Bryce A. Basques, MD

ABSTRACT

This study uses a prospective surgical registry to characterize the timing of 10 postoperative adverse events following geriatric hip fracture surgery. There were 19,873 patients identified who were ≥70 years undergoing surgery for hip fracture as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). The median postoperative day of diagnosis (and interquartile range) for myocardial infarction was 3 (1-5), cardiac arrest requiring cardiopulmonary resuscitation 3 (0-8), stroke 3 (1-10), pneumonia 4 (2-10), pulmonary embolism 4 (2-11), urinary tract infection 7 (2-13), deep vein thrombosis 9 (4-16), sepsis 9 (4-18), mortality 11 (6-19), and surgical site infection 16 (11-22). For the earliest diagnosed adverse events, the rate of adverse events had diminished by postoperative day 30. For the later diagnosed adverse events, the rate of adverse events remained high at postoperative day 30. Findings help to enable more targeted clinical surveillance, inform patient counseling, and determine the duration of follow-up required to study specific adverse events effectively. Orthopedic surgeons should have the lowest threshold for testing for each adverse event during the time period of greatest risk.

Continue to: Geriatric hip fracture surgery is associated with...

Geriatric hip fracture surgery is associated with a higher rate of occurrence of postoperative adverse events than any other commonly performed orthopedic procedure.^1-4 Indeed, the 90-day mortality rate following a geriatric hip fracture surgery may be as high as 15%² and the 30-day morbidity rate as high as 30%.³ Furthermore, more than half of postoperative mortalities following orthopedic procedures occur after surgery for hip fracture.⁴ Therefore, extensive research has been conducted regarding interventions to reduce the rates of adverse events following a hip fracture surgery.^5-12 For example, randomized trials have been conducted involving venous thromboembolism prophylaxis,^5,6nutritional supplementation,⁷ delirium prevention,^8-10 anemia correction,¹¹ geriatrics consultation,⁹ and anesthetic technique.¹²

Despite these extensive research efforts, there is currently little information in the literature regarding when postoperative adverse events occur. A clear depiction of the timing of adverse events could help target clinical surveillance, inform patient counseling, and determine the duration of follow-up required for studies. The reason that the timing of adverse events has not been previously characterized may be that the sample sizes available through standard single- or multi-institutional studies may be insufficient to accurately characterize the timing of rare adverse events (eg, myocardial infarction, stroke, etc.). Moreover, although administrative datasets have become common data sources for investigation of rare postoperative adverse events,^13-16 such data sources often do not contain data on the timing of diagnosis.

The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) is a relatively new and growing surgical registry.^1,3,13-22 The registry follows up patients undergoing surgical procedures at several hundred community and academic institutions nationwide. Unlike the administrative datasets discussed above, the ACS-NSQIP characterizes the postoperative day of diagnosis of well-defined adverse events during the first 30 postoperative days.²²

In this study, data collected by the ACS-NSQIP are used to characterize the timing of 10 specific postoperative adverse events following a geriatric hip fracture surgery.

Continue to: METHODS...

 

 

METHODS

A retrospective analysis of data collected prospectively through the ACS-NSQIP was conducted. Geriatric patients who underwent hip fracture surgery during 2010 to 2013 were identified. Specific inclusion criteria were (1) International Classification of Diseases, Ninth Revision, diagnosis code 820, (2) primary Current Procedural Terminology codes 27125, 27130, 27235, 27236, 27244, or 27245, and (3) age ≥70 years.

The ACS-NSQIP captures patient demographic, comorbidity, and procedural characteristics at baseline.²² At the end of the 30-day follow-up period, the ACS-NSQIP personnel review both inpatient and outpatient charts to characterize the occurrence vs nonoccurrence of specific postoperative adverse events.^22-25 When an adverse event does occur, the postoperative day of diagnosis is recorded.

For this study, the following adverse event categories were investigated: myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation, stroke, pneumonia, pulmonary embolism, urinary tract infection, deep vein thrombosis, sepsis (either with or without shock), mortality, and surgical site infection (including superficial surgical site infection, deep surgical site infection, and organ or space surgical site infection). Detailed definitions of each adverse event are provided in ACS-NSQIP materials.²²

First, the 30-day incidence (and the associated 95% confidence interval) was determined for each adverse event. Second, the median postoperative day of diagnosis (and the associated interquartile range) was determined for each adverse event. Third, the postoperative length of stay was used to estimate the proportion of diagnoses occurring prior to vs following discharge for each adverse event. Finally, multivariate Cox proportional hazards models were used to identify independent risk factors for earlier occurrence of postoperative adverse events. The final models were selected using a backward stepwise process that sequentially eliminated variables with the weakest associations until all variables had P < .05.

Because the ACS-NSQIP reports timing data in calendar days, when the postoperative length of stay was equivalent to the postoperative day of diagnosis, it was not possible to ascertain whether the diagnosis occurred prior to or following discharge. For this study, when the postoperative length of stay was equivalent to the postoperative day of diagnosis, the adverse event was considered to have been diagnosed following discharge. The rationale for this is that for most of the adverse events, it was thought to be unlikely that an inpatient would be discharged before the end of the same day as an inpatient diagnosis. However, there was one exception to this rule; when the postoperative day of discharge, the postoperative length of stay, and the postoperative day of death were all equivalent, the adverse event was considered to have occurred prior to discharge. This is because when a patient dies during the initial inpatient stay, the ACS-NSQIP considers the postoperative length of stay to be equivalent to the postoperative day of death. This makes it much more likely that a diagnosis on the final hospital day had occurred in a patient who had not been discharged.

The mandatory ACS-NSQIP statement is “The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS-NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.”²⁶

Continue to: RESULTS...

 

 

RESULTS

In total, 19,873 geriatric patients undergoing a hip fracture surgery were identified (Table 1). The rates of adverse events ranged from 6.7% for urinary tract infection to 0.6% for pulmonary embolism (Table 2).

Table 1. Patient Population

	Number	Percent
Total	19,873	100.0%
Age
70-74 years	1852	9.3%
75-79 years	2764	13.9%
80-84 years	4328	21.8%
85-89 years	5525	27.8%
≥90 years	5404	27.2%
Sex
Male	5359	27.0%
Female	14,514	73.0%
Body mass index
<30 kg/m2	17,733	89.2%
≥30 kg/m2	2140	10.8%
Functional status
Independent	14,348	72.2%
Dependent	5525	27.8%
Diabetes	3321	16.7%
Congestive heart failure	738	3.7%
Dyspnea on exertion	1542	7.8%
Hypertension	14,265	71.8%
End-stage renal disease	322	1.6%
COPD	2239	11.3%
Current smoker	1506	7.6%

Abbreviation: COPD, chronic obstructive pulmonary disease.

Table 2. Patients with Adverse Events Diagnosed During the First 30 postoperative days (N = 19,873)

Adverse Event	Number	Percent	95% CI
Urinary tract infection	1321	6.7%	6.3%-7.0%
Mortality	1240	6.2%	5.9%-6.6%
Pneumonia	771	3.9%	3.6%-4.2%
Sepsis	428	2.2%	2.0%-2.4%
Myocardial infarction	347	1.8%	1.6%-1.9%
Surgical site infection	247	1.2%	1.1%-1.4%
Deep vein thrombosis	199	1.0%	0.9%-1.1%
Stroke	144	0.7%	0.6%-0.8%
Cardiac arrest	136	0.7%	0.6%-0.8%
Pulmonary embolism	126	0.6%	0.5%-0.7%

Abbreviation: CI, confidence interval.

Figure 1 depicts the timing of postoperative adverse events in detail in histograms and timing curves. For the earliest diagnosed adverse events, the rate of adverse events had diminished by postoperative day 30. For the later diagnosed adverse events, the rate of adverse events remained high at postoperative day 30.

Figure 2 provides the summary statistics for adverse events diagnosed in the first 30 postoperative days. The median postoperative day of diagnosis (and the interquartile range) was 3 (1-5) for myocardial infarction, 3 (0-8) for cardiac arrest requiring cardiopulmonary resuscitation, 3 (1-10) for stroke, 4 (2-10) for pneumonia, 4 (2-11) for pulmonary embolism, 7 (2-13) for urinary tract infection, 9 (4-16) for deep vein thrombosis, 9 (4-18) for sepsis, 11 (6-19) for mortality, and 16 (11-22) for surgical site infection.

Figure 3 depicts the timing of adverse events relative to discharge. The proportions of adverse events diagnosed prior to discharge were 81.0% for myocardial infarction, 77.8% for stroke, 76.1% for cardiac arrest requiring cardiopulmonary resuscitation, 71.9% for pulmonary embolism, 71.1% for pneumonia, 58.0% for urinary tract infection, 52.1% for sepsis, 46.9% for deep vein thrombosis, 44.3% for mortality, and 27.6% for surgical site infection.

Table 3 shows the independent risk factors for earlier occurrence of adverse events. Following multivariate stepwise selection of final models, at least 1 patient characteristic was independently associated with the timing of cardiac arrest, stroke, urinary tract infection, deep vein thrombosis, and death. In contrast, no patient characteristics were independently associated with the timing of myocardial infarction, pneumonia, pulmonary embolism, sepsis, and surgical site infection.

Table 3. Timing of Diagnosis of Adverse Events^a

Adverse events and associated baseline characteristic(s)	Median postoperative day of diagnosis with vs without baseline characteristic	P-valueb
Cardiac arrest
End-stage renal disease	1 vs 3	.005
Stroke
Hypertension	4 vs 2	.025
Dependent functional status	2 vs 4	.027
Urinary tract infection
Female sex	6 vs 8	.009
Deep vein thrombosis
Body mass index ≥30 kg/m2	5 vs 10	.015
Death
End-stage renal disease	10 vs 11	.031

^aBaseline characteristics that were independently associated with the timing of each adverse event were identified through a backwards stepwise selection process initially including all characteristics listed in Table 1, and sequentially excluding characteristics with the weakest associations until only characteristics with P < .05 remained. Independent associations with the timing of cardiac arrest, stroke, urinary tract infection, deep vein thrombosis, and death are shown. There were no characteristics independently associated with timing of myocardial infarction, pneumonia, pulmonary embolism, sepsis, or surgical site infection; hence, these adverse events are not listed in the table.

^bFrom final Cox proportional hazards models identified through multivariate stepwise selection.

Continue to: DISCUSSION...

 

 

DISCUSSION

Adverse events are extremely common following a geriatric hip fracture surgery.^1-4 Despite extensive investigation regarding methods to prevent these events,^5-12 there is limited published description of the timing at which such events occur. This study used a large prospectively followed up cohort of geriatric patients undergoing a hip fracture surgery to deliver a better description of the timing of adverse events than was previously available. The findings of this study should enable more targeted clinical surveillance, inform patient counseling, and help determine the duration of follow-up required for studies on adverse events.

There was wide variability in the timing at which the different postoperative adverse events were diagnosed (Figures 1, 2). Myocardial infarction was diagnosed the earliest, with more than three-fourth of diagnoses in the first postoperative week. Other relatively early-diagnosed adverse events included cardiac arrest requiring cardiopulmonary resuscitation, stroke, pneumonia, and pulmonary embolism.

The latest-diagnosed adverse event was surgical site infection (Figures 1, 2). Surgical site infection was actually the only adverse event with a rate of diagnosis during the first week that was lower than the rate of diagnosis later in the month (as can be seen by the inflection in the timing curve for surgical site infection in Figure 1). Mortality showed a relatively consistent rate of diagnosis throughout the entire first postoperative month. Other relatively late-diagnosed postoperative events, including sepsis, deep vein thrombosis, and urinary tract infection, showed varying degrees of decreased rate of diagnosis near the end of the first postoperative month. Of note, for the later-diagnosed adverse events, the estimated median and interquartile ranges (Figure 2) were presumably quite biased toward earlier diagnosis, as the 30-day follow-up period clearly failed to capture a large proportion of later-occurring adverse events (Figure 1).

Certain risk factors were independently associated with earlier occurrence of adverse events. Perhaps most strikingly, body mass index in the obese range was associated with substantially earlier occurrence of deep vein thrombosis (median of 5 vs 10 days). This finding suggests that clinical monitoring for deep vein thrombosis should be performed earlier in patients with greater body mass index. Also notable is the earlier occurrence of cardiac arrest and death among patients with end-stage renal disease than among those without. Patients with end-stage renal disease may have a greater risk for these adverse events immediately following the cardiac stresses of surgery.²⁷ Similarly, such patients may be more prone to early electrolyte abnormalities and arrhythmia.

Continue to: In addition to its clinical implications, this study...

 

 

In addition to its clinical implications, this study informs about the interpretation of the many studies of adverse events following hip fracture procedures that have been conducted using retrospective data. Several such studies have relied on inpatient-only administrative databases.^{4,13,14,28-35} As clearly demonstrated in Figure 3, for most of the commonly studied adverse events, inpatient-only databases failed to capture a large proportion of adverse events occurring in the first postoperative month. This highlights a substantial limitation of this commonly published type of study that is often not emphasized in the literature.

There has also been an increase in the publication of studies of adverse events following a hip fracture surgery using the ACS-NSQIP data.^{3,13,14,17,18,21} As discussed, the ACS-NSQIP provides data on 30-days of follow-up. This relatively extended follow-up is often touted as a distinct advantage. However, this study demonstrates that even the 30-day follow-up afforded by the ACS-NSQIP is limited in its ability to enable investigation of the later-occurring adverse events (Figure 1). In particular, the rate of surgical site infection shows little sign of slowing by postoperative day 30. Similarly, the rates of mortality, sepsis, deep vein thrombosis, and urinary tract infection remain substantial.

This study does have limitations. First, as discussed, the duration of follow-up is a limitation of any ACS-NSQIP-based investigation, including this study. Second, the ACS-NSQIP does not capture relevant orthopedic-specific outcomes (eg, screw cutout). In addition, it could not be determined with certainty whether adverse events occurring on the final hospital day occurred prior to or following discharge. However, only a small proportion of most of the adverse events was diagnosed on the final hospital day. Finally, the ACS-NSQIP reports on days from the operation until diagnosis of the adverse event. Although some adverse events are probably diagnosed quickly after they have occurred (eg, myocardial infarction and cardiac arrest), other adverse events may have a delayed diagnosis (eg, surgical site infection may be identified days after its initial occurrence during a follow-up examination). Therefore, it is important to note the subtle distinction between occurrence and diagnosis throughout the article. This article reports on the timing of diagnosis, not actual occurrence.

CONCLUSION

The timing of postoperative adverse events has been understudied in the past. This may be due to an inability of standard single- or multi-institutional investigations to achieve sample sizes adequate to study the less commonly occurring adverse events. Using a relatively new prospective surgical registry, this study provides a far more detailed description of the timing of adverse events following surgery than was previously available. The authors anticipate that these data can be used to inform patient counseling, target clinical surveillance, and direct clinical research. The authors chose to study the timing of postoperative adverse events following geriatric hip fracture surgery because of the high rate of adverse events associated with the procedure. However, future ACS-NSQIP studies may involve characterization of the timing of adverse events following other orthopedic and non-orthopedic procedures.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

This study uses a prospective surgical registry to characterize the timing of 10 postoperative adverse events following geriatric hip fracture surgery. There were 19,873 patients identified who were ≥70 years undergoing surgery for hip fracture as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). The median postoperative day of diagnosis (and interquartile range) for myocardial infarction was 3 (1-5), cardiac arrest requiring cardiopulmonary resuscitation 3 (0-8), stroke 3 (1-10), pneumonia 4 (2-10), pulmonary embolism 4 (2-11), urinary tract infection 7 (2-13), deep vein thrombosis 9 (4-16), sepsis 9 (4-18), mortality 11 (6-19), and surgical site infection 16 (11-22). For the earliest diagnosed adverse events, the rate of adverse events had diminished by postoperative day 30. For the later diagnosed adverse events, the rate of adverse events remained high at postoperative day 30. Findings help to enable more targeted clinical surveillance, inform patient counseling, and determine the duration of follow-up required to study specific adverse events effectively. Orthopedic surgeons should have the lowest threshold for testing for each adverse event during the time period of greatest risk.

Continue to: Geriatric hip fracture surgery is associated with...

Geriatric hip fracture surgery is associated with a higher rate of occurrence of postoperative adverse events than any other commonly performed orthopedic procedure.^1-4 Indeed, the 90-day mortality rate following a geriatric hip fracture surgery may be as high as 15%² and the 30-day morbidity rate as high as 30%.³ Furthermore, more than half of postoperative mortalities following orthopedic procedures occur after surgery for hip fracture.⁴ Therefore, extensive research has been conducted regarding interventions to reduce the rates of adverse events following a hip fracture surgery.^5-12 For example, randomized trials have been conducted involving venous thromboembolism prophylaxis,^5,6nutritional supplementation,⁷ delirium prevention,^8-10 anemia correction,¹¹ geriatrics consultation,⁹ and anesthetic technique.¹²

Despite these extensive research efforts, there is currently little information in the literature regarding when postoperative adverse events occur. A clear depiction of the timing of adverse events could help target clinical surveillance, inform patient counseling, and determine the duration of follow-up required for studies. The reason that the timing of adverse events has not been previously characterized may be that the sample sizes available through standard single- or multi-institutional studies may be insufficient to accurately characterize the timing of rare adverse events (eg, myocardial infarction, stroke, etc.). Moreover, although administrative datasets have become common data sources for investigation of rare postoperative adverse events,^13-16 such data sources often do not contain data on the timing of diagnosis.

The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) is a relatively new and growing surgical registry.^1,3,13-22 The registry follows up patients undergoing surgical procedures at several hundred community and academic institutions nationwide. Unlike the administrative datasets discussed above, the ACS-NSQIP characterizes the postoperative day of diagnosis of well-defined adverse events during the first 30 postoperative days.²²

In this study, data collected by the ACS-NSQIP are used to characterize the timing of 10 specific postoperative adverse events following a geriatric hip fracture surgery.

Continue to: METHODS...

 

 

METHODS

A retrospective analysis of data collected prospectively through the ACS-NSQIP was conducted. Geriatric patients who underwent hip fracture surgery during 2010 to 2013 were identified. Specific inclusion criteria were (1) International Classification of Diseases, Ninth Revision, diagnosis code 820, (2) primary Current Procedural Terminology codes 27125, 27130, 27235, 27236, 27244, or 27245, and (3) age ≥70 years.

The ACS-NSQIP captures patient demographic, comorbidity, and procedural characteristics at baseline.²² At the end of the 30-day follow-up period, the ACS-NSQIP personnel review both inpatient and outpatient charts to characterize the occurrence vs nonoccurrence of specific postoperative adverse events.^22-25 When an adverse event does occur, the postoperative day of diagnosis is recorded.

For this study, the following adverse event categories were investigated: myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation, stroke, pneumonia, pulmonary embolism, urinary tract infection, deep vein thrombosis, sepsis (either with or without shock), mortality, and surgical site infection (including superficial surgical site infection, deep surgical site infection, and organ or space surgical site infection). Detailed definitions of each adverse event are provided in ACS-NSQIP materials.²²

First, the 30-day incidence (and the associated 95% confidence interval) was determined for each adverse event. Second, the median postoperative day of diagnosis (and the associated interquartile range) was determined for each adverse event. Third, the postoperative length of stay was used to estimate the proportion of diagnoses occurring prior to vs following discharge for each adverse event. Finally, multivariate Cox proportional hazards models were used to identify independent risk factors for earlier occurrence of postoperative adverse events. The final models were selected using a backward stepwise process that sequentially eliminated variables with the weakest associations until all variables had P < .05.

Because the ACS-NSQIP reports timing data in calendar days, when the postoperative length of stay was equivalent to the postoperative day of diagnosis, it was not possible to ascertain whether the diagnosis occurred prior to or following discharge. For this study, when the postoperative length of stay was equivalent to the postoperative day of diagnosis, the adverse event was considered to have been diagnosed following discharge. The rationale for this is that for most of the adverse events, it was thought to be unlikely that an inpatient would be discharged before the end of the same day as an inpatient diagnosis. However, there was one exception to this rule; when the postoperative day of discharge, the postoperative length of stay, and the postoperative day of death were all equivalent, the adverse event was considered to have occurred prior to discharge. This is because when a patient dies during the initial inpatient stay, the ACS-NSQIP considers the postoperative length of stay to be equivalent to the postoperative day of death. This makes it much more likely that a diagnosis on the final hospital day had occurred in a patient who had not been discharged.

The mandatory ACS-NSQIP statement is “The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS-NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.”²⁶

Continue to: RESULTS...

 

 

RESULTS

In total, 19,873 geriatric patients undergoing a hip fracture surgery were identified (Table 1). The rates of adverse events ranged from 6.7% for urinary tract infection to 0.6% for pulmonary embolism (Table 2).

Table 1. Patient Population

	Number	Percent
Total	19,873	100.0%
Age
70-74 years	1852	9.3%
75-79 years	2764	13.9%
80-84 years	4328	21.8%
85-89 years	5525	27.8%
≥90 years	5404	27.2%
Sex
Male	5359	27.0%
Female	14,514	73.0%
Body mass index
<30 kg/m2	17,733	89.2%
≥30 kg/m2	2140	10.8%
Functional status
Independent	14,348	72.2%
Dependent	5525	27.8%
Diabetes	3321	16.7%
Congestive heart failure	738	3.7%
Dyspnea on exertion	1542	7.8%
Hypertension	14,265	71.8%
End-stage renal disease	322	1.6%
COPD	2239	11.3%
Current smoker	1506	7.6%

Abbreviation: COPD, chronic obstructive pulmonary disease.

Table 2. Patients with Adverse Events Diagnosed During the First 30 postoperative days (N = 19,873)

Adverse Event	Number	Percent	95% CI
Urinary tract infection	1321	6.7%	6.3%-7.0%
Mortality	1240	6.2%	5.9%-6.6%
Pneumonia	771	3.9%	3.6%-4.2%
Sepsis	428	2.2%	2.0%-2.4%
Myocardial infarction	347	1.8%	1.6%-1.9%
Surgical site infection	247	1.2%	1.1%-1.4%
Deep vein thrombosis	199	1.0%	0.9%-1.1%
Stroke	144	0.7%	0.6%-0.8%
Cardiac arrest	136	0.7%	0.6%-0.8%
Pulmonary embolism	126	0.6%	0.5%-0.7%

Abbreviation: CI, confidence interval.

Figure 1 depicts the timing of postoperative adverse events in detail in histograms and timing curves. For the earliest diagnosed adverse events, the rate of adverse events had diminished by postoperative day 30. For the later diagnosed adverse events, the rate of adverse events remained high at postoperative day 30.

Figure 2 provides the summary statistics for adverse events diagnosed in the first 30 postoperative days. The median postoperative day of diagnosis (and the interquartile range) was 3 (1-5) for myocardial infarction, 3 (0-8) for cardiac arrest requiring cardiopulmonary resuscitation, 3 (1-10) for stroke, 4 (2-10) for pneumonia, 4 (2-11) for pulmonary embolism, 7 (2-13) for urinary tract infection, 9 (4-16) for deep vein thrombosis, 9 (4-18) for sepsis, 11 (6-19) for mortality, and 16 (11-22) for surgical site infection.

Figure 3 depicts the timing of adverse events relative to discharge. The proportions of adverse events diagnosed prior to discharge were 81.0% for myocardial infarction, 77.8% for stroke, 76.1% for cardiac arrest requiring cardiopulmonary resuscitation, 71.9% for pulmonary embolism, 71.1% for pneumonia, 58.0% for urinary tract infection, 52.1% for sepsis, 46.9% for deep vein thrombosis, 44.3% for mortality, and 27.6% for surgical site infection.

Table 3 shows the independent risk factors for earlier occurrence of adverse events. Following multivariate stepwise selection of final models, at least 1 patient characteristic was independently associated with the timing of cardiac arrest, stroke, urinary tract infection, deep vein thrombosis, and death. In contrast, no patient characteristics were independently associated with the timing of myocardial infarction, pneumonia, pulmonary embolism, sepsis, and surgical site infection.

Table 3. Timing of Diagnosis of Adverse Events^a

Adverse events and associated baseline characteristic(s)	Median postoperative day of diagnosis with vs without baseline characteristic	P-valueb
Cardiac arrest
End-stage renal disease	1 vs 3	.005
Stroke
Hypertension	4 vs 2	.025
Dependent functional status	2 vs 4	.027
Urinary tract infection
Female sex	6 vs 8	.009
Deep vein thrombosis
Body mass index ≥30 kg/m2	5 vs 10	.015
Death
End-stage renal disease	10 vs 11	.031

^aBaseline characteristics that were independently associated with the timing of each adverse event were identified through a backwards stepwise selection process initially including all characteristics listed in Table 1, and sequentially excluding characteristics with the weakest associations until only characteristics with P < .05 remained. Independent associations with the timing of cardiac arrest, stroke, urinary tract infection, deep vein thrombosis, and death are shown. There were no characteristics independently associated with timing of myocardial infarction, pneumonia, pulmonary embolism, sepsis, or surgical site infection; hence, these adverse events are not listed in the table.

^bFrom final Cox proportional hazards models identified through multivariate stepwise selection.

Continue to: DISCUSSION...

 

 

DISCUSSION

Adverse events are extremely common following a geriatric hip fracture surgery.^1-4 Despite extensive investigation regarding methods to prevent these events,^5-12 there is limited published description of the timing at which such events occur. This study used a large prospectively followed up cohort of geriatric patients undergoing a hip fracture surgery to deliver a better description of the timing of adverse events than was previously available. The findings of this study should enable more targeted clinical surveillance, inform patient counseling, and help determine the duration of follow-up required for studies on adverse events.

There was wide variability in the timing at which the different postoperative adverse events were diagnosed (Figures 1, 2). Myocardial infarction was diagnosed the earliest, with more than three-fourth of diagnoses in the first postoperative week. Other relatively early-diagnosed adverse events included cardiac arrest requiring cardiopulmonary resuscitation, stroke, pneumonia, and pulmonary embolism.

The latest-diagnosed adverse event was surgical site infection (Figures 1, 2). Surgical site infection was actually the only adverse event with a rate of diagnosis during the first week that was lower than the rate of diagnosis later in the month (as can be seen by the inflection in the timing curve for surgical site infection in Figure 1). Mortality showed a relatively consistent rate of diagnosis throughout the entire first postoperative month. Other relatively late-diagnosed postoperative events, including sepsis, deep vein thrombosis, and urinary tract infection, showed varying degrees of decreased rate of diagnosis near the end of the first postoperative month. Of note, for the later-diagnosed adverse events, the estimated median and interquartile ranges (Figure 2) were presumably quite biased toward earlier diagnosis, as the 30-day follow-up period clearly failed to capture a large proportion of later-occurring adverse events (Figure 1).

Certain risk factors were independently associated with earlier occurrence of adverse events. Perhaps most strikingly, body mass index in the obese range was associated with substantially earlier occurrence of deep vein thrombosis (median of 5 vs 10 days). This finding suggests that clinical monitoring for deep vein thrombosis should be performed earlier in patients with greater body mass index. Also notable is the earlier occurrence of cardiac arrest and death among patients with end-stage renal disease than among those without. Patients with end-stage renal disease may have a greater risk for these adverse events immediately following the cardiac stresses of surgery.²⁷ Similarly, such patients may be more prone to early electrolyte abnormalities and arrhythmia.

Continue to: In addition to its clinical implications, this study...

 

 

In addition to its clinical implications, this study informs about the interpretation of the many studies of adverse events following hip fracture procedures that have been conducted using retrospective data. Several such studies have relied on inpatient-only administrative databases.^{4,13,14,28-35} As clearly demonstrated in Figure 3, for most of the commonly studied adverse events, inpatient-only databases failed to capture a large proportion of adverse events occurring in the first postoperative month. This highlights a substantial limitation of this commonly published type of study that is often not emphasized in the literature.

There has also been an increase in the publication of studies of adverse events following a hip fracture surgery using the ACS-NSQIP data.^{3,13,14,17,18,21} As discussed, the ACS-NSQIP provides data on 30-days of follow-up. This relatively extended follow-up is often touted as a distinct advantage. However, this study demonstrates that even the 30-day follow-up afforded by the ACS-NSQIP is limited in its ability to enable investigation of the later-occurring adverse events (Figure 1). In particular, the rate of surgical site infection shows little sign of slowing by postoperative day 30. Similarly, the rates of mortality, sepsis, deep vein thrombosis, and urinary tract infection remain substantial.

This study does have limitations. First, as discussed, the duration of follow-up is a limitation of any ACS-NSQIP-based investigation, including this study. Second, the ACS-NSQIP does not capture relevant orthopedic-specific outcomes (eg, screw cutout). In addition, it could not be determined with certainty whether adverse events occurring on the final hospital day occurred prior to or following discharge. However, only a small proportion of most of the adverse events was diagnosed on the final hospital day. Finally, the ACS-NSQIP reports on days from the operation until diagnosis of the adverse event. Although some adverse events are probably diagnosed quickly after they have occurred (eg, myocardial infarction and cardiac arrest), other adverse events may have a delayed diagnosis (eg, surgical site infection may be identified days after its initial occurrence during a follow-up examination). Therefore, it is important to note the subtle distinction between occurrence and diagnosis throughout the article. This article reports on the timing of diagnosis, not actual occurrence.

CONCLUSION

The timing of postoperative adverse events has been understudied in the past. This may be due to an inability of standard single- or multi-institutional investigations to achieve sample sizes adequate to study the less commonly occurring adverse events. Using a relatively new prospective surgical registry, this study provides a far more detailed description of the timing of adverse events following surgery than was previously available. The authors anticipate that these data can be used to inform patient counseling, target clinical surveillance, and direct clinical research. The authors chose to study the timing of postoperative adverse events following geriatric hip fracture surgery because of the high rate of adverse events associated with the procedure. However, future ACS-NSQIP studies may involve characterization of the timing of adverse events following other orthopedic and non-orthopedic procedures.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

This study uses a prospective surgical registry to characterize the timing of 10 postoperative adverse events following geriatric hip fracture surgery. There were 19,873 patients identified who were ≥70 years undergoing surgery for hip fracture as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). The median postoperative day of diagnosis (and interquartile range) for myocardial infarction was 3 (1-5), cardiac arrest requiring cardiopulmonary resuscitation 3 (0-8), stroke 3 (1-10), pneumonia 4 (2-10), pulmonary embolism 4 (2-11), urinary tract infection 7 (2-13), deep vein thrombosis 9 (4-16), sepsis 9 (4-18), mortality 11 (6-19), and surgical site infection 16 (11-22). For the earliest diagnosed adverse events, the rate of adverse events had diminished by postoperative day 30. For the later diagnosed adverse events, the rate of adverse events remained high at postoperative day 30. Findings help to enable more targeted clinical surveillance, inform patient counseling, and determine the duration of follow-up required to study specific adverse events effectively. Orthopedic surgeons should have the lowest threshold for testing for each adverse event during the time period of greatest risk.

Continue to: Geriatric hip fracture surgery is associated with...

Geriatric hip fracture surgery is associated with a higher rate of occurrence of postoperative adverse events than any other commonly performed orthopedic procedure.^1-4 Indeed, the 90-day mortality rate following a geriatric hip fracture surgery may be as high as 15%² and the 30-day morbidity rate as high as 30%.³ Furthermore, more than half of postoperative mortalities following orthopedic procedures occur after surgery for hip fracture.⁴ Therefore, extensive research has been conducted regarding interventions to reduce the rates of adverse events following a hip fracture surgery.^5-12 For example, randomized trials have been conducted involving venous thromboembolism prophylaxis,^5,6nutritional supplementation,⁷ delirium prevention,^8-10 anemia correction,¹¹ geriatrics consultation,⁹ and anesthetic technique.¹²

Despite these extensive research efforts, there is currently little information in the literature regarding when postoperative adverse events occur. A clear depiction of the timing of adverse events could help target clinical surveillance, inform patient counseling, and determine the duration of follow-up required for studies. The reason that the timing of adverse events has not been previously characterized may be that the sample sizes available through standard single- or multi-institutional studies may be insufficient to accurately characterize the timing of rare adverse events (eg, myocardial infarction, stroke, etc.). Moreover, although administrative datasets have become common data sources for investigation of rare postoperative adverse events,^13-16 such data sources often do not contain data on the timing of diagnosis.

The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) is a relatively new and growing surgical registry.^1,3,13-22 The registry follows up patients undergoing surgical procedures at several hundred community and academic institutions nationwide. Unlike the administrative datasets discussed above, the ACS-NSQIP characterizes the postoperative day of diagnosis of well-defined adverse events during the first 30 postoperative days.²²

In this study, data collected by the ACS-NSQIP are used to characterize the timing of 10 specific postoperative adverse events following a geriatric hip fracture surgery.

Continue to: METHODS...

 

 

METHODS

A retrospective analysis of data collected prospectively through the ACS-NSQIP was conducted. Geriatric patients who underwent hip fracture surgery during 2010 to 2013 were identified. Specific inclusion criteria were (1) International Classification of Diseases, Ninth Revision, diagnosis code 820, (2) primary Current Procedural Terminology codes 27125, 27130, 27235, 27236, 27244, or 27245, and (3) age ≥70 years.

The ACS-NSQIP captures patient demographic, comorbidity, and procedural characteristics at baseline.²² At the end of the 30-day follow-up period, the ACS-NSQIP personnel review both inpatient and outpatient charts to characterize the occurrence vs nonoccurrence of specific postoperative adverse events.^22-25 When an adverse event does occur, the postoperative day of diagnosis is recorded.

For this study, the following adverse event categories were investigated: myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation, stroke, pneumonia, pulmonary embolism, urinary tract infection, deep vein thrombosis, sepsis (either with or without shock), mortality, and surgical site infection (including superficial surgical site infection, deep surgical site infection, and organ or space surgical site infection). Detailed definitions of each adverse event are provided in ACS-NSQIP materials.²²

First, the 30-day incidence (and the associated 95% confidence interval) was determined for each adverse event. Second, the median postoperative day of diagnosis (and the associated interquartile range) was determined for each adverse event. Third, the postoperative length of stay was used to estimate the proportion of diagnoses occurring prior to vs following discharge for each adverse event. Finally, multivariate Cox proportional hazards models were used to identify independent risk factors for earlier occurrence of postoperative adverse events. The final models were selected using a backward stepwise process that sequentially eliminated variables with the weakest associations until all variables had P < .05.

Because the ACS-NSQIP reports timing data in calendar days, when the postoperative length of stay was equivalent to the postoperative day of diagnosis, it was not possible to ascertain whether the diagnosis occurred prior to or following discharge. For this study, when the postoperative length of stay was equivalent to the postoperative day of diagnosis, the adverse event was considered to have been diagnosed following discharge. The rationale for this is that for most of the adverse events, it was thought to be unlikely that an inpatient would be discharged before the end of the same day as an inpatient diagnosis. However, there was one exception to this rule; when the postoperative day of discharge, the postoperative length of stay, and the postoperative day of death were all equivalent, the adverse event was considered to have occurred prior to discharge. This is because when a patient dies during the initial inpatient stay, the ACS-NSQIP considers the postoperative length of stay to be equivalent to the postoperative day of death. This makes it much more likely that a diagnosis on the final hospital day had occurred in a patient who had not been discharged.

The mandatory ACS-NSQIP statement is “The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS-NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.”²⁶

Continue to: RESULTS...

 

 

RESULTS

In total, 19,873 geriatric patients undergoing a hip fracture surgery were identified (Table 1). The rates of adverse events ranged from 6.7% for urinary tract infection to 0.6% for pulmonary embolism (Table 2).

Table 1. Patient Population

	Number	Percent
Total	19,873	100.0%
Age
70-74 years	1852	9.3%
75-79 years	2764	13.9%
80-84 years	4328	21.8%
85-89 years	5525	27.8%
≥90 years	5404	27.2%
Sex
Male	5359	27.0%
Female	14,514	73.0%
Body mass index
<30 kg/m2	17,733	89.2%
≥30 kg/m2	2140	10.8%
Functional status
Independent	14,348	72.2%
Dependent	5525	27.8%
Diabetes	3321	16.7%
Congestive heart failure	738	3.7%
Dyspnea on exertion	1542	7.8%
Hypertension	14,265	71.8%
End-stage renal disease	322	1.6%
COPD	2239	11.3%
Current smoker	1506	7.6%

Abbreviation: COPD, chronic obstructive pulmonary disease.

Table 2. Patients with Adverse Events Diagnosed During the First 30 postoperative days (N = 19,873)

Adverse Event	Number	Percent	95% CI
Urinary tract infection	1321	6.7%	6.3%-7.0%
Mortality	1240	6.2%	5.9%-6.6%
Pneumonia	771	3.9%	3.6%-4.2%
Sepsis	428	2.2%	2.0%-2.4%
Myocardial infarction	347	1.8%	1.6%-1.9%
Surgical site infection	247	1.2%	1.1%-1.4%
Deep vein thrombosis	199	1.0%	0.9%-1.1%
Stroke	144	0.7%	0.6%-0.8%
Cardiac arrest	136	0.7%	0.6%-0.8%
Pulmonary embolism	126	0.6%	0.5%-0.7%

Abbreviation: CI, confidence interval.

Figure 1 depicts the timing of postoperative adverse events in detail in histograms and timing curves. For the earliest diagnosed adverse events, the rate of adverse events had diminished by postoperative day 30. For the later diagnosed adverse events, the rate of adverse events remained high at postoperative day 30.

Figure 2 provides the summary statistics for adverse events diagnosed in the first 30 postoperative days. The median postoperative day of diagnosis (and the interquartile range) was 3 (1-5) for myocardial infarction, 3 (0-8) for cardiac arrest requiring cardiopulmonary resuscitation, 3 (1-10) for stroke, 4 (2-10) for pneumonia, 4 (2-11) for pulmonary embolism, 7 (2-13) for urinary tract infection, 9 (4-16) for deep vein thrombosis, 9 (4-18) for sepsis, 11 (6-19) for mortality, and 16 (11-22) for surgical site infection.

Figure 3 depicts the timing of adverse events relative to discharge. The proportions of adverse events diagnosed prior to discharge were 81.0% for myocardial infarction, 77.8% for stroke, 76.1% for cardiac arrest requiring cardiopulmonary resuscitation, 71.9% for pulmonary embolism, 71.1% for pneumonia, 58.0% for urinary tract infection, 52.1% for sepsis, 46.9% for deep vein thrombosis, 44.3% for mortality, and 27.6% for surgical site infection.

Table 3 shows the independent risk factors for earlier occurrence of adverse events. Following multivariate stepwise selection of final models, at least 1 patient characteristic was independently associated with the timing of cardiac arrest, stroke, urinary tract infection, deep vein thrombosis, and death. In contrast, no patient characteristics were independently associated with the timing of myocardial infarction, pneumonia, pulmonary embolism, sepsis, and surgical site infection.

Table 3. Timing of Diagnosis of Adverse Events^a

Adverse events and associated baseline characteristic(s)	Median postoperative day of diagnosis with vs without baseline characteristic	P-valueb
Cardiac arrest
End-stage renal disease	1 vs 3	.005
Stroke
Hypertension	4 vs 2	.025
Dependent functional status	2 vs 4	.027
Urinary tract infection
Female sex	6 vs 8	.009
Deep vein thrombosis
Body mass index ≥30 kg/m2	5 vs 10	.015
Death
End-stage renal disease	10 vs 11	.031

^aBaseline characteristics that were independently associated with the timing of each adverse event were identified through a backwards stepwise selection process initially including all characteristics listed in Table 1, and sequentially excluding characteristics with the weakest associations until only characteristics with P < .05 remained. Independent associations with the timing of cardiac arrest, stroke, urinary tract infection, deep vein thrombosis, and death are shown. There were no characteristics independently associated with timing of myocardial infarction, pneumonia, pulmonary embolism, sepsis, or surgical site infection; hence, these adverse events are not listed in the table.

^bFrom final Cox proportional hazards models identified through multivariate stepwise selection.

Continue to: DISCUSSION...

 

 

DISCUSSION

Adverse events are extremely common following a geriatric hip fracture surgery.^1-4 Despite extensive investigation regarding methods to prevent these events,^5-12 there is limited published description of the timing at which such events occur. This study used a large prospectively followed up cohort of geriatric patients undergoing a hip fracture surgery to deliver a better description of the timing of adverse events than was previously available. The findings of this study should enable more targeted clinical surveillance, inform patient counseling, and help determine the duration of follow-up required for studies on adverse events.

There was wide variability in the timing at which the different postoperative adverse events were diagnosed (Figures 1, 2). Myocardial infarction was diagnosed the earliest, with more than three-fourth of diagnoses in the first postoperative week. Other relatively early-diagnosed adverse events included cardiac arrest requiring cardiopulmonary resuscitation, stroke, pneumonia, and pulmonary embolism.

The latest-diagnosed adverse event was surgical site infection (Figures 1, 2). Surgical site infection was actually the only adverse event with a rate of diagnosis during the first week that was lower than the rate of diagnosis later in the month (as can be seen by the inflection in the timing curve for surgical site infection in Figure 1). Mortality showed a relatively consistent rate of diagnosis throughout the entire first postoperative month. Other relatively late-diagnosed postoperative events, including sepsis, deep vein thrombosis, and urinary tract infection, showed varying degrees of decreased rate of diagnosis near the end of the first postoperative month. Of note, for the later-diagnosed adverse events, the estimated median and interquartile ranges (Figure 2) were presumably quite biased toward earlier diagnosis, as the 30-day follow-up period clearly failed to capture a large proportion of later-occurring adverse events (Figure 1).

Certain risk factors were independently associated with earlier occurrence of adverse events. Perhaps most strikingly, body mass index in the obese range was associated with substantially earlier occurrence of deep vein thrombosis (median of 5 vs 10 days). This finding suggests that clinical monitoring for deep vein thrombosis should be performed earlier in patients with greater body mass index. Also notable is the earlier occurrence of cardiac arrest and death among patients with end-stage renal disease than among those without. Patients with end-stage renal disease may have a greater risk for these adverse events immediately following the cardiac stresses of surgery.²⁷ Similarly, such patients may be more prone to early electrolyte abnormalities and arrhythmia.

Continue to: In addition to its clinical implications, this study...

 

 

In addition to its clinical implications, this study informs about the interpretation of the many studies of adverse events following hip fracture procedures that have been conducted using retrospective data. Several such studies have relied on inpatient-only administrative databases.^{4,13,14,28-35} As clearly demonstrated in Figure 3, for most of the commonly studied adverse events, inpatient-only databases failed to capture a large proportion of adverse events occurring in the first postoperative month. This highlights a substantial limitation of this commonly published type of study that is often not emphasized in the literature.

There has also been an increase in the publication of studies of adverse events following a hip fracture surgery using the ACS-NSQIP data.^{3,13,14,17,18,21} As discussed, the ACS-NSQIP provides data on 30-days of follow-up. This relatively extended follow-up is often touted as a distinct advantage. However, this study demonstrates that even the 30-day follow-up afforded by the ACS-NSQIP is limited in its ability to enable investigation of the later-occurring adverse events (Figure 1). In particular, the rate of surgical site infection shows little sign of slowing by postoperative day 30. Similarly, the rates of mortality, sepsis, deep vein thrombosis, and urinary tract infection remain substantial.

This study does have limitations. First, as discussed, the duration of follow-up is a limitation of any ACS-NSQIP-based investigation, including this study. Second, the ACS-NSQIP does not capture relevant orthopedic-specific outcomes (eg, screw cutout). In addition, it could not be determined with certainty whether adverse events occurring on the final hospital day occurred prior to or following discharge. However, only a small proportion of most of the adverse events was diagnosed on the final hospital day. Finally, the ACS-NSQIP reports on days from the operation until diagnosis of the adverse event. Although some adverse events are probably diagnosed quickly after they have occurred (eg, myocardial infarction and cardiac arrest), other adverse events may have a delayed diagnosis (eg, surgical site infection may be identified days after its initial occurrence during a follow-up examination). Therefore, it is important to note the subtle distinction between occurrence and diagnosis throughout the article. This article reports on the timing of diagnosis, not actual occurrence.

CONCLUSION

The timing of postoperative adverse events has been understudied in the past. This may be due to an inability of standard single- or multi-institutional investigations to achieve sample sizes adequate to study the less commonly occurring adverse events. Using a relatively new prospective surgical registry, this study provides a far more detailed description of the timing of adverse events following surgery than was previously available. The authors anticipate that these data can be used to inform patient counseling, target clinical surveillance, and direct clinical research. The authors chose to study the timing of postoperative adverse events following geriatric hip fracture surgery because of the high rate of adverse events associated with the procedure. However, future ACS-NSQIP studies may involve characterization of the timing of adverse events following other orthopedic and non-orthopedic procedures.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

It has been suggested that the utilization of joint arthroplasty in patients with rheumatoid arthritis (RA) is decreasing; however, this observation is largely based upon evidence pertaining to lower-extremity joint arthroplasty. It remains unknown if these observed trends also hold true for shoulder arthroplasty. The purpose of this study is to utilize a nationally representative population database in the US to identify trends in the utilization of shoulder arthroplasty among patients with RA. Secondarily, we sought to determine the rate of early adverse events, length of stay, and hospitalization costs associated with RA patients undergoing shoulder arthroplasty and to compare these outcomes to those of patients without a diagnosis of RA undergoing shoulder arthroplasty. Using a large population database in the US, we determined the annual rates of shoulder arthroplasty (overall and individual) in RA patients between 2002 and 2011. Early adverse events, length of stay, and hospitalization costs were determined and compared with those of non-RA patients undergoing shoulder arthroplasty. Overall, we identified 332,593 patients who underwent shoulder arthroplasty between 2002 and 2011, of whom 17,883 patients (5.4%) had a diagnosis of RA. Over the study period, there was a significant increase in the utilization of shoulder arthroplasty in RA patients, particularly total shoulder arthroplasty. Over the same period, there was a significant increase in the number of RA patients who underwent shoulder arthroplasty with a diagnosis of rotator cuff disease. There were no significant differences in adverse events or mean hospitalization costs between RA and non-RA patients. Non-RA patients had a significantly shorter length of stay; however, the difference did not appear to be clinically significant. In conclusion, the utilization of shoulder arthroplasty in patients with RA significantly increased from 2002 to 2011, which may partly reflect a trend toward management of rotator cuff disease with arthroplasty rather than repair.

Continue to: It has been suggested...

It has been suggested that the utilization of total joint arthroplasty (TJA) in patients with rheumatoid arthritis (RA) is decreasing over time;¹ however, this observation is largely based upon evidence pertaining to lower extremity TJA.² It remains unknown if these observed trends also hold true for shoulder arthroplasty, whereby the utilization of shoulder arthroplasty in RA patients is not limited to the management of end-stage inflammatory arthropathy. In this study, we used a nationally representative population database in the US to identify trends in the utilization of shoulder arthroplasty among patients with RA. As a secondary objective, we sought to determine the rate of early adverse events, length of stay, and hospitalization costs associated with RA patients undergoing shoulder arthroplasty and compare these outcomes to those of patients without a diagnosis of RA undergoing shoulder arthroplasty. We hypothesize that the utilization of shoulder arthroplasty in RA patients would be decreasing, but adverse events, length of stay, and hospitalization costs would not differ between patients with and without RA undergoing shoulder arthroplasty.

METHODS

We conducted a retrospective cohort study using the Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS) from 2002 to 2011.³ The NIS comprises a 20% stratified sample of all hospital discharges in the US. The NIS includes information about patient characteristics (age, sex, insurance status, and medical comorbidities) and hospitalization outcomes (adverse events, costs, and length of stay). The NIS allows identification of hospitalizations according to procedures and diagnoses using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Given the anonymity of this study, it was exempt from Institutional Review Board ethics approval.

Hospitalizations were selected for the study based on ICD-9-CM procedural codes for hemiarthroplasty (81.81), anatomic total shoulder arthroplasty (TSA) (81.80), and reverse TSA (81.88). These patients were then stratified by an ICD-9-CM diagnosis of RA (714.X). We also utilized ICD-9-CM diagnosis codes to determine the presence of rotator cuff pathology at the time of shoulder arthroplasty (726.13, 727.61, 840.4) and to exclude patients with a history of trauma (812.X, 716.11, 733.8X). In a separate analysis, all patients in the NIS database with an ICD-9-CM diagnosis of RA were identified for each calendar year of the study, and a national estimate of RA patients was generated annually to assess overall and individual utilization rates of shoulder arthroplasty in this population (the national estimate served as the denominator).

Preoperative patient data withdrawn from the NIS included age, sex, insurance status, and medical comorbidities. An Elixhauser Comorbidity Index (ECI) was generated for each patient based on the presence of 29 comorbid conditions. The ECI was chosen because of its capacity to accurately predict mortality and represent the patient burden of comorbidities in similar administrative database studies.^4-6

Early adverse events were also chosen based on ICD-9-CM diagnosis codes (Appendix A), and included the following: death, acute kidney injury, cardiac arrest, thromboembolic event, myocardial infarction, peripheral nerve injury, pneumonia, sepsis, stroke, surgical site infection, urinary tract infection, and wound dehiscence. The overall adverse event rate was defined as the occurrence of ≥1 of the above adverse events in a patient.

Appendix A. ICD-9-CM Codes Corresponding to Postoperative Adverse Events

Abbreviation: ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification

Length of stay and total hospital charges were available for each patient. Length of stay represents the number of calendar days a patient stayed in the hospital. All hospital charges were converted to hospitalization costs using the HCUP Cost-to-Charge Ratio Files. All hospitalization costs were adjusted for inflation using the US Bureau of Labor statistics yearly inflation calculator to represent charges in the year 2011, which was the final and most recent year in this study.

Continue to: Statistical analysis...

STATISTICAL ANALYSIS

Statistical analyses were conducted using Stata version 13.1 (StataCorp, LP). All analyses took into account the complex survey design of the NIS. Discharge weights, strata, and cluster variables were included to correctly estimate variance and to produce national estimates from the stratified sample. Pearson’s chi-squared test was used to compare age, sex, ECI, and insurance status between RA and non-RA patients undergoing shoulder arthroplasty.

Bivariate and multivariate logistic regressions were subsequently used to compare the rates of adverse events between RA and non-RA patients undergoing shoulder arthroplasty (non-RA cases were used as the reference). Multivariate linear regressions were used to compare hospital length of stay and hospitalization costs between RA and non-RA patients undergoing shoulder arthroplasty. The multivariate regressions were adjusted for baseline differences in age, sex, ECI, and insurance status. Cochran-Armitage tests for trend were used to assess trends over time. All tests were 2-tailed, and the statistical difference was established at a 2-sided α level of 0.05 (P < .05).

RESULTS

Overall, we identified 332,593 patients who underwent shoulder arthroplasty in the US between 2002 and 2011, of which 17,883 patients (5.4%) had a diagnosis of RA. In comparison with non-RA patients undergoing shoulder arthroplasty, patients with RA at the time of shoulder arthroplasty were significantly younger (65.2 ± 12.5 years vs 68.4 ± 11.0 years, P < .001), included a significantly greater proportion of female patients (76.7% vs 53.8%, P < .001), and included a significantly higher proportion of patients with Medicaid insurance (3.6% vs 2.3%, P < .001). There were no significant differences in the mean ECI between patients with and without a diagnosis of RA (Table 1). As depicted in Table 1, there were significant differences in the utilization of specific shoulder arthroplasty types between patients with and without RA, whereby a significantly greater proportion of RA patients underwent hemiarthroplasty (HA) (31.6% vs 29.3%, P = .002) and reverse TSA (7.7% vs 6.6%, P = .002), whereas a significantly greater proportion of non-RA patients underwent anatomic SA (64.0% vs 60.8%, P = .002).

Over the study period from 2002 to 2011, there was a significant increase in the overall utilization of shoulder arthroplasty in RA patients, as indicated by both the absolute number and the proportion of patients with a diagnosis of RA (P < .001) (Table 2, Figure). More specifically, 0.39% of RA patients underwent shoulder arthroplasty in 2002, as compared with 0.58% of RA patients in 2011 (P < .001) (Table 2). With respect to specific arthroplasty types, there was an exponential rise in the utilization of reverse TSA beginning in 2010 and a corresponding decrease in the rates of both HA and anatomic TSA (Table 2, Figure). In addition to changes in shoulder arthroplasty utilization over time among RA patients, we also observed a significant increase in the number of RA patients undergoing shoulder arthroplasty with a corresponding diagnosis of rotator cuff disease (9.7% in 2002 to 15.2% in 2011, P < .001).

Table 2. The Annual Utilization of Shoulder Arthroplasty Among Patients with a Diagnosis of Rheumatoid Arthritis.

^aRate determined as number of RA patients undergoing shoulder arthroplasty compared to the number of patients with an RA diagnosis in the stated calendar year.

Abbreviations: HA, hemiarthroplasty; RA, rheumatoid arthritis; TSA, total shoulder arthroplasty.

Continue to: Among patients with RA...

Among patients with RA undergoing shoulder arthroplasty, the overall rate of early adverse events was 3.12%, of which the most common early adverse events were urinary tract infections (1.8%), acute kidney injury (0.66%), and pneumonia (0.38%) (Table 3). As compared with patients without a diagnosis of RA undergoing shoulder arthroplasty, there were no significant differences in the overall and individual rates of early adverse events (Table 3).

Table 3. A Comparison of Early Adverse Events, Length of Stay, and Cost Between Patients With and Without Rheumatoid Arthritis (RA) Undergoing Shoulder Arthroplasty

^aReported in days. ^bReported in 2011 US dollars, adjusted for inflation.

The mean length of stay following shoulder arthroplasty in RA patients was 2.4 ± 1.6 days, and the mean hospitalization cost was $14,787 ± $7625 (Table 3). As compared with non-RA patients undergoing shoulder arthroplasty, there were no significant differences in the mean hospitalization costs; however, non-RA patients had a significantly shorter length of stay by 0.1 days (P = .002) (Table 3).

DISCUSSION

In this study, we observed that the utilization of shoulder arthroplasty in patients with RA increased significantly in the decade from 2002 to 2011, largely related to a rise in TSA. Interestingly, we also observed a corresponding rise in the proportion of RA patients undergoing shoulder arthroplasty with a diagnosis of rotator cuff disease, and we believe that this may partly account for the recent increase in the use of the reverse TSA in this patient population. Additionally, we found shoulder arthroplasty in RA patients to be safe in the early postoperative period, with no significant increase in cost as compared with patients undergoing shoulder arthroplasty without a diagnosis of RA. Although we did observe a significant increase in length of stay among RA patients as compared with non-RA patients, the absolute difference was only 0.1 days, and given the aforementioned similarities in cost between RA and non-RA patients, we do not believe this difference to be clinically significant.

It has been theorized that the utilization of TJA in RA patients has been decreasing with improvements in medical management; however, this is largely based upon literature pertaining to lower extremity TJA.² On the contrary, past research pertaining to the utilization of shoulder arthroplasty in RA patients has been highly variable. For instance, a Swedish study demonstrated a statistically significant decrease in admissions associated with RA-related upper limb surgery and a stable rate of shoulder arthroplasty between 1998 and 2004.⁷ Similarly, a Finnish study demonstrated that the annual incidence of primary joint arthroplasty in RA patients had declined from 1995 to 2010, with a greater decline for upper-limb arthroplasty as compared with lower-limb arthroplasty.⁸ Despite these European observations, Jain and colleagues⁹ reported an increasing rate of TSA among RA patients in the US between the years 1992 and 2005. In this study, we demonstrate a clear increase in the utilization of shoulder arthroplasty among RA patients between 2002 and 2011. What was most striking about our observation was that the rise in utilization appeared to be driven by an increase in TSA, whereas the utilization of HA decreased over time. This change in practice likely reflects several factors, including the multitude of studies that have demonstrated improved outcomes with anatomic TSA as compared with HA in RA patients.^10-14

Perhaps the most interesting aspect of our data was the recent exponential rise in the utilization of the reverse TSA. Despite improved outcomes following TSA as compared with HA in RA patients, these outcomes all appear to be highly dependent upon the integrity of the rotator cuff.¹⁰ In fact, there is evidence that failure of the rotator cuff could be as high as 75% within 10 years of TSA in patients with RA,¹⁵ which ultimately could jeopardize the long-term durability of the TSA implant in this patient population.¹¹ For this reason, interest in the reverse TSA for the RA patient population has increased since its introduction in the US in 2004;¹⁶ in fact, in RA patients with end-stage inflammatory arthropathy and a damaged rotator cuff, the reverse TSA has demonstrated excellent results.^17-20 Based upon this evidence, it is not surprising that we found an exponential rise in the use of the reverse TSA since 2010, which corresponds to the introduction of an ICD-9 code for this implant.²¹ Prior to 2010, it is likely that many implanted reverse TSAs were coded as TSA, and for this reason, we believe that the observed rise in the utilization of TSA in RA patients prior to 2010 may have been partly fueled by an increase in the use of the reverse TSA. To further support this theory, there was a dramatic decrease in the use of anatomic TSA following 2010, and we believe this was related to increased awareness of the newly introduced reverse TSA code among surgeons.

Another consideration when examining the utilization of shoulder arthroplasty in RA patients is its versatility in managing different disease states, including rotator cuff disease. As has been documented in the literature, outcomes of rotator cuff repair in RA patients are discouraging.²² For this reason, it is reasonable for surgeons and patients with RA to consider alternatives to rotator cuff repair when nonoperative management has failed to provide adequate improvement in symptoms. One alternative may be shoulder arthroplasty, namely the reverse TSA. In this study, we observed a significant increase in the rate of diagnosis of rotator cuff disease among RA patients undergoing shoulder arthroplasty from 2002 to 2011 (9.7% in 2002 to 15.2% in 2011, P < .001), and it is our belief that the simultaneous increase in the diagnosis of rotator cuff disease and use of TSA is not coincidental. More specifically, there is likely an emerging trend among surgeons toward using the reverse TSA to manage rotator cuff tears in the RA population, rather than undertaking a rotator cuff repair that carries a high rate of failure. Going forward, there is a need to not only identify this trend more clearly but to also compare the outcomes between reverse TSA and rotator cuff repair in the management of rotator cuff tears in RA patients.

Continue to: In this study, we observed...

In this study, we observed that RA patients undergoing shoulder arthroplasty were significantly younger than non-RA patients undergoing shoulder arthroplasty. At first, this observation seems to counter recent literature suggesting that the age of patients with inflammatory arthropathy undergoing TJA is increasing over time;¹ however, looking more closely at the data, it becomes clearer that the mean age we report is actually a relative increase as compared with past clinical studies pertaining to RA patients undergoing shoulder arthroplasty (mean ages of 47 years,²³ 55 years,²⁴ 60 years,¹⁰ and 62 years²⁵). On the other hand, the continued existence of an age gap between RA and non-RA patients undergoing shoulder arthroplasty may be the result of several possible phenomena. First, this may reflect issues with patient access to and coverage of expensive biologic antirheumatic medication that would otherwise mitigate disease progression. For instance, the out-of-pocket expense for biologic medication through Medicaid and Medicare is substantial,²⁶ which has direct implications on over two-thirds of our RA cohort. Second, it may be skewed by the proportion of RA patients who have previously been or continue to be poorly managed, enabling disease progression to end-stage arthropathy at a younger age. Ultimately, further investigation is needed to determine the reasons for this continued age disparity.

In comparing RA and non-RA patients undergoing shoulder arthroplasty, we did not find a significant difference in the overall nor the individual rates of early adverse events. This finding appears to be unique, as similar studies pertaining to total knee arthroplasty (TKA) demonstrated a significantly higher incidence of postoperative pneumonia and bleeding requiring transfusion among RA patients as compared with non-RA patients.²⁷ In patients with RA being treated with biologic medication and undergoing shoulder arthroplasty, the frequent concern in the postoperative period is the integrity of the wound and the potential for infection.²⁸ In this study, we did not find a significant difference in the rate of early infection, and although the difference in the rate of early wound dehiscence approached significance, it did not meet the threshold of 0.05 (P = .09). This finding is in keeping with the aforementioned NIS study pertaining to TKA, and we believe that it likely reflects the short duration of follow-up for patients in both studies. Given the nature of the database we utilized, we were only privy to complications that arose during the inpatient hospital stay, and it is likely that the clear majority of patients who develop a postoperative infection or wound dehiscence do so in the postoperative setting following discharge. A second concern regarding postoperative wound complications is the management of biologic medication in the perioperative period, which we cannot determine using this database. Despite all these limitations specific to this database, a past systematic review of reverse TSA in RA patients found a low rate of deep infection after reverse TSA in RA patients (3.3%),¹⁷ which was not higher than that after shoulder arthroplasty performed in non-RA patients.

A final demonstration from this study is that the hospital length of stay was significantly longer for RA patients than non-RA patients undergoing shoulder arthroplasty; however, given that the difference was only 0.1 days, and there was no significant difference in hospitalization cost, we are inclined to believe that statistical significance may not translate into clinical significance in this scenario. Ultimately, we do believe that length of stay is an important consideration in the current healthcare system, and given our finding that shoulder arthroplasty in the RA patient is safe in the early postoperative period, that a prolonged postoperative hospitalization is not warranted on the sole basis of a patient’s history of RA.

As with all studies using data from a search of an administrative database, such as the NIS database, this study has limitations. First, this type of research is limited by the reliability of both diagnosis and procedural coding. Although the NIS database has demonstrated high reliability,³ it is still possible that events may have been miscoded. Second, the tracking period for adverse events is limited to the inpatient hospital stay, which may be too short to detect certain postoperative complications. As such, the rates we report are likely underestimates of the true incidence of these complications, but this is true for both the RA and non-RA populations. Third, the comparisons we draw between RA and non-RA patients are limited to the scope of the NIS database and the available data; as such, we could not draw comparisons between preoperative disease stage, intraoperative findings, and postoperative course following hospital discharge. Lastly, our data are limited to a distinct period between 2002 and 2011 and may not reflect current practice. Ultimately, our findings may underestimate current trends in shoulder arthroplasty utilization among RA patients, particularly for the reverse TSA.

CONCLUSION

In this study, we found that the utilization of shoulder arthroplasty in patients with RA increased significantly from 2002 to 2011, largely related to a rise in the utilization of TSA. Similarly, we observed a rise in the proportion of RA patients undergoing shoulder arthroplasty with a corresponding diagnosis of rotator cuff disease, and we believe the increased utilization of shoulder arthroplasty among RA patients resulted from management of both end-stage inflammatory arthropathy and rotator cuff disease. Although we did not find a significant difference between RA and non-RA patients in the rates of early adverse events and overall hospitalization costs following shoulder arthroplasty, length of stay was significantly longer among RA patients; however, the absolute difference does not appear to be clinically significant.

ABSTRACT

It has been suggested that the utilization of joint arthroplasty in patients with rheumatoid arthritis (RA) is decreasing; however, this observation is largely based upon evidence pertaining to lower-extremity joint arthroplasty. It remains unknown if these observed trends also hold true for shoulder arthroplasty. The purpose of this study is to utilize a nationally representative population database in the US to identify trends in the utilization of shoulder arthroplasty among patients with RA. Secondarily, we sought to determine the rate of early adverse events, length of stay, and hospitalization costs associated with RA patients undergoing shoulder arthroplasty and to compare these outcomes to those of patients without a diagnosis of RA undergoing shoulder arthroplasty. Using a large population database in the US, we determined the annual rates of shoulder arthroplasty (overall and individual) in RA patients between 2002 and 2011. Early adverse events, length of stay, and hospitalization costs were determined and compared with those of non-RA patients undergoing shoulder arthroplasty. Overall, we identified 332,593 patients who underwent shoulder arthroplasty between 2002 and 2011, of whom 17,883 patients (5.4%) had a diagnosis of RA. Over the study period, there was a significant increase in the utilization of shoulder arthroplasty in RA patients, particularly total shoulder arthroplasty. Over the same period, there was a significant increase in the number of RA patients who underwent shoulder arthroplasty with a diagnosis of rotator cuff disease. There were no significant differences in adverse events or mean hospitalization costs between RA and non-RA patients. Non-RA patients had a significantly shorter length of stay; however, the difference did not appear to be clinically significant. In conclusion, the utilization of shoulder arthroplasty in patients with RA significantly increased from 2002 to 2011, which may partly reflect a trend toward management of rotator cuff disease with arthroplasty rather than repair.

Continue to: It has been suggested...

It has been suggested that the utilization of total joint arthroplasty (TJA) in patients with rheumatoid arthritis (RA) is decreasing over time;¹ however, this observation is largely based upon evidence pertaining to lower extremity TJA.² It remains unknown if these observed trends also hold true for shoulder arthroplasty, whereby the utilization of shoulder arthroplasty in RA patients is not limited to the management of end-stage inflammatory arthropathy. In this study, we used a nationally representative population database in the US to identify trends in the utilization of shoulder arthroplasty among patients with RA. As a secondary objective, we sought to determine the rate of early adverse events, length of stay, and hospitalization costs associated with RA patients undergoing shoulder arthroplasty and compare these outcomes to those of patients without a diagnosis of RA undergoing shoulder arthroplasty. We hypothesize that the utilization of shoulder arthroplasty in RA patients would be decreasing, but adverse events, length of stay, and hospitalization costs would not differ between patients with and without RA undergoing shoulder arthroplasty.

METHODS

We conducted a retrospective cohort study using the Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS) from 2002 to 2011.³ The NIS comprises a 20% stratified sample of all hospital discharges in the US. The NIS includes information about patient characteristics (age, sex, insurance status, and medical comorbidities) and hospitalization outcomes (adverse events, costs, and length of stay). The NIS allows identification of hospitalizations according to procedures and diagnoses using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Given the anonymity of this study, it was exempt from Institutional Review Board ethics approval.

Hospitalizations were selected for the study based on ICD-9-CM procedural codes for hemiarthroplasty (81.81), anatomic total shoulder arthroplasty (TSA) (81.80), and reverse TSA (81.88). These patients were then stratified by an ICD-9-CM diagnosis of RA (714.X). We also utilized ICD-9-CM diagnosis codes to determine the presence of rotator cuff pathology at the time of shoulder arthroplasty (726.13, 727.61, 840.4) and to exclude patients with a history of trauma (812.X, 716.11, 733.8X). In a separate analysis, all patients in the NIS database with an ICD-9-CM diagnosis of RA were identified for each calendar year of the study, and a national estimate of RA patients was generated annually to assess overall and individual utilization rates of shoulder arthroplasty in this population (the national estimate served as the denominator).

Preoperative patient data withdrawn from the NIS included age, sex, insurance status, and medical comorbidities. An Elixhauser Comorbidity Index (ECI) was generated for each patient based on the presence of 29 comorbid conditions. The ECI was chosen because of its capacity to accurately predict mortality and represent the patient burden of comorbidities in similar administrative database studies.^4-6

Early adverse events were also chosen based on ICD-9-CM diagnosis codes (Appendix A), and included the following: death, acute kidney injury, cardiac arrest, thromboembolic event, myocardial infarction, peripheral nerve injury, pneumonia, sepsis, stroke, surgical site infection, urinary tract infection, and wound dehiscence. The overall adverse event rate was defined as the occurrence of ≥1 of the above adverse events in a patient.

Appendix A. ICD-9-CM Codes Corresponding to Postoperative Adverse Events

Abbreviation: ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification

Length of stay and total hospital charges were available for each patient. Length of stay represents the number of calendar days a patient stayed in the hospital. All hospital charges were converted to hospitalization costs using the HCUP Cost-to-Charge Ratio Files. All hospitalization costs were adjusted for inflation using the US Bureau of Labor statistics yearly inflation calculator to represent charges in the year 2011, which was the final and most recent year in this study.

Continue to: Statistical analysis...

STATISTICAL ANALYSIS

Statistical analyses were conducted using Stata version 13.1 (StataCorp, LP). All analyses took into account the complex survey design of the NIS. Discharge weights, strata, and cluster variables were included to correctly estimate variance and to produce national estimates from the stratified sample. Pearson’s chi-squared test was used to compare age, sex, ECI, and insurance status between RA and non-RA patients undergoing shoulder arthroplasty.

Bivariate and multivariate logistic regressions were subsequently used to compare the rates of adverse events between RA and non-RA patients undergoing shoulder arthroplasty (non-RA cases were used as the reference). Multivariate linear regressions were used to compare hospital length of stay and hospitalization costs between RA and non-RA patients undergoing shoulder arthroplasty. The multivariate regressions were adjusted for baseline differences in age, sex, ECI, and insurance status. Cochran-Armitage tests for trend were used to assess trends over time. All tests were 2-tailed, and the statistical difference was established at a 2-sided α level of 0.05 (P < .05).

RESULTS

Overall, we identified 332,593 patients who underwent shoulder arthroplasty in the US between 2002 and 2011, of which 17,883 patients (5.4%) had a diagnosis of RA. In comparison with non-RA patients undergoing shoulder arthroplasty, patients with RA at the time of shoulder arthroplasty were significantly younger (65.2 ± 12.5 years vs 68.4 ± 11.0 years, P < .001), included a significantly greater proportion of female patients (76.7% vs 53.8%, P < .001), and included a significantly higher proportion of patients with Medicaid insurance (3.6% vs 2.3%, P < .001). There were no significant differences in the mean ECI between patients with and without a diagnosis of RA (Table 1). As depicted in Table 1, there were significant differences in the utilization of specific shoulder arthroplasty types between patients with and without RA, whereby a significantly greater proportion of RA patients underwent hemiarthroplasty (HA) (31.6% vs 29.3%, P = .002) and reverse TSA (7.7% vs 6.6%, P = .002), whereas a significantly greater proportion of non-RA patients underwent anatomic SA (64.0% vs 60.8%, P = .002).

Over the study period from 2002 to 2011, there was a significant increase in the overall utilization of shoulder arthroplasty in RA patients, as indicated by both the absolute number and the proportion of patients with a diagnosis of RA (P < .001) (Table 2, Figure). More specifically, 0.39% of RA patients underwent shoulder arthroplasty in 2002, as compared with 0.58% of RA patients in 2011 (P < .001) (Table 2). With respect to specific arthroplasty types, there was an exponential rise in the utilization of reverse TSA beginning in 2010 and a corresponding decrease in the rates of both HA and anatomic TSA (Table 2, Figure). In addition to changes in shoulder arthroplasty utilization over time among RA patients, we also observed a significant increase in the number of RA patients undergoing shoulder arthroplasty with a corresponding diagnosis of rotator cuff disease (9.7% in 2002 to 15.2% in 2011, P < .001).

Table 2. The Annual Utilization of Shoulder Arthroplasty Among Patients with a Diagnosis of Rheumatoid Arthritis.

^aRate determined as number of RA patients undergoing shoulder arthroplasty compared to the number of patients with an RA diagnosis in the stated calendar year.

Abbreviations: HA, hemiarthroplasty; RA, rheumatoid arthritis; TSA, total shoulder arthroplasty.

Continue to: Among patients with RA...

Among patients with RA undergoing shoulder arthroplasty, the overall rate of early adverse events was 3.12%, of which the most common early adverse events were urinary tract infections (1.8%), acute kidney injury (0.66%), and pneumonia (0.38%) (Table 3). As compared with patients without a diagnosis of RA undergoing shoulder arthroplasty, there were no significant differences in the overall and individual rates of early adverse events (Table 3).

Table 3. A Comparison of Early Adverse Events, Length of Stay, and Cost Between Patients With and Without Rheumatoid Arthritis (RA) Undergoing Shoulder Arthroplasty

^aReported in days. ^bReported in 2011 US dollars, adjusted for inflation.

The mean length of stay following shoulder arthroplasty in RA patients was 2.4 ± 1.6 days, and the mean hospitalization cost was $14,787 ± $7625 (Table 3). As compared with non-RA patients undergoing shoulder arthroplasty, there were no significant differences in the mean hospitalization costs; however, non-RA patients had a significantly shorter length of stay by 0.1 days (P = .002) (Table 3).

DISCUSSION

In this study, we observed that the utilization of shoulder arthroplasty in patients with RA increased significantly in the decade from 2002 to 2011, largely related to a rise in TSA. Interestingly, we also observed a corresponding rise in the proportion of RA patients undergoing shoulder arthroplasty with a diagnosis of rotator cuff disease, and we believe that this may partly account for the recent increase in the use of the reverse TSA in this patient population. Additionally, we found shoulder arthroplasty in RA patients to be safe in the early postoperative period, with no significant increase in cost as compared with patients undergoing shoulder arthroplasty without a diagnosis of RA. Although we did observe a significant increase in length of stay among RA patients as compared with non-RA patients, the absolute difference was only 0.1 days, and given the aforementioned similarities in cost between RA and non-RA patients, we do not believe this difference to be clinically significant.

It has been theorized that the utilization of TJA in RA patients has been decreasing with improvements in medical management; however, this is largely based upon literature pertaining to lower extremity TJA.² On the contrary, past research pertaining to the utilization of shoulder arthroplasty in RA patients has been highly variable. For instance, a Swedish study demonstrated a statistically significant decrease in admissions associated with RA-related upper limb surgery and a stable rate of shoulder arthroplasty between 1998 and 2004.⁷ Similarly, a Finnish study demonstrated that the annual incidence of primary joint arthroplasty in RA patients had declined from 1995 to 2010, with a greater decline for upper-limb arthroplasty as compared with lower-limb arthroplasty.⁸ Despite these European observations, Jain and colleagues⁹ reported an increasing rate of TSA among RA patients in the US between the years 1992 and 2005. In this study, we demonstrate a clear increase in the utilization of shoulder arthroplasty among RA patients between 2002 and 2011. What was most striking about our observation was that the rise in utilization appeared to be driven by an increase in TSA, whereas the utilization of HA decreased over time. This change in practice likely reflects several factors, including the multitude of studies that have demonstrated improved outcomes with anatomic TSA as compared with HA in RA patients.^10-14

Perhaps the most interesting aspect of our data was the recent exponential rise in the utilization of the reverse TSA. Despite improved outcomes following TSA as compared with HA in RA patients, these outcomes all appear to be highly dependent upon the integrity of the rotator cuff.¹⁰ In fact, there is evidence that failure of the rotator cuff could be as high as 75% within 10 years of TSA in patients with RA,¹⁵ which ultimately could jeopardize the long-term durability of the TSA implant in this patient population.¹¹ For this reason, interest in the reverse TSA for the RA patient population has increased since its introduction in the US in 2004;¹⁶ in fact, in RA patients with end-stage inflammatory arthropathy and a damaged rotator cuff, the reverse TSA has demonstrated excellent results.^17-20 Based upon this evidence, it is not surprising that we found an exponential rise in the use of the reverse TSA since 2010, which corresponds to the introduction of an ICD-9 code for this implant.²¹ Prior to 2010, it is likely that many implanted reverse TSAs were coded as TSA, and for this reason, we believe that the observed rise in the utilization of TSA in RA patients prior to 2010 may have been partly fueled by an increase in the use of the reverse TSA. To further support this theory, there was a dramatic decrease in the use of anatomic TSA following 2010, and we believe this was related to increased awareness of the newly introduced reverse TSA code among surgeons.

Another consideration when examining the utilization of shoulder arthroplasty in RA patients is its versatility in managing different disease states, including rotator cuff disease. As has been documented in the literature, outcomes of rotator cuff repair in RA patients are discouraging.²² For this reason, it is reasonable for surgeons and patients with RA to consider alternatives to rotator cuff repair when nonoperative management has failed to provide adequate improvement in symptoms. One alternative may be shoulder arthroplasty, namely the reverse TSA. In this study, we observed a significant increase in the rate of diagnosis of rotator cuff disease among RA patients undergoing shoulder arthroplasty from 2002 to 2011 (9.7% in 2002 to 15.2% in 2011, P < .001), and it is our belief that the simultaneous increase in the diagnosis of rotator cuff disease and use of TSA is not coincidental. More specifically, there is likely an emerging trend among surgeons toward using the reverse TSA to manage rotator cuff tears in the RA population, rather than undertaking a rotator cuff repair that carries a high rate of failure. Going forward, there is a need to not only identify this trend more clearly but to also compare the outcomes between reverse TSA and rotator cuff repair in the management of rotator cuff tears in RA patients.

Continue to: In this study, we observed...

In this study, we observed that RA patients undergoing shoulder arthroplasty were significantly younger than non-RA patients undergoing shoulder arthroplasty. At first, this observation seems to counter recent literature suggesting that the age of patients with inflammatory arthropathy undergoing TJA is increasing over time;¹ however, looking more closely at the data, it becomes clearer that the mean age we report is actually a relative increase as compared with past clinical studies pertaining to RA patients undergoing shoulder arthroplasty (mean ages of 47 years,²³ 55 years,²⁴ 60 years,¹⁰ and 62 years²⁵). On the other hand, the continued existence of an age gap between RA and non-RA patients undergoing shoulder arthroplasty may be the result of several possible phenomena. First, this may reflect issues with patient access to and coverage of expensive biologic antirheumatic medication that would otherwise mitigate disease progression. For instance, the out-of-pocket expense for biologic medication through Medicaid and Medicare is substantial,²⁶ which has direct implications on over two-thirds of our RA cohort. Second, it may be skewed by the proportion of RA patients who have previously been or continue to be poorly managed, enabling disease progression to end-stage arthropathy at a younger age. Ultimately, further investigation is needed to determine the reasons for this continued age disparity.

In comparing RA and non-RA patients undergoing shoulder arthroplasty, we did not find a significant difference in the overall nor the individual rates of early adverse events. This finding appears to be unique, as similar studies pertaining to total knee arthroplasty (TKA) demonstrated a significantly higher incidence of postoperative pneumonia and bleeding requiring transfusion among RA patients as compared with non-RA patients.²⁷ In patients with RA being treated with biologic medication and undergoing shoulder arthroplasty, the frequent concern in the postoperative period is the integrity of the wound and the potential for infection.²⁸ In this study, we did not find a significant difference in the rate of early infection, and although the difference in the rate of early wound dehiscence approached significance, it did not meet the threshold of 0.05 (P = .09). This finding is in keeping with the aforementioned NIS study pertaining to TKA, and we believe that it likely reflects the short duration of follow-up for patients in both studies. Given the nature of the database we utilized, we were only privy to complications that arose during the inpatient hospital stay, and it is likely that the clear majority of patients who develop a postoperative infection or wound dehiscence do so in the postoperative setting following discharge. A second concern regarding postoperative wound complications is the management of biologic medication in the perioperative period, which we cannot determine using this database. Despite all these limitations specific to this database, a past systematic review of reverse TSA in RA patients found a low rate of deep infection after reverse TSA in RA patients (3.3%),¹⁷ which was not higher than that after shoulder arthroplasty performed in non-RA patients.

A final demonstration from this study is that the hospital length of stay was significantly longer for RA patients than non-RA patients undergoing shoulder arthroplasty; however, given that the difference was only 0.1 days, and there was no significant difference in hospitalization cost, we are inclined to believe that statistical significance may not translate into clinical significance in this scenario. Ultimately, we do believe that length of stay is an important consideration in the current healthcare system, and given our finding that shoulder arthroplasty in the RA patient is safe in the early postoperative period, that a prolonged postoperative hospitalization is not warranted on the sole basis of a patient’s history of RA.

As with all studies using data from a search of an administrative database, such as the NIS database, this study has limitations. First, this type of research is limited by the reliability of both diagnosis and procedural coding. Although the NIS database has demonstrated high reliability,³ it is still possible that events may have been miscoded. Second, the tracking period for adverse events is limited to the inpatient hospital stay, which may be too short to detect certain postoperative complications. As such, the rates we report are likely underestimates of the true incidence of these complications, but this is true for both the RA and non-RA populations. Third, the comparisons we draw between RA and non-RA patients are limited to the scope of the NIS database and the available data; as such, we could not draw comparisons between preoperative disease stage, intraoperative findings, and postoperative course following hospital discharge. Lastly, our data are limited to a distinct period between 2002 and 2011 and may not reflect current practice. Ultimately, our findings may underestimate current trends in shoulder arthroplasty utilization among RA patients, particularly for the reverse TSA.

CONCLUSION

In this study, we found that the utilization of shoulder arthroplasty in patients with RA increased significantly from 2002 to 2011, largely related to a rise in the utilization of TSA. Similarly, we observed a rise in the proportion of RA patients undergoing shoulder arthroplasty with a corresponding diagnosis of rotator cuff disease, and we believe the increased utilization of shoulder arthroplasty among RA patients resulted from management of both end-stage inflammatory arthropathy and rotator cuff disease. Although we did not find a significant difference between RA and non-RA patients in the rates of early adverse events and overall hospitalization costs following shoulder arthroplasty, length of stay was significantly longer among RA patients; however, the absolute difference does not appear to be clinically significant.

ABSTRACT

It has been suggested that the utilization of joint arthroplasty in patients with rheumatoid arthritis (RA) is decreasing; however, this observation is largely based upon evidence pertaining to lower-extremity joint arthroplasty. It remains unknown if these observed trends also hold true for shoulder arthroplasty. The purpose of this study is to utilize a nationally representative population database in the US to identify trends in the utilization of shoulder arthroplasty among patients with RA. Secondarily, we sought to determine the rate of early adverse events, length of stay, and hospitalization costs associated with RA patients undergoing shoulder arthroplasty and to compare these outcomes to those of patients without a diagnosis of RA undergoing shoulder arthroplasty. Using a large population database in the US, we determined the annual rates of shoulder arthroplasty (overall and individual) in RA patients between 2002 and 2011. Early adverse events, length of stay, and hospitalization costs were determined and compared with those of non-RA patients undergoing shoulder arthroplasty. Overall, we identified 332,593 patients who underwent shoulder arthroplasty between 2002 and 2011, of whom 17,883 patients (5.4%) had a diagnosis of RA. Over the study period, there was a significant increase in the utilization of shoulder arthroplasty in RA patients, particularly total shoulder arthroplasty. Over the same period, there was a significant increase in the number of RA patients who underwent shoulder arthroplasty with a diagnosis of rotator cuff disease. There were no significant differences in adverse events or mean hospitalization costs between RA and non-RA patients. Non-RA patients had a significantly shorter length of stay; however, the difference did not appear to be clinically significant. In conclusion, the utilization of shoulder arthroplasty in patients with RA significantly increased from 2002 to 2011, which may partly reflect a trend toward management of rotator cuff disease with arthroplasty rather than repair.

Continue to: It has been suggested...

It has been suggested that the utilization of total joint arthroplasty (TJA) in patients with rheumatoid arthritis (RA) is decreasing over time;¹ however, this observation is largely based upon evidence pertaining to lower extremity TJA.² It remains unknown if these observed trends also hold true for shoulder arthroplasty, whereby the utilization of shoulder arthroplasty in RA patients is not limited to the management of end-stage inflammatory arthropathy. In this study, we used a nationally representative population database in the US to identify trends in the utilization of shoulder arthroplasty among patients with RA. As a secondary objective, we sought to determine the rate of early adverse events, length of stay, and hospitalization costs associated with RA patients undergoing shoulder arthroplasty and compare these outcomes to those of patients without a diagnosis of RA undergoing shoulder arthroplasty. We hypothesize that the utilization of shoulder arthroplasty in RA patients would be decreasing, but adverse events, length of stay, and hospitalization costs would not differ between patients with and without RA undergoing shoulder arthroplasty.

METHODS

We conducted a retrospective cohort study using the Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS) from 2002 to 2011.³ The NIS comprises a 20% stratified sample of all hospital discharges in the US. The NIS includes information about patient characteristics (age, sex, insurance status, and medical comorbidities) and hospitalization outcomes (adverse events, costs, and length of stay). The NIS allows identification of hospitalizations according to procedures and diagnoses using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Given the anonymity of this study, it was exempt from Institutional Review Board ethics approval.

Hospitalizations were selected for the study based on ICD-9-CM procedural codes for hemiarthroplasty (81.81), anatomic total shoulder arthroplasty (TSA) (81.80), and reverse TSA (81.88). These patients were then stratified by an ICD-9-CM diagnosis of RA (714.X). We also utilized ICD-9-CM diagnosis codes to determine the presence of rotator cuff pathology at the time of shoulder arthroplasty (726.13, 727.61, 840.4) and to exclude patients with a history of trauma (812.X, 716.11, 733.8X). In a separate analysis, all patients in the NIS database with an ICD-9-CM diagnosis of RA were identified for each calendar year of the study, and a national estimate of RA patients was generated annually to assess overall and individual utilization rates of shoulder arthroplasty in this population (the national estimate served as the denominator).

Preoperative patient data withdrawn from the NIS included age, sex, insurance status, and medical comorbidities. An Elixhauser Comorbidity Index (ECI) was generated for each patient based on the presence of 29 comorbid conditions. The ECI was chosen because of its capacity to accurately predict mortality and represent the patient burden of comorbidities in similar administrative database studies.^4-6

Early adverse events were also chosen based on ICD-9-CM diagnosis codes (Appendix A), and included the following: death, acute kidney injury, cardiac arrest, thromboembolic event, myocardial infarction, peripheral nerve injury, pneumonia, sepsis, stroke, surgical site infection, urinary tract infection, and wound dehiscence. The overall adverse event rate was defined as the occurrence of ≥1 of the above adverse events in a patient.

Appendix A. ICD-9-CM Codes Corresponding to Postoperative Adverse Events

Abbreviation: ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification

Length of stay and total hospital charges were available for each patient. Length of stay represents the number of calendar days a patient stayed in the hospital. All hospital charges were converted to hospitalization costs using the HCUP Cost-to-Charge Ratio Files. All hospitalization costs were adjusted for inflation using the US Bureau of Labor statistics yearly inflation calculator to represent charges in the year 2011, which was the final and most recent year in this study.

Continue to: Statistical analysis...

STATISTICAL ANALYSIS

Statistical analyses were conducted using Stata version 13.1 (StataCorp, LP). All analyses took into account the complex survey design of the NIS. Discharge weights, strata, and cluster variables were included to correctly estimate variance and to produce national estimates from the stratified sample. Pearson’s chi-squared test was used to compare age, sex, ECI, and insurance status between RA and non-RA patients undergoing shoulder arthroplasty.

Bivariate and multivariate logistic regressions were subsequently used to compare the rates of adverse events between RA and non-RA patients undergoing shoulder arthroplasty (non-RA cases were used as the reference). Multivariate linear regressions were used to compare hospital length of stay and hospitalization costs between RA and non-RA patients undergoing shoulder arthroplasty. The multivariate regressions were adjusted for baseline differences in age, sex, ECI, and insurance status. Cochran-Armitage tests for trend were used to assess trends over time. All tests were 2-tailed, and the statistical difference was established at a 2-sided α level of 0.05 (P < .05).

RESULTS

Overall, we identified 332,593 patients who underwent shoulder arthroplasty in the US between 2002 and 2011, of which 17,883 patients (5.4%) had a diagnosis of RA. In comparison with non-RA patients undergoing shoulder arthroplasty, patients with RA at the time of shoulder arthroplasty were significantly younger (65.2 ± 12.5 years vs 68.4 ± 11.0 years, P < .001), included a significantly greater proportion of female patients (76.7% vs 53.8%, P < .001), and included a significantly higher proportion of patients with Medicaid insurance (3.6% vs 2.3%, P < .001). There were no significant differences in the mean ECI between patients with and without a diagnosis of RA (Table 1). As depicted in Table 1, there were significant differences in the utilization of specific shoulder arthroplasty types between patients with and without RA, whereby a significantly greater proportion of RA patients underwent hemiarthroplasty (HA) (31.6% vs 29.3%, P = .002) and reverse TSA (7.7% vs 6.6%, P = .002), whereas a significantly greater proportion of non-RA patients underwent anatomic SA (64.0% vs 60.8%, P = .002).

Over the study period from 2002 to 2011, there was a significant increase in the overall utilization of shoulder arthroplasty in RA patients, as indicated by both the absolute number and the proportion of patients with a diagnosis of RA (P < .001) (Table 2, Figure). More specifically, 0.39% of RA patients underwent shoulder arthroplasty in 2002, as compared with 0.58% of RA patients in 2011 (P < .001) (Table 2). With respect to specific arthroplasty types, there was an exponential rise in the utilization of reverse TSA beginning in 2010 and a corresponding decrease in the rates of both HA and anatomic TSA (Table 2, Figure). In addition to changes in shoulder arthroplasty utilization over time among RA patients, we also observed a significant increase in the number of RA patients undergoing shoulder arthroplasty with a corresponding diagnosis of rotator cuff disease (9.7% in 2002 to 15.2% in 2011, P < .001).

Table 2. The Annual Utilization of Shoulder Arthroplasty Among Patients with a Diagnosis of Rheumatoid Arthritis.

^aRate determined as number of RA patients undergoing shoulder arthroplasty compared to the number of patients with an RA diagnosis in the stated calendar year.

Abbreviations: HA, hemiarthroplasty; RA, rheumatoid arthritis; TSA, total shoulder arthroplasty.

Continue to: Among patients with RA...

Among patients with RA undergoing shoulder arthroplasty, the overall rate of early adverse events was 3.12%, of which the most common early adverse events were urinary tract infections (1.8%), acute kidney injury (0.66%), and pneumonia (0.38%) (Table 3). As compared with patients without a diagnosis of RA undergoing shoulder arthroplasty, there were no significant differences in the overall and individual rates of early adverse events (Table 3).

Table 3. A Comparison of Early Adverse Events, Length of Stay, and Cost Between Patients With and Without Rheumatoid Arthritis (RA) Undergoing Shoulder Arthroplasty

^aReported in days. ^bReported in 2011 US dollars, adjusted for inflation.

The mean length of stay following shoulder arthroplasty in RA patients was 2.4 ± 1.6 days, and the mean hospitalization cost was $14,787 ± $7625 (Table 3). As compared with non-RA patients undergoing shoulder arthroplasty, there were no significant differences in the mean hospitalization costs; however, non-RA patients had a significantly shorter length of stay by 0.1 days (P = .002) (Table 3).

DISCUSSION

In this study, we observed that the utilization of shoulder arthroplasty in patients with RA increased significantly in the decade from 2002 to 2011, largely related to a rise in TSA. Interestingly, we also observed a corresponding rise in the proportion of RA patients undergoing shoulder arthroplasty with a diagnosis of rotator cuff disease, and we believe that this may partly account for the recent increase in the use of the reverse TSA in this patient population. Additionally, we found shoulder arthroplasty in RA patients to be safe in the early postoperative period, with no significant increase in cost as compared with patients undergoing shoulder arthroplasty without a diagnosis of RA. Although we did observe a significant increase in length of stay among RA patients as compared with non-RA patients, the absolute difference was only 0.1 days, and given the aforementioned similarities in cost between RA and non-RA patients, we do not believe this difference to be clinically significant.

It has been theorized that the utilization of TJA in RA patients has been decreasing with improvements in medical management; however, this is largely based upon literature pertaining to lower extremity TJA.² On the contrary, past research pertaining to the utilization of shoulder arthroplasty in RA patients has been highly variable. For instance, a Swedish study demonstrated a statistically significant decrease in admissions associated with RA-related upper limb surgery and a stable rate of shoulder arthroplasty between 1998 and 2004.⁷ Similarly, a Finnish study demonstrated that the annual incidence of primary joint arthroplasty in RA patients had declined from 1995 to 2010, with a greater decline for upper-limb arthroplasty as compared with lower-limb arthroplasty.⁸ Despite these European observations, Jain and colleagues⁹ reported an increasing rate of TSA among RA patients in the US between the years 1992 and 2005. In this study, we demonstrate a clear increase in the utilization of shoulder arthroplasty among RA patients between 2002 and 2011. What was most striking about our observation was that the rise in utilization appeared to be driven by an increase in TSA, whereas the utilization of HA decreased over time. This change in practice likely reflects several factors, including the multitude of studies that have demonstrated improved outcomes with anatomic TSA as compared with HA in RA patients.^10-14

Perhaps the most interesting aspect of our data was the recent exponential rise in the utilization of the reverse TSA. Despite improved outcomes following TSA as compared with HA in RA patients, these outcomes all appear to be highly dependent upon the integrity of the rotator cuff.¹⁰ In fact, there is evidence that failure of the rotator cuff could be as high as 75% within 10 years of TSA in patients with RA,¹⁵ which ultimately could jeopardize the long-term durability of the TSA implant in this patient population.¹¹ For this reason, interest in the reverse TSA for the RA patient population has increased since its introduction in the US in 2004;¹⁶ in fact, in RA patients with end-stage inflammatory arthropathy and a damaged rotator cuff, the reverse TSA has demonstrated excellent results.^17-20 Based upon this evidence, it is not surprising that we found an exponential rise in the use of the reverse TSA since 2010, which corresponds to the introduction of an ICD-9 code for this implant.²¹ Prior to 2010, it is likely that many implanted reverse TSAs were coded as TSA, and for this reason, we believe that the observed rise in the utilization of TSA in RA patients prior to 2010 may have been partly fueled by an increase in the use of the reverse TSA. To further support this theory, there was a dramatic decrease in the use of anatomic TSA following 2010, and we believe this was related to increased awareness of the newly introduced reverse TSA code among surgeons.

Another consideration when examining the utilization of shoulder arthroplasty in RA patients is its versatility in managing different disease states, including rotator cuff disease. As has been documented in the literature, outcomes of rotator cuff repair in RA patients are discouraging.²² For this reason, it is reasonable for surgeons and patients with RA to consider alternatives to rotator cuff repair when nonoperative management has failed to provide adequate improvement in symptoms. One alternative may be shoulder arthroplasty, namely the reverse TSA. In this study, we observed a significant increase in the rate of diagnosis of rotator cuff disease among RA patients undergoing shoulder arthroplasty from 2002 to 2011 (9.7% in 2002 to 15.2% in 2011, P < .001), and it is our belief that the simultaneous increase in the diagnosis of rotator cuff disease and use of TSA is not coincidental. More specifically, there is likely an emerging trend among surgeons toward using the reverse TSA to manage rotator cuff tears in the RA population, rather than undertaking a rotator cuff repair that carries a high rate of failure. Going forward, there is a need to not only identify this trend more clearly but to also compare the outcomes between reverse TSA and rotator cuff repair in the management of rotator cuff tears in RA patients.

Continue to: In this study, we observed...

In this study, we observed that RA patients undergoing shoulder arthroplasty were significantly younger than non-RA patients undergoing shoulder arthroplasty. At first, this observation seems to counter recent literature suggesting that the age of patients with inflammatory arthropathy undergoing TJA is increasing over time;¹ however, looking more closely at the data, it becomes clearer that the mean age we report is actually a relative increase as compared with past clinical studies pertaining to RA patients undergoing shoulder arthroplasty (mean ages of 47 years,²³ 55 years,²⁴ 60 years,¹⁰ and 62 years²⁵). On the other hand, the continued existence of an age gap between RA and non-RA patients undergoing shoulder arthroplasty may be the result of several possible phenomena. First, this may reflect issues with patient access to and coverage of expensive biologic antirheumatic medication that would otherwise mitigate disease progression. For instance, the out-of-pocket expense for biologic medication through Medicaid and Medicare is substantial,²⁶ which has direct implications on over two-thirds of our RA cohort. Second, it may be skewed by the proportion of RA patients who have previously been or continue to be poorly managed, enabling disease progression to end-stage arthropathy at a younger age. Ultimately, further investigation is needed to determine the reasons for this continued age disparity.

In comparing RA and non-RA patients undergoing shoulder arthroplasty, we did not find a significant difference in the overall nor the individual rates of early adverse events. This finding appears to be unique, as similar studies pertaining to total knee arthroplasty (TKA) demonstrated a significantly higher incidence of postoperative pneumonia and bleeding requiring transfusion among RA patients as compared with non-RA patients.²⁷ In patients with RA being treated with biologic medication and undergoing shoulder arthroplasty, the frequent concern in the postoperative period is the integrity of the wound and the potential for infection.²⁸ In this study, we did not find a significant difference in the rate of early infection, and although the difference in the rate of early wound dehiscence approached significance, it did not meet the threshold of 0.05 (P = .09). This finding is in keeping with the aforementioned NIS study pertaining to TKA, and we believe that it likely reflects the short duration of follow-up for patients in both studies. Given the nature of the database we utilized, we were only privy to complications that arose during the inpatient hospital stay, and it is likely that the clear majority of patients who develop a postoperative infection or wound dehiscence do so in the postoperative setting following discharge. A second concern regarding postoperative wound complications is the management of biologic medication in the perioperative period, which we cannot determine using this database. Despite all these limitations specific to this database, a past systematic review of reverse TSA in RA patients found a low rate of deep infection after reverse TSA in RA patients (3.3%),¹⁷ which was not higher than that after shoulder arthroplasty performed in non-RA patients.

A final demonstration from this study is that the hospital length of stay was significantly longer for RA patients than non-RA patients undergoing shoulder arthroplasty; however, given that the difference was only 0.1 days, and there was no significant difference in hospitalization cost, we are inclined to believe that statistical significance may not translate into clinical significance in this scenario. Ultimately, we do believe that length of stay is an important consideration in the current healthcare system, and given our finding that shoulder arthroplasty in the RA patient is safe in the early postoperative period, that a prolonged postoperative hospitalization is not warranted on the sole basis of a patient’s history of RA.

As with all studies using data from a search of an administrative database, such as the NIS database, this study has limitations. First, this type of research is limited by the reliability of both diagnosis and procedural coding. Although the NIS database has demonstrated high reliability,³ it is still possible that events may have been miscoded. Second, the tracking period for adverse events is limited to the inpatient hospital stay, which may be too short to detect certain postoperative complications. As such, the rates we report are likely underestimates of the true incidence of these complications, but this is true for both the RA and non-RA populations. Third, the comparisons we draw between RA and non-RA patients are limited to the scope of the NIS database and the available data; as such, we could not draw comparisons between preoperative disease stage, intraoperative findings, and postoperative course following hospital discharge. Lastly, our data are limited to a distinct period between 2002 and 2011 and may not reflect current practice. Ultimately, our findings may underestimate current trends in shoulder arthroplasty utilization among RA patients, particularly for the reverse TSA.

CONCLUSION

In this study, we found that the utilization of shoulder arthroplasty in patients with RA increased significantly from 2002 to 2011, largely related to a rise in the utilization of TSA. Similarly, we observed a rise in the proportion of RA patients undergoing shoulder arthroplasty with a corresponding diagnosis of rotator cuff disease, and we believe the increased utilization of shoulder arthroplasty among RA patients resulted from management of both end-stage inflammatory arthropathy and rotator cuff disease. Although we did not find a significant difference between RA and non-RA patients in the rates of early adverse events and overall hospitalization costs following shoulder arthroplasty, length of stay was significantly longer among RA patients; however, the absolute difference does not appear to be clinically significant.

Take-Home Points

• Studies of AEs after orthopedic surgery commonly use composite AE outcomes.
• These types of outcomes treat AEs with different clinical significance similarly.
• This study created a single severity-weighted outcome that can be used to characterize the overall severity of a given patient’s postoperative course.
• Future studies may benefit from using this new severity-weighted outcome score.

Recently there has been an increase in the use of national databases for orthopedic surgery research.^1-4 Studies commonly compare rates of postoperative adverse events (AEs) across different demographic, comorbidity, and procedural characteristics.^5-23 Their conclusions often highlight different modifiable and/or nonmodifiable risk factors associated with the occurrence of postoperative events.

The several dozen AEs that have been investigated range from very severe (eg, death, myocardial infarction, coma) to less severe (eg, urinary tract infection [UTI], anemia requiring blood transfusion). A common approach for these studies is to consider many AEs together in the same analysis, asking a question such as, “What are risk factors for the occurrence of ‘adverse events’ after spine surgery?” Such studies test for associations with the occurrence of “any adverse event,” the occurrence of any “serious adverse event,” or similar composite outcomes. How common this type of study has become is indicated by the fact that in 2013 and 2014, at least 12 such studies were published in Clinical Orthopaedics and Related Research and the Journal of Bone and Joint Surgery,^5-14,21-23 and many more in other orthopedic journals.^15-20 However, there is a problem in using this type of composite outcome to perform such analyses: AEs with highly varying degrees of severity have identical impacts on the outcome variable, changing it from negative (“no adverse event”) to positive (“at least one adverse event”). As a result, the system may treat a very severe AE such as death and a very minor AE such as UTI similarly. Even in studies that use the slightly more specific composite outcome of “serious adverse events,” death and a nonlethal thromboembolic event would be treated similarly. Failure to differentiate these AEs in terms of their clinical significance detracts from the clinical applicability of conclusions drawn from studies using these types of composite AE outcomes.

In one of many examples that can be considered, a retrospective cohort study compared general and spinal anesthesia used in total knee arthroplasty.¹⁰ The rate of any AEs was higher with general anesthesia than with spinal anesthesia (12.34% vs 10.72%; P = .003). However, the only 2 specific AEs that had statistically significant differences were anemia requiring blood transfusion (6.07% vs 5.02%; P = .009) and superficial surgical-site infection (SSI; 0.92% vs 0.68%; P < .001). These 2 AEs are of relatively low severity; nevertheless, because these AEs are common, their differences constituted the majority of the difference in the rate of any AEs. In contrast, differences in the more severe AEs, such as death (0.11% vs 0.22%; P > .05), septic shock (0.14% vs 0.12%; P > .05), and myocardial infarction (0.20% vs 0.20%; P > .05), were small and not statistically significant. Had more weight been given to these more severe events, the outcome of the study likely would have been “no difference.”

To address this shortcoming in orthopedic research methodology, we created a severity-weighted outcome score that can be used to determine the overall “severity” of any given patient’s postoperative course. We also tested this novel outcome score for correlation with procedure type and patient characteristics using orthopedic patients from the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP). Our intention is for database investigators to be able to use this outcome score in place of the composite outcomes that are dominating this type of research.

Methods

Generation of Severity Weights

Our method is described generally as utility weighting, assigning value weights reflective of overall impact to differing outcome states.²⁴ Parallel methods have been used to generate the disability weights used to determine disability-adjusted life years for the Global Burden of Disease project²⁵ and many other areas of health, economic, and policy research.

All orthopedic faculty members at 2 geographically disparate, large US academic institutions were invited to participate in a severity-weighting exercise. Each surgeon who agreed to participate performed the exercise independently.

• STEP 1: Please reorder the AE cards by your perception of “severity” for a patient experiencing that event after an orthopedic procedure.
• STEP 2: Once your cards are in order, please determine how many postoperative occurrences of each event you would “trade” for 1 patient experiencing postoperative death. Place this number of occurrences in the box in the upper right corner of each card.
• NOTES: As you consider each AE:
• Please consider an “average” occurrence of that AE, but note that in no case does the AE result in perioperative death.
• Please consider only the “severity” for the patient. (Do not consider the extent to which the event may be related to surgical error.)
• Please consider that the numbers you assign are relative to each other. Hence, if you would trade 20 of “event A” for 1 death, and if you would trade 40 of “event B” for 1 death, the implication is that you would trade 20 of “event A” for 40 of “event B.”
• You may readjust the order of your cards at any point.

Participants’ responses were recorded. For each number provided by each participant, the inverse (reciprocal) was taken and multiplied by 100%. This new number was taken to be the percentage severity of death that the given participant considered the given AE to embody. For example, as a hypothetical on one end of the spectrum, if a participant reported 1 (he/she would trade 1 AE X for 1 death), then the severity would be 1/1 × 100% = 100% of death, a very severe AE. Conversely, if a participant reported a very large number like 100,000 (he/she would trade 100,000 AEs X for 1 death), then the severity would be 1/100,000 × 100% = 0.001% of death, a very minor AE. More commonly, a participant will report a number like 25, which would translate to 4% of death (1/25 × 100% = 4%). For each AE, weights were then averaged across participants to derive a mean severity weight to be used to generate a novel composite outcome score.

Definition of Novel Composite Outcome Score

The novel composite outcome score would be expressed as a percentage to be interpreted as percentage severity of death, which we termed severity-weighted outcome relative to death (SWORD). For each patient, SWORD was defined as no AE (0%) or postoperative death (100%), with other AEs assigned mean severity weights based on faculty members’ survey responses. A patient with multiple AEs would be assigned the weight for the more severe AE. This method was chosen over summing the AE weights because in many cases the AEs were thought to overlap; hence, summing would be inappropriate. For example, generally a deep SSI would result in a return to the operating room, and one would not want to double-count this AE. Similarly, it would not make sense for a patient who died of a complication to have a SWORD of >100%, which would be the summing result.

Application to ACS-NSQIP Patients

ACS-NSQIP is a surgical registry that prospectively identifies patients undergoing major surgery at any of >500 institutions nationwide.^26,27 Patients are characterized at baseline and are followed for AEs over the first 30 postoperative days.

First, mean SWORD was calculated and reported for patients undergoing each of the 8 procedures. Analysis of variance (ANOVA) was used to test for associations of mean SWORD with type of procedure both before and after multivariate adjustment for demographics (sex; age in years, <40, 40-49, 50-59, 60-69, 70-79, 80-89, ≥90) and comorbidities (diabetes, hypertension, chronic obstructive pulmonary disease, exertional dyspnea, end-stage renal disease, congestive heart failure).

Second, patients undergoing the procedure with the highest mean SWORD (hip fracture surgery) were examined in depth. Among only these patients, multivariate ANOVA was used to test for associations of mean SWORD with the same demographics and comorbidities.

All statistical tests were 2-tailed. Significance was set at α = 0.05 (P < .05).

All 23 institution A faculty members (100%) and 24 (89%) of the 27 institution B faculty members completed the exercise.

Results

Figure 1 shows mean severity weights and standard errors generated from faculty responses. Mean (standard error) severity weight for UTI was 0.23% (0.08%); blood transfusion, 0.28% (0.09%); pneumonia, 0.55% (0.15%); hospital readmission, 0.59% (0.23%); wound dehiscence, 0.64% (0.17%); deep vein thrombosis, 0.64% (0.19%); superficial SSI, 0.68% (0.23%); return to operating room, 0.91% (0.29%); progressive renal insufficiency, 0.93% (0.27%); graft/prosthesis/flap failure, 1.20% (0.34%); unplanned intubation, 1.38% (0.53%); deep SSI, 1.45% (0.38%); failure to wean from ventilator, 1.45% (0.48%); organ/space SSI, 1.76% (0.46%); sepsis without shock, 1.77% (0.42%); peripheral nerve injury, 1.83% (0.47%); pulmonary embolism, 2.99% (0.76%); acute renal failure, 3.95% (0.85%); myocardial infarction, 4.16% (0.98%); septic shock, 7.17% (1.36%); stroke, 8.73% (1.74%); cardiac arrest requiring cardiopulmonary resuscitation, 9.97% (2.46%); and coma, 15.14% (3.04%).

Among ACS-NSQIP patients, mean SWORD ranged from 0.2% (elective anterior cervical decompression and fusion) to 6.0% (hip fracture surgery) (Figure 2).

Discussion

The use of national databases in studies has become increasingly common in orthopedic surgery.^1-4

The academic orthopedic surgeons who participated in our severity-weighting exercise thought the various AEs have markedly different severities. The least severe AE (UTI) was considered 0.23% as severe as postoperative death, with other events spanning the range up to 15.14% as severe as death. This wide range of severities demonstrates the problem with composite outcomes that implicitly consider all AEs similarly severe. Use of these markedly disparate weights in the development of SWORD enables this outcome to be more clinically applicable than outcomes such as “any adverse events.”

SWORD was highly associated with procedure type both before and after adjustment for demographics and comorbidities. Among patients undergoing the highest SWORD procedure (hip fracture surgery), SWORD was also associated with age, sex, and 4 of 6 tested comorbidities. Together, our findings show how SWORD is intended to be used in studies: to identify demographic, comorbidity, and procedural risk factors for an adverse postoperative course. We propose that researchers use our weighted outcome as their primary outcome—it is more meaningful than the simpler composite outcomes commonly used.

Outside orthopedic surgery, a small series of studies has addressed severity weighting of postoperative AEs.^25,28-30 However, their approach was very different, as they were not designed to generate weights that could be transferred to future studies; rather, they simply compared severities of postoperative courses for patients within each individual study. In each study, a review of each original patient record was required, as the severity of each patient’s postoperative course was characterized according to the degree of any postoperative intervention—from no intervention to minor interventions such as placement of an intravenous catheter and major interventions such as endoscopic, radiologic, and surgical procedures. Only after the degree of intervention was defined could an outcome score be assigned to a given patient. However, databases do not depict the degree of intervention with nearly enough detail for this type of approach; they typically identify only occurrence or nonoccurrence of each event. Our work, which arose independently from this body of literature, enables an entirely different type of analysis. SWORD, which is not based on degree of intervention but on perceived severity of an “average” event, enables direct application of severity weights to large databases that store simple information on occurrence and nonoccurrence of specific AEs.

This study had several limitations. Most significantly, the generated severity weights were based on the surgeons’ subjective perceptions of severity, not on definitive assessments of the impacts of specific AEs on actual patients. We did not query the specialists who treat the complications or who present data on the costs and disabilities that may arise from these AEs. In addition, to develop our severity weighting scale, we queried faculty at only 2 institutions. A survey of surgeons throughout the United States would be more representative and would minimize selection bias. This is a potential research area. Another limitation is that scoring was subjective, based on surgeons’ perceptions of patients—in contrast to the Global Burden of Disease project, in which severity was based more objectively on epidemiologic data from >150 countries.

Orthopedic database research itself has often-noted limitations, including inability to sufficiently control for confounders, potential inaccuracies in data coding, limited follow-up, and lack of orthopedic-specific outcomes.^1-4,31-33 However, this research also has much to offer, has increased tremendously over the past several years, and is expected to continue to expand. Many of the limitations of database studies cannot be entirely reversed. In providing a system for weighting postoperative AEs, our study fills a methodologic void. Future studies in orthopedics may benefit from using the severity-weighted outcome score presented here. Other fields with growth in database research may consider using similar methods to create severity-weighting systems of their own.

Am J Orthop. 2017;46(4):E235-E243. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Studies of AEs after orthopedic surgery commonly use composite AE outcomes.
• These types of outcomes treat AEs with different clinical significance similarly.
• This study created a single severity-weighted outcome that can be used to characterize the overall severity of a given patient’s postoperative course.
• Future studies may benefit from using this new severity-weighted outcome score.

Recently there has been an increase in the use of national databases for orthopedic surgery research.^1-4 Studies commonly compare rates of postoperative adverse events (AEs) across different demographic, comorbidity, and procedural characteristics.^5-23 Their conclusions often highlight different modifiable and/or nonmodifiable risk factors associated with the occurrence of postoperative events.

The several dozen AEs that have been investigated range from very severe (eg, death, myocardial infarction, coma) to less severe (eg, urinary tract infection [UTI], anemia requiring blood transfusion). A common approach for these studies is to consider many AEs together in the same analysis, asking a question such as, “What are risk factors for the occurrence of ‘adverse events’ after spine surgery?” Such studies test for associations with the occurrence of “any adverse event,” the occurrence of any “serious adverse event,” or similar composite outcomes. How common this type of study has become is indicated by the fact that in 2013 and 2014, at least 12 such studies were published in Clinical Orthopaedics and Related Research and the Journal of Bone and Joint Surgery,^5-14,21-23 and many more in other orthopedic journals.^15-20 However, there is a problem in using this type of composite outcome to perform such analyses: AEs with highly varying degrees of severity have identical impacts on the outcome variable, changing it from negative (“no adverse event”) to positive (“at least one adverse event”). As a result, the system may treat a very severe AE such as death and a very minor AE such as UTI similarly. Even in studies that use the slightly more specific composite outcome of “serious adverse events,” death and a nonlethal thromboembolic event would be treated similarly. Failure to differentiate these AEs in terms of their clinical significance detracts from the clinical applicability of conclusions drawn from studies using these types of composite AE outcomes.

In one of many examples that can be considered, a retrospective cohort study compared general and spinal anesthesia used in total knee arthroplasty.¹⁰ The rate of any AEs was higher with general anesthesia than with spinal anesthesia (12.34% vs 10.72%; P = .003). However, the only 2 specific AEs that had statistically significant differences were anemia requiring blood transfusion (6.07% vs 5.02%; P = .009) and superficial surgical-site infection (SSI; 0.92% vs 0.68%; P < .001). These 2 AEs are of relatively low severity; nevertheless, because these AEs are common, their differences constituted the majority of the difference in the rate of any AEs. In contrast, differences in the more severe AEs, such as death (0.11% vs 0.22%; P > .05), septic shock (0.14% vs 0.12%; P > .05), and myocardial infarction (0.20% vs 0.20%; P > .05), were small and not statistically significant. Had more weight been given to these more severe events, the outcome of the study likely would have been “no difference.”

To address this shortcoming in orthopedic research methodology, we created a severity-weighted outcome score that can be used to determine the overall “severity” of any given patient’s postoperative course. We also tested this novel outcome score for correlation with procedure type and patient characteristics using orthopedic patients from the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP). Our intention is for database investigators to be able to use this outcome score in place of the composite outcomes that are dominating this type of research.

Methods

Generation of Severity Weights

Our method is described generally as utility weighting, assigning value weights reflective of overall impact to differing outcome states.²⁴ Parallel methods have been used to generate the disability weights used to determine disability-adjusted life years for the Global Burden of Disease project²⁵ and many other areas of health, economic, and policy research.

All orthopedic faculty members at 2 geographically disparate, large US academic institutions were invited to participate in a severity-weighting exercise. Each surgeon who agreed to participate performed the exercise independently.

• STEP 1: Please reorder the AE cards by your perception of “severity” for a patient experiencing that event after an orthopedic procedure.
• STEP 2: Once your cards are in order, please determine how many postoperative occurrences of each event you would “trade” for 1 patient experiencing postoperative death. Place this number of occurrences in the box in the upper right corner of each card.
• NOTES: As you consider each AE:
• Please consider an “average” occurrence of that AE, but note that in no case does the AE result in perioperative death.
• Please consider only the “severity” for the patient. (Do not consider the extent to which the event may be related to surgical error.)
• Please consider that the numbers you assign are relative to each other. Hence, if you would trade 20 of “event A” for 1 death, and if you would trade 40 of “event B” for 1 death, the implication is that you would trade 20 of “event A” for 40 of “event B.”
• You may readjust the order of your cards at any point.

Participants’ responses were recorded. For each number provided by each participant, the inverse (reciprocal) was taken and multiplied by 100%. This new number was taken to be the percentage severity of death that the given participant considered the given AE to embody. For example, as a hypothetical on one end of the spectrum, if a participant reported 1 (he/she would trade 1 AE X for 1 death), then the severity would be 1/1 × 100% = 100% of death, a very severe AE. Conversely, if a participant reported a very large number like 100,000 (he/she would trade 100,000 AEs X for 1 death), then the severity would be 1/100,000 × 100% = 0.001% of death, a very minor AE. More commonly, a participant will report a number like 25, which would translate to 4% of death (1/25 × 100% = 4%). For each AE, weights were then averaged across participants to derive a mean severity weight to be used to generate a novel composite outcome score.

Definition of Novel Composite Outcome Score

The novel composite outcome score would be expressed as a percentage to be interpreted as percentage severity of death, which we termed severity-weighted outcome relative to death (SWORD). For each patient, SWORD was defined as no AE (0%) or postoperative death (100%), with other AEs assigned mean severity weights based on faculty members’ survey responses. A patient with multiple AEs would be assigned the weight for the more severe AE. This method was chosen over summing the AE weights because in many cases the AEs were thought to overlap; hence, summing would be inappropriate. For example, generally a deep SSI would result in a return to the operating room, and one would not want to double-count this AE. Similarly, it would not make sense for a patient who died of a complication to have a SWORD of >100%, which would be the summing result.

Application to ACS-NSQIP Patients

ACS-NSQIP is a surgical registry that prospectively identifies patients undergoing major surgery at any of >500 institutions nationwide.^26,27 Patients are characterized at baseline and are followed for AEs over the first 30 postoperative days.

First, mean SWORD was calculated and reported for patients undergoing each of the 8 procedures. Analysis of variance (ANOVA) was used to test for associations of mean SWORD with type of procedure both before and after multivariate adjustment for demographics (sex; age in years, <40, 40-49, 50-59, 60-69, 70-79, 80-89, ≥90) and comorbidities (diabetes, hypertension, chronic obstructive pulmonary disease, exertional dyspnea, end-stage renal disease, congestive heart failure).

Second, patients undergoing the procedure with the highest mean SWORD (hip fracture surgery) were examined in depth. Among only these patients, multivariate ANOVA was used to test for associations of mean SWORD with the same demographics and comorbidities.

All statistical tests were 2-tailed. Significance was set at α = 0.05 (P < .05).

All 23 institution A faculty members (100%) and 24 (89%) of the 27 institution B faculty members completed the exercise.

Results

Figure 1 shows mean severity weights and standard errors generated from faculty responses. Mean (standard error) severity weight for UTI was 0.23% (0.08%); blood transfusion, 0.28% (0.09%); pneumonia, 0.55% (0.15%); hospital readmission, 0.59% (0.23%); wound dehiscence, 0.64% (0.17%); deep vein thrombosis, 0.64% (0.19%); superficial SSI, 0.68% (0.23%); return to operating room, 0.91% (0.29%); progressive renal insufficiency, 0.93% (0.27%); graft/prosthesis/flap failure, 1.20% (0.34%); unplanned intubation, 1.38% (0.53%); deep SSI, 1.45% (0.38%); failure to wean from ventilator, 1.45% (0.48%); organ/space SSI, 1.76% (0.46%); sepsis without shock, 1.77% (0.42%); peripheral nerve injury, 1.83% (0.47%); pulmonary embolism, 2.99% (0.76%); acute renal failure, 3.95% (0.85%); myocardial infarction, 4.16% (0.98%); septic shock, 7.17% (1.36%); stroke, 8.73% (1.74%); cardiac arrest requiring cardiopulmonary resuscitation, 9.97% (2.46%); and coma, 15.14% (3.04%).

Among ACS-NSQIP patients, mean SWORD ranged from 0.2% (elective anterior cervical decompression and fusion) to 6.0% (hip fracture surgery) (Figure 2).

Discussion

The use of national databases in studies has become increasingly common in orthopedic surgery.^1-4

The academic orthopedic surgeons who participated in our severity-weighting exercise thought the various AEs have markedly different severities. The least severe AE (UTI) was considered 0.23% as severe as postoperative death, with other events spanning the range up to 15.14% as severe as death. This wide range of severities demonstrates the problem with composite outcomes that implicitly consider all AEs similarly severe. Use of these markedly disparate weights in the development of SWORD enables this outcome to be more clinically applicable than outcomes such as “any adverse events.”

SWORD was highly associated with procedure type both before and after adjustment for demographics and comorbidities. Among patients undergoing the highest SWORD procedure (hip fracture surgery), SWORD was also associated with age, sex, and 4 of 6 tested comorbidities. Together, our findings show how SWORD is intended to be used in studies: to identify demographic, comorbidity, and procedural risk factors for an adverse postoperative course. We propose that researchers use our weighted outcome as their primary outcome—it is more meaningful than the simpler composite outcomes commonly used.

Outside orthopedic surgery, a small series of studies has addressed severity weighting of postoperative AEs.^25,28-30 However, their approach was very different, as they were not designed to generate weights that could be transferred to future studies; rather, they simply compared severities of postoperative courses for patients within each individual study. In each study, a review of each original patient record was required, as the severity of each patient’s postoperative course was characterized according to the degree of any postoperative intervention—from no intervention to minor interventions such as placement of an intravenous catheter and major interventions such as endoscopic, radiologic, and surgical procedures. Only after the degree of intervention was defined could an outcome score be assigned to a given patient. However, databases do not depict the degree of intervention with nearly enough detail for this type of approach; they typically identify only occurrence or nonoccurrence of each event. Our work, which arose independently from this body of literature, enables an entirely different type of analysis. SWORD, which is not based on degree of intervention but on perceived severity of an “average” event, enables direct application of severity weights to large databases that store simple information on occurrence and nonoccurrence of specific AEs.

This study had several limitations. Most significantly, the generated severity weights were based on the surgeons’ subjective perceptions of severity, not on definitive assessments of the impacts of specific AEs on actual patients. We did not query the specialists who treat the complications or who present data on the costs and disabilities that may arise from these AEs. In addition, to develop our severity weighting scale, we queried faculty at only 2 institutions. A survey of surgeons throughout the United States would be more representative and would minimize selection bias. This is a potential research area. Another limitation is that scoring was subjective, based on surgeons’ perceptions of patients—in contrast to the Global Burden of Disease project, in which severity was based more objectively on epidemiologic data from >150 countries.

Orthopedic database research itself has often-noted limitations, including inability to sufficiently control for confounders, potential inaccuracies in data coding, limited follow-up, and lack of orthopedic-specific outcomes.^1-4,31-33 However, this research also has much to offer, has increased tremendously over the past several years, and is expected to continue to expand. Many of the limitations of database studies cannot be entirely reversed. In providing a system for weighting postoperative AEs, our study fills a methodologic void. Future studies in orthopedics may benefit from using the severity-weighted outcome score presented here. Other fields with growth in database research may consider using similar methods to create severity-weighting systems of their own.

Am J Orthop. 2017;46(4):E235-E243. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Studies of AEs after orthopedic surgery commonly use composite AE outcomes.
• These types of outcomes treat AEs with different clinical significance similarly.
• This study created a single severity-weighted outcome that can be used to characterize the overall severity of a given patient’s postoperative course.
• Future studies may benefit from using this new severity-weighted outcome score.

Recently there has been an increase in the use of national databases for orthopedic surgery research.^1-4 Studies commonly compare rates of postoperative adverse events (AEs) across different demographic, comorbidity, and procedural characteristics.^5-23 Their conclusions often highlight different modifiable and/or nonmodifiable risk factors associated with the occurrence of postoperative events.

The several dozen AEs that have been investigated range from very severe (eg, death, myocardial infarction, coma) to less severe (eg, urinary tract infection [UTI], anemia requiring blood transfusion). A common approach for these studies is to consider many AEs together in the same analysis, asking a question such as, “What are risk factors for the occurrence of ‘adverse events’ after spine surgery?” Such studies test for associations with the occurrence of “any adverse event,” the occurrence of any “serious adverse event,” or similar composite outcomes. How common this type of study has become is indicated by the fact that in 2013 and 2014, at least 12 such studies were published in Clinical Orthopaedics and Related Research and the Journal of Bone and Joint Surgery,^5-14,21-23 and many more in other orthopedic journals.^15-20 However, there is a problem in using this type of composite outcome to perform such analyses: AEs with highly varying degrees of severity have identical impacts on the outcome variable, changing it from negative (“no adverse event”) to positive (“at least one adverse event”). As a result, the system may treat a very severe AE such as death and a very minor AE such as UTI similarly. Even in studies that use the slightly more specific composite outcome of “serious adverse events,” death and a nonlethal thromboembolic event would be treated similarly. Failure to differentiate these AEs in terms of their clinical significance detracts from the clinical applicability of conclusions drawn from studies using these types of composite AE outcomes.

In one of many examples that can be considered, a retrospective cohort study compared general and spinal anesthesia used in total knee arthroplasty.¹⁰ The rate of any AEs was higher with general anesthesia than with spinal anesthesia (12.34% vs 10.72%; P = .003). However, the only 2 specific AEs that had statistically significant differences were anemia requiring blood transfusion (6.07% vs 5.02%; P = .009) and superficial surgical-site infection (SSI; 0.92% vs 0.68%; P < .001). These 2 AEs are of relatively low severity; nevertheless, because these AEs are common, their differences constituted the majority of the difference in the rate of any AEs. In contrast, differences in the more severe AEs, such as death (0.11% vs 0.22%; P > .05), septic shock (0.14% vs 0.12%; P > .05), and myocardial infarction (0.20% vs 0.20%; P > .05), were small and not statistically significant. Had more weight been given to these more severe events, the outcome of the study likely would have been “no difference.”

To address this shortcoming in orthopedic research methodology, we created a severity-weighted outcome score that can be used to determine the overall “severity” of any given patient’s postoperative course. We also tested this novel outcome score for correlation with procedure type and patient characteristics using orthopedic patients from the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP). Our intention is for database investigators to be able to use this outcome score in place of the composite outcomes that are dominating this type of research.

Methods

Generation of Severity Weights

Our method is described generally as utility weighting, assigning value weights reflective of overall impact to differing outcome states.²⁴ Parallel methods have been used to generate the disability weights used to determine disability-adjusted life years for the Global Burden of Disease project²⁵ and many other areas of health, economic, and policy research.

All orthopedic faculty members at 2 geographically disparate, large US academic institutions were invited to participate in a severity-weighting exercise. Each surgeon who agreed to participate performed the exercise independently.

• STEP 1: Please reorder the AE cards by your perception of “severity” for a patient experiencing that event after an orthopedic procedure.
• STEP 2: Once your cards are in order, please determine how many postoperative occurrences of each event you would “trade” for 1 patient experiencing postoperative death. Place this number of occurrences in the box in the upper right corner of each card.
• NOTES: As you consider each AE:
• Please consider an “average” occurrence of that AE, but note that in no case does the AE result in perioperative death.
• Please consider only the “severity” for the patient. (Do not consider the extent to which the event may be related to surgical error.)
• Please consider that the numbers you assign are relative to each other. Hence, if you would trade 20 of “event A” for 1 death, and if you would trade 40 of “event B” for 1 death, the implication is that you would trade 20 of “event A” for 40 of “event B.”
• You may readjust the order of your cards at any point.

Participants’ responses were recorded. For each number provided by each participant, the inverse (reciprocal) was taken and multiplied by 100%. This new number was taken to be the percentage severity of death that the given participant considered the given AE to embody. For example, as a hypothetical on one end of the spectrum, if a participant reported 1 (he/she would trade 1 AE X for 1 death), then the severity would be 1/1 × 100% = 100% of death, a very severe AE. Conversely, if a participant reported a very large number like 100,000 (he/she would trade 100,000 AEs X for 1 death), then the severity would be 1/100,000 × 100% = 0.001% of death, a very minor AE. More commonly, a participant will report a number like 25, which would translate to 4% of death (1/25 × 100% = 4%). For each AE, weights were then averaged across participants to derive a mean severity weight to be used to generate a novel composite outcome score.

Definition of Novel Composite Outcome Score

The novel composite outcome score would be expressed as a percentage to be interpreted as percentage severity of death, which we termed severity-weighted outcome relative to death (SWORD). For each patient, SWORD was defined as no AE (0%) or postoperative death (100%), with other AEs assigned mean severity weights based on faculty members’ survey responses. A patient with multiple AEs would be assigned the weight for the more severe AE. This method was chosen over summing the AE weights because in many cases the AEs were thought to overlap; hence, summing would be inappropriate. For example, generally a deep SSI would result in a return to the operating room, and one would not want to double-count this AE. Similarly, it would not make sense for a patient who died of a complication to have a SWORD of >100%, which would be the summing result.

Application to ACS-NSQIP Patients

ACS-NSQIP is a surgical registry that prospectively identifies patients undergoing major surgery at any of >500 institutions nationwide.^26,27 Patients are characterized at baseline and are followed for AEs over the first 30 postoperative days.

First, mean SWORD was calculated and reported for patients undergoing each of the 8 procedures. Analysis of variance (ANOVA) was used to test for associations of mean SWORD with type of procedure both before and after multivariate adjustment for demographics (sex; age in years, <40, 40-49, 50-59, 60-69, 70-79, 80-89, ≥90) and comorbidities (diabetes, hypertension, chronic obstructive pulmonary disease, exertional dyspnea, end-stage renal disease, congestive heart failure).

Second, patients undergoing the procedure with the highest mean SWORD (hip fracture surgery) were examined in depth. Among only these patients, multivariate ANOVA was used to test for associations of mean SWORD with the same demographics and comorbidities.

All statistical tests were 2-tailed. Significance was set at α = 0.05 (P < .05).

All 23 institution A faculty members (100%) and 24 (89%) of the 27 institution B faculty members completed the exercise.

Results

Figure 1 shows mean severity weights and standard errors generated from faculty responses. Mean (standard error) severity weight for UTI was 0.23% (0.08%); blood transfusion, 0.28% (0.09%); pneumonia, 0.55% (0.15%); hospital readmission, 0.59% (0.23%); wound dehiscence, 0.64% (0.17%); deep vein thrombosis, 0.64% (0.19%); superficial SSI, 0.68% (0.23%); return to operating room, 0.91% (0.29%); progressive renal insufficiency, 0.93% (0.27%); graft/prosthesis/flap failure, 1.20% (0.34%); unplanned intubation, 1.38% (0.53%); deep SSI, 1.45% (0.38%); failure to wean from ventilator, 1.45% (0.48%); organ/space SSI, 1.76% (0.46%); sepsis without shock, 1.77% (0.42%); peripheral nerve injury, 1.83% (0.47%); pulmonary embolism, 2.99% (0.76%); acute renal failure, 3.95% (0.85%); myocardial infarction, 4.16% (0.98%); septic shock, 7.17% (1.36%); stroke, 8.73% (1.74%); cardiac arrest requiring cardiopulmonary resuscitation, 9.97% (2.46%); and coma, 15.14% (3.04%).

Among ACS-NSQIP patients, mean SWORD ranged from 0.2% (elective anterior cervical decompression and fusion) to 6.0% (hip fracture surgery) (Figure 2).

Discussion

The use of national databases in studies has become increasingly common in orthopedic surgery.^1-4

The academic orthopedic surgeons who participated in our severity-weighting exercise thought the various AEs have markedly different severities. The least severe AE (UTI) was considered 0.23% as severe as postoperative death, with other events spanning the range up to 15.14% as severe as death. This wide range of severities demonstrates the problem with composite outcomes that implicitly consider all AEs similarly severe. Use of these markedly disparate weights in the development of SWORD enables this outcome to be more clinically applicable than outcomes such as “any adverse events.”

SWORD was highly associated with procedure type both before and after adjustment for demographics and comorbidities. Among patients undergoing the highest SWORD procedure (hip fracture surgery), SWORD was also associated with age, sex, and 4 of 6 tested comorbidities. Together, our findings show how SWORD is intended to be used in studies: to identify demographic, comorbidity, and procedural risk factors for an adverse postoperative course. We propose that researchers use our weighted outcome as their primary outcome—it is more meaningful than the simpler composite outcomes commonly used.

Outside orthopedic surgery, a small series of studies has addressed severity weighting of postoperative AEs.^25,28-30 However, their approach was very different, as they were not designed to generate weights that could be transferred to future studies; rather, they simply compared severities of postoperative courses for patients within each individual study. In each study, a review of each original patient record was required, as the severity of each patient’s postoperative course was characterized according to the degree of any postoperative intervention—from no intervention to minor interventions such as placement of an intravenous catheter and major interventions such as endoscopic, radiologic, and surgical procedures. Only after the degree of intervention was defined could an outcome score be assigned to a given patient. However, databases do not depict the degree of intervention with nearly enough detail for this type of approach; they typically identify only occurrence or nonoccurrence of each event. Our work, which arose independently from this body of literature, enables an entirely different type of analysis. SWORD, which is not based on degree of intervention but on perceived severity of an “average” event, enables direct application of severity weights to large databases that store simple information on occurrence and nonoccurrence of specific AEs.

This study had several limitations. Most significantly, the generated severity weights were based on the surgeons’ subjective perceptions of severity, not on definitive assessments of the impacts of specific AEs on actual patients. We did not query the specialists who treat the complications or who present data on the costs and disabilities that may arise from these AEs. In addition, to develop our severity weighting scale, we queried faculty at only 2 institutions. A survey of surgeons throughout the United States would be more representative and would minimize selection bias. This is a potential research area. Another limitation is that scoring was subjective, based on surgeons’ perceptions of patients—in contrast to the Global Burden of Disease project, in which severity was based more objectively on epidemiologic data from >150 countries.

Orthopedic database research itself has often-noted limitations, including inability to sufficiently control for confounders, potential inaccuracies in data coding, limited follow-up, and lack of orthopedic-specific outcomes.^1-4,31-33 However, this research also has much to offer, has increased tremendously over the past several years, and is expected to continue to expand. Many of the limitations of database studies cannot be entirely reversed. In providing a system for weighting postoperative AEs, our study fills a methodologic void. Future studies in orthopedics may benefit from using the severity-weighted outcome score presented here. Other fields with growth in database research may consider using similar methods to create severity-weighting systems of their own.

Am J Orthop. 2017;46(4):E235-E243. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• The prevalence of preoperative pneumonia is 1.2% among hip fracture patients aged >65 years.
• Preoperative pneumonia is an independent risk factor for mortality and adverse events including renal failure, prolonged ventilator dependence, and prolonged altered mental status after geriatric hip fracture surgery.
• Underweight BMI (<18.5 kg/m²) was associated with higher mortality within 30 days among hip fracture patients admitted with pneumonia.
• The mortality rate normalized to that of patients without pneumonia within 2 weeks of hip fracture surgery.
• Time from admission to surgery was not associated with adverse events or mortality among hip fracture patients admitted with pneumonia.

Preoperative pneumonia remains relatively unexplored as a risk factor for adverse outcomes in geriatric hip fracture surgery. Dated studies report a 0.3% to 3.2% prevalence of “recent pneumonia” in patients presenting with hip fracture but provide neither a definition of pneumonia based on clinical criteria nor a subset analysis of outcomes in the pneumonia group.^1-3 Although active pneumonia has been identified as a preoperative optimization target in the management guidelines for geriatric hip fracture,⁴ we are unaware of any studies that have reported on differences in demographics, comorbidities, delay to surgery, or adverse outcomes between hip fracture patients with and without preoperative pneumonia.

This paucity of information on the effect of preoperative pneumonia in the hip fracture population may be related to low prevalence of preoperative pneumonia and a cadre of variable definitions, which limit identification of a cohort of patients with preoperative pneumonia large enough from which to draw meaningful results. Database studies, especially those using surgical registries rather than administrative or reimbursement data, offer particular advantages for investigation of such rare clinical entities.⁵Medical care of patients with pneumonia alone is known to be facilitated by assessments of mortality risk from clinical and laboratory data. The modified British Thoracic Society rule/CURB-65 (confusion, urea, respiratory rate, blood pressure) score is strongly predictive of mortality in hospitalized adults with pneumonia (odds ratio [OR], 4.59; 95% confidence interval [CI], 1.42-14.85; P = .011) and may guide antibiotic therapy, laboratory investigations, and the decision to intubate in a patient with pneumonia.^6-8 This score is predictive of adverse events (AEs), hospital length of stay, and use of intensive care services.^6,7,9-13 We hypothesized that preoperative clinical indicators assessed by pneumonia severity scores as well as patient demographics and baseline comorbidities may also have prognostic value for risk of AEs in a cohort of geriatric hip fracture surgery patients with preoperative pneumonia.

In this article, we first describe the prevalence of preoperative pneumonia in geriatric hip fracture surgery patients as well as demographic and operative differences between patients with and without the disease. We then ask 3 questions: Is preoperative pneumonia an independent risk factor for mortality and adverse outcomes in geriatric hip fracture surgery? Is there a postoperative interval during which the unadjusted mortality rate is higher among patients with preoperative pneumonia? In patients with preoperative pneumonia, what are the predictors of morbidity and mortality?

Methods

Yale University’s Human Investigations Committee approved this retrospective cohort study, which used the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database for the period 2005 to 2012. ACS-NSQIP is a prospective, multi-institutional outcomes program that collects data on preoperative comorbidities, intraoperative variables, and 30-day postoperative outcomes for patients undergoing surgical procedures in inpatient and outpatient settings.¹⁴

Unlike administrative databases, which are based on reimbursement data, ACS-NSQIP data are collected by trained surgical clinical reviewers for the purposes of quality improvement and clinical research, and data quality is ensured with routine auditing.¹⁵ The program has gained a high degree of respect as a powerful and valid data source in both general¹⁶ and orthopedic¹⁷ surgery literature. The database offers a particular advantage with respect to the study of preoperative pneumonia: Only patients with new or recently diagnosed pneumonia on antibiotic therapy who meet strict criteria for characteristic findings on chest radiography, clinical signs and symptoms of respiratory illness, and positive cultures are coded as having actively treated pneumonia at time of surgery.¹⁵

To identify hip fracture patients over the age of 65 years who underwent operative fixation of a hip fracture, we used Current Procedural Terminology (CPT) hip fracture codes, including 27235 (percutaneous screw fixation), 27236 or 27244 (plate-and-screw fixation), and 27245 (intramedullary device), as well as 27125 (hemiarthroplasty) and 27130 (arthroplasty) for patients with a postoperative International Classification of Disease, Ninth Revision (ICD-9) diagnosis code (820.x, 820.2x, or 820.8) consistent with acute hip fracture.^18,19 Procedure type, anesthesia type, and delay from admission to surgery were captured for all procedures.

Preoperative demographics included age, sex, transfer origin, functional status, and body mass index (BMI) category. Binary comorbidities were classified as preoperative anemia (hematocrit, <0.41 for men, <0.36 for women), confusion, dyspnea at rest, uremia (blood urea nitrogen, >6.8 mmol/L), history of cardiovascular disease (congestive heart failure, myocardial infarction, percutaneous coronary intervention, angina pectoris, medically treated hypertension, peripheral vascular disease, or resting claudication), chronic obstructive pulmonary disease, diabetes, renal disease (renal failure or dialysis), and cigarette use in preceding 12 months.^20,21 Although preoperative hypotension and respiratory rate are often considered in patients with pneumonia, these variables were not available from the ACS-NSQIP data.^6,22Pearson χ² test for categorical variables was used to compare baseline demographics and operative characteristics between patients with and without pneumonia, and Student t test was used to compare intervals from hospital admission to hip fracture surgery, surgery start to surgery stop, and surgery to discharge between patients with and without preoperative pneumonia.

Binary outcome measures were compared between patients with and without preoperative pneumonia. “Any AE” included any serious AE (SAE) or any minor AE. SAEs included death, acute renal failure, ventilator use >48 hours, unplanned intubation, septic shock, sepsis, return to operating room, coma >24 hours, cardiac arrest requiring cardiopulmonary resuscitation, myocardial infarction, thromboembolic event (deep vein thrombosis or pulmonary embolism), and stroke/cerebrovascular accident. Minor AEs included progressive renal insufficiency, urinary tract infection, organ/space infection, superficial surgical-site infection, deep surgical-site infection, and wound dehiscence. Other binary outcome measures included discharge destination and unplanned readmission within 30 days after hip fracture surgery.²³Poisson regression with robust error variance as described by Zou²⁴ was used to compare the rates of any, minor, and individual AEs, and any SAEs, between patients with and without pneumonia. Multivariate analysis accounted for the baseline variables in Table 1. AEs that occurred more than once in each group were included in the analyses.

Results

Of the 7128 geriatric hip fracture patients in this study, 82 (1.2%) had active pneumonia at time of surgery (Table 1). Age, BMI, preoperative uremia, history of cardiovascular disease, diabetes, renal disease, and smoking were similar between groups. In addition, there was no difference in anesthesia type or fixation procedure between the pneumonia and no-pneumonia groups. Patients with preoperative pneumonia differed significantly with respect to sex, transfer from facility, preoperative functional dependence, anemia, confusion, dyspnea at rest, and history of chronic obstructive pulmonary disease (Table 1).

Interval from admission to surgery was longer (P < .001) for geriatric hip fracture patients with preoperative pneumonia (mean, 6.8 days; 95% CI, 2.5-11.1 days) than for those without pneumonia (mean, 1.5 days; CI, 1.4-1.5 days). There was no difference (P = .124) in operative time between the pneumonia group (mean, 72.8 min; CI, 64.0-81.5 min) and the no-pneumonia group (mean, 66.1 min; CI, 61.2-67.0 min). Interval from surgery to discharge was longer (P < .001) for patients with preoperative pneumonia (mean, 10.1 days; CI, 6.9-13.4 days) than for those without pneumonia (mean, 6.3 days; CI, 6.1-6.4 days).

Adverse outcomes of geriatric hip fracture surgery are listed in Table 2. In the multivariate analysis, preoperative pneumonia was significantly associated with any AE (relative risk [RR]) = 1.44) and any SAE (RR = 1.79).

Discussion

In the general US population, pneumonia accounts for 1.4% of deaths in people 65 years to 74 years old, 2.1% in people 75 years to 84 years, and 3.1% in people 85 years or older. In total, 3.4% of hospital inpatient deaths are attributed to pneumonia.²⁵ In hospitalized general orthopedic surgical patients as well as hip fracture patients, pneumonia is strongly associated with increased mortality.^26,27

We identified a preoperative pneumonia prevalence of 1.2%, which is comparable to the rates reported in the literature (0.3%-3.2%).^1-3 To our knowledge, our study represents the largest series of patients with preoperative pneumonia at time of hip fracture repair, and the first to independently associate preoperative pneumonia with increased incidence of AEs, including death.

This study had its limitations. First, the ACS-NSQIP morbidity and mortality data, which are limited to the first 30 postoperative days, may be skewed because AEs that occurred after that interval are not captured. Second, coding of pneumonia in ACS-NSQIP does not convey specific information about the disease and its severity—infectious organism(s) responsible; acquisition setting (healthcare or community); treatment given, including antibiotic(s) selection, steroid use, dosing, and duration; and measures of treatment efficacy—limiting interpretation of the difference in delay to surgery. We cannot say whether the longer interval in patients with pneumonia reflects medical optimization, or whether the delay itself or any interventions during that time positively or negatively affected outcomes. In addition, despite using a large national database, we obtained a relatively small sample of patients (82) who had pneumonia before surgical hip fracture repair.

Multivariate analysis controlling for baseline demographics and comorbidities revealed that multiple SAEs were independently associated with preoperative pneumonia (overall SAE, RR = 1.79). Postoperative use of ventilator support for longer than 48 hours (RR = 6.48) and coma longer than 24 hours (RR = 7.31) are expected given the severity of pulmonary compromise in the study cohort.^28,29 Acute renal failure (RR = 14.61) can occur in both hip fracture patients and community-acquired pneumonia patients and may be a multifactorial complication of the pulmonary infection, of the anesthesia, or of the surgical intervention in this cohort.^30-32Unadjusted mortality in hip fracture takes months to a year to normalize to that of age-matched controls.^32-34 In our series, the unadjusted death rate in the pneumonia cohort (Figure) was transiently elevated during the first weeks after surgery but then drew nearer the rate in the nondiseased hip fracture cohort by the end of the first month. Early death in the pneumonia group likely was multifactorial, potentially influenced by the increased burden of comorbidities in the pneumonia group at baseline, and the longer delay to surgery,^35-38 as well as by the natural history of treated pneumonia in hospital patients, who, compared with age-matched hospitalized controls, also exhibit higher mortality during only the first 2 to 4 months of hospitalization for pneumonia.³⁹ We regret that quality improvement strategies in the treatment of geriatric hip fracture surgery with pneumonia cannot be extrapolated from these results.

Similarly, the utility of BMI <18.5 kg/m² as an actionable preoperative finding cannot be assessed from these results. However, we propose that underweight geriatric hip fracture patients with pneumonia may benefit from more aggressive preoperative optimization that does not delay surgery. Higher acuity of postoperative care, including more intensive nursing care and early coordination of care with respiratory therapists and medical comanagement teams, may also be beneficial.

Anesthesia type did not differ between patients with and without preoperative pneumonia and was not associated with AEs in patients with preoperative pneumonia. Consistent with our findings, multiple studies have reported no significant differences in short-term outcomes of hip fracture repair between general and spinal anesthesia, though no other study has compared the benefits of general and spinal anesthesia for patients with preoperative pneumonia.^40-44 Although spinal anesthesia (relative to general anesthesia) has been reported to have benefits in hip and knee arthroplasty, these benefits appear not to translate to hip fracture repair.^45-50 The results of the present study suggest that general and spinal anesthesia may be equivalent in terms of risk for the geriatric hip fracture patient with preoperative pneumonia.^43,44Our attempt to evaluate the CURB-65 pneumonia severity score as a prognosticator of AEs was thwarted by the absence of required variables in the ACS-NSQIP dataset (confusion, uremia, dyspnea, and age were available; hypotension and blood pressure were not). In our analysis, we did include, individually, variables previously found to predict AEs in the medical pneumonia population (confusion, uremia, dyspnea at rest, anemia).^9-11,32 However, these clinical findings are nonspecific in hip fracture patients, who may become anemic, confused, dyspneic, or uremic from a multitude of factors related to their injury and unrelated to pneumonia, including but not limited to hemorrhage, muscle damage, renal injury, and pulmonary embolism. It is not surprising that confusion, uremia, dyspnea at rest, and anemia were not individually predictive of AEs or death within 30 days after surgery in the cohort of geriatric hip fracture patients with pneumonia.

There is no literature that argues for or against delaying hip fracture surgery in geriatric hip fracture patients with pneumonia. The surgical delay observed in this population is ostensibly related to medical optimization of the pneumonia and/or underlying comorbidities. However, we did not find a morbidity or mortality detriment or benefit in delaying surgery by 1 to 4 days in this population. Delay of surgery is a poor covariate, given extensive confounding by medical management and preoperative optimizing of comorbid conditions (reflected in our independent variable and covariates) as well as institutional and surgeon variations in policy and behavior and other unaccounted influences. Although some authors have found no difference in mortality or major AEs between hip fracture patients who had a surgical delay and those who did not,^31,51-53 other series and meta-analyses have suggested a mortality detriment in a surgical delay of more than 2 days^36,54 or 4 days⁵⁵ from admission. Given our data, we cannot recommend against immediate hip fracture repair in the subpopulation of geriatric hip fracture patients with pneumonia.

Our study findings suggest that preoperative pneumonia is a rare independent risk factor for AEs after hip fracture surgery in geriatric patients. Underweight BMI is predictive of death in geriatric hip fracture surgery patients who present with pneumonia, whereas early surgical repair appears not to be associated with adverse outcomes. Further investigation is warranted to determine if such patients benefit from specific preoperative and postoperative strategies for optimizing medical and surgical care based on these findings.

Am J Orthop. 2017;46(3):E177-E185. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• The prevalence of preoperative pneumonia is 1.2% among hip fracture patients aged >65 years.
• Preoperative pneumonia is an independent risk factor for mortality and adverse events including renal failure, prolonged ventilator dependence, and prolonged altered mental status after geriatric hip fracture surgery.
• Underweight BMI (<18.5 kg/m²) was associated with higher mortality within 30 days among hip fracture patients admitted with pneumonia.
• The mortality rate normalized to that of patients without pneumonia within 2 weeks of hip fracture surgery.
• Time from admission to surgery was not associated with adverse events or mortality among hip fracture patients admitted with pneumonia.

Preoperative pneumonia remains relatively unexplored as a risk factor for adverse outcomes in geriatric hip fracture surgery. Dated studies report a 0.3% to 3.2% prevalence of “recent pneumonia” in patients presenting with hip fracture but provide neither a definition of pneumonia based on clinical criteria nor a subset analysis of outcomes in the pneumonia group.^1-3 Although active pneumonia has been identified as a preoperative optimization target in the management guidelines for geriatric hip fracture,⁴ we are unaware of any studies that have reported on differences in demographics, comorbidities, delay to surgery, or adverse outcomes between hip fracture patients with and without preoperative pneumonia.

This paucity of information on the effect of preoperative pneumonia in the hip fracture population may be related to low prevalence of preoperative pneumonia and a cadre of variable definitions, which limit identification of a cohort of patients with preoperative pneumonia large enough from which to draw meaningful results. Database studies, especially those using surgical registries rather than administrative or reimbursement data, offer particular advantages for investigation of such rare clinical entities.⁵Medical care of patients with pneumonia alone is known to be facilitated by assessments of mortality risk from clinical and laboratory data. The modified British Thoracic Society rule/CURB-65 (confusion, urea, respiratory rate, blood pressure) score is strongly predictive of mortality in hospitalized adults with pneumonia (odds ratio [OR], 4.59; 95% confidence interval [CI], 1.42-14.85; P = .011) and may guide antibiotic therapy, laboratory investigations, and the decision to intubate in a patient with pneumonia.^6-8 This score is predictive of adverse events (AEs), hospital length of stay, and use of intensive care services.^6,7,9-13 We hypothesized that preoperative clinical indicators assessed by pneumonia severity scores as well as patient demographics and baseline comorbidities may also have prognostic value for risk of AEs in a cohort of geriatric hip fracture surgery patients with preoperative pneumonia.

In this article, we first describe the prevalence of preoperative pneumonia in geriatric hip fracture surgery patients as well as demographic and operative differences between patients with and without the disease. We then ask 3 questions: Is preoperative pneumonia an independent risk factor for mortality and adverse outcomes in geriatric hip fracture surgery? Is there a postoperative interval during which the unadjusted mortality rate is higher among patients with preoperative pneumonia? In patients with preoperative pneumonia, what are the predictors of morbidity and mortality?

Methods

Yale University’s Human Investigations Committee approved this retrospective cohort study, which used the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database for the period 2005 to 2012. ACS-NSQIP is a prospective, multi-institutional outcomes program that collects data on preoperative comorbidities, intraoperative variables, and 30-day postoperative outcomes for patients undergoing surgical procedures in inpatient and outpatient settings.¹⁴

Unlike administrative databases, which are based on reimbursement data, ACS-NSQIP data are collected by trained surgical clinical reviewers for the purposes of quality improvement and clinical research, and data quality is ensured with routine auditing.¹⁵ The program has gained a high degree of respect as a powerful and valid data source in both general¹⁶ and orthopedic¹⁷ surgery literature. The database offers a particular advantage with respect to the study of preoperative pneumonia: Only patients with new or recently diagnosed pneumonia on antibiotic therapy who meet strict criteria for characteristic findings on chest radiography, clinical signs and symptoms of respiratory illness, and positive cultures are coded as having actively treated pneumonia at time of surgery.¹⁵

To identify hip fracture patients over the age of 65 years who underwent operative fixation of a hip fracture, we used Current Procedural Terminology (CPT) hip fracture codes, including 27235 (percutaneous screw fixation), 27236 or 27244 (plate-and-screw fixation), and 27245 (intramedullary device), as well as 27125 (hemiarthroplasty) and 27130 (arthroplasty) for patients with a postoperative International Classification of Disease, Ninth Revision (ICD-9) diagnosis code (820.x, 820.2x, or 820.8) consistent with acute hip fracture.^18,19 Procedure type, anesthesia type, and delay from admission to surgery were captured for all procedures.

Preoperative demographics included age, sex, transfer origin, functional status, and body mass index (BMI) category. Binary comorbidities were classified as preoperative anemia (hematocrit, <0.41 for men, <0.36 for women), confusion, dyspnea at rest, uremia (blood urea nitrogen, >6.8 mmol/L), history of cardiovascular disease (congestive heart failure, myocardial infarction, percutaneous coronary intervention, angina pectoris, medically treated hypertension, peripheral vascular disease, or resting claudication), chronic obstructive pulmonary disease, diabetes, renal disease (renal failure or dialysis), and cigarette use in preceding 12 months.^20,21 Although preoperative hypotension and respiratory rate are often considered in patients with pneumonia, these variables were not available from the ACS-NSQIP data.^6,22Pearson χ² test for categorical variables was used to compare baseline demographics and operative characteristics between patients with and without pneumonia, and Student t test was used to compare intervals from hospital admission to hip fracture surgery, surgery start to surgery stop, and surgery to discharge between patients with and without preoperative pneumonia.

Binary outcome measures were compared between patients with and without preoperative pneumonia. “Any AE” included any serious AE (SAE) or any minor AE. SAEs included death, acute renal failure, ventilator use >48 hours, unplanned intubation, septic shock, sepsis, return to operating room, coma >24 hours, cardiac arrest requiring cardiopulmonary resuscitation, myocardial infarction, thromboembolic event (deep vein thrombosis or pulmonary embolism), and stroke/cerebrovascular accident. Minor AEs included progressive renal insufficiency, urinary tract infection, organ/space infection, superficial surgical-site infection, deep surgical-site infection, and wound dehiscence. Other binary outcome measures included discharge destination and unplanned readmission within 30 days after hip fracture surgery.²³Poisson regression with robust error variance as described by Zou²⁴ was used to compare the rates of any, minor, and individual AEs, and any SAEs, between patients with and without pneumonia. Multivariate analysis accounted for the baseline variables in Table 1. AEs that occurred more than once in each group were included in the analyses.

Results

Of the 7128 geriatric hip fracture patients in this study, 82 (1.2%) had active pneumonia at time of surgery (Table 1). Age, BMI, preoperative uremia, history of cardiovascular disease, diabetes, renal disease, and smoking were similar between groups. In addition, there was no difference in anesthesia type or fixation procedure between the pneumonia and no-pneumonia groups. Patients with preoperative pneumonia differed significantly with respect to sex, transfer from facility, preoperative functional dependence, anemia, confusion, dyspnea at rest, and history of chronic obstructive pulmonary disease (Table 1).

Interval from admission to surgery was longer (P < .001) for geriatric hip fracture patients with preoperative pneumonia (mean, 6.8 days; 95% CI, 2.5-11.1 days) than for those without pneumonia (mean, 1.5 days; CI, 1.4-1.5 days). There was no difference (P = .124) in operative time between the pneumonia group (mean, 72.8 min; CI, 64.0-81.5 min) and the no-pneumonia group (mean, 66.1 min; CI, 61.2-67.0 min). Interval from surgery to discharge was longer (P < .001) for patients with preoperative pneumonia (mean, 10.1 days; CI, 6.9-13.4 days) than for those without pneumonia (mean, 6.3 days; CI, 6.1-6.4 days).

Adverse outcomes of geriatric hip fracture surgery are listed in Table 2. In the multivariate analysis, preoperative pneumonia was significantly associated with any AE (relative risk [RR]) = 1.44) and any SAE (RR = 1.79).

Discussion

In the general US population, pneumonia accounts for 1.4% of deaths in people 65 years to 74 years old, 2.1% in people 75 years to 84 years, and 3.1% in people 85 years or older. In total, 3.4% of hospital inpatient deaths are attributed to pneumonia.²⁵ In hospitalized general orthopedic surgical patients as well as hip fracture patients, pneumonia is strongly associated with increased mortality.^26,27

We identified a preoperative pneumonia prevalence of 1.2%, which is comparable to the rates reported in the literature (0.3%-3.2%).^1-3 To our knowledge, our study represents the largest series of patients with preoperative pneumonia at time of hip fracture repair, and the first to independently associate preoperative pneumonia with increased incidence of AEs, including death.

This study had its limitations. First, the ACS-NSQIP morbidity and mortality data, which are limited to the first 30 postoperative days, may be skewed because AEs that occurred after that interval are not captured. Second, coding of pneumonia in ACS-NSQIP does not convey specific information about the disease and its severity—infectious organism(s) responsible; acquisition setting (healthcare or community); treatment given, including antibiotic(s) selection, steroid use, dosing, and duration; and measures of treatment efficacy—limiting interpretation of the difference in delay to surgery. We cannot say whether the longer interval in patients with pneumonia reflects medical optimization, or whether the delay itself or any interventions during that time positively or negatively affected outcomes. In addition, despite using a large national database, we obtained a relatively small sample of patients (82) who had pneumonia before surgical hip fracture repair.

Multivariate analysis controlling for baseline demographics and comorbidities revealed that multiple SAEs were independently associated with preoperative pneumonia (overall SAE, RR = 1.79). Postoperative use of ventilator support for longer than 48 hours (RR = 6.48) and coma longer than 24 hours (RR = 7.31) are expected given the severity of pulmonary compromise in the study cohort.^28,29 Acute renal failure (RR = 14.61) can occur in both hip fracture patients and community-acquired pneumonia patients and may be a multifactorial complication of the pulmonary infection, of the anesthesia, or of the surgical intervention in this cohort.^30-32Unadjusted mortality in hip fracture takes months to a year to normalize to that of age-matched controls.^32-34 In our series, the unadjusted death rate in the pneumonia cohort (Figure) was transiently elevated during the first weeks after surgery but then drew nearer the rate in the nondiseased hip fracture cohort by the end of the first month. Early death in the pneumonia group likely was multifactorial, potentially influenced by the increased burden of comorbidities in the pneumonia group at baseline, and the longer delay to surgery,^35-38 as well as by the natural history of treated pneumonia in hospital patients, who, compared with age-matched hospitalized controls, also exhibit higher mortality during only the first 2 to 4 months of hospitalization for pneumonia.³⁹ We regret that quality improvement strategies in the treatment of geriatric hip fracture surgery with pneumonia cannot be extrapolated from these results.

Similarly, the utility of BMI <18.5 kg/m² as an actionable preoperative finding cannot be assessed from these results. However, we propose that underweight geriatric hip fracture patients with pneumonia may benefit from more aggressive preoperative optimization that does not delay surgery. Higher acuity of postoperative care, including more intensive nursing care and early coordination of care with respiratory therapists and medical comanagement teams, may also be beneficial.

Anesthesia type did not differ between patients with and without preoperative pneumonia and was not associated with AEs in patients with preoperative pneumonia. Consistent with our findings, multiple studies have reported no significant differences in short-term outcomes of hip fracture repair between general and spinal anesthesia, though no other study has compared the benefits of general and spinal anesthesia for patients with preoperative pneumonia.^40-44 Although spinal anesthesia (relative to general anesthesia) has been reported to have benefits in hip and knee arthroplasty, these benefits appear not to translate to hip fracture repair.^45-50 The results of the present study suggest that general and spinal anesthesia may be equivalent in terms of risk for the geriatric hip fracture patient with preoperative pneumonia.^43,44Our attempt to evaluate the CURB-65 pneumonia severity score as a prognosticator of AEs was thwarted by the absence of required variables in the ACS-NSQIP dataset (confusion, uremia, dyspnea, and age were available; hypotension and blood pressure were not). In our analysis, we did include, individually, variables previously found to predict AEs in the medical pneumonia population (confusion, uremia, dyspnea at rest, anemia).^9-11,32 However, these clinical findings are nonspecific in hip fracture patients, who may become anemic, confused, dyspneic, or uremic from a multitude of factors related to their injury and unrelated to pneumonia, including but not limited to hemorrhage, muscle damage, renal injury, and pulmonary embolism. It is not surprising that confusion, uremia, dyspnea at rest, and anemia were not individually predictive of AEs or death within 30 days after surgery in the cohort of geriatric hip fracture patients with pneumonia.

There is no literature that argues for or against delaying hip fracture surgery in geriatric hip fracture patients with pneumonia. The surgical delay observed in this population is ostensibly related to medical optimization of the pneumonia and/or underlying comorbidities. However, we did not find a morbidity or mortality detriment or benefit in delaying surgery by 1 to 4 days in this population. Delay of surgery is a poor covariate, given extensive confounding by medical management and preoperative optimizing of comorbid conditions (reflected in our independent variable and covariates) as well as institutional and surgeon variations in policy and behavior and other unaccounted influences. Although some authors have found no difference in mortality or major AEs between hip fracture patients who had a surgical delay and those who did not,^31,51-53 other series and meta-analyses have suggested a mortality detriment in a surgical delay of more than 2 days^36,54 or 4 days⁵⁵ from admission. Given our data, we cannot recommend against immediate hip fracture repair in the subpopulation of geriatric hip fracture patients with pneumonia.

Our study findings suggest that preoperative pneumonia is a rare independent risk factor for AEs after hip fracture surgery in geriatric patients. Underweight BMI is predictive of death in geriatric hip fracture surgery patients who present with pneumonia, whereas early surgical repair appears not to be associated with adverse outcomes. Further investigation is warranted to determine if such patients benefit from specific preoperative and postoperative strategies for optimizing medical and surgical care based on these findings.

Am J Orthop. 2017;46(3):E177-E185. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• The prevalence of preoperative pneumonia is 1.2% among hip fracture patients aged >65 years.
• Preoperative pneumonia is an independent risk factor for mortality and adverse events including renal failure, prolonged ventilator dependence, and prolonged altered mental status after geriatric hip fracture surgery.
• Underweight BMI (<18.5 kg/m²) was associated with higher mortality within 30 days among hip fracture patients admitted with pneumonia.
• The mortality rate normalized to that of patients without pneumonia within 2 weeks of hip fracture surgery.
• Time from admission to surgery was not associated with adverse events or mortality among hip fracture patients admitted with pneumonia.

Preoperative pneumonia remains relatively unexplored as a risk factor for adverse outcomes in geriatric hip fracture surgery. Dated studies report a 0.3% to 3.2% prevalence of “recent pneumonia” in patients presenting with hip fracture but provide neither a definition of pneumonia based on clinical criteria nor a subset analysis of outcomes in the pneumonia group.^1-3 Although active pneumonia has been identified as a preoperative optimization target in the management guidelines for geriatric hip fracture,⁴ we are unaware of any studies that have reported on differences in demographics, comorbidities, delay to surgery, or adverse outcomes between hip fracture patients with and without preoperative pneumonia.

This paucity of information on the effect of preoperative pneumonia in the hip fracture population may be related to low prevalence of preoperative pneumonia and a cadre of variable definitions, which limit identification of a cohort of patients with preoperative pneumonia large enough from which to draw meaningful results. Database studies, especially those using surgical registries rather than administrative or reimbursement data, offer particular advantages for investigation of such rare clinical entities.⁵Medical care of patients with pneumonia alone is known to be facilitated by assessments of mortality risk from clinical and laboratory data. The modified British Thoracic Society rule/CURB-65 (confusion, urea, respiratory rate, blood pressure) score is strongly predictive of mortality in hospitalized adults with pneumonia (odds ratio [OR], 4.59; 95% confidence interval [CI], 1.42-14.85; P = .011) and may guide antibiotic therapy, laboratory investigations, and the decision to intubate in a patient with pneumonia.^6-8 This score is predictive of adverse events (AEs), hospital length of stay, and use of intensive care services.^6,7,9-13 We hypothesized that preoperative clinical indicators assessed by pneumonia severity scores as well as patient demographics and baseline comorbidities may also have prognostic value for risk of AEs in a cohort of geriatric hip fracture surgery patients with preoperative pneumonia.

In this article, we first describe the prevalence of preoperative pneumonia in geriatric hip fracture surgery patients as well as demographic and operative differences between patients with and without the disease. We then ask 3 questions: Is preoperative pneumonia an independent risk factor for mortality and adverse outcomes in geriatric hip fracture surgery? Is there a postoperative interval during which the unadjusted mortality rate is higher among patients with preoperative pneumonia? In patients with preoperative pneumonia, what are the predictors of morbidity and mortality?

Methods

Yale University’s Human Investigations Committee approved this retrospective cohort study, which used the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database for the period 2005 to 2012. ACS-NSQIP is a prospective, multi-institutional outcomes program that collects data on preoperative comorbidities, intraoperative variables, and 30-day postoperative outcomes for patients undergoing surgical procedures in inpatient and outpatient settings.¹⁴

Unlike administrative databases, which are based on reimbursement data, ACS-NSQIP data are collected by trained surgical clinical reviewers for the purposes of quality improvement and clinical research, and data quality is ensured with routine auditing.¹⁵ The program has gained a high degree of respect as a powerful and valid data source in both general¹⁶ and orthopedic¹⁷ surgery literature. The database offers a particular advantage with respect to the study of preoperative pneumonia: Only patients with new or recently diagnosed pneumonia on antibiotic therapy who meet strict criteria for characteristic findings on chest radiography, clinical signs and symptoms of respiratory illness, and positive cultures are coded as having actively treated pneumonia at time of surgery.¹⁵

To identify hip fracture patients over the age of 65 years who underwent operative fixation of a hip fracture, we used Current Procedural Terminology (CPT) hip fracture codes, including 27235 (percutaneous screw fixation), 27236 or 27244 (plate-and-screw fixation), and 27245 (intramedullary device), as well as 27125 (hemiarthroplasty) and 27130 (arthroplasty) for patients with a postoperative International Classification of Disease, Ninth Revision (ICD-9) diagnosis code (820.x, 820.2x, or 820.8) consistent with acute hip fracture.^18,19 Procedure type, anesthesia type, and delay from admission to surgery were captured for all procedures.

Preoperative demographics included age, sex, transfer origin, functional status, and body mass index (BMI) category. Binary comorbidities were classified as preoperative anemia (hematocrit, <0.41 for men, <0.36 for women), confusion, dyspnea at rest, uremia (blood urea nitrogen, >6.8 mmol/L), history of cardiovascular disease (congestive heart failure, myocardial infarction, percutaneous coronary intervention, angina pectoris, medically treated hypertension, peripheral vascular disease, or resting claudication), chronic obstructive pulmonary disease, diabetes, renal disease (renal failure or dialysis), and cigarette use in preceding 12 months.^20,21 Although preoperative hypotension and respiratory rate are often considered in patients with pneumonia, these variables were not available from the ACS-NSQIP data.^6,22Pearson χ² test for categorical variables was used to compare baseline demographics and operative characteristics between patients with and without pneumonia, and Student t test was used to compare intervals from hospital admission to hip fracture surgery, surgery start to surgery stop, and surgery to discharge between patients with and without preoperative pneumonia.

Binary outcome measures were compared between patients with and without preoperative pneumonia. “Any AE” included any serious AE (SAE) or any minor AE. SAEs included death, acute renal failure, ventilator use >48 hours, unplanned intubation, septic shock, sepsis, return to operating room, coma >24 hours, cardiac arrest requiring cardiopulmonary resuscitation, myocardial infarction, thromboembolic event (deep vein thrombosis or pulmonary embolism), and stroke/cerebrovascular accident. Minor AEs included progressive renal insufficiency, urinary tract infection, organ/space infection, superficial surgical-site infection, deep surgical-site infection, and wound dehiscence. Other binary outcome measures included discharge destination and unplanned readmission within 30 days after hip fracture surgery.²³Poisson regression with robust error variance as described by Zou²⁴ was used to compare the rates of any, minor, and individual AEs, and any SAEs, between patients with and without pneumonia. Multivariate analysis accounted for the baseline variables in Table 1. AEs that occurred more than once in each group were included in the analyses.

Results

Of the 7128 geriatric hip fracture patients in this study, 82 (1.2%) had active pneumonia at time of surgery (Table 1). Age, BMI, preoperative uremia, history of cardiovascular disease, diabetes, renal disease, and smoking were similar between groups. In addition, there was no difference in anesthesia type or fixation procedure between the pneumonia and no-pneumonia groups. Patients with preoperative pneumonia differed significantly with respect to sex, transfer from facility, preoperative functional dependence, anemia, confusion, dyspnea at rest, and history of chronic obstructive pulmonary disease (Table 1).

Interval from admission to surgery was longer (P < .001) for geriatric hip fracture patients with preoperative pneumonia (mean, 6.8 days; 95% CI, 2.5-11.1 days) than for those without pneumonia (mean, 1.5 days; CI, 1.4-1.5 days). There was no difference (P = .124) in operative time between the pneumonia group (mean, 72.8 min; CI, 64.0-81.5 min) and the no-pneumonia group (mean, 66.1 min; CI, 61.2-67.0 min). Interval from surgery to discharge was longer (P < .001) for patients with preoperative pneumonia (mean, 10.1 days; CI, 6.9-13.4 days) than for those without pneumonia (mean, 6.3 days; CI, 6.1-6.4 days).

Adverse outcomes of geriatric hip fracture surgery are listed in Table 2. In the multivariate analysis, preoperative pneumonia was significantly associated with any AE (relative risk [RR]) = 1.44) and any SAE (RR = 1.79).

Discussion

In the general US population, pneumonia accounts for 1.4% of deaths in people 65 years to 74 years old, 2.1% in people 75 years to 84 years, and 3.1% in people 85 years or older. In total, 3.4% of hospital inpatient deaths are attributed to pneumonia.²⁵ In hospitalized general orthopedic surgical patients as well as hip fracture patients, pneumonia is strongly associated with increased mortality.^26,27

We identified a preoperative pneumonia prevalence of 1.2%, which is comparable to the rates reported in the literature (0.3%-3.2%).^1-3 To our knowledge, our study represents the largest series of patients with preoperative pneumonia at time of hip fracture repair, and the first to independently associate preoperative pneumonia with increased incidence of AEs, including death.

This study had its limitations. First, the ACS-NSQIP morbidity and mortality data, which are limited to the first 30 postoperative days, may be skewed because AEs that occurred after that interval are not captured. Second, coding of pneumonia in ACS-NSQIP does not convey specific information about the disease and its severity—infectious organism(s) responsible; acquisition setting (healthcare or community); treatment given, including antibiotic(s) selection, steroid use, dosing, and duration; and measures of treatment efficacy—limiting interpretation of the difference in delay to surgery. We cannot say whether the longer interval in patients with pneumonia reflects medical optimization, or whether the delay itself or any interventions during that time positively or negatively affected outcomes. In addition, despite using a large national database, we obtained a relatively small sample of patients (82) who had pneumonia before surgical hip fracture repair.

Multivariate analysis controlling for baseline demographics and comorbidities revealed that multiple SAEs were independently associated with preoperative pneumonia (overall SAE, RR = 1.79). Postoperative use of ventilator support for longer than 48 hours (RR = 6.48) and coma longer than 24 hours (RR = 7.31) are expected given the severity of pulmonary compromise in the study cohort.^28,29 Acute renal failure (RR = 14.61) can occur in both hip fracture patients and community-acquired pneumonia patients and may be a multifactorial complication of the pulmonary infection, of the anesthesia, or of the surgical intervention in this cohort.^30-32Unadjusted mortality in hip fracture takes months to a year to normalize to that of age-matched controls.^32-34 In our series, the unadjusted death rate in the pneumonia cohort (Figure) was transiently elevated during the first weeks after surgery but then drew nearer the rate in the nondiseased hip fracture cohort by the end of the first month. Early death in the pneumonia group likely was multifactorial, potentially influenced by the increased burden of comorbidities in the pneumonia group at baseline, and the longer delay to surgery,^35-38 as well as by the natural history of treated pneumonia in hospital patients, who, compared with age-matched hospitalized controls, also exhibit higher mortality during only the first 2 to 4 months of hospitalization for pneumonia.³⁹ We regret that quality improvement strategies in the treatment of geriatric hip fracture surgery with pneumonia cannot be extrapolated from these results.

Similarly, the utility of BMI <18.5 kg/m² as an actionable preoperative finding cannot be assessed from these results. However, we propose that underweight geriatric hip fracture patients with pneumonia may benefit from more aggressive preoperative optimization that does not delay surgery. Higher acuity of postoperative care, including more intensive nursing care and early coordination of care with respiratory therapists and medical comanagement teams, may also be beneficial.

Anesthesia type did not differ between patients with and without preoperative pneumonia and was not associated with AEs in patients with preoperative pneumonia. Consistent with our findings, multiple studies have reported no significant differences in short-term outcomes of hip fracture repair between general and spinal anesthesia, though no other study has compared the benefits of general and spinal anesthesia for patients with preoperative pneumonia.^40-44 Although spinal anesthesia (relative to general anesthesia) has been reported to have benefits in hip and knee arthroplasty, these benefits appear not to translate to hip fracture repair.^45-50 The results of the present study suggest that general and spinal anesthesia may be equivalent in terms of risk for the geriatric hip fracture patient with preoperative pneumonia.^43,44Our attempt to evaluate the CURB-65 pneumonia severity score as a prognosticator of AEs was thwarted by the absence of required variables in the ACS-NSQIP dataset (confusion, uremia, dyspnea, and age were available; hypotension and blood pressure were not). In our analysis, we did include, individually, variables previously found to predict AEs in the medical pneumonia population (confusion, uremia, dyspnea at rest, anemia).^9-11,32 However, these clinical findings are nonspecific in hip fracture patients, who may become anemic, confused, dyspneic, or uremic from a multitude of factors related to their injury and unrelated to pneumonia, including but not limited to hemorrhage, muscle damage, renal injury, and pulmonary embolism. It is not surprising that confusion, uremia, dyspnea at rest, and anemia were not individually predictive of AEs or death within 30 days after surgery in the cohort of geriatric hip fracture patients with pneumonia.

There is no literature that argues for or against delaying hip fracture surgery in geriatric hip fracture patients with pneumonia. The surgical delay observed in this population is ostensibly related to medical optimization of the pneumonia and/or underlying comorbidities. However, we did not find a morbidity or mortality detriment or benefit in delaying surgery by 1 to 4 days in this population. Delay of surgery is a poor covariate, given extensive confounding by medical management and preoperative optimizing of comorbid conditions (reflected in our independent variable and covariates) as well as institutional and surgeon variations in policy and behavior and other unaccounted influences. Although some authors have found no difference in mortality or major AEs between hip fracture patients who had a surgical delay and those who did not,^31,51-53 other series and meta-analyses have suggested a mortality detriment in a surgical delay of more than 2 days^36,54 or 4 days⁵⁵ from admission. Given our data, we cannot recommend against immediate hip fracture repair in the subpopulation of geriatric hip fracture patients with pneumonia.

Our study findings suggest that preoperative pneumonia is a rare independent risk factor for AEs after hip fracture surgery in geriatric patients. Underweight BMI is predictive of death in geriatric hip fracture surgery patients who present with pneumonia, whereas early surgical repair appears not to be associated with adverse outcomes. Further investigation is warranted to determine if such patients benefit from specific preoperative and postoperative strategies for optimizing medical and surgical care based on these findings.

Am J Orthop. 2017;46(3):E177-E185. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• With EHR implementation there are small changes in the level of billing coding.
• Although these changes may be statistically significant they are relatively minor.
• In the general internal medicine department, level 4 coding increased by 1.2% while level 3 coding decreased by 0.5%.
• In the orthopedics department, level 4 coding increased by 3.3% while level 3 coding decreased by 3.1%.
• Reports in the lay media regarding dramatic up-coding after EHR implementation may be misleading.

The Health Information Technology for Economic and Clinical Health (HITECH) Act, which was signed into law in 2009, mandated that hospitals that care for Medicare patients either begin using electronic health records (EHRs) or pay a nontrivial penalty.¹ By now, the majority of orthopedic surgeons have implemented EHRs in their practices.² Despite ongoing debate in the orthopedic literature,³ EHRs are expected to improve coordination of care, reduce duplicate testing, and reduce costs over the long term as healthcare insurance coverage is extended to millions more Americans.

In early coverage, however, media reported that EHR implementation at some hospitals was correlated with substantial increases in Medicare payments.⁴ Journalists suggested the billion dollars more paid by Medicare to hospitals in 2010 than in 2005 were partly attributable to up-coding facilitated by EHRs.⁵ The secretary of the Department of Health and Human Services (DHHS) and the attorney general of the Department of Justice also weighed in on this controversy by expressing their concerns in a letter to the presidents of 5 hospital associations.⁶ The inspector general of DHHS also published a report critical of Medicare officials’ oversight of EHRs.⁷Responding to the critical reception of EHR implementations, investigators studied the validity of the early reports and anecdotes. Some initial reports cited the emergency department (ED) as an area at high risk for using the convenience of EHRs to up-code visits.⁵ The DHHS Office of the Inspector General noted that, between 2001 and 2010, the proportion of claims for lower reimbursement categories of American Medical Association Current Procedural Terminology (CPT) codes decreased while the proportion for higher-paid billing codes increased for all visit types.⁸ Addressing these concerns, the American Hospital Association⁹ issued a brief that noted that any observed coding increases were more likely attributable to more ED use by Medicare patients and increased average illness severity. In a thoughtful perspective, Pitts¹⁰ conceded that, though utilization and illness severity may explain part of the trend, the trend may also be related to technological innovations and changes in culture and practice style in the ED.

Because these studies and reports variously suggested that EHR implementation affects patient volume and up-coding, and because none of the reports specifically addressed orthopedics, we conducted a study to determine whether any significant up-coding or change in patient volumes occurred around the time of EHR implementation in ambulatory practices at our academic medical center. In a recent national study, Adler-Milstein and Jha¹¹ compared billing data of hospitals that adopted EHRs and hospitals that did not. Although both groups showed increased billing trends, the increases were not significantly different between the EHR adopters and nonadopters. To more effectively control for the confounding differences between groups of EHR adopters and nonadopters, we studied individual departments during EHR implementation at our institution.

Methods

In 2011, our academic medical center began the transition to EHRs (Epic). We examined our center’s trends in patient volumes and billing coding around the time of the transition in the outpatient practice of the general internal medicine (GIM) department (EHR transition, October 2011) and the outpatient practice of the orthopedics department (EHR transition, March 2012). These departments were chosen because they are representative of a GIM practice and a subspecialty practice, and because a recent study found that GIM practitioners and orthopedic surgeons were among those specialists who used EHRs the most.¹²

After this study was approved by our Human Investigations Committee, we began using CPT codes to identify all outpatient visits (new, consultation, and return) on a monthly basis. We compared the volume of patient visits and the billing coding level in the GIM and orthopedics departments before and after EHR implementation. Pearson χ² test was used when appropriate, and statistical analyses were performed with SPSS for Windows Version 16.0.

Results

In the GIM department, mean monthly volume of patient visits in the 12 months before EHR implementation was similar to that in the 12 months afterward (613 vs 587; P = .439). Even when normalized for changes in provider availability (maternity leave), the decrease in volume of patient visits after EHR implementation in the GIM department was not significant (6.9%; P = .107). Likewise, in the orthopedics department, mean monthly volume of patient visits in the 17 months before EHR implementation was similar to that in the 7 months afterward (2157 vs 2317; P = .156). In fact, patient volumes remained constant during the EHR transition (Figure 1).

EHR implementation brought small changes in billing coding levels. In the GIM department, the largest change was a 1.2% increase in level 4 billing coding—an increase accompanied by a 0.5% decrease in level 3 coding.

Discussion

It is remarkable that the volumes of patient visits in the GIM and orthopedics departments at our academic center were not affected by EHR implementation.

Rather than reduce scheduling during the EHR transition, surgeons in our practice either added or lengthened clinic sessions, and the level of ancillary staffing was adjusted accordingly. As staffing costs at any given time are multifactorial and vary widely, estimating the cost of these staffing changes during the EHR transition is difficult. We should note that extending ancillary staff hours during the transition very likely increased costs, and it is unclear whether they were higher or lower than the costs that would have been incurred had we reduced scheduling or tried some combination of these strategies.

Although billing coding levels changed with EHR implementation, the changes were small. In the GIM department, level 4 CPT coded visits as percentages of all visits increased to 59.5% from 58.3%, and level 5 visits increased to 6.2% from 6.0%; in the orthopedics department, level 4 visits increased to 40.2% from 37.1%, and level 5 visits increased to 5.5% from 3.8% (Table). The 1.2% and 0.2% absolute increases in level 4 and level 5 visits in the GIM department represent 2.1% and 3.3% relative increases in level 4 and level 5 visits, and the 3.3% and 1.7% absolute increases in the orthopedics department represent 8.4% and 44.7% relative increases in level 4 and level 5 visits after EHR implementation.

Although the absolute increases in level 4 and level 5 visits were relatively minor, popular media have raised the alarm about 43% and 82% relative increases in level 5 visits after EHR implementation in some hospitals’ EDs.⁴ Although our orthopedics department showed a 44.7% relative increase in level 5 visits after EHR implementation, this represented an increase of only 1.7% of patient visits overall. Our findings therefore indicate that lay media reports could be misleading. Nevertheless, the small changes we found were statistically significant.

One explanation for these small changes is that EHRs facilitate better documentation of services provided. Therefore, what seem to be billing coding changes could be more accurate reports of high-level care that is the same as before. In addition, because of meaningful use mandates that coincided with the requirement to implement EHRs, additional data elements are now being consistently collected and reviewed (these may not necessarily have been collected and reviewed before). In some patient encounters, these additional data elements may have contributed to higher levels of service, and this effect could be especially apparent in EDs.

Some have suggested a potential for large-scale up-coding during EHR transitions. Others have contended that coding level increases are a consequence of a time-intensive data entry process, collection and review of additional data, and more accurate reporting of services already being provided. We are not convinced that large coding changes are attributable solely to EHR implementation, as the changes at our center have been relatively small.

Nevertheless, minor coding level changes could translate to large changes in healthcare costs when scaled nationally. Although causes may be innocuous, any increases in national healthcare costs are concerning in our time of limited budgets and scrutinized healthcare utilization.

This study had its limitations. First, including billing data from only 2 departments at a single center may limit the generalizability of findings. However, we specifically selected a GIM department and a specialty (orthopedics) department in an attempt to capture a representative sample of practices. Another limitation is that we investigated billing codes over only 2 years, around the implementation of EHRs in these departments, and therefore may have captured only short-term changes. However, as patient volumes and billing are subject to many factors, including staffing changes (eg, new partners, new hires, retirements, other departures), we attempted to limit the effect of confounding variables by limiting the period of analysis.

Overall, changes in patient volume and coded level of service during EHR implementation at our institution were relatively small. Although the trend toward higher billing coding levels was statistically significant, these 0.2% and 1.7% increases in level 5 coding hardly deserve the negative attention from lay media. These small increases are unlikely caused by intentional up-coding, and more likely reflect better documentation of an already high level of care. We hope these findings allay the concern that up-coding increased dramatically with EHR implementation.

Am J Orthop. 2017;46(3):E172-E176. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• With EHR implementation there are small changes in the level of billing coding.
• Although these changes may be statistically significant they are relatively minor.
• In the general internal medicine department, level 4 coding increased by 1.2% while level 3 coding decreased by 0.5%.
• In the orthopedics department, level 4 coding increased by 3.3% while level 3 coding decreased by 3.1%.
• Reports in the lay media regarding dramatic up-coding after EHR implementation may be misleading.

The Health Information Technology for Economic and Clinical Health (HITECH) Act, which was signed into law in 2009, mandated that hospitals that care for Medicare patients either begin using electronic health records (EHRs) or pay a nontrivial penalty.¹ By now, the majority of orthopedic surgeons have implemented EHRs in their practices.² Despite ongoing debate in the orthopedic literature,³ EHRs are expected to improve coordination of care, reduce duplicate testing, and reduce costs over the long term as healthcare insurance coverage is extended to millions more Americans.

In early coverage, however, media reported that EHR implementation at some hospitals was correlated with substantial increases in Medicare payments.⁴ Journalists suggested the billion dollars more paid by Medicare to hospitals in 2010 than in 2005 were partly attributable to up-coding facilitated by EHRs.⁵ The secretary of the Department of Health and Human Services (DHHS) and the attorney general of the Department of Justice also weighed in on this controversy by expressing their concerns in a letter to the presidents of 5 hospital associations.⁶ The inspector general of DHHS also published a report critical of Medicare officials’ oversight of EHRs.⁷Responding to the critical reception of EHR implementations, investigators studied the validity of the early reports and anecdotes. Some initial reports cited the emergency department (ED) as an area at high risk for using the convenience of EHRs to up-code visits.⁵ The DHHS Office of the Inspector General noted that, between 2001 and 2010, the proportion of claims for lower reimbursement categories of American Medical Association Current Procedural Terminology (CPT) codes decreased while the proportion for higher-paid billing codes increased for all visit types.⁸ Addressing these concerns, the American Hospital Association⁹ issued a brief that noted that any observed coding increases were more likely attributable to more ED use by Medicare patients and increased average illness severity. In a thoughtful perspective, Pitts¹⁰ conceded that, though utilization and illness severity may explain part of the trend, the trend may also be related to technological innovations and changes in culture and practice style in the ED.

Because these studies and reports variously suggested that EHR implementation affects patient volume and up-coding, and because none of the reports specifically addressed orthopedics, we conducted a study to determine whether any significant up-coding or change in patient volumes occurred around the time of EHR implementation in ambulatory practices at our academic medical center. In a recent national study, Adler-Milstein and Jha¹¹ compared billing data of hospitals that adopted EHRs and hospitals that did not. Although both groups showed increased billing trends, the increases were not significantly different between the EHR adopters and nonadopters. To more effectively control for the confounding differences between groups of EHR adopters and nonadopters, we studied individual departments during EHR implementation at our institution.

Methods

In 2011, our academic medical center began the transition to EHRs (Epic). We examined our center’s trends in patient volumes and billing coding around the time of the transition in the outpatient practice of the general internal medicine (GIM) department (EHR transition, October 2011) and the outpatient practice of the orthopedics department (EHR transition, March 2012). These departments were chosen because they are representative of a GIM practice and a subspecialty practice, and because a recent study found that GIM practitioners and orthopedic surgeons were among those specialists who used EHRs the most.¹²

After this study was approved by our Human Investigations Committee, we began using CPT codes to identify all outpatient visits (new, consultation, and return) on a monthly basis. We compared the volume of patient visits and the billing coding level in the GIM and orthopedics departments before and after EHR implementation. Pearson χ² test was used when appropriate, and statistical analyses were performed with SPSS for Windows Version 16.0.

Results

In the GIM department, mean monthly volume of patient visits in the 12 months before EHR implementation was similar to that in the 12 months afterward (613 vs 587; P = .439). Even when normalized for changes in provider availability (maternity leave), the decrease in volume of patient visits after EHR implementation in the GIM department was not significant (6.9%; P = .107). Likewise, in the orthopedics department, mean monthly volume of patient visits in the 17 months before EHR implementation was similar to that in the 7 months afterward (2157 vs 2317; P = .156). In fact, patient volumes remained constant during the EHR transition (Figure 1).

EHR implementation brought small changes in billing coding levels. In the GIM department, the largest change was a 1.2% increase in level 4 billing coding—an increase accompanied by a 0.5% decrease in level 3 coding.

Discussion

It is remarkable that the volumes of patient visits in the GIM and orthopedics departments at our academic center were not affected by EHR implementation.

Rather than reduce scheduling during the EHR transition, surgeons in our practice either added or lengthened clinic sessions, and the level of ancillary staffing was adjusted accordingly. As staffing costs at any given time are multifactorial and vary widely, estimating the cost of these staffing changes during the EHR transition is difficult. We should note that extending ancillary staff hours during the transition very likely increased costs, and it is unclear whether they were higher or lower than the costs that would have been incurred had we reduced scheduling or tried some combination of these strategies.

Although billing coding levels changed with EHR implementation, the changes were small. In the GIM department, level 4 CPT coded visits as percentages of all visits increased to 59.5% from 58.3%, and level 5 visits increased to 6.2% from 6.0%; in the orthopedics department, level 4 visits increased to 40.2% from 37.1%, and level 5 visits increased to 5.5% from 3.8% (Table). The 1.2% and 0.2% absolute increases in level 4 and level 5 visits in the GIM department represent 2.1% and 3.3% relative increases in level 4 and level 5 visits, and the 3.3% and 1.7% absolute increases in the orthopedics department represent 8.4% and 44.7% relative increases in level 4 and level 5 visits after EHR implementation.

Although the absolute increases in level 4 and level 5 visits were relatively minor, popular media have raised the alarm about 43% and 82% relative increases in level 5 visits after EHR implementation in some hospitals’ EDs.⁴ Although our orthopedics department showed a 44.7% relative increase in level 5 visits after EHR implementation, this represented an increase of only 1.7% of patient visits overall. Our findings therefore indicate that lay media reports could be misleading. Nevertheless, the small changes we found were statistically significant.

One explanation for these small changes is that EHRs facilitate better documentation of services provided. Therefore, what seem to be billing coding changes could be more accurate reports of high-level care that is the same as before. In addition, because of meaningful use mandates that coincided with the requirement to implement EHRs, additional data elements are now being consistently collected and reviewed (these may not necessarily have been collected and reviewed before). In some patient encounters, these additional data elements may have contributed to higher levels of service, and this effect could be especially apparent in EDs.

Some have suggested a potential for large-scale up-coding during EHR transitions. Others have contended that coding level increases are a consequence of a time-intensive data entry process, collection and review of additional data, and more accurate reporting of services already being provided. We are not convinced that large coding changes are attributable solely to EHR implementation, as the changes at our center have been relatively small.

Nevertheless, minor coding level changes could translate to large changes in healthcare costs when scaled nationally. Although causes may be innocuous, any increases in national healthcare costs are concerning in our time of limited budgets and scrutinized healthcare utilization.

This study had its limitations. First, including billing data from only 2 departments at a single center may limit the generalizability of findings. However, we specifically selected a GIM department and a specialty (orthopedics) department in an attempt to capture a representative sample of practices. Another limitation is that we investigated billing codes over only 2 years, around the implementation of EHRs in these departments, and therefore may have captured only short-term changes. However, as patient volumes and billing are subject to many factors, including staffing changes (eg, new partners, new hires, retirements, other departures), we attempted to limit the effect of confounding variables by limiting the period of analysis.

Overall, changes in patient volume and coded level of service during EHR implementation at our institution were relatively small. Although the trend toward higher billing coding levels was statistically significant, these 0.2% and 1.7% increases in level 5 coding hardly deserve the negative attention from lay media. These small increases are unlikely caused by intentional up-coding, and more likely reflect better documentation of an already high level of care. We hope these findings allay the concern that up-coding increased dramatically with EHR implementation.

Am J Orthop. 2017;46(3):E172-E176. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• With EHR implementation there are small changes in the level of billing coding.
• Although these changes may be statistically significant they are relatively minor.
• In the general internal medicine department, level 4 coding increased by 1.2% while level 3 coding decreased by 0.5%.
• In the orthopedics department, level 4 coding increased by 3.3% while level 3 coding decreased by 3.1%.
• Reports in the lay media regarding dramatic up-coding after EHR implementation may be misleading.

The Health Information Technology for Economic and Clinical Health (HITECH) Act, which was signed into law in 2009, mandated that hospitals that care for Medicare patients either begin using electronic health records (EHRs) or pay a nontrivial penalty.¹ By now, the majority of orthopedic surgeons have implemented EHRs in their practices.² Despite ongoing debate in the orthopedic literature,³ EHRs are expected to improve coordination of care, reduce duplicate testing, and reduce costs over the long term as healthcare insurance coverage is extended to millions more Americans.

In early coverage, however, media reported that EHR implementation at some hospitals was correlated with substantial increases in Medicare payments.⁴ Journalists suggested the billion dollars more paid by Medicare to hospitals in 2010 than in 2005 were partly attributable to up-coding facilitated by EHRs.⁵ The secretary of the Department of Health and Human Services (DHHS) and the attorney general of the Department of Justice also weighed in on this controversy by expressing their concerns in a letter to the presidents of 5 hospital associations.⁶ The inspector general of DHHS also published a report critical of Medicare officials’ oversight of EHRs.⁷Responding to the critical reception of EHR implementations, investigators studied the validity of the early reports and anecdotes. Some initial reports cited the emergency department (ED) as an area at high risk for using the convenience of EHRs to up-code visits.⁵ The DHHS Office of the Inspector General noted that, between 2001 and 2010, the proportion of claims for lower reimbursement categories of American Medical Association Current Procedural Terminology (CPT) codes decreased while the proportion for higher-paid billing codes increased for all visit types.⁸ Addressing these concerns, the American Hospital Association⁹ issued a brief that noted that any observed coding increases were more likely attributable to more ED use by Medicare patients and increased average illness severity. In a thoughtful perspective, Pitts¹⁰ conceded that, though utilization and illness severity may explain part of the trend, the trend may also be related to technological innovations and changes in culture and practice style in the ED.

Because these studies and reports variously suggested that EHR implementation affects patient volume and up-coding, and because none of the reports specifically addressed orthopedics, we conducted a study to determine whether any significant up-coding or change in patient volumes occurred around the time of EHR implementation in ambulatory practices at our academic medical center. In a recent national study, Adler-Milstein and Jha¹¹ compared billing data of hospitals that adopted EHRs and hospitals that did not. Although both groups showed increased billing trends, the increases were not significantly different between the EHR adopters and nonadopters. To more effectively control for the confounding differences between groups of EHR adopters and nonadopters, we studied individual departments during EHR implementation at our institution.

Methods

In 2011, our academic medical center began the transition to EHRs (Epic). We examined our center’s trends in patient volumes and billing coding around the time of the transition in the outpatient practice of the general internal medicine (GIM) department (EHR transition, October 2011) and the outpatient practice of the orthopedics department (EHR transition, March 2012). These departments were chosen because they are representative of a GIM practice and a subspecialty practice, and because a recent study found that GIM practitioners and orthopedic surgeons were among those specialists who used EHRs the most.¹²

After this study was approved by our Human Investigations Committee, we began using CPT codes to identify all outpatient visits (new, consultation, and return) on a monthly basis. We compared the volume of patient visits and the billing coding level in the GIM and orthopedics departments before and after EHR implementation. Pearson χ² test was used when appropriate, and statistical analyses were performed with SPSS for Windows Version 16.0.

Results

In the GIM department, mean monthly volume of patient visits in the 12 months before EHR implementation was similar to that in the 12 months afterward (613 vs 587; P = .439). Even when normalized for changes in provider availability (maternity leave), the decrease in volume of patient visits after EHR implementation in the GIM department was not significant (6.9%; P = .107). Likewise, in the orthopedics department, mean monthly volume of patient visits in the 17 months before EHR implementation was similar to that in the 7 months afterward (2157 vs 2317; P = .156). In fact, patient volumes remained constant during the EHR transition (Figure 1).

EHR implementation brought small changes in billing coding levels. In the GIM department, the largest change was a 1.2% increase in level 4 billing coding—an increase accompanied by a 0.5% decrease in level 3 coding.

Discussion

It is remarkable that the volumes of patient visits in the GIM and orthopedics departments at our academic center were not affected by EHR implementation.

Rather than reduce scheduling during the EHR transition, surgeons in our practice either added or lengthened clinic sessions, and the level of ancillary staffing was adjusted accordingly. As staffing costs at any given time are multifactorial and vary widely, estimating the cost of these staffing changes during the EHR transition is difficult. We should note that extending ancillary staff hours during the transition very likely increased costs, and it is unclear whether they were higher or lower than the costs that would have been incurred had we reduced scheduling or tried some combination of these strategies.

Although billing coding levels changed with EHR implementation, the changes were small. In the GIM department, level 4 CPT coded visits as percentages of all visits increased to 59.5% from 58.3%, and level 5 visits increased to 6.2% from 6.0%; in the orthopedics department, level 4 visits increased to 40.2% from 37.1%, and level 5 visits increased to 5.5% from 3.8% (Table). The 1.2% and 0.2% absolute increases in level 4 and level 5 visits in the GIM department represent 2.1% and 3.3% relative increases in level 4 and level 5 visits, and the 3.3% and 1.7% absolute increases in the orthopedics department represent 8.4% and 44.7% relative increases in level 4 and level 5 visits after EHR implementation.

Although the absolute increases in level 4 and level 5 visits were relatively minor, popular media have raised the alarm about 43% and 82% relative increases in level 5 visits after EHR implementation in some hospitals’ EDs.⁴ Although our orthopedics department showed a 44.7% relative increase in level 5 visits after EHR implementation, this represented an increase of only 1.7% of patient visits overall. Our findings therefore indicate that lay media reports could be misleading. Nevertheless, the small changes we found were statistically significant.

One explanation for these small changes is that EHRs facilitate better documentation of services provided. Therefore, what seem to be billing coding changes could be more accurate reports of high-level care that is the same as before. In addition, because of meaningful use mandates that coincided with the requirement to implement EHRs, additional data elements are now being consistently collected and reviewed (these may not necessarily have been collected and reviewed before). In some patient encounters, these additional data elements may have contributed to higher levels of service, and this effect could be especially apparent in EDs.

Some have suggested a potential for large-scale up-coding during EHR transitions. Others have contended that coding level increases are a consequence of a time-intensive data entry process, collection and review of additional data, and more accurate reporting of services already being provided. We are not convinced that large coding changes are attributable solely to EHR implementation, as the changes at our center have been relatively small.

Nevertheless, minor coding level changes could translate to large changes in healthcare costs when scaled nationally. Although causes may be innocuous, any increases in national healthcare costs are concerning in our time of limited budgets and scrutinized healthcare utilization.

This study had its limitations. First, including billing data from only 2 departments at a single center may limit the generalizability of findings. However, we specifically selected a GIM department and a specialty (orthopedics) department in an attempt to capture a representative sample of practices. Another limitation is that we investigated billing codes over only 2 years, around the implementation of EHRs in these departments, and therefore may have captured only short-term changes. However, as patient volumes and billing are subject to many factors, including staffing changes (eg, new partners, new hires, retirements, other departures), we attempted to limit the effect of confounding variables by limiting the period of analysis.

Overall, changes in patient volume and coded level of service during EHR implementation at our institution were relatively small. Although the trend toward higher billing coding levels was statistically significant, these 0.2% and 1.7% increases in level 5 coding hardly deserve the negative attention from lay media. These small increases are unlikely caused by intentional up-coding, and more likely reflect better documentation of an already high level of care. We hope these findings allay the concern that up-coding increased dramatically with EHR implementation.

Am J Orthop. 2017;46(3):E172-E176. Copyright Frontline Medical Communications Inc. 2017. All rights reserved.

Take-Home Points

• Adolescent growth spurt, height-to-weight ratio, and perioperative weight loss are risk factors associated with SMA syndrome following pediatric spine surgery.
• Must recognize nonspecific symptoms such as abdominal pain, tenderness, distention, bilious or projectile vomiting, hypoactive bowel sounds, and anorexia postoperatively.
• Complications of SMA syndrome can potentially lead to aspiration pneumonia, acute gastric rupture, or cardiovascular collapse and death.

Superior mesenteric artery (SMA) syndrome resulting from surgical treatment of scoliosis has been recognized in the medical literature since 1752.¹ Throughout the literature, SMA syndrome variably has been referred to as cast syndrome, Wilkie syndrome, arteriomesenteric duodenal obstruction, and chronic duodenal ileus.² We now recognize numerous etiologies of SMA syndrome, as several sources can externally compress the duodenum. Classic acute symptoms of bowel obstruction include bilious vomiting, nausea, and epigastric pain. Chronic manifestations of SMA syndrome may include weight loss and decreased appetite. Our literature review revealed that adolescent growth spurt, height-to-weight ratio, and perioperative weight loss are risk factors associated with SMA syndrome after pediatric spine surgery.

We report the case of a 14-year-old boy who developed SMA syndrome after undergoing scoliosis surgery. The patient and his mother provided written informed consent for print and electronic publication of this case report.

Case Report

A 14-year-old boy with a history of idiopathic scoliosis presented to Cohen Children’s Hospital (Long Island Jewish Medical Center) with bilious vomiting that had persisted for 7 days after posterior T9–L4 fusion with instrumentation.

Discussion

SMA syndrome is attributed to the anatomical orientation of the third part of the duodenum, which passes between the aorta and the SMA (Figure 4).

Adolescents are particularly vulnerable to this condition. Faster adolescent bone growth relative to visceral growth is accompanied by a decrease in SMA angle.³ Occasionally, body casts are used after surgery to immobilize the vertebrae and augment healing. Cast syndrome occurs when pressure from a body cast causes a bowel obstruction secondary to spinal hyperextension and amplified spinal lordosis.² This finding, dating to the 19th century, was reported by Willet⁴ when a patient died 48 hours after application of a body cast. In 1950, the term cast syndrome was coined after a motorcyclist’s injuries were treated with a hip spica cast and the patient died of cardiovascular collapse secondary to persistent vomiting.⁵

Table 1 summarizes various evaluation, diagnosis, and treatment algorithms designed to optimize nutrition and weight in patients developing signs and symptoms of SMA syndrome after posterior spinal instrumentation and fusion for adolescent idiopathic scoliosis (AIS).

Take-Home Points

• Adolescent growth spurt, height-to-weight ratio, and perioperative weight loss are risk factors associated with SMA syndrome following pediatric spine surgery.
• Must recognize nonspecific symptoms such as abdominal pain, tenderness, distention, bilious or projectile vomiting, hypoactive bowel sounds, and anorexia postoperatively.
• Complications of SMA syndrome can potentially lead to aspiration pneumonia, acute gastric rupture, or cardiovascular collapse and death.

Superior mesenteric artery (SMA) syndrome resulting from surgical treatment of scoliosis has been recognized in the medical literature since 1752.¹ Throughout the literature, SMA syndrome variably has been referred to as cast syndrome, Wilkie syndrome, arteriomesenteric duodenal obstruction, and chronic duodenal ileus.² We now recognize numerous etiologies of SMA syndrome, as several sources can externally compress the duodenum. Classic acute symptoms of bowel obstruction include bilious vomiting, nausea, and epigastric pain. Chronic manifestations of SMA syndrome may include weight loss and decreased appetite. Our literature review revealed that adolescent growth spurt, height-to-weight ratio, and perioperative weight loss are risk factors associated with SMA syndrome after pediatric spine surgery.

We report the case of a 14-year-old boy who developed SMA syndrome after undergoing scoliosis surgery. The patient and his mother provided written informed consent for print and electronic publication of this case report.

Case Report

A 14-year-old boy with a history of idiopathic scoliosis presented to Cohen Children’s Hospital (Long Island Jewish Medical Center) with bilious vomiting that had persisted for 7 days after posterior T9–L4 fusion with instrumentation.

Discussion

SMA syndrome is attributed to the anatomical orientation of the third part of the duodenum, which passes between the aorta and the SMA (Figure 4).

Adolescents are particularly vulnerable to this condition. Faster adolescent bone growth relative to visceral growth is accompanied by a decrease in SMA angle.³ Occasionally, body casts are used after surgery to immobilize the vertebrae and augment healing. Cast syndrome occurs when pressure from a body cast causes a bowel obstruction secondary to spinal hyperextension and amplified spinal lordosis.² This finding, dating to the 19th century, was reported by Willet⁴ when a patient died 48 hours after application of a body cast. In 1950, the term cast syndrome was coined after a motorcyclist’s injuries were treated with a hip spica cast and the patient died of cardiovascular collapse secondary to persistent vomiting.⁵

Table 1 summarizes various evaluation, diagnosis, and treatment algorithms designed to optimize nutrition and weight in patients developing signs and symptoms of SMA syndrome after posterior spinal instrumentation and fusion for adolescent idiopathic scoliosis (AIS).

Take-Home Points

• Adolescent growth spurt, height-to-weight ratio, and perioperative weight loss are risk factors associated with SMA syndrome following pediatric spine surgery.
• Must recognize nonspecific symptoms such as abdominal pain, tenderness, distention, bilious or projectile vomiting, hypoactive bowel sounds, and anorexia postoperatively.
• Complications of SMA syndrome can potentially lead to aspiration pneumonia, acute gastric rupture, or cardiovascular collapse and death.

Superior mesenteric artery (SMA) syndrome resulting from surgical treatment of scoliosis has been recognized in the medical literature since 1752.¹ Throughout the literature, SMA syndrome variably has been referred to as cast syndrome, Wilkie syndrome, arteriomesenteric duodenal obstruction, and chronic duodenal ileus.² We now recognize numerous etiologies of SMA syndrome, as several sources can externally compress the duodenum. Classic acute symptoms of bowel obstruction include bilious vomiting, nausea, and epigastric pain. Chronic manifestations of SMA syndrome may include weight loss and decreased appetite. Our literature review revealed that adolescent growth spurt, height-to-weight ratio, and perioperative weight loss are risk factors associated with SMA syndrome after pediatric spine surgery.

We report the case of a 14-year-old boy who developed SMA syndrome after undergoing scoliosis surgery. The patient and his mother provided written informed consent for print and electronic publication of this case report.

Case Report

A 14-year-old boy with a history of idiopathic scoliosis presented to Cohen Children’s Hospital (Long Island Jewish Medical Center) with bilious vomiting that had persisted for 7 days after posterior T9–L4 fusion with instrumentation.

Discussion

SMA syndrome is attributed to the anatomical orientation of the third part of the duodenum, which passes between the aorta and the SMA (Figure 4).

Adolescents are particularly vulnerable to this condition. Faster adolescent bone growth relative to visceral growth is accompanied by a decrease in SMA angle.³ Occasionally, body casts are used after surgery to immobilize the vertebrae and augment healing. Cast syndrome occurs when pressure from a body cast causes a bowel obstruction secondary to spinal hyperextension and amplified spinal lordosis.² This finding, dating to the 19th century, was reported by Willet⁴ when a patient died 48 hours after application of a body cast. In 1950, the term cast syndrome was coined after a motorcyclist’s injuries were treated with a hip spica cast and the patient died of cardiovascular collapse secondary to persistent vomiting.⁵

Table 1 summarizes various evaluation, diagnosis, and treatment algorithms designed to optimize nutrition and weight in patients developing signs and symptoms of SMA syndrome after posterior spinal instrumentation and fusion for adolescent idiopathic scoliosis (AIS).