. It should be noted that the ACS-NSQIP data does not report any events occurring outside of the 30-day postoperative period, representing potential limitation of the use of this database. As was previously advocated by Roxbury and colleagues, procedure/subspecialty specific outcome data collection would also improve outcome analysis of airway and other otolaryngologic procedures [9]. In the setting of airway reconstruction, this would include data pertaining to Cotton-Meyer grading systems well as postoperative voice and swallow outcomes.

In addition to safety profile, endoscopic procedures were associated with shorter LOS when compared with open reconstruction, representing another potential source of cost savings with this less invasive method. This is especially significant given that open reconstruction patients spend much of their inpatient stay in an ICU setting. In patients who are candidates for endoscopic procedures, this lower-risk, lower-cost profile of endoscopic dilation has the opportunity to improve value in health care and may be the source of future improvement initiatives.

In addition to comparing overall outcomes between endoscopic and open management of laryngotracheal stenosis, our study aimed to identify factors that were associated with varied outcomes in patients undergoing primary endoscopic dilation. We found that children younger than 1 year of age were 5.8 times more likely to undergo an unplanned reoperation after an endoscopic dilation than children over 1 year. A similar finding was reported in open airway surgeries, with increased reoperation rates in children < 3 years old [9]. The justification of a dichotomization at 1 year was made as expert opinion recognizes that the infant airway is less forgiving to intervention given its small size. Young age was also a factor in prolonged LOS as was determined by linear regression. It is likely that this increased LOS may be in part due to associations of young age and the neonatal ICU population. One must balance the increased risk of surgery in the young with that of tracheostomy, which has a published complication rate of 18% to 50% and direct mortality rate of 1% to 2% in the pediatric population [15–18]. Understanding these relative risks may help guide the airway surgeon in preoperative counseling with families and medical decision-making.

As discussed above, the limitation of data to a 30-day period is a relative weakness of ACS-NSQIP database use for studies of airway reconstruction, as the ultimate outcome—a stable, decannulated airway—may occur outside of this time period. As many quality metrics utilize data from the 30-day postoperative period, knowledge of these outcomes remains valuable in surgical decision-making. Ultimately, collection of data in a large, long-term dataset would allow broader generalizations to be made about the differences between open and endoscopic procedures and would also give a more comprehensive picture of the outcomes of endoscopic dilation.

In conclusion, this study is the first to analyze 30-day postoperative outcomes in pediatric endoscopic airway dilations using data aggregated by ACS-NSQIP from institutions across the United States. This data indicates that endoscopic airway dilation is a relatively safe procedure, especially compared with open reconstruction; however, additional data on disease severity and other outcomes is necessary to draw final conclusions of superiority of technique. Future improvement initiatives could be aimed at the impact of this lower-risk, lower-cost procedure in the appropriately selected patient. Outcomes of endoscopic dilation are poorer in those less than 1 year of age, as they are associated with increased reoperation rates and increased length of stay compared to older children. One must balance these risks in the very young with the risks associated with tracheostomy and other alternative airway management modalities.

 

Note: This work was presented in a paper at the AAO-HNS 2017 meeting, Chicago, IL, 10 Sep 2017.

Corresponding author: Jennifer Lavin, MD, MS, 225 E Chicago Ave., Box 25, Chicago, IL 60611, [email protected].

Financial disclosures: None.

From the Northwestern University, Feinberg School of Medicine, Chicago, IL (Mr. Bavishi, Dr. Lavin), the Johns Hopkins University, Baltimore, MD (Dr. Boss), Children’s National Medical Center, Washington, DC (Dr. Shah), and Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL (Dr. Lavin).

 

Abstract

• Background: Endoscopic management of pediatric subglottic stenosis is common; however, no multiinstitutional studies have assessed its perioperative outcomes. The American College of Surgeon’s National Surgical Quality Improvement Program – Pediatric (ACS-NSQIP-P) represents a source of such data.
• Objective: To investigate 30-day outcomes of endoscopic dilation of the pediatric airway and to compare these outcomes to those seen with open reconstruction techniques.
• Methods: Current procedural terminology (CPT) codes were queried for endoscopic or open airway reconstruction in the 2015 ACS-NSQIP-P Public Use File (PUF). Demo­graphics and 30-day events were abstracted to compare open to endoscopic techniques and to assess for risk factors for varied outcomes after endoscopic dilation. Outcome measures included length of stay (LOS), 30-day rates of reintubation, readmission, and reoperation.
• Results: 171 endoscopic and 116 open procedures were identified. Mean age at endoscopic and open procedures was 4.1 (SEM = 0.37) and 5.4 years (SEM = 0.40). Mean LOS was shorter after endoscopic procedures (5.5 days, SEM = 1.13 vs. 11.3 days SEM = 1.01, P < 0.001). Open procedures had higher rates of reintubation (OR = 7.41, P = 0.026) and reoperation (OR = 3.09, P = 0.009). In patients undergoing endoscopic dilation, children < 1 year were more likely to require readmission (OR = 4.21, P = 0.03) and reoperation (OR = 4.39, P = 0.03) when compared with older children.
• Conclusion: Open airway reconstruction is associated with longer LOS and increased reintubations and reoperations, suggesting a possible opportunity to improve value in health care in the appropriately selected patient. Reoperations and readmissions following endoscopic dilation are more prevalent in children younger than 1 year.

Keywords: airway stenosis; subglottic stenosis; endoscopic dilation; pediatrics; outcomes.

 

Historically, pediatric laryngotracheal stenosis was managed using open reconstruction techniques, including laryngoplasty, tracheal resection, and cervical tracheoplasty. Initial reports of endoscopic dilation were described in the 1980s as a means to salvage re-stenosis after open reconstruction [1]. Currently, primary endoscopic dilation has become commonplace in otolaryngology due to its less invasive nature as well as—in cases of balloon dilation—minimization of tissue damage [2]. The advancements made in endoscopic balloon dilation have reduced the frequency with which open reconstruction is performed.

Systematic reviews and case series investigating endoscopic dilation indicate a 70% to 80% success rate in preventing future open surgery or tracheostomy [2–5]. While increased severity of stenosis has been associated with poorer outcomes in endoscopic procedures, few other risk factors that influence surgical success have been identified [4,5]. In a single study in the adult literature, open surgical management of idiopathic subglottic stenosis was associated with improved outcomes when compared to endoscopic techniques [5]. Such findings suggest a need to identify these factors for the purpose of optimizing clinical decision-making.

As laryngotracheal stenosis is rare, postoperative outcomes and risk factors are best identified on a multi­institutional level. Due to its participation from 80 hospitals and its accurate and reliable reporting of both demographic and risk-stratified 30-day outcomes data, the American College of Surgeon’s National Surgical Quality Improvement Program – Pediatric (ACS NSQIP-P) provides such a platform [6–8]. Thirty-day outcomes and risk factors for open reconstruction utilizing the ACS NSQIP-P database have previously been reported; however, no such outcomes for endoscopic dilation have been described, and no comparison between endoscopic and open procedures has been made [9]. The purpose of this study was to utilize the 2015 ACS-NSQIP-P database to investigate 30-day outcomes of endoscopic dilation of the pediatric airway and to compare these outcomes to open reconstruction techniques. Secondarily, we aimed to determine if any demographic factors or medical comorbidities are associated with varied outcomes in endoscopic reconstruction. While these data reflect safety and quality of this procedure in the United States, findings may potentially be applied across international settings.

Methods

Data Source

Data was obtained from the 2015 ACS-NSQIP-P Public Use File (PUF). Due to the de-identified and public nature of these data, this research was exempt from review by the Ann & Robert H. Lurie Children’s Hospital of Chicago review board. Data collection methods for ACS-NSQIP-P have previously been described [10]. In brief, data was collected from 80 hospitals on approximately 120 preoperative, intraoperative, and postoperative variables. Cases are systematically sampled on an 8-day cycle basis, where the first 35 cases meeting the inclusion criteria in each hospital in each cycle are submitted to ACS-NSQIP-P.

Variables and Outcomes

Airway procedures for endoscopic dilations and open reconstructions were obtained by CPT code. Endoscopic dilations (CPT 31528) were compared to open reconstructions, which included laryngoplasty (31580, 31582), cervical tracheoplasty (31750), cricoid split (31587), and tracheal resection (31780). Demographic variables included age, sex, race, and history of prematurity. Presence of specific comorbid diseases were also collected and tested for significance.

Dependent outcomes of interest were unplanned 30-day postoperative events grouped as reoperation, unplanned readmission, and postoperative reintubation. In the case of endoscopic procedures, the presence of salvage open reconstruction or tracheostomy within 30 days of surgery was also recorded. Length of stay (LOS) after the procedure was collected. Specific postoperative complications and reasons for readmission were recorded within the limitations of data available in the PUF.

Analysis

Analysis was performed using descriptive statistics and frequency analysis where appropriate. Chi-square analysis was used to compare adverse events between open and endoscopic procedures. Logistic regression with calculation of odds ratio (OR) was performed to determine predictive factors for reoperation, readmission, and reintubation in all pediatric airway reconstructive procedures in adjusted and unadjusted models. T-test and linear regression was performed on the continuous outcome of length of stay. For all analyses, a p value of < 0.05 was considered statistically significant. All variable recoding and statistical analyses were performed in SAS/STAT software (Cary, NC).

Results

A total of 84,056 pediatric procedures were extracted from the 2015 NSQIP-P PUFs. Using the above CPT codes, 171 endoscopic dilations and 116 open airway reconstructions were identified, with patient age ranging from 0 days to 17.6 years. Average age of patients undergoing endoscopic dilation and open reconstruction was 4.1 and 5.4 years, respectively (Table 1).

Potential confounders were tested with univariate logistic regression to determine if they had a significant impact on readmission, reintubation, or reoperation rates. These variables (Table 2) 

included age, tracheostomy status, prematurity, sex, race, congenital malformations, prior cardiac surgery, underlying pulmonary disease, immune disease, brain disease, gastrointestinal disease, cardiac disease. Only age was significantly associated with reoperation rates. These variables were also tested against length of stay using univariate linear regression: age, pulmonary disease and tracheostomy were significantly associated with change in length of stay. All models were adjusted for these 3 variables accordingly.

In patients undergoing endoscopic dilation, average length of stay was 5.5 days (SEM = 1.13), with 79 (48.5%) patients having a length of stay of zero days. Of all patients who had endoscopic dilations, 70 (40.1%) had a pre-existing tracheostomy and these accounted for the majority (73%) of patients who had zero days as their LOS. LOS after endoscopic management was significantly shorter than the mean of 11.3 days (SEM = 1.01) reported in those undergoing open reconstruction (P < 0.001).

With respect to 30-day adverse events, 2 patients in the endoscopic group (1.1%) required reintubation. Thirteen endoscopic dilation cases (7.6%) had an unplanned readmission, four (2.3%) of which were associated with reoperation within 30 days of the primary surgical procedure. There were 9 other reoperations unassociated with unplanned readmission. Three of these reoperations were due to failed endoscopic dilations, resulting in 2 tracheostomies and one open airway reconstruction. There was one patient death, in a 0-day old with tetralogy of Fallot, trachea-esophageal fistula, and ventilator dependence who underwent emergent endoscopic dilation and died the same day.

Open procedures were associated with 11 unplanned readmissions (9.5%), 7 re-intubations (6%) and 18 reoperations (15.5%). Of patents undergoing reoperation, one patient undergoing open reconstruction underwent tracheostomy within 30 days of surgery.

When comparing open reconstruction to endoscopic dilation, there was a significant increase in reintubation (OR = 7.41, P = 0.026) and reoperation (OR = 3.09, P = 0.009) for open procedures, even with adjustment for age, tracheostomy status, and pulmonary disease. There was no significant difference between the two for unplanned readmissions (OR = 1.19, P = 0.79) (Figure).

Younger age was also found to be significantly associated with reoperation rates, in an adjusted logistic model that accounted for tracheostomy status, type of surgery, and pulmonary disease. Per year of life, younger children had higher reoperation rates than older children (OR = 1.91, P = 0.017). When endoscopic dilation was individually examined, children younger than 1 year of age were more likely to undergo reoperation after an endoscopic dilation than children older than 1 (OR = 4.39, P = 0.03). Children under age 1 were also more likely to have an unplanned readmission after an endoscopic dilation (OR = 4.21, P = 0.03). The relationship between age and re-intubation was not significant (OR = 0, P = 0.95). For open reconstruction, this age dichotomization was not associated with any increased reoperation (OR = 2.3, P = 0.52), readmission (OR = 0, P = 0.97), or reintubation (OR = 0, P = 0.94).

T-test analysis was performed to determine if children < 1 year old also had significantly longer hospital stays after endoscopic dilation than older children (mean 14.1 days vs 1.9 days, P < 0.001). This relationship held true in a linear regression after adjustment for pulmonary disease and tracheostomy, with length of stay decreasing by 0.48 days per year of life (P = 0.03). For endoscopic dilations, the same relationship held true, where length of stay decreased by 0.75 days per year of life.

 

Discussion

Endoscopic dilation for primary management of pediatric laryngotracheal stenosis has become commonplace. Despite this, outcomes of this procedure have only been described in case series and meta-analyses [2–5]. The relative rarity of pediatric laryngotracheal stenosis suggests the need for large, multi-institutional data for purposes of patient selection and medical decision-making.

This study utilized the ACS-NSQIP-Pediatric database to highlight 30-day outcomes of endoscopic dilation and to compare these outcomes to those of open airway reconstruction procedures. The ACS-NSQIP database has been endorsed by multiple organizations, including the Center for Medicare and Medicaid Services, the Joint Commission, and the American Medical Association. It has been shown to have higher sensitivity and a lower false-positive rate when compared to administrative data, in part due to data collection from trained personnel [11]. Furthermore, ACS-NSQIP use has the additional benefit of reporting an unplanned admission—a feature unavailable in review of claims data [12].

With respect to adverse events, our study demonstrates that endoscopic dilation is associated with an equally high rate of unplanned readmission when compared to open reconstruction. The high prevalence of comorbid disease such as chronic lung disease (32% of endoscopic dilation and 43% of open reconstruction) can account for some of the morbidity associated with any airway procedures.

Despite high rates of unplanned readmission, patients undergoing endoscopic dilation were less likely to have reoperations within 30 days of initial surgery when compared to those undergoing open reconstruction. While differences in disease severity may be partially responsible for this difference in the reoperation rate, this finding is notable given the health care costs associated with multiple operations as well as safety concerns with multiple anesthetics in the very young [13,14].

The ACS-NSQIP platform does not distinguish unplanned from planned reoperations. In the setting of airway surgery, where multiple planned reoperations are commonplace, this metric is a suboptimal stand-alone indicator of adverse outcomes. Other markers available in the database—such as reintubations and performance of tracheostomy or open airway reconstruction within 30-days of surgery—are more indicative of surgical outcome in the setting of airway surgery. We found that both reintubations and salvage open reconstruction within 30-days were rare. It should be noted that the ACS-NSQIP data does not report any events occurring outside of the 30-day postoperative period, representing potential limitation of the use of this database. As was previously advocated by Roxbury and colleagues, procedure/subspecialty specific outcome data collection would also improve outcome analysis of airway and other otolaryngologic procedures [9]. In the setting of airway reconstruction, this would include data pertaining to Cotton-Meyer grading systems well as postoperative voice and swallow outcomes.

In addition to safety profile, endoscopic procedures were associated with shorter LOS when compared with open reconstruction, representing another potential source of cost savings with this less invasive method. This is especially significant given that open reconstruction patients spend much of their inpatient stay in an ICU setting. In patients who are candidates for endoscopic procedures, this lower-risk, lower-cost profile of endoscopic dilation has the opportunity to improve value in health care and may be the source of future improvement initiatives.

In addition to comparing overall outcomes between endoscopic and open management of laryngotracheal stenosis, our study aimed to identify factors that were associated with varied outcomes in patients undergoing primary endoscopic dilation. We found that children younger than 1 year of age were 5.8 times more likely to undergo an unplanned reoperation after an endoscopic dilation than children over 1 year. A similar finding was reported in open airway surgeries, with increased reoperation rates in children < 3 years old [9]. The justification of a dichotomization at 1 year was made as expert opinion recognizes that the infant airway is less forgiving to intervention given its small size. Young age was also a factor in prolonged LOS as was determined by linear regression. It is likely that this increased LOS may be in part due to associations of young age and the neonatal ICU population. One must balance the increased risk of surgery in the young with that of tracheostomy, which has a published complication rate of 18% to 50% and direct mortality rate of 1% to 2% in the pediatric population [15–18]. Understanding these relative risks may help guide the airway surgeon in preoperative counseling with families and medical decision-making.

As discussed above, the limitation of data to a 30-day period is a relative weakness of ACS-NSQIP database use for studies of airway reconstruction, as the ultimate outcome—a stable, decannulated airway—may occur outside of this time period. As many quality metrics utilize data from the 30-day postoperative period, knowledge of these outcomes remains valuable in surgical decision-making. Ultimately, collection of data in a large, long-term dataset would allow broader generalizations to be made about the differences between open and endoscopic procedures and would also give a more comprehensive picture of the outcomes of endoscopic dilation.

In conclusion, this study is the first to analyze 30-day postoperative outcomes in pediatric endoscopic airway dilations using data aggregated by ACS-NSQIP from institutions across the United States. This data indicates that endoscopic airway dilation is a relatively safe procedure, especially compared with open reconstruction; however, additional data on disease severity and other outcomes is necessary to draw final conclusions of superiority of technique. Future improvement initiatives could be aimed at the impact of this lower-risk, lower-cost procedure in the appropriately selected patient. Outcomes of endoscopic dilation are poorer in those less than 1 year of age, as they are associated with increased reoperation rates and increased length of stay compared to older children. One must balance these risks in the very young with the risks associated with tracheostomy and other alternative airway management modalities.

 

Note: This work was presented in a paper at the AAO-HNS 2017 meeting, Chicago, IL, 10 Sep 2017.

Corresponding author: Jennifer Lavin, MD, MS, 225 E Chicago Ave., Box 25, Chicago, IL 60611, [email protected].

Financial disclosures: None.

From the Northwestern University, Feinberg School of Medicine, Chicago, IL (Mr. Bavishi, Dr. Lavin), the Johns Hopkins University, Baltimore, MD (Dr. Boss), Children’s National Medical Center, Washington, DC (Dr. Shah), and Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL (Dr. Lavin).

 

Abstract

• Background: Endoscopic management of pediatric subglottic stenosis is common; however, no multiinstitutional studies have assessed its perioperative outcomes. The American College of Surgeon’s National Surgical Quality Improvement Program – Pediatric (ACS-NSQIP-P) represents a source of such data.
• Objective: To investigate 30-day outcomes of endoscopic dilation of the pediatric airway and to compare these outcomes to those seen with open reconstruction techniques.
• Methods: Current procedural terminology (CPT) codes were queried for endoscopic or open airway reconstruction in the 2015 ACS-NSQIP-P Public Use File (PUF). Demo­graphics and 30-day events were abstracted to compare open to endoscopic techniques and to assess for risk factors for varied outcomes after endoscopic dilation. Outcome measures included length of stay (LOS), 30-day rates of reintubation, readmission, and reoperation.
• Results: 171 endoscopic and 116 open procedures were identified. Mean age at endoscopic and open procedures was 4.1 (SEM = 0.37) and 5.4 years (SEM = 0.40). Mean LOS was shorter after endoscopic procedures (5.5 days, SEM = 1.13 vs. 11.3 days SEM = 1.01, P < 0.001). Open procedures had higher rates of reintubation (OR = 7.41, P = 0.026) and reoperation (OR = 3.09, P = 0.009). In patients undergoing endoscopic dilation, children < 1 year were more likely to require readmission (OR = 4.21, P = 0.03) and reoperation (OR = 4.39, P = 0.03) when compared with older children.
• Conclusion: Open airway reconstruction is associated with longer LOS and increased reintubations and reoperations, suggesting a possible opportunity to improve value in health care in the appropriately selected patient. Reoperations and readmissions following endoscopic dilation are more prevalent in children younger than 1 year.

Keywords: airway stenosis; subglottic stenosis; endoscopic dilation; pediatrics; outcomes.

 

Historically, pediatric laryngotracheal stenosis was managed using open reconstruction techniques, including laryngoplasty, tracheal resection, and cervical tracheoplasty. Initial reports of endoscopic dilation were described in the 1980s as a means to salvage re-stenosis after open reconstruction [1]. Currently, primary endoscopic dilation has become commonplace in otolaryngology due to its less invasive nature as well as—in cases of balloon dilation—minimization of tissue damage [2]. The advancements made in endoscopic balloon dilation have reduced the frequency with which open reconstruction is performed.

Systematic reviews and case series investigating endoscopic dilation indicate a 70% to 80% success rate in preventing future open surgery or tracheostomy [2–5]. While increased severity of stenosis has been associated with poorer outcomes in endoscopic procedures, few other risk factors that influence surgical success have been identified [4,5]. In a single study in the adult literature, open surgical management of idiopathic subglottic stenosis was associated with improved outcomes when compared to endoscopic techniques [5]. Such findings suggest a need to identify these factors for the purpose of optimizing clinical decision-making.

As laryngotracheal stenosis is rare, postoperative outcomes and risk factors are best identified on a multi­institutional level. Due to its participation from 80 hospitals and its accurate and reliable reporting of both demographic and risk-stratified 30-day outcomes data, the American College of Surgeon’s National Surgical Quality Improvement Program – Pediatric (ACS NSQIP-P) provides such a platform [6–8]. Thirty-day outcomes and risk factors for open reconstruction utilizing the ACS NSQIP-P database have previously been reported; however, no such outcomes for endoscopic dilation have been described, and no comparison between endoscopic and open procedures has been made [9]. The purpose of this study was to utilize the 2015 ACS-NSQIP-P database to investigate 30-day outcomes of endoscopic dilation of the pediatric airway and to compare these outcomes to open reconstruction techniques. Secondarily, we aimed to determine if any demographic factors or medical comorbidities are associated with varied outcomes in endoscopic reconstruction. While these data reflect safety and quality of this procedure in the United States, findings may potentially be applied across international settings.

Methods

Data Source

Data was obtained from the 2015 ACS-NSQIP-P Public Use File (PUF). Due to the de-identified and public nature of these data, this research was exempt from review by the Ann & Robert H. Lurie Children’s Hospital of Chicago review board. Data collection methods for ACS-NSQIP-P have previously been described [10]. In brief, data was collected from 80 hospitals on approximately 120 preoperative, intraoperative, and postoperative variables. Cases are systematically sampled on an 8-day cycle basis, where the first 35 cases meeting the inclusion criteria in each hospital in each cycle are submitted to ACS-NSQIP-P.

Variables and Outcomes

Airway procedures for endoscopic dilations and open reconstructions were obtained by CPT code. Endoscopic dilations (CPT 31528) were compared to open reconstructions, which included laryngoplasty (31580, 31582), cervical tracheoplasty (31750), cricoid split (31587), and tracheal resection (31780). Demographic variables included age, sex, race, and history of prematurity. Presence of specific comorbid diseases were also collected and tested for significance.

Dependent outcomes of interest were unplanned 30-day postoperative events grouped as reoperation, unplanned readmission, and postoperative reintubation. In the case of endoscopic procedures, the presence of salvage open reconstruction or tracheostomy within 30 days of surgery was also recorded. Length of stay (LOS) after the procedure was collected. Specific postoperative complications and reasons for readmission were recorded within the limitations of data available in the PUF.

Analysis

Analysis was performed using descriptive statistics and frequency analysis where appropriate. Chi-square analysis was used to compare adverse events between open and endoscopic procedures. Logistic regression with calculation of odds ratio (OR) was performed to determine predictive factors for reoperation, readmission, and reintubation in all pediatric airway reconstructive procedures in adjusted and unadjusted models. T-test and linear regression was performed on the continuous outcome of length of stay. For all analyses, a p value of < 0.05 was considered statistically significant. All variable recoding and statistical analyses were performed in SAS/STAT software (Cary, NC).

Results

A total of 84,056 pediatric procedures were extracted from the 2015 NSQIP-P PUFs. Using the above CPT codes, 171 endoscopic dilations and 116 open airway reconstructions were identified, with patient age ranging from 0 days to 17.6 years. Average age of patients undergoing endoscopic dilation and open reconstruction was 4.1 and 5.4 years, respectively (Table 1).

Potential confounders were tested with univariate logistic regression to determine if they had a significant impact on readmission, reintubation, or reoperation rates. These variables (Table 2) 

included age, tracheostomy status, prematurity, sex, race, congenital malformations, prior cardiac surgery, underlying pulmonary disease, immune disease, brain disease, gastrointestinal disease, cardiac disease. Only age was significantly associated with reoperation rates. These variables were also tested against length of stay using univariate linear regression: age, pulmonary disease and tracheostomy were significantly associated with change in length of stay. All models were adjusted for these 3 variables accordingly.

In patients undergoing endoscopic dilation, average length of stay was 5.5 days (SEM = 1.13), with 79 (48.5%) patients having a length of stay of zero days. Of all patients who had endoscopic dilations, 70 (40.1%) had a pre-existing tracheostomy and these accounted for the majority (73%) of patients who had zero days as their LOS. LOS after endoscopic management was significantly shorter than the mean of 11.3 days (SEM = 1.01) reported in those undergoing open reconstruction (P < 0.001).

With respect to 30-day adverse events, 2 patients in the endoscopic group (1.1%) required reintubation. Thirteen endoscopic dilation cases (7.6%) had an unplanned readmission, four (2.3%) of which were associated with reoperation within 30 days of the primary surgical procedure. There were 9 other reoperations unassociated with unplanned readmission. Three of these reoperations were due to failed endoscopic dilations, resulting in 2 tracheostomies and one open airway reconstruction. There was one patient death, in a 0-day old with tetralogy of Fallot, trachea-esophageal fistula, and ventilator dependence who underwent emergent endoscopic dilation and died the same day.

Open procedures were associated with 11 unplanned readmissions (9.5%), 7 re-intubations (6%) and 18 reoperations (15.5%). Of patents undergoing reoperation, one patient undergoing open reconstruction underwent tracheostomy within 30 days of surgery.

When comparing open reconstruction to endoscopic dilation, there was a significant increase in reintubation (OR = 7.41, P = 0.026) and reoperation (OR = 3.09, P = 0.009) for open procedures, even with adjustment for age, tracheostomy status, and pulmonary disease. There was no significant difference between the two for unplanned readmissions (OR = 1.19, P = 0.79) (Figure).

Younger age was also found to be significantly associated with reoperation rates, in an adjusted logistic model that accounted for tracheostomy status, type of surgery, and pulmonary disease. Per year of life, younger children had higher reoperation rates than older children (OR = 1.91, P = 0.017). When endoscopic dilation was individually examined, children younger than 1 year of age were more likely to undergo reoperation after an endoscopic dilation than children older than 1 (OR = 4.39, P = 0.03). Children under age 1 were also more likely to have an unplanned readmission after an endoscopic dilation (OR = 4.21, P = 0.03). The relationship between age and re-intubation was not significant (OR = 0, P = 0.95). For open reconstruction, this age dichotomization was not associated with any increased reoperation (OR = 2.3, P = 0.52), readmission (OR = 0, P = 0.97), or reintubation (OR = 0, P = 0.94).

T-test analysis was performed to determine if children < 1 year old also had significantly longer hospital stays after endoscopic dilation than older children (mean 14.1 days vs 1.9 days, P < 0.001). This relationship held true in a linear regression after adjustment for pulmonary disease and tracheostomy, with length of stay decreasing by 0.48 days per year of life (P = 0.03). For endoscopic dilations, the same relationship held true, where length of stay decreased by 0.75 days per year of life.

 

Discussion

Endoscopic dilation for primary management of pediatric laryngotracheal stenosis has become commonplace. Despite this, outcomes of this procedure have only been described in case series and meta-analyses [2–5]. The relative rarity of pediatric laryngotracheal stenosis suggests the need for large, multi-institutional data for purposes of patient selection and medical decision-making.

This study utilized the ACS-NSQIP-Pediatric database to highlight 30-day outcomes of endoscopic dilation and to compare these outcomes to those of open airway reconstruction procedures. The ACS-NSQIP database has been endorsed by multiple organizations, including the Center for Medicare and Medicaid Services, the Joint Commission, and the American Medical Association. It has been shown to have higher sensitivity and a lower false-positive rate when compared to administrative data, in part due to data collection from trained personnel [11]. Furthermore, ACS-NSQIP use has the additional benefit of reporting an unplanned admission—a feature unavailable in review of claims data [12].

With respect to adverse events, our study demonstrates that endoscopic dilation is associated with an equally high rate of unplanned readmission when compared to open reconstruction. The high prevalence of comorbid disease such as chronic lung disease (32% of endoscopic dilation and 43% of open reconstruction) can account for some of the morbidity associated with any airway procedures.

Despite high rates of unplanned readmission, patients undergoing endoscopic dilation were less likely to have reoperations within 30 days of initial surgery when compared to those undergoing open reconstruction. While differences in disease severity may be partially responsible for this difference in the reoperation rate, this finding is notable given the health care costs associated with multiple operations as well as safety concerns with multiple anesthetics in the very young [13,14].

The ACS-NSQIP platform does not distinguish unplanned from planned reoperations. In the setting of airway surgery, where multiple planned reoperations are commonplace, this metric is a suboptimal stand-alone indicator of adverse outcomes. Other markers available in the database—such as reintubations and performance of tracheostomy or open airway reconstruction within 30-days of surgery—are more indicative of surgical outcome in the setting of airway surgery. We found that both reintubations and salvage open reconstruction within 30-days were rare. It should be noted that the ACS-NSQIP data does not report any events occurring outside of the 30-day postoperative period, representing potential limitation of the use of this database. As was previously advocated by Roxbury and colleagues, procedure/subspecialty specific outcome data collection would also improve outcome analysis of airway and other otolaryngologic procedures [9]. In the setting of airway reconstruction, this would include data pertaining to Cotton-Meyer grading systems well as postoperative voice and swallow outcomes.

In addition to safety profile, endoscopic procedures were associated with shorter LOS when compared with open reconstruction, representing another potential source of cost savings with this less invasive method. This is especially significant given that open reconstruction patients spend much of their inpatient stay in an ICU setting. In patients who are candidates for endoscopic procedures, this lower-risk, lower-cost profile of endoscopic dilation has the opportunity to improve value in health care and may be the source of future improvement initiatives.

In addition to comparing overall outcomes between endoscopic and open management of laryngotracheal stenosis, our study aimed to identify factors that were associated with varied outcomes in patients undergoing primary endoscopic dilation. We found that children younger than 1 year of age were 5.8 times more likely to undergo an unplanned reoperation after an endoscopic dilation than children over 1 year. A similar finding was reported in open airway surgeries, with increased reoperation rates in children < 3 years old [9]. The justification of a dichotomization at 1 year was made as expert opinion recognizes that the infant airway is less forgiving to intervention given its small size. Young age was also a factor in prolonged LOS as was determined by linear regression. It is likely that this increased LOS may be in part due to associations of young age and the neonatal ICU population. One must balance the increased risk of surgery in the young with that of tracheostomy, which has a published complication rate of 18% to 50% and direct mortality rate of 1% to 2% in the pediatric population [15–18]. Understanding these relative risks may help guide the airway surgeon in preoperative counseling with families and medical decision-making.

As discussed above, the limitation of data to a 30-day period is a relative weakness of ACS-NSQIP database use for studies of airway reconstruction, as the ultimate outcome—a stable, decannulated airway—may occur outside of this time period. As many quality metrics utilize data from the 30-day postoperative period, knowledge of these outcomes remains valuable in surgical decision-making. Ultimately, collection of data in a large, long-term dataset would allow broader generalizations to be made about the differences between open and endoscopic procedures and would also give a more comprehensive picture of the outcomes of endoscopic dilation.

In conclusion, this study is the first to analyze 30-day postoperative outcomes in pediatric endoscopic airway dilations using data aggregated by ACS-NSQIP from institutions across the United States. This data indicates that endoscopic airway dilation is a relatively safe procedure, especially compared with open reconstruction; however, additional data on disease severity and other outcomes is necessary to draw final conclusions of superiority of technique. Future improvement initiatives could be aimed at the impact of this lower-risk, lower-cost procedure in the appropriately selected patient. Outcomes of endoscopic dilation are poorer in those less than 1 year of age, as they are associated with increased reoperation rates and increased length of stay compared to older children. One must balance these risks in the very young with the risks associated with tracheostomy and other alternative airway management modalities.

 

Note: This work was presented in a paper at the AAO-HNS 2017 meeting, Chicago, IL, 10 Sep 2017.

Corresponding author: Jennifer Lavin, MD, MS, 225 E Chicago Ave., Box 25, Chicago, IL 60611, [email protected].

Financial disclosures: None.

Colorectal cancer (CRC) is an important concern for the VA, and colonoscopy is one primary screening, surveillance, and diagnostic modality used. The observed reductions in CRC incidence and mortality over the past decade largely have been attributed to the widespread use of CRC screening options.^1,2 Colonoscopy quality is critical to CRC prevention in veterans. However, endoscopy skills to detect and remove colorectal polyps using colonoscopy vary in practice.^3-5

Quality benchmarks, linked to patient outcomes, have been established by specialty societies and proposed by the Centers for Medicare and Medicaid Services as reportable quality metrics.⁶ Colonoscopy quality metrics have been shown to be associated with patient outcomes, such as the risk of developing CRC after colonoscopy. The adenoma detection rate (ADR), defined as the proportion of average-risk screening colonoscopies in which 1 or more adenomas are detected, has the strongest association to interval or “missed” CRC after screening colonoscopy and has been linked to a risk for fatal CRC despite colonoscopy.³

In a landmark study of 314,872 examinations performed by 136 gastroenterologists, the ADR ranged from 7.4% to 52.5%.³ Among patients with ADRs in the highest quintile compared with patients in the lowest, the adjusted hazard ratios (HRs) for any interval cancer was 0.52 (95% confidence interval [CI], 0.39-0.69) and for fatal interval cancers was 0.38 (95% CI, 0.22-0.65).³ Another pooled analysis from 8 surveillance studies that followed more than 800 participants with adenoma(s) after a baseline colonoscopy showed 52% of incident cancers as probable missed lesions, 19% as possibly related to incomplete resection of an earlier, noninvasive lesion, and only 24% as probable new lesions.⁷ These interval cancers highlight the current imperfections of colonoscopy and the focus on measurement and reporting of quality indicators for colonoscopy.^8-12

According to VHA Directive 1015, in December 2014, colonoscopy quality should be monitored as part of an ongoing quality assurance program.¹³ A recent report from the VA Office of the Inspector General (OIG) highlighted colonoscopy-quality deficiencies.¹⁴ The OIG report strongly recommended that the “Acting Under Secretary for Health require standardized documentation of quality indicators based on professional society guidelines and published literature.”¹⁴However, no currently standardized and readily available VHA resource measures, reports, and ensures colonoscopy quality.

The authors hypothesized that colonoscopy quality assurance programs vary widely across VHA sites. The objective of this survey was to assess the measurement and reporting practices for colonoscopy quality and identify both strengths and areas for improvement to facilitate implementation of quality assurance programs across the VA health care system.

Methods

The authors performed an online survey of VA sites to assess current colonoscopy quality assurance practices. The institutional review boards (IRBs) at the University of Utah and VA Salt Lake City Health Care System and University of California, San Francisco and San Francisco VA Health Care System classified the study as a quality improvement project that did not qualify for human subjects’ research requiring IRB review.

The authors iteratively developed and refined the questionnaire with a survey methodologist and 2 clinical domain experts. The National Program Director for Gastroenterology, and the National Gastroenterology Field Advisory Committee reviewed the survey content and pretested the survey instrument prior to final data collection. The National Program Office for Gastroenterology provided an e-mail list of all known VA gastroenterology section chiefs. The authors administered the final survey via e-mail, using the Research Electronic Data Capture (REDCap; Vanderbilt University Medical Center) platform beginning January 9, 2017.¹⁵

A follow-up reminder e-mail was sent to nonresponders after 2 weeks. After this second invitation, sites were contacted by telephone to verify that the correct contact information had been captured. Subsequently, 50 contacts were updated if e-mails bounced back or the correct contact was obtained. Points of contact received a total of 3 reminder e-mails until the final closeout of the survey on March 28, 2017; 65 of 89 (73%) of the original contacts completed the survey vs 31 of 50 (62%) of the updated contacts.

Analysis

Descriptive statistics of the responses were calculated to determine the overall proportion of VA sites measuring colonoscopy quality metrics and identification of areas in need of quality improvement. The response rate for the survey was defined as the total number of responses obtained as a proportion of the total number of points of contact. This corresponds to the American Association of Public Opinion Research’s RR1, or minimum response rate, formula.¹⁶ All categoric responses are presented as proportions. Statistical analyses were performed using STATA SE12.0 (College Station, TX).

Results

Of the 139 points of contact invited, 96 completed the survey (response rate of 69.0%), representing 93 VA facilities (of 141 possible facilities) in 44 different states. Three sites had 2 responses. Sites used various and often a combination of methods to measure quality (Table 1). 

The majority of sites reported using manual chart review (79.2%) to collect colonoscopy quality metrics, although only 39.6% reported using endoscopic software to collect quality metrics. A variety of personnel collected quality metrics, the most common being nursing staff (44.8% of participating sites), division/section chiefs (38.5%), multiple other staff (29.2%), and attending physicians (22.9%). The most common reporting frequencies were quarterly (41.7%), monthly (17.7%), and biannually (13.5%).

 

A majority of sites’ (63.5%) quality reports represented individual provider data, whereas fewer provided quality reports for physician groups (22.9%) or for the entire facility (40.6%). Provider quality information was de-identified in 43.8% of reporting sites’ quality reports and identifiable in 37.5% of reporting sites’ quality reports. A majority of sites (74.0%) reported that the local gastroenterology section chief or quality manager has access to the quality reports. Fewer sites reported providing data to individual endoscopists (44.8% for personal and peer data and 32.3% for personal data only). One site (1%) responded that quality reports were available for public access. Survey respondents also were asked to provide the estimated time (hours required per month) to collect the data for quality metrics. Of 75 respondents providing data for this question, 28 (29.2%) and 17 (17.7%), estimated between 1 to 5 and 6 to 10 hours per month, respectively. Ten sites estimated spending between 11 to 20 hours, and 7 sites estimated spending more than 20 hours per month collecting quality metrics. A total of 13 respondents (13.5%) stated uncertainty about the time burden.

As shown in the Figure,  numerous quality metrics were collected across sites with more than 80% of sites collecting information on bowel preparation quality (88.5%), cecal intubation rate (87.5%), and complications (83.3%). A majority of sites also reported collecting data on appropriateness of surveillance intervals (62.5%), colonoscopy withdrawal times (62.5%), and ADRs (61.5%). Seven sites (7.3%) did not collect quality metrics.

Information also was collected on colonoscopy procedure documentation to inform future efforts at standardization. A small majority (53.1%) of sites reported using endoscopic software to generate colonoscopy procedure documentation. Within these sites, 6 different types of endoscopic note writing software were used to generate procedure notes (Table 2). 

Remaining sites used a variety of methods to generate procedure notes, including typing free-text notes (17.7%) and dictation (11.5%); 35.4% of sites use a template note in the Computerized Patient Record System. Sites also used a variety of methods for photo documentation in the electronic medical record (eg, pictures of cecal intubation and polyps).

Most sites (85.4%) were aware of VHA Directive 1015 recommendations for colonoscopy quality assurance programs. A significant majority (89.5%) of respondents also indicated interest in a centralized automatic reporting system to measure and report colonoscopy quality in some form, either with aggregate data, provider data, or both (Table 3).

 

 

Discussion

This survey on colonoscopy quality assurance programs is the first assessment of the VHA’s efforts to measure and report colonoscopy quality indicators. The findings indicated that the majority of VA sites are measuring and reporting at least some measures of colonoscopy quality. However, the programs are significantly variable in terms of methods used to collect quality metrics, specific quality measures obtained, and how quality is reported.

The authors’ work is novel in that this is the first report of the status of colonoscopy quality assurance programs in a large U.S. health care system. The VA health care system is the largest integrated health system in the U.S., serving more than 9 million veterans annually. This survey’s high response rate further strengthens the findings. Specifically, the survey found that VA sites are making a strong concerted effort to measure and report colonoscopy quality. However, there is significant variability in documentation, measurement, and reporting practices. Moreover, the majority of VA sites do not have formal performance improvement plans in place for endoscopists who do not meet thresholds for colonoscopy quality.

Screening colonoscopy for CRC offers known mortality benefits to patients.^1,17-19 Significant prior work has described and validated the importance of colonoscopy quality metrics, including bowel preparation quality, cecal intubation rate, and ADR and their association with interval colorectal cancer and death.^20-23 Gastroenterology professional societies, including the American College of Gastroenterology and the American Society for Gastrointestinal Endoscopy, have recommended and endorsed measurement and reporting of colonoscopy metrics.²⁴ There is general agreement among endoscopists that colonoscopy quality is an important aspect of performing the procedure.

The lack of formal performance improvement programs is a key finding of this survey. Recent studies have shown that improvements in quality metrics, such as the ADR, by individual endoscopists result in reductions in interval colorectal cancer and death.²⁵ Kahi and colleagues previously showed that providing a quarterly report card improves colonoscopy quality.²⁶ Keswani and colleagues studied a combination of a report card and implementation of standards of practice with resultant improvement in colonoscopy quality.²⁷ Most recently, in a large prospective cohort study of individuals who underwent a screening colonoscopy, 294 of the screening endoscopists received annual feedback and quality benchmark indicators to improve colonoscopy performance.²⁵ The majority of the endoscopists (74.5%) increased their annual ADR category over the study period. Moreover, patients examined by endoscopists who reached or maintained the highest ADR quintile (> 24.6%) had significantly lower risk of interval CRC and death. The lack of formal performance improvement programs across the VHA is concerning but reveals a significant opportunity to improve veteran health outcomes on a large scale.

This study’s findings also highlight the intense resources necessary to measure and report colonoscopy quality. The ability to measure and report quality metrics requires having adequate documentation and data to obtain quality metrics. Administrative databases from electronic health records offer some potential for routine monitoring of quality metrics.²⁸ However, most administrative databases, including the VA Corporate Data Warehouse (CDW), contain administrative billing codes (ICD and CPT) linked to limited patient data, including demographics and structured medical record data. The actual data required for quality reporting of important metrics (bowel preparation quality, cecal intubation rates, and ADRs) are usually found in clinical text notes or endoscopic note documentation and not available as structured data. Due to this issue, the majority of VA sites (79.2%) are using manual chart review to collect quality metric data, resulting in widely variable estimates on time burden. A minority of sites in this study (39.6%) reported using automated endoscopic software reporting capability that can help with the time burden. However, even in the VA, an integrated health system, a wide variety of software brands, documentation practices, and photo documentation was found.

Future endoscopy budget and purchase decisions for the individual VA sites should take into account how new technology and software can more easily facilitate accurate quality reporting. A specific policy recommendation would be for the VA to consider a uniform endoscopic note writer for procedure notes. Pathology data, which is necessary for the calculation of ADR, also should be available as structured data in the CDW to more easily measure colonoscopy quality. Continuous measurement and reporting of quality also requires ongoing information technology infrastructure and quality control of the measurement process.

 

Limitations

This survey was a cross-section of VA sites’ points of contact regarding colonoscopy quality assurance programs, so the results are descriptive in nature. However, the instrument was carefully developed, using both subject matter and survey method expertise. The questionnaire also was refined through pretesting prior to data collection. The initial contact list was found to have errors, and the list had to be updated after launching the survey. Updated information for most of the contacts was available.

Another limitation was the inability to survey nongastroenterologist-run endoscopy centers, because many centers use surgeons or other nongastroenterology providers. The authors speculate that quality monitoring may be less likely to be present at these facilities as they may not be aware of the gastroenterology professional society recommendations. The authors did not require or insist that all questions be answered, so some data were missing from sites. However, 93.7% of respondents completed the entire survey.

Conclusion

The authors have described the status of colonoscopy quality assurance programs across the VA health care system. Many sites are making robust efforts to measure and report quality especially of process measures. However, there are significant time and manual workforce efforts required, and this work is likely associated with the variability in programs. Importantly, ADR, which is the quality metric that has been most strongly associated with risk of colon cancer mortality, is not being measured by 38% of sites.

These results reinforce a critical need for a centralized, automated quality reporting infrastructure to standardize colonoscopy quality reporting, reduce workload, and ensure veterans receive high-quality colonoscopy.

Acknowledgments
The authors acknowledge the support and feedback of the National Gastroenterology Program Field Advisory Committee for survey development and testing. The authors coordinated the survey through the Salt Lake City Specialty Care Center of Innovation in partnership with the National Gastroenterology Program Office and the Quality Enhancement Research Initiative: Quality Enhancement Research Initiative, Measurement Science Program, QUE15-283. The work also was partially supported by the National Center for Advancing Translational Sciences of the National Institutes of Health Award UL1TR001067 and Merit Review Award 1 I01 HX001574-01A1 from the United States Department of Veterans Affairs Health Services Research & Development Service of the VA Office of Research and Development.

Colorectal cancer (CRC) is an important concern for the VA, and colonoscopy is one primary screening, surveillance, and diagnostic modality used. The observed reductions in CRC incidence and mortality over the past decade largely have been attributed to the widespread use of CRC screening options.^1,2 Colonoscopy quality is critical to CRC prevention in veterans. However, endoscopy skills to detect and remove colorectal polyps using colonoscopy vary in practice.^3-5

Quality benchmarks, linked to patient outcomes, have been established by specialty societies and proposed by the Centers for Medicare and Medicaid Services as reportable quality metrics.⁶ Colonoscopy quality metrics have been shown to be associated with patient outcomes, such as the risk of developing CRC after colonoscopy. The adenoma detection rate (ADR), defined as the proportion of average-risk screening colonoscopies in which 1 or more adenomas are detected, has the strongest association to interval or “missed” CRC after screening colonoscopy and has been linked to a risk for fatal CRC despite colonoscopy.³

In a landmark study of 314,872 examinations performed by 136 gastroenterologists, the ADR ranged from 7.4% to 52.5%.³ Among patients with ADRs in the highest quintile compared with patients in the lowest, the adjusted hazard ratios (HRs) for any interval cancer was 0.52 (95% confidence interval [CI], 0.39-0.69) and for fatal interval cancers was 0.38 (95% CI, 0.22-0.65).³ Another pooled analysis from 8 surveillance studies that followed more than 800 participants with adenoma(s) after a baseline colonoscopy showed 52% of incident cancers as probable missed lesions, 19% as possibly related to incomplete resection of an earlier, noninvasive lesion, and only 24% as probable new lesions.⁷ These interval cancers highlight the current imperfections of colonoscopy and the focus on measurement and reporting of quality indicators for colonoscopy.^8-12

According to VHA Directive 1015, in December 2014, colonoscopy quality should be monitored as part of an ongoing quality assurance program.¹³ A recent report from the VA Office of the Inspector General (OIG) highlighted colonoscopy-quality deficiencies.¹⁴ The OIG report strongly recommended that the “Acting Under Secretary for Health require standardized documentation of quality indicators based on professional society guidelines and published literature.”¹⁴However, no currently standardized and readily available VHA resource measures, reports, and ensures colonoscopy quality.

The authors hypothesized that colonoscopy quality assurance programs vary widely across VHA sites. The objective of this survey was to assess the measurement and reporting practices for colonoscopy quality and identify both strengths and areas for improvement to facilitate implementation of quality assurance programs across the VA health care system.

Methods

The authors performed an online survey of VA sites to assess current colonoscopy quality assurance practices. The institutional review boards (IRBs) at the University of Utah and VA Salt Lake City Health Care System and University of California, San Francisco and San Francisco VA Health Care System classified the study as a quality improvement project that did not qualify for human subjects’ research requiring IRB review.

The authors iteratively developed and refined the questionnaire with a survey methodologist and 2 clinical domain experts. The National Program Director for Gastroenterology, and the National Gastroenterology Field Advisory Committee reviewed the survey content and pretested the survey instrument prior to final data collection. The National Program Office for Gastroenterology provided an e-mail list of all known VA gastroenterology section chiefs. The authors administered the final survey via e-mail, using the Research Electronic Data Capture (REDCap; Vanderbilt University Medical Center) platform beginning January 9, 2017.¹⁵

A follow-up reminder e-mail was sent to nonresponders after 2 weeks. After this second invitation, sites were contacted by telephone to verify that the correct contact information had been captured. Subsequently, 50 contacts were updated if e-mails bounced back or the correct contact was obtained. Points of contact received a total of 3 reminder e-mails until the final closeout of the survey on March 28, 2017; 65 of 89 (73%) of the original contacts completed the survey vs 31 of 50 (62%) of the updated contacts.

Analysis

Descriptive statistics of the responses were calculated to determine the overall proportion of VA sites measuring colonoscopy quality metrics and identification of areas in need of quality improvement. The response rate for the survey was defined as the total number of responses obtained as a proportion of the total number of points of contact. This corresponds to the American Association of Public Opinion Research’s RR1, or minimum response rate, formula.¹⁶ All categoric responses are presented as proportions. Statistical analyses were performed using STATA SE12.0 (College Station, TX).

Results

Of the 139 points of contact invited, 96 completed the survey (response rate of 69.0%), representing 93 VA facilities (of 141 possible facilities) in 44 different states. Three sites had 2 responses. Sites used various and often a combination of methods to measure quality (Table 1). 

The majority of sites reported using manual chart review (79.2%) to collect colonoscopy quality metrics, although only 39.6% reported using endoscopic software to collect quality metrics. A variety of personnel collected quality metrics, the most common being nursing staff (44.8% of participating sites), division/section chiefs (38.5%), multiple other staff (29.2%), and attending physicians (22.9%). The most common reporting frequencies were quarterly (41.7%), monthly (17.7%), and biannually (13.5%).

 

A majority of sites’ (63.5%) quality reports represented individual provider data, whereas fewer provided quality reports for physician groups (22.9%) or for the entire facility (40.6%). Provider quality information was de-identified in 43.8% of reporting sites’ quality reports and identifiable in 37.5% of reporting sites’ quality reports. A majority of sites (74.0%) reported that the local gastroenterology section chief or quality manager has access to the quality reports. Fewer sites reported providing data to individual endoscopists (44.8% for personal and peer data and 32.3% for personal data only). One site (1%) responded that quality reports were available for public access. Survey respondents also were asked to provide the estimated time (hours required per month) to collect the data for quality metrics. Of 75 respondents providing data for this question, 28 (29.2%) and 17 (17.7%), estimated between 1 to 5 and 6 to 10 hours per month, respectively. Ten sites estimated spending between 11 to 20 hours, and 7 sites estimated spending more than 20 hours per month collecting quality metrics. A total of 13 respondents (13.5%) stated uncertainty about the time burden.

As shown in the Figure,  numerous quality metrics were collected across sites with more than 80% of sites collecting information on bowel preparation quality (88.5%), cecal intubation rate (87.5%), and complications (83.3%). A majority of sites also reported collecting data on appropriateness of surveillance intervals (62.5%), colonoscopy withdrawal times (62.5%), and ADRs (61.5%). Seven sites (7.3%) did not collect quality metrics.

Information also was collected on colonoscopy procedure documentation to inform future efforts at standardization. A small majority (53.1%) of sites reported using endoscopic software to generate colonoscopy procedure documentation. Within these sites, 6 different types of endoscopic note writing software were used to generate procedure notes (Table 2). 

Remaining sites used a variety of methods to generate procedure notes, including typing free-text notes (17.7%) and dictation (11.5%); 35.4% of sites use a template note in the Computerized Patient Record System. Sites also used a variety of methods for photo documentation in the electronic medical record (eg, pictures of cecal intubation and polyps).

Most sites (85.4%) were aware of VHA Directive 1015 recommendations for colonoscopy quality assurance programs. A significant majority (89.5%) of respondents also indicated interest in a centralized automatic reporting system to measure and report colonoscopy quality in some form, either with aggregate data, provider data, or both (Table 3).

 

 

Discussion

This survey on colonoscopy quality assurance programs is the first assessment of the VHA’s efforts to measure and report colonoscopy quality indicators. The findings indicated that the majority of VA sites are measuring and reporting at least some measures of colonoscopy quality. However, the programs are significantly variable in terms of methods used to collect quality metrics, specific quality measures obtained, and how quality is reported.

The authors’ work is novel in that this is the first report of the status of colonoscopy quality assurance programs in a large U.S. health care system. The VA health care system is the largest integrated health system in the U.S., serving more than 9 million veterans annually. This survey’s high response rate further strengthens the findings. Specifically, the survey found that VA sites are making a strong concerted effort to measure and report colonoscopy quality. However, there is significant variability in documentation, measurement, and reporting practices. Moreover, the majority of VA sites do not have formal performance improvement plans in place for endoscopists who do not meet thresholds for colonoscopy quality.

Screening colonoscopy for CRC offers known mortality benefits to patients.^1,17-19 Significant prior work has described and validated the importance of colonoscopy quality metrics, including bowel preparation quality, cecal intubation rate, and ADR and their association with interval colorectal cancer and death.^20-23 Gastroenterology professional societies, including the American College of Gastroenterology and the American Society for Gastrointestinal Endoscopy, have recommended and endorsed measurement and reporting of colonoscopy metrics.²⁴ There is general agreement among endoscopists that colonoscopy quality is an important aspect of performing the procedure.

The lack of formal performance improvement programs is a key finding of this survey. Recent studies have shown that improvements in quality metrics, such as the ADR, by individual endoscopists result in reductions in interval colorectal cancer and death.²⁵ Kahi and colleagues previously showed that providing a quarterly report card improves colonoscopy quality.²⁶ Keswani and colleagues studied a combination of a report card and implementation of standards of practice with resultant improvement in colonoscopy quality.²⁷ Most recently, in a large prospective cohort study of individuals who underwent a screening colonoscopy, 294 of the screening endoscopists received annual feedback and quality benchmark indicators to improve colonoscopy performance.²⁵ The majority of the endoscopists (74.5%) increased their annual ADR category over the study period. Moreover, patients examined by endoscopists who reached or maintained the highest ADR quintile (> 24.6%) had significantly lower risk of interval CRC and death. The lack of formal performance improvement programs across the VHA is concerning but reveals a significant opportunity to improve veteran health outcomes on a large scale.

This study’s findings also highlight the intense resources necessary to measure and report colonoscopy quality. The ability to measure and report quality metrics requires having adequate documentation and data to obtain quality metrics. Administrative databases from electronic health records offer some potential for routine monitoring of quality metrics.²⁸ However, most administrative databases, including the VA Corporate Data Warehouse (CDW), contain administrative billing codes (ICD and CPT) linked to limited patient data, including demographics and structured medical record data. The actual data required for quality reporting of important metrics (bowel preparation quality, cecal intubation rates, and ADRs) are usually found in clinical text notes or endoscopic note documentation and not available as structured data. Due to this issue, the majority of VA sites (79.2%) are using manual chart review to collect quality metric data, resulting in widely variable estimates on time burden. A minority of sites in this study (39.6%) reported using automated endoscopic software reporting capability that can help with the time burden. However, even in the VA, an integrated health system, a wide variety of software brands, documentation practices, and photo documentation was found.

Future endoscopy budget and purchase decisions for the individual VA sites should take into account how new technology and software can more easily facilitate accurate quality reporting. A specific policy recommendation would be for the VA to consider a uniform endoscopic note writer for procedure notes. Pathology data, which is necessary for the calculation of ADR, also should be available as structured data in the CDW to more easily measure colonoscopy quality. Continuous measurement and reporting of quality also requires ongoing information technology infrastructure and quality control of the measurement process.

 

Limitations

This survey was a cross-section of VA sites’ points of contact regarding colonoscopy quality assurance programs, so the results are descriptive in nature. However, the instrument was carefully developed, using both subject matter and survey method expertise. The questionnaire also was refined through pretesting prior to data collection. The initial contact list was found to have errors, and the list had to be updated after launching the survey. Updated information for most of the contacts was available.

Another limitation was the inability to survey nongastroenterologist-run endoscopy centers, because many centers use surgeons or other nongastroenterology providers. The authors speculate that quality monitoring may be less likely to be present at these facilities as they may not be aware of the gastroenterology professional society recommendations. The authors did not require or insist that all questions be answered, so some data were missing from sites. However, 93.7% of respondents completed the entire survey.

Conclusion

The authors have described the status of colonoscopy quality assurance programs across the VA health care system. Many sites are making robust efforts to measure and report quality especially of process measures. However, there are significant time and manual workforce efforts required, and this work is likely associated with the variability in programs. Importantly, ADR, which is the quality metric that has been most strongly associated with risk of colon cancer mortality, is not being measured by 38% of sites.

These results reinforce a critical need for a centralized, automated quality reporting infrastructure to standardize colonoscopy quality reporting, reduce workload, and ensure veterans receive high-quality colonoscopy.

Acknowledgments
The authors acknowledge the support and feedback of the National Gastroenterology Program Field Advisory Committee for survey development and testing. The authors coordinated the survey through the Salt Lake City Specialty Care Center of Innovation in partnership with the National Gastroenterology Program Office and the Quality Enhancement Research Initiative: Quality Enhancement Research Initiative, Measurement Science Program, QUE15-283. The work also was partially supported by the National Center for Advancing Translational Sciences of the National Institutes of Health Award UL1TR001067 and Merit Review Award 1 I01 HX001574-01A1 from the United States Department of Veterans Affairs Health Services Research & Development Service of the VA Office of Research and Development.

Colorectal cancer (CRC) is an important concern for the VA, and colonoscopy is one primary screening, surveillance, and diagnostic modality used. The observed reductions in CRC incidence and mortality over the past decade largely have been attributed to the widespread use of CRC screening options.^1,2 Colonoscopy quality is critical to CRC prevention in veterans. However, endoscopy skills to detect and remove colorectal polyps using colonoscopy vary in practice.^3-5

Quality benchmarks, linked to patient outcomes, have been established by specialty societies and proposed by the Centers for Medicare and Medicaid Services as reportable quality metrics.⁶ Colonoscopy quality metrics have been shown to be associated with patient outcomes, such as the risk of developing CRC after colonoscopy. The adenoma detection rate (ADR), defined as the proportion of average-risk screening colonoscopies in which 1 or more adenomas are detected, has the strongest association to interval or “missed” CRC after screening colonoscopy and has been linked to a risk for fatal CRC despite colonoscopy.³

In a landmark study of 314,872 examinations performed by 136 gastroenterologists, the ADR ranged from 7.4% to 52.5%.³ Among patients with ADRs in the highest quintile compared with patients in the lowest, the adjusted hazard ratios (HRs) for any interval cancer was 0.52 (95% confidence interval [CI], 0.39-0.69) and for fatal interval cancers was 0.38 (95% CI, 0.22-0.65).³ Another pooled analysis from 8 surveillance studies that followed more than 800 participants with adenoma(s) after a baseline colonoscopy showed 52% of incident cancers as probable missed lesions, 19% as possibly related to incomplete resection of an earlier, noninvasive lesion, and only 24% as probable new lesions.⁷ These interval cancers highlight the current imperfections of colonoscopy and the focus on measurement and reporting of quality indicators for colonoscopy.^8-12

According to VHA Directive 1015, in December 2014, colonoscopy quality should be monitored as part of an ongoing quality assurance program.¹³ A recent report from the VA Office of the Inspector General (OIG) highlighted colonoscopy-quality deficiencies.¹⁴ The OIG report strongly recommended that the “Acting Under Secretary for Health require standardized documentation of quality indicators based on professional society guidelines and published literature.”¹⁴However, no currently standardized and readily available VHA resource measures, reports, and ensures colonoscopy quality.

The authors hypothesized that colonoscopy quality assurance programs vary widely across VHA sites. The objective of this survey was to assess the measurement and reporting practices for colonoscopy quality and identify both strengths and areas for improvement to facilitate implementation of quality assurance programs across the VA health care system.

Methods

The authors performed an online survey of VA sites to assess current colonoscopy quality assurance practices. The institutional review boards (IRBs) at the University of Utah and VA Salt Lake City Health Care System and University of California, San Francisco and San Francisco VA Health Care System classified the study as a quality improvement project that did not qualify for human subjects’ research requiring IRB review.

The authors iteratively developed and refined the questionnaire with a survey methodologist and 2 clinical domain experts. The National Program Director for Gastroenterology, and the National Gastroenterology Field Advisory Committee reviewed the survey content and pretested the survey instrument prior to final data collection. The National Program Office for Gastroenterology provided an e-mail list of all known VA gastroenterology section chiefs. The authors administered the final survey via e-mail, using the Research Electronic Data Capture (REDCap; Vanderbilt University Medical Center) platform beginning January 9, 2017.¹⁵

A follow-up reminder e-mail was sent to nonresponders after 2 weeks. After this second invitation, sites were contacted by telephone to verify that the correct contact information had been captured. Subsequently, 50 contacts were updated if e-mails bounced back or the correct contact was obtained. Points of contact received a total of 3 reminder e-mails until the final closeout of the survey on March 28, 2017; 65 of 89 (73%) of the original contacts completed the survey vs 31 of 50 (62%) of the updated contacts.

Analysis

Descriptive statistics of the responses were calculated to determine the overall proportion of VA sites measuring colonoscopy quality metrics and identification of areas in need of quality improvement. The response rate for the survey was defined as the total number of responses obtained as a proportion of the total number of points of contact. This corresponds to the American Association of Public Opinion Research’s RR1, or minimum response rate, formula.¹⁶ All categoric responses are presented as proportions. Statistical analyses were performed using STATA SE12.0 (College Station, TX).

Results

Of the 139 points of contact invited, 96 completed the survey (response rate of 69.0%), representing 93 VA facilities (of 141 possible facilities) in 44 different states. Three sites had 2 responses. Sites used various and often a combination of methods to measure quality (Table 1). 

The majority of sites reported using manual chart review (79.2%) to collect colonoscopy quality metrics, although only 39.6% reported using endoscopic software to collect quality metrics. A variety of personnel collected quality metrics, the most common being nursing staff (44.8% of participating sites), division/section chiefs (38.5%), multiple other staff (29.2%), and attending physicians (22.9%). The most common reporting frequencies were quarterly (41.7%), monthly (17.7%), and biannually (13.5%).

 

A majority of sites’ (63.5%) quality reports represented individual provider data, whereas fewer provided quality reports for physician groups (22.9%) or for the entire facility (40.6%). Provider quality information was de-identified in 43.8% of reporting sites’ quality reports and identifiable in 37.5% of reporting sites’ quality reports. A majority of sites (74.0%) reported that the local gastroenterology section chief or quality manager has access to the quality reports. Fewer sites reported providing data to individual endoscopists (44.8% for personal and peer data and 32.3% for personal data only). One site (1%) responded that quality reports were available for public access. Survey respondents also were asked to provide the estimated time (hours required per month) to collect the data for quality metrics. Of 75 respondents providing data for this question, 28 (29.2%) and 17 (17.7%), estimated between 1 to 5 and 6 to 10 hours per month, respectively. Ten sites estimated spending between 11 to 20 hours, and 7 sites estimated spending more than 20 hours per month collecting quality metrics. A total of 13 respondents (13.5%) stated uncertainty about the time burden.

As shown in the Figure,  numerous quality metrics were collected across sites with more than 80% of sites collecting information on bowel preparation quality (88.5%), cecal intubation rate (87.5%), and complications (83.3%). A majority of sites also reported collecting data on appropriateness of surveillance intervals (62.5%), colonoscopy withdrawal times (62.5%), and ADRs (61.5%). Seven sites (7.3%) did not collect quality metrics.

Information also was collected on colonoscopy procedure documentation to inform future efforts at standardization. A small majority (53.1%) of sites reported using endoscopic software to generate colonoscopy procedure documentation. Within these sites, 6 different types of endoscopic note writing software were used to generate procedure notes (Table 2). 

Remaining sites used a variety of methods to generate procedure notes, including typing free-text notes (17.7%) and dictation (11.5%); 35.4% of sites use a template note in the Computerized Patient Record System. Sites also used a variety of methods for photo documentation in the electronic medical record (eg, pictures of cecal intubation and polyps).

Most sites (85.4%) were aware of VHA Directive 1015 recommendations for colonoscopy quality assurance programs. A significant majority (89.5%) of respondents also indicated interest in a centralized automatic reporting system to measure and report colonoscopy quality in some form, either with aggregate data, provider data, or both (Table 3).

 

 

Discussion

This survey on colonoscopy quality assurance programs is the first assessment of the VHA’s efforts to measure and report colonoscopy quality indicators. The findings indicated that the majority of VA sites are measuring and reporting at least some measures of colonoscopy quality. However, the programs are significantly variable in terms of methods used to collect quality metrics, specific quality measures obtained, and how quality is reported.

The authors’ work is novel in that this is the first report of the status of colonoscopy quality assurance programs in a large U.S. health care system. The VA health care system is the largest integrated health system in the U.S., serving more than 9 million veterans annually. This survey’s high response rate further strengthens the findings. Specifically, the survey found that VA sites are making a strong concerted effort to measure and report colonoscopy quality. However, there is significant variability in documentation, measurement, and reporting practices. Moreover, the majority of VA sites do not have formal performance improvement plans in place for endoscopists who do not meet thresholds for colonoscopy quality.

Screening colonoscopy for CRC offers known mortality benefits to patients.^1,17-19 Significant prior work has described and validated the importance of colonoscopy quality metrics, including bowel preparation quality, cecal intubation rate, and ADR and their association with interval colorectal cancer and death.^20-23 Gastroenterology professional societies, including the American College of Gastroenterology and the American Society for Gastrointestinal Endoscopy, have recommended and endorsed measurement and reporting of colonoscopy metrics.²⁴ There is general agreement among endoscopists that colonoscopy quality is an important aspect of performing the procedure.

The lack of formal performance improvement programs is a key finding of this survey. Recent studies have shown that improvements in quality metrics, such as the ADR, by individual endoscopists result in reductions in interval colorectal cancer and death.²⁵ Kahi and colleagues previously showed that providing a quarterly report card improves colonoscopy quality.²⁶ Keswani and colleagues studied a combination of a report card and implementation of standards of practice with resultant improvement in colonoscopy quality.²⁷ Most recently, in a large prospective cohort study of individuals who underwent a screening colonoscopy, 294 of the screening endoscopists received annual feedback and quality benchmark indicators to improve colonoscopy performance.²⁵ The majority of the endoscopists (74.5%) increased their annual ADR category over the study period. Moreover, patients examined by endoscopists who reached or maintained the highest ADR quintile (> 24.6%) had significantly lower risk of interval CRC and death. The lack of formal performance improvement programs across the VHA is concerning but reveals a significant opportunity to improve veteran health outcomes on a large scale.

This study’s findings also highlight the intense resources necessary to measure and report colonoscopy quality. The ability to measure and report quality metrics requires having adequate documentation and data to obtain quality metrics. Administrative databases from electronic health records offer some potential for routine monitoring of quality metrics.²⁸ However, most administrative databases, including the VA Corporate Data Warehouse (CDW), contain administrative billing codes (ICD and CPT) linked to limited patient data, including demographics and structured medical record data. The actual data required for quality reporting of important metrics (bowel preparation quality, cecal intubation rates, and ADRs) are usually found in clinical text notes or endoscopic note documentation and not available as structured data. Due to this issue, the majority of VA sites (79.2%) are using manual chart review to collect quality metric data, resulting in widely variable estimates on time burden. A minority of sites in this study (39.6%) reported using automated endoscopic software reporting capability that can help with the time burden. However, even in the VA, an integrated health system, a wide variety of software brands, documentation practices, and photo documentation was found.

Future endoscopy budget and purchase decisions for the individual VA sites should take into account how new technology and software can more easily facilitate accurate quality reporting. A specific policy recommendation would be for the VA to consider a uniform endoscopic note writer for procedure notes. Pathology data, which is necessary for the calculation of ADR, also should be available as structured data in the CDW to more easily measure colonoscopy quality. Continuous measurement and reporting of quality also requires ongoing information technology infrastructure and quality control of the measurement process.

 

Limitations

This survey was a cross-section of VA sites’ points of contact regarding colonoscopy quality assurance programs, so the results are descriptive in nature. However, the instrument was carefully developed, using both subject matter and survey method expertise. The questionnaire also was refined through pretesting prior to data collection. The initial contact list was found to have errors, and the list had to be updated after launching the survey. Updated information for most of the contacts was available.

Another limitation was the inability to survey nongastroenterologist-run endoscopy centers, because many centers use surgeons or other nongastroenterology providers. The authors speculate that quality monitoring may be less likely to be present at these facilities as they may not be aware of the gastroenterology professional society recommendations. The authors did not require or insist that all questions be answered, so some data were missing from sites. However, 93.7% of respondents completed the entire survey.

Conclusion

The authors have described the status of colonoscopy quality assurance programs across the VA health care system. Many sites are making robust efforts to measure and report quality especially of process measures. However, there are significant time and manual workforce efforts required, and this work is likely associated with the variability in programs. Importantly, ADR, which is the quality metric that has been most strongly associated with risk of colon cancer mortality, is not being measured by 38% of sites.

These results reinforce a critical need for a centralized, automated quality reporting infrastructure to standardize colonoscopy quality reporting, reduce workload, and ensure veterans receive high-quality colonoscopy.

Acknowledgments
The authors acknowledge the support and feedback of the National Gastroenterology Program Field Advisory Committee for survey development and testing. The authors coordinated the survey through the Salt Lake City Specialty Care Center of Innovation in partnership with the National Gastroenterology Program Office and the Quality Enhancement Research Initiative: Quality Enhancement Research Initiative, Measurement Science Program, QUE15-283. The work also was partially supported by the National Center for Advancing Translational Sciences of the National Institutes of Health Award UL1TR001067 and Merit Review Award 1 I01 HX001574-01A1 from the United States Department of Veterans Affairs Health Services Research & Development Service of the VA Office of Research and Development.

ABSTRACT

Glenoid bone deficiency in the setting of shoulder replacement surgery is far more common than originally reported. The frequency and severity of the glenoid defects are noted to be more common and severe with the advent of computer-assisted surgery. The results of an anatomic total shoulder arthroplasty (aTSA) with glenoid deficiency have been reported to be inferior to aTSA patients without a glenoid deficiency. Options for treating the glenoid deficiency include eccentric reaming, bone grafting, and the use of augmented glenoid components. The purpose of this article is to present the indications, technique, and results of augmented glenoids for both aTSA and reverse TSA (RTSA).

Augments for both aTSA and RTSA are viable options. They preserve subchondral bone at the same time as optimizing the joint line without the need for bone grafts. Complications, revisions and results are as good as compared to shoulder arthroplasties without glenoid wear.

Continue to: Glenoid bone deficiency...

 

Glenoid bone deficiency in arthritic or cuff-deficient shoulder has been reported in up to 50% of shoulder defect cases.^1,2 The type and severity of glenoid deformities vary depending on the underlying pathology and time of manifestation. Osteoarthritis with bone loss typically results in posterior or posterior inferior glenoid wear and is commonly classified as Walch types B1 or B2 (biconcave). In cases of severe erosion, B3 classification has been proposed; in this classification, bone loss becomes extremely severe, progressing to resemble a type C glenoid. Unlike primary osteoarthritis, inflammatory arthropathy more commonly causes central loss of glenoid bone (Walch A2). With the rotator cuff insufficiency, superior migration of the humeral head occurs. As these conditions progress, cuff tear arthropathy (CTA) changes result in superior or posterior-superior bone loss.¹ Anterior bone loss (type D) will be rarely encountered due to recurrent anterior instability.³

Classically, with anatomic total shoulder arthroplasty (aTSA), the surgeon considers several options for managing glenoid deficiencies. The most commonly employed technique involves eccentrically reaming the glenoid and correcting the deformity. This procedure is relatively easy but features significant drawbacks, such as sacrificing the subchondral bone, medializing the glenohumeral joint line, and secondarily shrinking the glenoid surface area. Other options include structural bone grafting behind the glenoid component. Most anatomic prosthetic glenoids prove to be unsuitable for fixation of structural bone graft. Therefore, the graft is first internally fixed, followed by placement of the glenoid component. Cement, which is commonly used for glenoid fixation, may potentially inhibit bone-graft healing. Reports using this technique documented high radiographic failure rate of up to 40% at midterm follow-up.⁴ Although leaving the glenoid component retroverted may be considered, surgeons should develop awareness of the possibility of peg penetration of the anterior glenoid neck. Additionally, retroversion in excess of 5°may increase the risk of recurrent posterior subluxation, resulting in early glenoid loosening.^5-7 Results of aTSA under significant glenoid deficiency are inferior to those of aTSA patients without glenoid deficiency.⁸ Such findings have been extremely inferior in patients with significant glenoid wear, prompting numerous surgeons to abandon aTSA in this population in favor of reverse TSA (RTSA) due to improved bony fixation.

In 2010, augmented anatomic glenoids were first introduced as a wedge (Exactech) and as a step shortly thereafter (DePuy Synthes; Figures 1A-1C). More recently, hemi-wedges have been introduced (Wright Medical Group). Augments have gained popularity due to improved range of motion vs reverse shoulder arthroplasty (RSA). However, debates remain regarding the use of posteriorly augmented components in the setting of posterior glenoid bone loss.⁸ Augments serve as another viable option for handling glenoid bone deficiency in aTSA.

Glenoid bone loss in RTSA presents similar options to aTSA. However, screw fixation of the glenoid component offers several distinct advantages. Baseplate fixation can readily be used with bone grafting and with a highly anticipated success rate. With multiple screw options, 100% support of the baseplate is not mandatory. Although bony increase offset RSAs (BIO-RSAs) have shown success, augmentation with allograft or autograft increases operative time and relies on osseous integration for long-term implant success.⁹ Metal augmented baseplates were first introduced in 2011 (Exactech) as a means of managing glenoid bone loss without structural grafting. Although initial results have been encouraging, additional studies are needed to assess the longevity of these implants (Figures 1A-1C).

aTSA AUGMENTS

aTSA augments were introduced as a means of correcting acquired glenoid bone deficiency, restoring native glenoid version, correcting humeral subluxation, and preserving the native subchondral bone. Compared with glenoid bone grafting, augmented glenoid components decrease operative time, allow for a technically easier operation, and require no bone healing for clinical success. Early and midterm results are encouraging, showing similar findings comparable to those of aTSA in non-glenoid deficient shoulders.^10-12

Continue to: INDICATIONS

 

 

INDICATIONS

Indications and limitations for augmented aTSA glenoids remain incompletely defined. The most common indication for an augmented aTSA is osteoarthritis with a B2 glenoid. We recommend augments in the occurrence of any indication of significant eccentric glenoid wear. With the expertise of surgeons, deformities of up to 20° to 25° of deformity can be readily handled with good predictability. More severe deformities can be managed with augmented aTSA components, but early failure rates may be high. The most severe acquired deformities remain best managed with RTSA. Currently, we prefer RTSA when glenoid bone loss exceeds 25°. With the widespread availability of computed tomography (CT) scans with 3-dimensional (3-D) reconstruction, glenoid bone defects are increasingly recognized. When correcting deformity, surgeons should strive to limit residual retroversion to a maximum of 5°.¹³ Preoperative planning software and computer-assisted surgery (ExactechGPS) may allow surgeons to better define the limits of augmented glenoid fixation prior to the date of surgery. We routinely utilize computer-guided glenoid preparation to control glenoid version to within 5° of neutral position.

The differences between B3 and a true type C glenoid must be recognized. Although B3 glenoids may still be a candidate for an augmented anatomic glenoid component, type C glenoids are not. Developmental abnormalities of type C glenoid occur simultaneously with humeral deformities, including medialized posterior rotator cuff musculature. Correction of the joint line to neutral version may not replicate the non-diseased state of a dysplastic type shoulder. Davis and colleagues¹⁴ have proposed treating these patients by leaving both the humerus and glenoid in their native version without correction.

TECHNIQUE

The implant that we have the most experience with is an 8° full-wedge augmented glenoid component. Such an implant is typically utilized for B2 glenoids. We recommend that a high-quality CT scan be performed for preoperative planning. As a general rule, the starting point often lies close to the ridge of B2 glenoid and more anterior than the apparent glenoid center, which is viewed intraoperatively due to asymmetric posterior wear. Full-wedge component is utilized to ream the ridge separating the neo and paleoglenoids to create a flat surface. This condition is best achieved by drilling a pilot hole at the planned glenoid central peg position to prevent the reamer from sliding anteriorly during reaming. Glenoid preparation begins with the smallest reamer until the ridge has been flattened, and the reamer makes full contact with the glenoid. The reamer diameter is then increased based on glenoid size. Slightly downsizing the glenoid implant will require less reaming to achieve full backside support. Once the glenoid is properly reamed, the central and peripheral peg holes are drilled using the appropriate guides. Holes are then dried, and all-polyethylene or composite glenoid component (either partially or completely cemented) is installed using favored cementing techniques. The advantage of composite glenoid component is that the central cage allows for bone ingrowth and may potentially improve long-term implant survival. Press fit of the central cage requires no waiting time for glenoid cement hardening before proceeding to the humerus. When placing an augmented component, adequate glenoid exposure is imperative to allow in-line placement and appropriate seating of the component without impingement on adjacent retractors.

When using the step-augmented glenoid, the paleoglenoid is prepared in a similar fashion to a standard aTSA. Once the paleoglenoid has been reamed to a neutral position, a protector plate is placed onto the paleoglenoid. and a step-cut saw is used to prepare the posterior stepped bone cut. Peripheral pegs are then drilled, and the component is installed in routine fashion. When using hemi-wedge augments, the paleoglenoid is again prepared in a similar fashion as a standard glenoid component over a cannulated guidewire. The neoglenoid is subsequently prepared using a specialized angled reamer with a positive stop to prevent over-reaming. These glenoid implants improve rotational force neutralization given the absence of flat back against the glenoid. All 3 designs preserve bone when compared with eccentric reaming alone,¹⁵ with the half-augmented wedge preserving the most bone.

Table 1. Results of Various Augmented Glenoid Components in Anatomic Total Shoulder

Arthroplasty

Augment		American Shoulder and Elbow Surgeons Score	Constant Score	Active Forward Flexion	Active External Rotation
8° cage (N = 21)	Preoperative Postoperative Change	41 91 50	42 73 31	98 142 44	27 51 24
All-polyethylene 8° (N = 45)	Preoperative Postoperative Change	43 89 46	39 74 35	98 142 44	14 45 31
All-polyethylene 16° (N = 7)	Preoperative Postoperative Change	41 74 33	38 59 21	100 128 28	9 47 38

 

RESULTS

In our institution, we first used all-polyethylene posteriorly augmented glenoid components in 2010. Between 2010 and 2015, 45 patients received an 8° all-polyethylene posterior augment, and 7 patients received a 16° augment. In 2015, we transitioned to the composite caged posterior augment. All patients in our database who received an augmented glenoid component experienced improvement in active forward elevation, external rotation, American Shoulder and Elbow Surgeons (ASES), and Constant scores (Table 1). Minimum follow-up was 1 year for patients receiving both an 8° cage (mean, 1.48 years) and an 8° all-polyethylene augment (mean, 3.18 years). Figures 2A-2C show a patient with significant posterior glenoid wear and humeral head subluxation treated with an 8° wedge composite posterior augment glenoid 3 years postoperative.

Continue to: COMPLICATIONS

 

 

COMPLICATIONS

Two complications developed in the group undergoing composite cage augment. One patient experienced glenoid loosening after a motor vehicle accident. Another patient sustained significant intraoperative tuberosity avulsion during implantation of the humeral component, requiring a change of implant and tuberosity fixation. Although no complications were noted in the 8° all-polyethylene group, 3 patients in the 16° augment group sustained complications. One of these patients suffered a cardiac event that was unrelated to the implant. Two complications in this group were both related to loosening of the glenoid component, requiring subsequent revision.

DISCUSSION

The first report on augmented aTSA was published in 2008, and it involved a 5° augmented, anatomic glenoid.¹² One study was based on a small series of augments; the poor results led the reporting surgeons to subsequently abandon the implant.¹² This early design produced a correction on the articular side of the implant rather than the pathologic bony side. By performing such correction, the component pegs remained anteriorly oriented, placing the component at risk of perforation through the anterior glenoid neck. All current augment designs feature pegs that are oriented down the glenoid vault, with corrections occurring on the bony surface. This condition requires 2 different axes for reaming the glenoid and drilling the pegs. This approach allows the pegs to be directed down the glenoid neck, and is a far superior solution to neutralizing shear forces when compared with the implants used in the 1990s.

Early to midterm results of modern aTSA augments have been extremely encouraging with low revision rates. The main concern of recurrent posterior subluxation has been rarely reported. The concerns over glenoid loosening due to high shear forces, similarly, have not been described to date. However, surgeons should remain cautious, as longer-term follow-up remains unavailable.

The main advantage of aTSA augments is their capacity to preserve bone compared with eccentric reaming and better long-term stability. Each of the augment designs requires varying amounts of bone removal. Through biomechanics and using finite element analysis, the 3 augment types act differently, with no design demonstrating remarkable biomechanical superiority.⁶ Favorito and colleagues¹⁶ performed a retrospective review of 22 patients who underwent aTSA using an all-polyethylene, posteriorly augmented, and stepped glenoid component for posterior bone loss. At an average follow-up of 36 months, all patients experienced improvements in active forward elevation, external rotation, visual analog scale, Short Form-36 Physical Component Summary, and Western Ontario Osteoarthritis of the Shoulder scores. The authors noted that 2 patients (9%) experienced complications: 1 with an anterior dislocation and the other with recurrent posterior instability requiring revision. Sandow and Schutz¹⁷ reported the preliminary results of 10 patients who underwent aTSA using trabecular metal augment with a minimum of 2-year follow-up. All patients received either a 15° or 30° posterior, metal-backed augment for severe glenoid bone loss (Walch grade B2 or C). At a minimum of 2-year follow-up, all patients received correction to within 10° of neutral glenoid version, without any complications nor implant failures.

Regardless of augment design, all current components restore the native glenoid version, improving the length and subsequent tension of rotator cuff musculature. Similarly, re-centering the humeral head decreases the forces on the glenoid and allows for optimal function with decreasing loss of vital subchondral bone.  

Continue to: RTSA AUGMENTS

 

 

RTSA AUGMENTS

Similar to anatomic augments, metal augments were introduced for use with RTSA in 2011. Unlike anatomic augments, those for RTSA were manufactured with metal. Given the difference in bony wear patterns in patients requiring RTSA, augments were available in a number of configurations. With CTA, wear is most commonly superior. Leaving a superiorly inclined baseplate must be avoided due to risks of notching, loosening, and early failure. However, correcting this tilt will require significant reaming of the inferior glenoid. A superior augment is ideally suited for this bone-loss pattern. If the glenoid is retroverted significantly, difficulty can also arise during glenoid preparation and baseplate placement. Posterior augments may ease this aspect of the procedure. Posterior augments feature the additional benefits of tensioning any remaining posterior rotator cuff, minimizing posterior inferior impingement, and technically easing the operation.¹⁸ As we improve our awareness of glenoid orientation using computer navigation, a posterior-superior augmented implant is commonly needed to simultaneously optimize the baseplate position and to minimize reaming (Figure 3). The posterior-superior augmented baseplate has become the most commonly used baseplate augment of choice in 90% of our RTSA cases that require an augment.

INDICATIONS

Augmented RTSA baseplates are indicated when adequate backside contact cannot be achieved with eccentric reaming, thus compromising potential fixation. In our practice, we preferably use augments at <50% contact with the backside of the baseplate. Excessive superior inclination is observed in a CTA setting, commonly indicating the use of superior augments. Similarly, severe primary osteoarthritis may contain elements of posterior bone loss, leading to increased retroversion, which is where we use posterior augments. When patients exhibit combined deformities, or when the surgeon wishes to tension the posterior rotator cuff, a posterior-superior augmented glenoid baseplate is used. For extremely severe defects, we have combined bone grafting and augments. In patients with a highly deficient glenoid but good quality of the remaining bone stock, an augment allows for better contact with less reaming although it is not fully supported when compared with a non-augmented baseplate. Bone grafts can function similarly, but the autograft humeral head is not constantly present in revision situations and requires increased operative time to allow for precision carpentry. Additionally, the success of BIO-RSA requires healing of bone graft on the native glenoid to support the baseplate.¹⁹ Jones and colleagues⁹ compared metal augmented RTSA with BIO-RSA and presented equivalent results.

To minimize reaming and to obtain appropriately inferior inclination, we have discovered preoperative templating and intraoperative, computer-guided glenoid preparation to be extremely valuable (ExactechGPS). These tools allow appropriate assessment of augments and for minimal bone removal when preparing the glenoid.

TECHNIQUE

When using an augment, a fine-cut CT scan is highly recommended to aid in surgery planning. We also find 3-D reconstructions to be helpful. Preoperative planning software also allows surgeons to maximize fixation of implant within the glenoid vault. The starting point for reaming is planned based on CT. Some surgeons using augments perform minimal or no reaming at all, electing to remove the remaining cartilage with a Cobb elevator. Different reaming and drilling axes are used when using augments. In cases of severe glenoid deformity and unavailability of computer assistance, a guide wire with inferior inclination can be installed based on CT scan. Penetration of this wire down the glenoid neck can be palpated and compared with the preoperative plan. We generally prefer at least 24 mm of bone containment for the central cage. Once the surgeon is satisfied with the placement of the wire, the appropriate augment guide is placed, followed by a second guide wire. This second wire acts as the reaming axis. The first wire is removed, and the glenoid is reamed with a cannulated reamer. Once reaming is completed, the original wire is replaced in the same hole and trajectory, and the reaming wire is removed. The first wire is then drilled with a cannulated drill for the central cage. The augmented baseplate is then impacted into place, and screw fixation is performed. Again, intraoperative computer guidance allows for precision screw placement with maximal bone attachment.

Table 2. Results of Reverse Total Shoulder Arthroplasty Augmented Baseplates

Augment		American Shoulder and Elbow Surgeons Score	Constant Score	Active Forward Flexion	Active External Rotation
Superior (N = 22)	Preoperative Postoperative Change	35 74 38	32 59 26	77 118 42	30 35 5
Posterior (N = 50)	Preoperative Postoperative Change	39 87 48	34 70 36	85 133 47	16 30 13
Posterosuperior (N = 67)	Preoperative Postoperative Change	36 80 44	34 66 32	82 132 51	17 35 18

RESULTS

Based on our experience, glenoid augments for RTSA have performed well at short- and mid-term follow-up. From October 2011 to July 2016, 139 patients undergoing RTSA received a posterior, superior, or posterior-superior augmented glenoid baseplate. All groups demonstrated improvements in functional outcome measures, including Constant, ASES, Shoulder Pain and Disability Index, and Simple Shoulder Test scores compared with baseline values (Table 2). The posterior-superior augment group experienced the most significant improvement in active forward flexion and external rotation, whereas the posterior augment group experienced the most significant improvement in ASES and Constant scores. Figures 4A-4C displays the radiographs of a patient with significant glenoid wear treated with a posterior-superior augment RTSA.

Continue to: COMPLICATIONS

 

 

COMPLICATIONS

In the superior augment group, 3 patients (13%) sustained 5 complications. One patient sustained 3 separate episodes of instability, eventually requiring revision of prosthesis. In the posterior augment group, 4 patients (8%) sustained complications. Two of the 4 patients presented postoperative humeral fractures related to traumatic events, whereas another patient sustained an intraoperative tuberosity fracture. The last complication in this group involved a postoperative draining wound that was treated with oral antibiotics.

Nine complications developed in the posterior-superior augment group (13%); these complications included aseptic baseplate loosening (5), glenoid fracture (1), humeral fracture (1), acromial stress fracture (1), and cerebrovascular accident (1).

DISCUSSION

As the use of augments in RTSA is relatively new, significantly scarce data exist regarding their outcomes and longevity. A few studies have focused on the short-term outcomes of these augments. Jones and colleagues⁹ performed a retrospective review of 80 patients who underwent RTSA and required either a structural bone graft or an augmented glenoid baseplate.⁹ They observed that although all patients showed improvements in pain, range of motion, and functional scores, the structural bone graft group incurred a 14.6% complication rate compared with none observed in the augment group. Additionally, Jones and colleagues⁹ noted that the augmented baseplate group exhibited a significantly lower rate of scapular notching compared with the bone-graft group (10% vs 18.5%) at similar follow-up intervals. A separate study by Wright and colleagues¹⁸ compared posterior vs superior augmented baseplates in RTSA. The posterior augment group demonstrated lower rates of scapular notching (6.3% vs 14.3%) and showed more significant improvements in Constant, ASES, and active forward elevation measures, compared with the superior augment group.

As more manufacturers develop augments for RTSA, and as ExactechGPS uses become more widespread, the use of RTSA baseplate augments will continually grow. Custom implants using massive metal augments are now also being introduced. Although currently too expensive for most cases, as technology drives the cost down, every patient may receive customized augmented implants in the future.

The advantages of augmented baseplate designs include minimized reaming and notching, improved tension of the remaining rotator cuff, and decreased operating room time. The disadvantages include increased cost and lack of mid- or long-term clinical data. The concerns with baseplate loosening with augments in RTSA are much less than those with augments for aTSA due to the outstanding baseplate fixation that can be achieved in RTSA.

Continue to: CONLCLUSION

 

 

CONCLUSION

Augments offer an excellent tool for surgeons performing both aTSA and RTSA with glenoid bone loss. Use of augments will become more common as more manufacturers develop them. Although clinical results fall short in full midterm, they have been positive for both augmented RTSA and aTSA. Concerns arise when performing augmented aTSA, as an upper limit of correction has not been defined with regard to component failure. Currently, no data support the maximum amount of correction that can be achieved. In our current practice, we face difficulty in correcting more than 25^° of version in young active patients with aTSA augment. Beyond this point, we perform a RTSA with an augment. In older patients or low-demand patients, we only correct minor deformities (<20^°) with an aTSA augment, opting instead for an augmented RTSA due to the lower midterm failure rates observed with this implant.

ABSTRACT

Glenoid bone deficiency in the setting of shoulder replacement surgery is far more common than originally reported. The frequency and severity of the glenoid defects are noted to be more common and severe with the advent of computer-assisted surgery. The results of an anatomic total shoulder arthroplasty (aTSA) with glenoid deficiency have been reported to be inferior to aTSA patients without a glenoid deficiency. Options for treating the glenoid deficiency include eccentric reaming, bone grafting, and the use of augmented glenoid components. The purpose of this article is to present the indications, technique, and results of augmented glenoids for both aTSA and reverse TSA (RTSA).

Augments for both aTSA and RTSA are viable options. They preserve subchondral bone at the same time as optimizing the joint line without the need for bone grafts. Complications, revisions and results are as good as compared to shoulder arthroplasties without glenoid wear.

Continue to: Glenoid bone deficiency...

 

Glenoid bone deficiency in arthritic or cuff-deficient shoulder has been reported in up to 50% of shoulder defect cases.^1,2 The type and severity of glenoid deformities vary depending on the underlying pathology and time of manifestation. Osteoarthritis with bone loss typically results in posterior or posterior inferior glenoid wear and is commonly classified as Walch types B1 or B2 (biconcave). In cases of severe erosion, B3 classification has been proposed; in this classification, bone loss becomes extremely severe, progressing to resemble a type C glenoid. Unlike primary osteoarthritis, inflammatory arthropathy more commonly causes central loss of glenoid bone (Walch A2). With the rotator cuff insufficiency, superior migration of the humeral head occurs. As these conditions progress, cuff tear arthropathy (CTA) changes result in superior or posterior-superior bone loss.¹ Anterior bone loss (type D) will be rarely encountered due to recurrent anterior instability.³

Classically, with anatomic total shoulder arthroplasty (aTSA), the surgeon considers several options for managing glenoid deficiencies. The most commonly employed technique involves eccentrically reaming the glenoid and correcting the deformity. This procedure is relatively easy but features significant drawbacks, such as sacrificing the subchondral bone, medializing the glenohumeral joint line, and secondarily shrinking the glenoid surface area. Other options include structural bone grafting behind the glenoid component. Most anatomic prosthetic glenoids prove to be unsuitable for fixation of structural bone graft. Therefore, the graft is first internally fixed, followed by placement of the glenoid component. Cement, which is commonly used for glenoid fixation, may potentially inhibit bone-graft healing. Reports using this technique documented high radiographic failure rate of up to 40% at midterm follow-up.⁴ Although leaving the glenoid component retroverted may be considered, surgeons should develop awareness of the possibility of peg penetration of the anterior glenoid neck. Additionally, retroversion in excess of 5°may increase the risk of recurrent posterior subluxation, resulting in early glenoid loosening.^5-7 Results of aTSA under significant glenoid deficiency are inferior to those of aTSA patients without glenoid deficiency.⁸ Such findings have been extremely inferior in patients with significant glenoid wear, prompting numerous surgeons to abandon aTSA in this population in favor of reverse TSA (RTSA) due to improved bony fixation.

In 2010, augmented anatomic glenoids were first introduced as a wedge (Exactech) and as a step shortly thereafter (DePuy Synthes; Figures 1A-1C). More recently, hemi-wedges have been introduced (Wright Medical Group). Augments have gained popularity due to improved range of motion vs reverse shoulder arthroplasty (RSA). However, debates remain regarding the use of posteriorly augmented components in the setting of posterior glenoid bone loss.⁸ Augments serve as another viable option for handling glenoid bone deficiency in aTSA.

Glenoid bone loss in RTSA presents similar options to aTSA. However, screw fixation of the glenoid component offers several distinct advantages. Baseplate fixation can readily be used with bone grafting and with a highly anticipated success rate. With multiple screw options, 100% support of the baseplate is not mandatory. Although bony increase offset RSAs (BIO-RSAs) have shown success, augmentation with allograft or autograft increases operative time and relies on osseous integration for long-term implant success.⁹ Metal augmented baseplates were first introduced in 2011 (Exactech) as a means of managing glenoid bone loss without structural grafting. Although initial results have been encouraging, additional studies are needed to assess the longevity of these implants (Figures 1A-1C).

aTSA AUGMENTS

aTSA augments were introduced as a means of correcting acquired glenoid bone deficiency, restoring native glenoid version, correcting humeral subluxation, and preserving the native subchondral bone. Compared with glenoid bone grafting, augmented glenoid components decrease operative time, allow for a technically easier operation, and require no bone healing for clinical success. Early and midterm results are encouraging, showing similar findings comparable to those of aTSA in non-glenoid deficient shoulders.^10-12

Continue to: INDICATIONS

 

 

INDICATIONS

Indications and limitations for augmented aTSA glenoids remain incompletely defined. The most common indication for an augmented aTSA is osteoarthritis with a B2 glenoid. We recommend augments in the occurrence of any indication of significant eccentric glenoid wear. With the expertise of surgeons, deformities of up to 20° to 25° of deformity can be readily handled with good predictability. More severe deformities can be managed with augmented aTSA components, but early failure rates may be high. The most severe acquired deformities remain best managed with RTSA. Currently, we prefer RTSA when glenoid bone loss exceeds 25°. With the widespread availability of computed tomography (CT) scans with 3-dimensional (3-D) reconstruction, glenoid bone defects are increasingly recognized. When correcting deformity, surgeons should strive to limit residual retroversion to a maximum of 5°.¹³ Preoperative planning software and computer-assisted surgery (ExactechGPS) may allow surgeons to better define the limits of augmented glenoid fixation prior to the date of surgery. We routinely utilize computer-guided glenoid preparation to control glenoid version to within 5° of neutral position.

The differences between B3 and a true type C glenoid must be recognized. Although B3 glenoids may still be a candidate for an augmented anatomic glenoid component, type C glenoids are not. Developmental abnormalities of type C glenoid occur simultaneously with humeral deformities, including medialized posterior rotator cuff musculature. Correction of the joint line to neutral version may not replicate the non-diseased state of a dysplastic type shoulder. Davis and colleagues¹⁴ have proposed treating these patients by leaving both the humerus and glenoid in their native version without correction.

TECHNIQUE

The implant that we have the most experience with is an 8° full-wedge augmented glenoid component. Such an implant is typically utilized for B2 glenoids. We recommend that a high-quality CT scan be performed for preoperative planning. As a general rule, the starting point often lies close to the ridge of B2 glenoid and more anterior than the apparent glenoid center, which is viewed intraoperatively due to asymmetric posterior wear. Full-wedge component is utilized to ream the ridge separating the neo and paleoglenoids to create a flat surface. This condition is best achieved by drilling a pilot hole at the planned glenoid central peg position to prevent the reamer from sliding anteriorly during reaming. Glenoid preparation begins with the smallest reamer until the ridge has been flattened, and the reamer makes full contact with the glenoid. The reamer diameter is then increased based on glenoid size. Slightly downsizing the glenoid implant will require less reaming to achieve full backside support. Once the glenoid is properly reamed, the central and peripheral peg holes are drilled using the appropriate guides. Holes are then dried, and all-polyethylene or composite glenoid component (either partially or completely cemented) is installed using favored cementing techniques. The advantage of composite glenoid component is that the central cage allows for bone ingrowth and may potentially improve long-term implant survival. Press fit of the central cage requires no waiting time for glenoid cement hardening before proceeding to the humerus. When placing an augmented component, adequate glenoid exposure is imperative to allow in-line placement and appropriate seating of the component without impingement on adjacent retractors.

When using the step-augmented glenoid, the paleoglenoid is prepared in a similar fashion to a standard aTSA. Once the paleoglenoid has been reamed to a neutral position, a protector plate is placed onto the paleoglenoid. and a step-cut saw is used to prepare the posterior stepped bone cut. Peripheral pegs are then drilled, and the component is installed in routine fashion. When using hemi-wedge augments, the paleoglenoid is again prepared in a similar fashion as a standard glenoid component over a cannulated guidewire. The neoglenoid is subsequently prepared using a specialized angled reamer with a positive stop to prevent over-reaming. These glenoid implants improve rotational force neutralization given the absence of flat back against the glenoid. All 3 designs preserve bone when compared with eccentric reaming alone,¹⁵ with the half-augmented wedge preserving the most bone.

Table 1. Results of Various Augmented Glenoid Components in Anatomic Total Shoulder

Arthroplasty

Augment		American Shoulder and Elbow Surgeons Score	Constant Score	Active Forward Flexion	Active External Rotation
8° cage (N = 21)	Preoperative Postoperative Change	41 91 50	42 73 31	98 142 44	27 51 24
All-polyethylene 8° (N = 45)	Preoperative Postoperative Change	43 89 46	39 74 35	98 142 44	14 45 31
All-polyethylene 16° (N = 7)	Preoperative Postoperative Change	41 74 33	38 59 21	100 128 28	9 47 38

 

RESULTS

In our institution, we first used all-polyethylene posteriorly augmented glenoid components in 2010. Between 2010 and 2015, 45 patients received an 8° all-polyethylene posterior augment, and 7 patients received a 16° augment. In 2015, we transitioned to the composite caged posterior augment. All patients in our database who received an augmented glenoid component experienced improvement in active forward elevation, external rotation, American Shoulder and Elbow Surgeons (ASES), and Constant scores (Table 1). Minimum follow-up was 1 year for patients receiving both an 8° cage (mean, 1.48 years) and an 8° all-polyethylene augment (mean, 3.18 years). Figures 2A-2C show a patient with significant posterior glenoid wear and humeral head subluxation treated with an 8° wedge composite posterior augment glenoid 3 years postoperative.

Continue to: COMPLICATIONS

 

 

COMPLICATIONS

Two complications developed in the group undergoing composite cage augment. One patient experienced glenoid loosening after a motor vehicle accident. Another patient sustained significant intraoperative tuberosity avulsion during implantation of the humeral component, requiring a change of implant and tuberosity fixation. Although no complications were noted in the 8° all-polyethylene group, 3 patients in the 16° augment group sustained complications. One of these patients suffered a cardiac event that was unrelated to the implant. Two complications in this group were both related to loosening of the glenoid component, requiring subsequent revision.

DISCUSSION

The first report on augmented aTSA was published in 2008, and it involved a 5° augmented, anatomic glenoid.¹² One study was based on a small series of augments; the poor results led the reporting surgeons to subsequently abandon the implant.¹² This early design produced a correction on the articular side of the implant rather than the pathologic bony side. By performing such correction, the component pegs remained anteriorly oriented, placing the component at risk of perforation through the anterior glenoid neck. All current augment designs feature pegs that are oriented down the glenoid vault, with corrections occurring on the bony surface. This condition requires 2 different axes for reaming the glenoid and drilling the pegs. This approach allows the pegs to be directed down the glenoid neck, and is a far superior solution to neutralizing shear forces when compared with the implants used in the 1990s.

Early to midterm results of modern aTSA augments have been extremely encouraging with low revision rates. The main concern of recurrent posterior subluxation has been rarely reported. The concerns over glenoid loosening due to high shear forces, similarly, have not been described to date. However, surgeons should remain cautious, as longer-term follow-up remains unavailable.

The main advantage of aTSA augments is their capacity to preserve bone compared with eccentric reaming and better long-term stability. Each of the augment designs requires varying amounts of bone removal. Through biomechanics and using finite element analysis, the 3 augment types act differently, with no design demonstrating remarkable biomechanical superiority.⁶ Favorito and colleagues¹⁶ performed a retrospective review of 22 patients who underwent aTSA using an all-polyethylene, posteriorly augmented, and stepped glenoid component for posterior bone loss. At an average follow-up of 36 months, all patients experienced improvements in active forward elevation, external rotation, visual analog scale, Short Form-36 Physical Component Summary, and Western Ontario Osteoarthritis of the Shoulder scores. The authors noted that 2 patients (9%) experienced complications: 1 with an anterior dislocation and the other with recurrent posterior instability requiring revision. Sandow and Schutz¹⁷ reported the preliminary results of 10 patients who underwent aTSA using trabecular metal augment with a minimum of 2-year follow-up. All patients received either a 15° or 30° posterior, metal-backed augment for severe glenoid bone loss (Walch grade B2 or C). At a minimum of 2-year follow-up, all patients received correction to within 10° of neutral glenoid version, without any complications nor implant failures.

Regardless of augment design, all current components restore the native glenoid version, improving the length and subsequent tension of rotator cuff musculature. Similarly, re-centering the humeral head decreases the forces on the glenoid and allows for optimal function with decreasing loss of vital subchondral bone.  

Continue to: RTSA AUGMENTS

 

 

RTSA AUGMENTS

Similar to anatomic augments, metal augments were introduced for use with RTSA in 2011. Unlike anatomic augments, those for RTSA were manufactured with metal. Given the difference in bony wear patterns in patients requiring RTSA, augments were available in a number of configurations. With CTA, wear is most commonly superior. Leaving a superiorly inclined baseplate must be avoided due to risks of notching, loosening, and early failure. However, correcting this tilt will require significant reaming of the inferior glenoid. A superior augment is ideally suited for this bone-loss pattern. If the glenoid is retroverted significantly, difficulty can also arise during glenoid preparation and baseplate placement. Posterior augments may ease this aspect of the procedure. Posterior augments feature the additional benefits of tensioning any remaining posterior rotator cuff, minimizing posterior inferior impingement, and technically easing the operation.¹⁸ As we improve our awareness of glenoid orientation using computer navigation, a posterior-superior augmented implant is commonly needed to simultaneously optimize the baseplate position and to minimize reaming (Figure 3). The posterior-superior augmented baseplate has become the most commonly used baseplate augment of choice in 90% of our RTSA cases that require an augment.

INDICATIONS

Augmented RTSA baseplates are indicated when adequate backside contact cannot be achieved with eccentric reaming, thus compromising potential fixation. In our practice, we preferably use augments at <50% contact with the backside of the baseplate. Excessive superior inclination is observed in a CTA setting, commonly indicating the use of superior augments. Similarly, severe primary osteoarthritis may contain elements of posterior bone loss, leading to increased retroversion, which is where we use posterior augments. When patients exhibit combined deformities, or when the surgeon wishes to tension the posterior rotator cuff, a posterior-superior augmented glenoid baseplate is used. For extremely severe defects, we have combined bone grafting and augments. In patients with a highly deficient glenoid but good quality of the remaining bone stock, an augment allows for better contact with less reaming although it is not fully supported when compared with a non-augmented baseplate. Bone grafts can function similarly, but the autograft humeral head is not constantly present in revision situations and requires increased operative time to allow for precision carpentry. Additionally, the success of BIO-RSA requires healing of bone graft on the native glenoid to support the baseplate.¹⁹ Jones and colleagues⁹ compared metal augmented RTSA with BIO-RSA and presented equivalent results.

To minimize reaming and to obtain appropriately inferior inclination, we have discovered preoperative templating and intraoperative, computer-guided glenoid preparation to be extremely valuable (ExactechGPS). These tools allow appropriate assessment of augments and for minimal bone removal when preparing the glenoid.

TECHNIQUE

When using an augment, a fine-cut CT scan is highly recommended to aid in surgery planning. We also find 3-D reconstructions to be helpful. Preoperative planning software also allows surgeons to maximize fixation of implant within the glenoid vault. The starting point for reaming is planned based on CT. Some surgeons using augments perform minimal or no reaming at all, electing to remove the remaining cartilage with a Cobb elevator. Different reaming and drilling axes are used when using augments. In cases of severe glenoid deformity and unavailability of computer assistance, a guide wire with inferior inclination can be installed based on CT scan. Penetration of this wire down the glenoid neck can be palpated and compared with the preoperative plan. We generally prefer at least 24 mm of bone containment for the central cage. Once the surgeon is satisfied with the placement of the wire, the appropriate augment guide is placed, followed by a second guide wire. This second wire acts as the reaming axis. The first wire is removed, and the glenoid is reamed with a cannulated reamer. Once reaming is completed, the original wire is replaced in the same hole and trajectory, and the reaming wire is removed. The first wire is then drilled with a cannulated drill for the central cage. The augmented baseplate is then impacted into place, and screw fixation is performed. Again, intraoperative computer guidance allows for precision screw placement with maximal bone attachment.

Table 2. Results of Reverse Total Shoulder Arthroplasty Augmented Baseplates

Augment		American Shoulder and Elbow Surgeons Score	Constant Score	Active Forward Flexion	Active External Rotation
Superior (N = 22)	Preoperative Postoperative Change	35 74 38	32 59 26	77 118 42	30 35 5
Posterior (N = 50)	Preoperative Postoperative Change	39 87 48	34 70 36	85 133 47	16 30 13
Posterosuperior (N = 67)	Preoperative Postoperative Change	36 80 44	34 66 32	82 132 51	17 35 18

RESULTS

Based on our experience, glenoid augments for RTSA have performed well at short- and mid-term follow-up. From October 2011 to July 2016, 139 patients undergoing RTSA received a posterior, superior, or posterior-superior augmented glenoid baseplate. All groups demonstrated improvements in functional outcome measures, including Constant, ASES, Shoulder Pain and Disability Index, and Simple Shoulder Test scores compared with baseline values (Table 2). The posterior-superior augment group experienced the most significant improvement in active forward flexion and external rotation, whereas the posterior augment group experienced the most significant improvement in ASES and Constant scores. Figures 4A-4C displays the radiographs of a patient with significant glenoid wear treated with a posterior-superior augment RTSA.

Continue to: COMPLICATIONS

 

 

COMPLICATIONS

In the superior augment group, 3 patients (13%) sustained 5 complications. One patient sustained 3 separate episodes of instability, eventually requiring revision of prosthesis. In the posterior augment group, 4 patients (8%) sustained complications. Two of the 4 patients presented postoperative humeral fractures related to traumatic events, whereas another patient sustained an intraoperative tuberosity fracture. The last complication in this group involved a postoperative draining wound that was treated with oral antibiotics.

Nine complications developed in the posterior-superior augment group (13%); these complications included aseptic baseplate loosening (5), glenoid fracture (1), humeral fracture (1), acromial stress fracture (1), and cerebrovascular accident (1).

DISCUSSION

As the use of augments in RTSA is relatively new, significantly scarce data exist regarding their outcomes and longevity. A few studies have focused on the short-term outcomes of these augments. Jones and colleagues⁹ performed a retrospective review of 80 patients who underwent RTSA and required either a structural bone graft or an augmented glenoid baseplate.⁹ They observed that although all patients showed improvements in pain, range of motion, and functional scores, the structural bone graft group incurred a 14.6% complication rate compared with none observed in the augment group. Additionally, Jones and colleagues⁹ noted that the augmented baseplate group exhibited a significantly lower rate of scapular notching compared with the bone-graft group (10% vs 18.5%) at similar follow-up intervals. A separate study by Wright and colleagues¹⁸ compared posterior vs superior augmented baseplates in RTSA. The posterior augment group demonstrated lower rates of scapular notching (6.3% vs 14.3%) and showed more significant improvements in Constant, ASES, and active forward elevation measures, compared with the superior augment group.

As more manufacturers develop augments for RTSA, and as ExactechGPS uses become more widespread, the use of RTSA baseplate augments will continually grow. Custom implants using massive metal augments are now also being introduced. Although currently too expensive for most cases, as technology drives the cost down, every patient may receive customized augmented implants in the future.

The advantages of augmented baseplate designs include minimized reaming and notching, improved tension of the remaining rotator cuff, and decreased operating room time. The disadvantages include increased cost and lack of mid- or long-term clinical data. The concerns with baseplate loosening with augments in RTSA are much less than those with augments for aTSA due to the outstanding baseplate fixation that can be achieved in RTSA.

Continue to: CONLCLUSION

 

 

CONCLUSION

Augments offer an excellent tool for surgeons performing both aTSA and RTSA with glenoid bone loss. Use of augments will become more common as more manufacturers develop them. Although clinical results fall short in full midterm, they have been positive for both augmented RTSA and aTSA. Concerns arise when performing augmented aTSA, as an upper limit of correction has not been defined with regard to component failure. Currently, no data support the maximum amount of correction that can be achieved. In our current practice, we face difficulty in correcting more than 25^° of version in young active patients with aTSA augment. Beyond this point, we perform a RTSA with an augment. In older patients or low-demand patients, we only correct minor deformities (<20^°) with an aTSA augment, opting instead for an augmented RTSA due to the lower midterm failure rates observed with this implant.