This cohort study examines whether Medicare beneficiaries who undergo emergency general surgery experience similar 1-year outcomes compared with patients admitted with acute medical conditions.
Key Points
Question
Do older adults who undergo emergency general surgery have similar rates of death, postdischarge hospital use, and days at home compared with those with acute medical conditions, such as pneumonia, heart failure, and acute myocardial infarction?
Findings
In this cohort study of 481 417 matched pairs of older Medicare beneficiaries, emergency general surgery was associated with similar odds of 1-year mortality, approximately 30% lower rates of hospital use during 1 year, and similar number of days at home compared with acute medical admission. However, in both groups, 1-year mortality was 29.7% or higher; more than 56% had a hospital encounter in the year after discharge and at least 56 days away from home.
Meaning
The findings suggest that similar to hospitalizations for pneumonia, heart failure, and acute myocardial infarction, emergency general surgery in older adults may be associated with high mortality and health care utilization after discharge and should also be considered a target for policies aimed at quality improvement.
Abstract
Importance
Emergency general surgery (EGS) represents 11% of hospitalizations, and almost half of these hospitalized patients are older adults. Older adults have high rates of mortality and readmissions after EGS, yet little is known as to how these outcomes compare with acute medical conditions that have been targets for quality improvement.
Objective
To examine whether Medicare beneficiaries who undergo EGS experience similar 1-year outcomes compared with patients admitted with acute medical conditions.
Design, Setting, and Participants
This population-based, retrospective cohort study using Medicare claims data from January 1, 2008, to December 31, 2014, included adults 65 years or older with at least 1 year of Medicare claims who had urgent or emergency admissions for 1 of the 5 highest-burden EGS procedures (partial colectomy, small-bowel resection, peptic ulcer disease surgery, lysis of adhesions, or laparotomy) or a primary diagnosis of an acute medical condition (pneumonia, heart failure, or acute myocardial infarction). Patients undergoing EGS and those with acute medical conditions were matched 1:1 in a 2-step algorithm: (1) exact match by hospital or (2) propensity score match with age, sex, race/ethnicity, Charlson Comorbidity Index, individual comorbid conditions, claims-based frailty index, year of admission, and any intensive care unit stay. Data analysis was performed from July 16, 2018, to November 13, 2019.
Exposures
Partial colectomy, small-bowel resection, peptic ulcer disease surgery, lysis of adhesions, or laparotomy or a primary diagnosis pneumonia, heart failure, or acute myocardial infarction.
Main Outcomes and Measures
One-year mortality, postdischarge health care utilization (emergency department visit, additional hospitalization, intensive care unit stay, or total hospital encounters), and days at home during 1 year.
Results
A total of 481 417 matched pairs (mean [SD] age, 78.9 [7.8] years; 272 482 [56.6%] female) with adequate covariate balance were included in the study. Patients undergoing EGS experienced higher 30-day mortality (60 683 [12.6%] vs 56 713 [11.8%], P < .001) yet lower 1-year mortality (142 846 [29.7%] vs 158 385 [32.9%], P < .001) compared with medical patients. Among 409 363 pairs who survived discharge, medical patients experienced higher rates of total hospital encounters in the year after discharge (4 vs 3 per person-year; incidence rate ratio, 1.31; 95% CI, 1.30-1.32) but had similar mean days at home compared with patients undergoing EGS (293 vs 309 days; incident rate ratio, 1.004; 95% CI, 1.004-1.004).
Conclusions and Relevance
In this study, older patients undergoing EGS had similarly high 1-year rates of mortality, hospital use, and days away from home as acutely ill medical patients. These findings suggest that EGS should also be targeted for national quality improvement programs.
Introduction
Directing quality improvement efforts for high-need, seriously ill, older patients is an urgent policy focus for patients, payers, and policy makers.1,2 Older patients undergoing emergency general surgery (EGS) are a rapidly increasing and seriously ill population with poor surgical outcomes.3 Although heterogeneous in origin, EGS procedures are typically unexpected, requiring urgent medical attention with limited time for preoperative optimization. Almost 40% of patients hospitalized with EGS conditions are older adults who also have high rates of comorbidity, frailty, cognitive impairment, and social vulnerability that contribute to especially poor outcomes.4,5,6,7,8,9 Up to 40% die within 30 days, complication rates exceed 50%, up to 50% are discharged to a facility, and nearly one-fifth are readmitted within 30 days.4,10,11,12,13,14 Given these poor outcomes, programs to improve the quality of EGS care have gained attention within the surgical community; however, quality initiatives led by payers such as the Centers for Medicare & Medicaid Services (CMS) do not currently target patients undergoing EGS.15,16
Understanding mortality and health care utilization in other medical and surgical conditions has identified opportunities to improve quality of care and prompted development of institutional care pathways, which have reduced morbidity, mortality, and readmissions.17,18,19,20,21 Conditions such as pneumonia, heart failure (HF), and acute myocardial infarction (AMI) have been foci for national quality improvement and payment reforms in the past decade, largely driven by data showing that these conditions are common indications for acute care hospitalization and that these patients experience variation in outcomes.22,23 However, despite the increasing prevalence, high mortality, and high rates of adverse outcomes among older patients undergoing EGS, much of the prior literature includes only elective or younger surgical patients as comparators. Therefore, we took a novel approach by contextualizing EGS outcomes in light of other high-risk, acute conditions common in older adults that have been the foci of national policies to improve outcomes.24,25 Although the American College of Surgeons’ National Surgical Quality Improvement Program includes a national sample of surgical patients, outcomes beyond 30 days have not been widely explored or compared with these acute medical admissions. However, these long-term outcomes are essential to demonstrate whether older patients who undergo the highest-burden EGS procedures are similarly high users of health care.26,27
To address this gap, we used national Medicare claims data to assess whether those with acute medical conditions are similarly ill as matched patients undergoing EGS, exhibit poor outcomes, and warrant attention for national quality improvement and care reform. Our primary objectives were to (1) examine 1-year mortality, health care utilization, and days at home among older patients who undergo the highest mortality–burden EGS procedures and (2) compare these outcomes with a propensity score–matched cohort of patients admitted with acute medical conditions, including pneumonia, HF, and AMI. We hypothesized that patients who undergo EGS would have similar outcomes compared with matched patients admitted with acute medical conditions.
Methods
Data Source
This retrospective cohort study of Medicare claims data included 100% fee-for-service Medicare beneficiaries from January 1, 2008, to December 31, 2014. The Master Beneficiary Summary File, denominator file, and MedPAR files were used to identify the study population. The inpatient, outpatient, carrier claims, durable medical equipment, hospice, and vital status files were available through the secure CMS Virtual Research Data Center Research Identifiable Files to identify study variables. Data analysis was performed from July 16, 2018, to November 13, 2019. The Partners Human Research Committee approved this study and waived the need for informed consent because data were deidentified.
Study Population
Patients 65 years or older with at least 12 months of continuous enrollment in Medicare Parts A and B were included. In the EGS group, the exposure of interest was an urgent or emergency hospitalization as coded in the Medicare inpatient file with a claim for the 5 EGS procedures with the highest-mortality burden in the United States (partial colectomy, small bowel resection, repair of peptic ulcer disease, lysis of adhesions, or laparotomy) identified using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes (eTable 1 in the Supplement).5 Patients with multiple qualifying EGS procedures during the same admission were classified based on the primary procedure. Patients with multiple hospitalizations during the study period with a qualifying EGS procedure were included if the hospitalizations occurred more than 1 year apart.
In the medical group, the exposure of interest was an urgent or emergency hospitalization with a primary diagnosis of pneumonia, HF, or AMI using ICD-9-CM diagnosis codes (eTable 1 in the Supplement). These conditions are nonsurgical emergencies commonly seen in older adults admitted to acute care hospitals and have been major targets for CMS quality improvement.22,23,24,25,28 Patients with multiple qualifying diagnoses during the same admission were classified based on the primary coded diagnosis. Similarly, patients with multiple qualifying acute medical admissions were included if the subsequent admission was at least 1 year after the first admission. Patients were excluded from this group if they had any history of an EGS procedure during the study period (January 1, 2008, to December 31, 2014). The exposure period in both groups was the time from admission date to 1 year, death (if <1 year), or the end of the study period (December 31, 2014).
Variables
Patient characteristics included age, sex, and race/ethnicity. Comorbidities included the Deyo Charlson Comorbidity Index,29 individual Charlson Comorbidity Index comorbidities (pulmonary disease, myocardial infarction, congestive heart failure, cerebrovascular disease, diabetes, renal failure, or cancer), dementia, and peripheral vascular disease. All diagnoses were obtained using ICD-9-CM diagnosis codes. Admission from a skilled nursing facility or long-term care facility was determined from the MedPAR file. Frailty was calculated using a claims-based frailty index (CFI) modeled off the Rockwood Frailty Index and previously validated using Medicare data.30,31 The CFI (range of 0 to 1, with higher scores indicating worse conditions) was determined by assigning defined weights to 93 administrative codes for durable medical equipment claims, comorbid conditions, and health care facility use in the 12 months preceding an admission. After summing the weights, a cutoff score of 0.25 or greater was considered to indicate frailty,32 and individuals were considered to be nonfrail (CFI<0.25) or frail (CFI≥0.25).
Admission characteristics extracted included the year of hospitalization and any intensive care unit (ICU) stay during the admission. Hospital characteristics, including Medicare region, bed size, and teaching status, were obtained after linking the 2013 American Hospital Association Database using the hospital provider number.
Outcomes
The primary outcomes were 30-day, 90-day, 180-day, and 1-year mortality among patients in the EGS and acute medical condition groups. Secondary outcomes included hospital length of stay, discharge destination, and health care utilization (subsequent hospitalization, emergency department [ED] visit, ICU stay, and total hospital encounters) in the year after discharge among pairs who lived past the index hospitalization. We defined total hospital encounters as an ED visit and/or subsequent hospitalization to represent a patient-centered utilization outcome because patients are unaware if an ED visit will result in hospitalization when they seek hospital care. In-hospital postoperative complications (eg, pulmonary, renal, infectious, cardiac, delirium, wound related, and procedure related) were obtained in the EGS group as previously described.27,33 Finally, we determined days at home during 1 year, a patient-centric measure that is the sum of days alive from discharge to death excluding days in a hospital or facility.34,35
Propensity Matching
To account for differences between the 2 groups, each EGS admission was matched 1:1 to an acute medical admission in a 2-step matching algorithm. First, the admissions were exact matched by hospital to account for hospital-level variation in outcomes. Second, propensity scores were derived using a logistic regression model that used the following variables to estimate the likelihood of receiving EGS: patient age, sex, race/ethnicity, Charlson Comorbidity Index score, comorbid conditions (history of myocardial infarction, pulmonary disease, congestive HF, or cancer), frailty, year of hospitalization, and presence of any ICU stay. Each EGS admission was matched by propensity scores using the nearest neighbor method with a caliper of 0.1.36 Standardized differences were compared between the 2 groups for each variable to evaluate the quality of matches.37 A standardized difference of less than 0.15 was considered to indicate good quality match.38,39
Statistical Analysis
The unit of analysis was the individual admission. Characteristics between the 2 groups were compared using the χ2 test for categorical variables, a 2-tailed, unpaired t test for normally distributed continuous variables, and the Wilcoxon test for nonnormally distributed continuous variables.
Crude mortality rates were calculated to examine mortality and then compared using bivariate analysis. A survival analysis, displayed as Kaplan-Meier curves, was performed to examine 1-year survival in the 2 groups. Because the proportionality assumption was not met, a time-varying Cox proportional hazards regression was performed to compare 1-year survival, which allows the rate ratio of mortality to change over time.
Crude incidence rates were calculated to determine the rate of health care utilization per person-year among pairs who lived past the index hospitalization. A Poisson regression model was performed to compare health care utilization and days at home per person-year between groups.
All regression analyses adjusted for patient-level clustering within hospitals and accounted for competing risk of death. Because the 2 groups were matched by hospital, accounting for patient clustering at the hospital level using the theory of generalized estimating equations gives correct SEs for the 2-level matching algorithm. All analyses were conducted using SAS software, version 9.4 (SAS Institute Inc) within the CMS Virtual Research Data Center secure environment. A 2-sided P < .05 was considered to be statistically significant.
Results
A total of 481 417 matched pairs (mean [SD] age, 78.9 [7.8] years; 272 482 [56.6%] female) with adequate covariate balance were included in the study. There were 517 931 EGS- and 3 156 925 non-EGS–eligible admissions (Table 1). Propensity score matches were obtained for 481 417 EGS admissions and an equal number of medical patients, with a match rate of 93% (eFigure in the Supplement). Table 1 details propensity score–matched patient and hospitalization characteristics. The study cohort was well matched, with a standard difference less than 0.15 for all variables. Characteristics of the index admissions are given in Table 2. In the EGS group, partial colectomy was the most common procedure (161 210 [33.5%]), followed by peptic ulcer disease surgery (135 641 [28.2%]). Pneumonia was the most common acute medical admission (290 230 [60.3%]). The most frequent admission diagnoses for both groups are given in eTable 2 in the Supplement. Patients in the EGS group experienced a higher proportion of discharge to a skilled nursing facility (122 203 [25.4%] vs 98 616 [20.5%], P < .001) and in-hospital mortality (43 977 [9.1%] vs 28 502 [5.9%], P < .001).
Table 1. Patient and Hospital Characteristics in a Cohort Admitted for EGS or Acute Medical Admission Before and After Matchinga.
| Characteristic | Unmatched | Propensity Score Matched | ||||
|---|---|---|---|---|---|---|
| EGS (n = 517 931) | Medical (n = 3 156 925) | Standardized Difference | EGS (n = 481 417) | Medical (n = 481 417) | Standardized Differenceb | |
| Age, mean (SD), y | 78.9 (7.7) | 81.3 (8.2) | 0.30 | 79.2 (7.7) | 78.7 (8.0) | 0.07 |
| Female | 299 664 (57.9) | 1 688 546 (53.5) | 0.09 | 274 303 (57.0) | 272 256 (56.6) | 0.01 |
| Race/ethnicity | ||||||
| White | 454 560 (87.8) | 2 797 684 (88.6) | 0.09 | 423 713 (88.0) | 423 581 (88.0) | 0.14 |
| Black | 42 159 (8.1) | 218 218 (6.9) | 37 899 (7.9) | 38 536 (8.0) | ||
| Other | 21 212 (4.1) | 141 023 (4.5) | 19 805 (4.1) | 19 300 (4.0) | ||
| Comorbid conditions | ||||||
| Charlson Comorbidity Index score ≥2 | 224 260 (43.3) | 1 771 876 (56.1) | 0.26 | 210 990 (43.8) | 217 075 (45.1) | 0.03 |
| Pulmonary disease | 109 203 (21.1) | 1 120 174 (35.5) | 0.32 | 107 070 (22.2) | 107 576 (22.3) | 0.003 |
| History of myocardial infarction | 37 519 (7.2) | 419 925 (13.3) | 0.20 | 36 981 (7.7) | 38 485 (8.0) | 0.01 |
| Congestive heart failure | 81 766 (15.8) | 1 025 445 (32.5) | 0.40 | 81 111 (16.8) | 80 543 (16.7) | 0.003 |
| Cerebrovascular disease | 47 442 (9.2) | 391 772 (12.4) | 0.10 | 45 526 (9.5) | 44 942 (9.3) | 0.004 |
| Diabetes mellitus | 119 104 (23.0) | 989 536 (31.3) | 0.19 | 112 444 (23.4) | 128 366 (26.7) | 0.08 |
| Renal disease | 79 947 (15.4) | 715 970 (22.7) | 0.19 | 76 930 (16.0) | 79 463 (16.5) | 0.01 |
| Dementia | 9202 (1.8) | 120 662 (3.8) | 0.12 | 8950 (1.9) | 9840 (2.0) | 0.01 |
| Peripheral vascular disease | 48 153 (9.3) | 380 112 (12.0) | 0.09 | 46 437 (9.6) | 42 343 (8.8) | 0.03 |
| Cancer | ||||||
| Nonmetastatic | 83 561 (16.1) | 427 713 (13.5) | 0.07 | 74 409 (15.5) | 77 937 (16.2) | 0.02 |
| Metastatic | 22 408 (4.3) | 103 062 (3.3) | 0.06 | 19 369 (4.0) | 21 130 (4.4) | 0.02 |
| Frailty | ||||||
| Nonfrail, CFI <0.25 | 435 865 (84.2) | 2 200 457 (69.8) | 0.45 | 400 555 (83.2) | 403 259 (83.8) | –0.015 |
| Frail, CFI ≥0.25 | 82 069 (15.9) | 956 468 (30.3) | 80 862 (16.8) | 78 158 (16.2) | ||
| Admitted from facilityc | 38 267 (7.4) | 332 751 (10.5) | 0.11 | 36 622 (7.6) | 36 079 (7.5) | –0.004 |
| Year of admission | ||||||
| 2008 | 74 912 (14.5) | 439 050 (13.9) | 0.08 | 68 810 (14.3) | 67 791 (14.1) | 0.04 |
| 2009 | 70 437 (13.6) | 427 878 (13.6) | 65 324 (13.6) | 64 565 (13.4) | ||
| 2010 | 68 864 (13.3) | 428 337 (13.6) | 64 387 (13.4) | 63 609 (13.2) | ||
| 2011 | 68 131 (13.2) | 428 561 (13.6) | 63 709 (13.2) | 63 096 (13.1) | ||
| 2012 | 65 480 (12.6) | 409 837 (13.0) | 61 256 (12.7) | 61 045 (12.7) | ||
| 2013 | 63 208 (12.2) | 391 027 (12.4) | 59 044 (12.3) | 59 583 (12.4) | ||
| 2014 | 61 986 (12.0) | 364 660 (11.6) | 57 348 (11.9) | 58 906 (12.2) | ||
| 2015 | 44 916 (8.7) | 267 575 (8.5) | 41 539 (8.6) | 42 822 (8.9) | ||
| Medicare region | ||||||
| South | 199 937 (38.6) | 1 256 461 (39.8) | 0.04 | 187 555 (39.0) | 186 926 (38.8) | 0 |
| Midwest | 129 747 (25.1) | 819 905 (26.0) | 121 621 (25.3) | 122 146 (25.4) | ||
| Northeast | 100 166 (19.3) | 623 671 (19.8) | 91 998 (19.1) | 92 470 (19.2) | ||
| West | 86 338 (16.7) | 446 012 (14.1) | 80 243 (16.7) | 79 875 (16.6) | ||
| Other | 1746 (0.3) | 10 876 (0.4) | 0 | 0 | ||
| Hospital bed size | ||||||
| <150 | 98 195 (19.0) | 855 131 (27.1) | 0.20 | 86 023 (17.9) | 86 023 (17.9) | 0 |
| 150-300 | 162 039 (31.3) | 935 777 (29.6) | 151 684 (31.5) | 151 684 (31.5) | ||
| >300 | 257 700 (49.8) | 1 366 017 (43.3) | 243 710 (50.6) | 243 710 (50.6) | ||
| Teaching hospital | 84 686 (16.4) | 415 289 (13.2) | 0.09 | 78 255 (16.3) | 78 255 (16.3) | 0 |
| Any ICU stay during admission | 294 468 (56.9) | 957 975 (30.3) | –0.554 | 260 312 (54.1) | 250 615 (52.1) | –0.04 |
Abbreviations: CFI, claims-based frailty index; EGS, emergency general surgery; ICU, intensive care unit.
Data are presented as number (percentage) of patients unless otherwise indicated. Two-step matching was performed: (1) exact match by hospital and (2) propensity score match based on age, sex, race/ethnicity, Charlson Comorbidity Index score, history of myocardial infarction, pulmonary disease, congestive heart failure, cancer, CFI, year of admission, and any ICU stay during admission.
Standard difference less than 0.15 indicates acceptable propensity score matching.
Facility defined as skilled nursing facility or long-term care facility.
Table 2. Admission Characteristics in a Matched Cohort of Patients Admitted for EGS or Acute Medical Admissiona.
| Characteristic | EGS (n = 481 417) | Medical (n = 481 417) | P Value |
|---|---|---|---|
| Reason for admission | |||
| Large-bowel resection | 161 210 (33.5) | NA | NA |
| Peptic ulcer disease surgery | 135 641 (28.2) | NA | |
| Lysis of adhesions | 86 009 (17.9) | NA | |
| Small-bowel resection | 82 927 (17.2) | NA | |
| Exploratory laparotomy | 15 630 (3.2) | NA | |
| Pneumonia | NA | 290 230 (60.3) | |
| Acute myocardial infarction | NA | 118 296 (24.6) | |
| Congestive heart failure | NA | 72 891 (15.1) | |
| Length of stay, median (IQR), d | 9.0 (9.00) | 5.0 (4.00) | <.001 |
| In-hospital complications | 251 314 (52.2) | NA | NA |
| Discharge group | |||
| Home | 173 938 (36.1) | 223 441 (46.4) | <.001 |
| With services | 84 837 (17.6) | 82 476 (17.1) | |
| SNF | 122 203 (25.4) | 98 616 (20.5) | |
| LTAC | 36 767 (7.6) | 25 765 (5.4) | |
| Expired | 43 977 (9.1) | 28 502 (5.9) | |
| Hospice | 13 149 (2.7) | 16 410 (3.4) | |
| Other | 6546 (1.4) | 6207 (1.3) |
Abbreviations: EGS, emergency general surgery; ICU, intensive care unit; IQR, interquartile range; LTAC, long-term acute care hospital; NA, not applicable; SNF, skilled nursing facility.
Data are presented as number (percentage) of patients unless otherwise indicated. Two-step matching was performed: (1) exact match by hospital and (2) propensity score match based on age, sex, race/ethnicity, Charlson Comorbidity Index score, history of myocardial infarction, pulmonary disease, congestive heart failure, cancer, frailty, year of admission, and any ICU stay during admission.
Mortality Outcomes
As indicated in Table 3, a greater proportion of patients undergoing EGS died at 30 days compared with medical patients (60 683 [12.6%] vs 56 713 [11.8%], P < .001). However, patients undergoing EGS experienced slightly lower 1-year mortality compared with medical patients (142 846 [29.7%] vs 158 385 [32.9%], P < .001). When divided into subgroups, patients who underwent exploratory laparotomy had the highest 1-year mortality (8325 [53.1%]) in the EGS group and patients with HF had the highest 1-year mortality in the medical group (26 710 [37.0%]) (eTable 3 in the Supplement).
Table 3. One-Year Mortality Outcomes in a Matched Cohort of Patients Admitted for EGS or Acute Medical Admissiona.
| Mortality | No. (%) of Patients | P Value | |
|---|---|---|---|
| EGS (n = 481 417) | Medical (n = 481 417) | ||
| 30 d | 60 683 (12.6) | 56 713 (11.8) | <.001 |
| 90 d | 97 015 (20.2) | 94 461 (19.6) | <.001 |
| 180 d | 116 800 (24.3) | 121 264 (25.2) | <.001 |
| 1 y | 142 846 (29.7) | 158 385 (32.9) | <.001 |
Abbreviation: EGS, emergency general surgery.
Two-step matching was performed: (1) exact match by hospital and (2) propensity score match based on age, sex, race/ethnicity, Charlson Comorbidity Index score, history of myocardial infarction, pulmonary disease, congestive heart failure, cancer, frailty, year of admission, and any intensive care unit stay during admission.
The Kaplan-Meier survival curves for both groups are shown in the Figure. The mean (SD) 1-year survival was 290 (140) days in the EGS group and 283 (138) days in the acute medical group (P < .001). The EGS group had lower survival compared with the medical group until day 105. After time-varying Cox proportional hazards regression, the hazard of mortality for EGS was 1.004 (95% CI, 0.995-1.013) between 0 and 105 days and 0.73 (95% CI, 0.72-0.74) between 105 and 365 days compared with the medical group.
Figure. One-Year Survival After Admission for Emergency General Surgery (EGS) or Acute Medical Condition in a Propensity Score–Matched Cohort of Medicare Beneficiaries.
Health Care Utilization During 1 Year After Discharge
Among 409 363 pairs who lived after discharge, 222 281 patients (54.3%) in the EGS group and 253 916 (62.0%) in the acute medical condition group experienced at least 1 ED visit or subsequent hospitalization in the year after discharge (eTable 4 in the Supplement). A total of 87 997 (21.5%) in the EGS group and 115 392 (28.2%) in the acute medical condition group experienced an ICU stay during the 1-year postdischarge period. As indicated in Table 4, medical patients experienced higher rates per person-year of hospital encounters in the year after discharge compared with patients undergoing EGS, including total hospital encounters (4.00 vs 3.05; incidence rate ratio, 1.31; 95% CI, 1.30-1.32). Medical patients experienced fewer mean days at home during 1 year compared with patients undergoing EGS (293 vs 309 days; incidence rate ratio, 1.004; 95% CI, 1.004-1.004).
Table 4. One-Year Postdischarge Outcomes in a Matched Cohort of Patients Admitted for EGS or Acute Medical Conditiona.
| Outcome | Incidence, per Person-Year | IRR (95% CI)b | |
|---|---|---|---|
| EGS (n = 409 363) | Medical (n = 409 363) | ||
| ED visits, No. | 2.4 | 3.3 | 1.39 (1.38-1.41) |
| Subsequent hospitalizations, No. | 1.5 | 2.1 | 1.39 (1.37-1.40) |
| Total hospital encounters, No.c | 3.0 | 4.0 | 1.31 (1.30-1.32) |
| ICU stays, No. | 0.43 | 0.7 | 1.55 (1.53-1.57) |
| Time at home, mean (SD), d | 309 (113) | 293 (125) | 1.004 (1.004-1.004) |
Abbreviations: ED, emergency department; EGS, emergency general surgery; ICU, intensive care unit; IRR, incidence rate ratio.
Two-step matching was performed: (1) exact match by hospital and (2) propensity score match based on age, sex, race/ethnicity, Charlson Comorbidity Index score, history of myocardial infarction, pulmonary disease, congestive heart failure, cancer, frailty (continuous), year of admission, and any ICU stay during admission.
Adjusted for patient-level clustering. P < .05 for all outcomes (reference is the EGS group).
Hospital encounters defined as the composite of ED visit and/or subsequent hospitalization.
Discussion
In this study, we compared 1-year outcomes after EGS with those after acute medical admissions that have been targets for national payment reform and found that nearly one-third of matched patients in both groups died within 1 year. Although patients admitted for acute medical conditions experienced approximately 30% higher rates of hospital use, more than half of patients in both groups experienced at least 1 hospital encounter in the year after discharge. Although long-term outcomes after acute medical admissions for pneumonia, HF, and AMI have been well described,40,41,42,43,44 long-term outcomes among older patients after EGS are understudied.13 Our findings suggest that, similar to older patients with acute medical conditions, older patients undergoing EGS are a high-risk group of patients, high users of health care, and potential targets for national quality improvement and reform.
Traditional value-based care initiatives have focused quality improvement and financial penalties for surgical hospitalizations on the 30 days after surgery. However, more recent bundled payment pilot programs target outcomes in the 90-day postoperative period.45 Mortality rates after EGS in this analysis were similar to those of prior studies26,27: 30-day mortality was 13%, 1-year mortality was 30%, and more than half of deaths occurred after 30 days of surgery. Similarly, Rangel et al26 found that more than 50% of deaths among older patients after emergency major abdominal surgery at a single center occurred after the 30-day postoperative period. Cooper et al27 examined 1-year mortality in the Health and Retirement Study linked to Medicare claims and found that 40% of older patients who underwent emergency major abdominal surgery died in the 31 to 180 days after discharge. Unlike those studies,26,27 which are limited by small sample size, the present study used a large national cohort of Medicare beneficiaries to assess whether mortality remained a significant risk up to a year after EGS. In this context, our findings are instructive to practitioners, payers, and policy makers, who must track and manage surgical outcomes beyond 30 days. In doing so, hospital care processes that are associated with improved mortality after discharge can be identified and reduce financial penalties that hospitals may incur. With increasing data supporting the public health burden of EGS among older adults, policy makers and practitioners who develop care pathways and EGS-specific quality improvement programs should also consider the 90-day period to measure postoperative outcomes.
Because of the high burden of mortality, hospital use, and costs among older adults, hospitalizations for acute medical conditions, such as pneumonia, HF, and AMI, are major targets for major policy initiatives, such as the CMS Hospital Value-Based Purchasing Program and Hospital Readmission Prevention Program, which impose financial penalties for poor outcomes.46 Contemporary studies41,42 revealed that acute admissions for HF are associated with a 1-year mortality rate of 36%, a 1-year readmission rate of 61.5%, and a 5-year mortality rate of 75%. Despite rapid advances in heart disease management, patients with AMI continue to be at risk for cardiovascular- and noncardiovascular-related readmission after discharge.47 Among older adults with pnenumonia, the 5-year mortality rate is approximately 50% and the rate of 30-day readmissions is approximately 15%.44,48 Our data suggest that patients undergoing EGS experience high rates of mortality and hospital encounters that are comparable to those of acute medical patients. In the present study, patients undergoing EGS also experienced higher mortality within 3 months of discharge and had high rates of post–acute care health care utilization. Therefore, existing quality improvement programs designed for medical patients may be missing opportunities to improve care for older adults with acute surgical conditions associated with similarly poor outcomes in the short term. In the United Kingdom, implementation of an evidence-based care bundle for emergency laparotomy at 28 National Health Service hospitals was associated with a 1% decrease in risk-adjusted mortality and 1 fewer hospital day after 2 years of implementation.49,50 In addition, measures of in-hospital care improved significantly, including reduced time to the operating room. These models suggest that nationwide care bundles can be implemented to improve outcomes for older patients undergoing EGS if there is concerted attention directed on a policy level.
Days at home have recently emerged as a patient-centered outcome to represent the burden of hospital- and facility-based care after discharge among seriously ill patients.34,51,52 Days at home have an added benefit of being easy for patients and families to understand when describing postdischarge health trajectories. We found that days at home were feasible to measure from administrative data. Although patients undergoing EGS spent a similar amount of days at home compared with medical patients, patients who lived after discharge experienced 8 weeks in a facility or hospital, which is a long duration. A previous study53 examined days at home among surgical patients and only within the 30-day postoperative period. Our findings provide novel insight into the long-term outcomes of an urgent or emergency EGS or medical admission in older adults and contextualize the frequency and duration of postdischarge hospital- and facility-based care. Identifying care coordination targets that increase days at home could be examined in future studies to improve the patient experience of acute care hospitalizations among older adults.
Limitations
This study has limitations. We did not include patients who received nonoperative management of EGS diagnoses or other lower mortality–burden EGS procedures; thus, these findings are not generalizable to all patients undergoing EGS. However, our intention was to focus on a high mortality–burden operative cohort that would be under the care of a surgical team. Further investigation is needed to determine whether a more heterogeneous range of EGS diagnoses or procedures have similarly poor long-term outcomes among older adults. This study is limited to older Medicare patients, prohibiting generalizability of these findings to younger or uninsured populations. Furthermore, claims data lack more granular clinical data that allow for a more detailed adjustment for physiologic confounders that may affect mortality and health care utilization after discharge. In addition, administrative claims data are reliant on coding and cannot be validated against actual patient data. We attempted to represent the patient experience by including patient-oriented outcomes, such as days at home and hospital encounters, in this study. However, other patient-oriented outcomes, such as functional status, patient satisfaction with care, or patient-reported quality of life, are not captured in Medicare data. Therefore, although mortality and health care utilization are valued outcomes for patients, they do not capture the entire patient experience for older patients, which is essential to measure value-based care from the patient perspective.54
Conclusions
The findings suggest that older Medicare patients who undergo EGS experience 1-year mortality comparable to that of similar patients with acute medical admissions and experience high rates of health care utilization in the year after discharge. As such, EGS may be a target condition for national quality improvement to reduce health care utilization, identify targets for resource allocation, and improve outcomes among older patients.
eTable 1. International Classification Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) Codes Used to Identify the Study Population
eTable 2. Top Primary Admission Diagnoses in the Cohort
eTable 3. One-Year Mortality Rates by Primary Admission Procedure or Diagnosis
eTable 4. Proportion of Patients With at Least One Hospital Encounter in the Year After Discharge Among a Matched Cohort of Patients Admitted for Emergency General Surgery (EGS) or Acute Medical Admission
eFigure. Flow Diagram of Medicare Beneficiaries Admitted for Emergency General Surgery (EGS) and Acute Medical Conditions Included in Matched Analyses
References
- 1.Blumenthal D, Chernof B, Fulmer T, Lumpkin J, Selberg J. Caring for high-need, high-cost patients: an urgent priority. N Engl J Med. 2016;375(10):909-911. doi: 10.1056/NEJMp1608511 [DOI] [PubMed] [Google Scholar]
- 2.National Academies of Sciences Engineering and Medicine Roundtable on Quality Care for People With Serious Illness Implementing Quality Measures for Accountability in Community-Based Care for People With Serious Illness: Proceedings of a Workshop. Washington, DC: National Academies Press; 2018. [PubMed] [Google Scholar]
- 3.Gale SC, Shafi S, Dombrovskiy VY, Arumugam D, Crystal JS. The public health burden of emergency general surgery in the United States: a 10-year analysis of the Nationwide Inpatient Sample–2001 to 2010. J Trauma Acute Care Surg. 2014;77(2):202-208. doi: 10.1097/TA.0000000000000362 [DOI] [PubMed] [Google Scholar]
- 4.Joseph B, Zangbar B, Pandit V, et al. Emergency general surgery in the elderly: too old or too frail? J Am Coll Surg. 2016;222(5):805-813. doi: 10.1016/j.jamcollsurg.2016.01.063 [DOI] [PubMed] [Google Scholar]
- 5.Scott JW, Olufajo OA, Brat GA, et al. Use of national burden to define operative emergency general surgery. JAMA Surg. 2016;151(6):e160480. doi: 10.1001/jamasurg.2016.0480 [DOI] [PubMed] [Google Scholar]
- 6.Robinson TN, Wu DS, Pointer L, Dunn CL, Cleveland JC Jr, Moss M. Simple frailty score predicts postoperative complications across surgical specialties. Am J Surg. 2013;206(4):544-550. doi: 10.1016/j.amjsurg.2013.03.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shah AA, Haider AH, Zogg CK, et al. National estimates of predictors of outcomes for emergency general surgery. J Trauma Acute Care Surg. 2015;78(3):482-490. doi: 10.1097/TA.0000000000000555 [DOI] [PubMed] [Google Scholar]
- 8.Berlin A, Hwang F, Singh R, Pentakota SR, Chernock B, Mosenthal AC. Outcomes and palliative care utilization in patients with dementia and acute abdominal emergency: opportunities for surgical quality improvement. Surgery. 2017. [DOI] [PubMed] [Google Scholar]
- 9.Cooper Z, Lilley EJ, Bollens-Lund E, et al. High burden of palliative care needs of older adults during emergency major abdominal surgery. J Am Geriatr Soc. 2018;66(11):2072-2078. doi: 10.1111/jgs.15516 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Olufajo OA, Reznor G, Lipsitz SR, et al. Preoperative assessment of surgical risk: creation of a scoring tool to estimate 1-year mortality after emergency abdominal surgery in the elderly patient. Am J Surg. 2017;213(4):771-777.e1. doi: 10.1016/j.amjsurg.2016.08.007 [DOI] [PubMed] [Google Scholar]
- 11.Gleason LJ, Schmitt EM, Kosar CM, et al. Effect of delirium and other major complications on outcomes after elective surgery in older adults. JAMA Surg. 2015;150(12):1134-1140. doi: 10.1001/jamasurg.2015.2606 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cauley CE, Panizales MT, Reznor G, et al. Outcomes after emergency abdominal surgery in patients with advanced cancer: opportunities to reduce complications and improve palliative care. J Trauma Acute Care Surg. 2015;79(3):399-406. doi: 10.1097/TA.0000000000000764 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cooper Z, Scott JW, Rosenthal RA, Mitchell SL. Emergency major abdominal surgical procedures in older adults: a systematic review of mortality and functional outcomes. J Am Geriatr Soc. 2015;63(12):2563-2571. doi: 10.1111/jgs.13818 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Havens JM, Olufajo OA, Cooper ZR, Haider AH, Shah AA, Salim A. Defining rates and risk factors for readmissions following emergency general surgery. JAMA Surg. 2016;151(4):330-336. doi: 10.1001/jamasurg.2015.4056 [DOI] [PubMed] [Google Scholar]
- 15.Havens JM, Neiman PU, Campbell BL, Croce MA, Spain DA, Napolitano LM. The future of emergency general surgery. Ann Surg. 2019;270(2):221-222. doi: 10.1097/SLA.0000000000003183 [DOI] [PubMed] [Google Scholar]
- 16.Wandling MW, Ko CY, Bankey PE, et al. Expanding the scope of quality measurement in surgery to include nonoperative care: results from the American College of Surgeons National Surgical Quality Improvement Program emergency general surgery pilot. J Trauma Acute Care Surg. 2017;83(5):837-845. doi: 10.1097/TA.0000000000001670 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hensley BJ, Cooney RN, Hellenthal NJ, et al. ; Upstate New York Surgical Quality Initiative Collaborative . Readmissions after colectomy: the Upstate New York Surgical Quality Initiative Experience. Dis Colon Rectum. 2016;59(5):419-425. doi: 10.1097/DCR.0000000000000566 [DOI] [PubMed] [Google Scholar]
- 18.Wasfy JH, Zigler CM, Choirat C, Wang Y, Dominici F, Yeh RW. Readmission rates after passage of the hospital readmissions reduction program: a pre-post analysis. Ann Intern Med. 2017;166(5):324-331. doi: 10.7326/M16-0185 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gonzalez AA, Abdelsattar ZM, Dimick JB, Dev S, Birkmeyer JD, Ghaferi AA. Time-to-readmission and mortality after high-risk surgery. Ann Surg. 2015;262(1):53-59. doi: 10.1097/SLA.0000000000000912 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Krumholz HM, Parent EM, Tu N, et al. Readmission after hospitalization for congestive heart failure among Medicare beneficiaries. Arch Intern Med. 1997;157(1):99-104. doi: 10.1001/archinte.1997.00440220103013 [DOI] [PubMed] [Google Scholar]
- 21.Shah PM, Johnston L, Sarosiek B, et al. Reducing readmissions while shortening length of stay: the positive impact of an enhanced recovery protocol in colorectal surgery. Dis Colon Rectum. 2017;60(2):219-227. doi: 10.1097/DCR.0000000000000748 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Angraal S, Khera R, Zhou S, et al. Trends in 30-day readmission rates for medicare and non-Medicare patients in the era of the Affordable Care Act. Am J Med. 2018;131(11):1324-1331.e14. doi: 10.1016/j.amjmed.2018.06.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Suter LG, Li SX, Grady JN, et al. National patterns of risk-standardized mortality and readmission after hospitalization for acute myocardial infarction, heart failure, and pneumonia: update on publicly reported outcomes measures based on the 2013 release. J Gen Intern Med. 2014;29(10):1333-1340. doi: 10.1007/s11606-014-2862-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Joynt KE, Orav EJ, Jha AK. Mortality rates for Medicare beneficiaries admitted to critical access and non-critical access hospitals, 2002-2010. JAMA. 2013;309(13):1379-1387. doi: 10.1001/jama.2013.2366 [DOI] [PubMed] [Google Scholar]
- 25.Dharmarajan K, Hsieh AF, Kulkarni VT, et al. Trajectories of risk after hospitalization for heart failure, acute myocardial infarction, or pneumonia: retrospective cohort study. BMJ. 2015;350:h411. doi: 10.1136/bmj.h411 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rangel EL, Cooper Z, Olufajo OA, et al. Mortality after emergency surgery continues to rise after discharge in the elderly: predictors of 1-year mortality. J Trauma Acute Care Surg. 2015;79(3):349-358. doi: 10.1097/TA.0000000000000773 [DOI] [PubMed] [Google Scholar]
- 27.Cooper Z, Mitchell SL, Gorges RJ, Rosenthal RA, Lipsitz SR, Kelley AS. Predictors of mortality up to 1 year after emergency major abdominal surgery in older adults. J Am Geriatr Soc. 2015;63(12):2572-2579. doi: 10.1111/jgs.13785 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Krumholz HM, Lin Z, Keenan PS, et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587-593. doi: 10.1001/jama.2013.333 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613-619. doi: 10.1016/0895-4356(92)90133-8 [DOI] [PubMed] [Google Scholar]
- 30.Kim DH, Schneeweiss S. Measuring frailty using claims data for pharmacoepidemiologic studies of mortality in older adults: evidence and recommendations. Pharmacoepidemiol Drug Saf. 2014;23(9):891-901. doi: 10.1002/pds.3674 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kim DH, Schneeweiss S, Glynn RJ, Lipsitz LA, Rockwood K, Avorn J. Measuring frailty in Medicare data: development and validation of a claims-based frailty index. J Gerontol A Biol Sci Med Sci. 2018;73(7):980-987. doi: 10.1093/gerona/glx229 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Rockwood K, Fox RA, Stolee P, Robertson D, Beattie BL. Frailty in elderly people: an evolving concept. CMAJ. 1994;150(4):489-495. [PMC free article] [PubMed] [Google Scholar]
- 33.Massarweh NN, Legner VJ, Symons RG, McCormick WC, Flum DR. Impact of advancing age on abdominal surgical outcomes. Arch Surg. 2009;144(12):1108-1114. doi: 10.1001/archsurg.2009.204 [DOI] [PubMed] [Google Scholar]
- 34.Greene SJ, O’Brien EC, Mentz RJ, et al. Home-time after discharge among patients hospitalized with heart failure. J Am Coll Cardiol. 2018;71(23):2643-2652. doi: 10.1016/j.jacc.2018.03.517 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Xian Y, Wu J, O’Brien EC, et al. Real world effectiveness of warfarin among ischemic stroke patients with atrial fibrillation: observational analysis from Patient-Centered Research Into Outcomes Stroke Patients Prefer and Effectiveness Research (PROSPER) study. BMJ. 2015;351:h3786. doi: 10.1136/bmj.h3786 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Austin PC. The use of propensity score methods with survival or time-to-event outcomes: reporting measures of effect similar to those used in randomized experiments. Stat Med. 2014;33(7):1242-1258. doi: 10.1002/sim.5984 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28(25):3083-3107. doi: 10.1002/sim.3697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Garrido MM, Kelley AS, Paris J, et al. Methods for constructing and assessing propensity scores. Health Serv Res. 2014;49(5):1701-1720. doi: 10.1111/1475-6773.12182 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Leisman DE. Ten pearls and pitfalls of propensity scores in critical care research: a guide for clinicians and researchers. Crit Care Med. 2019;47(2):176-185. doi: 10.1097/CCM.0000000000003567 [DOI] [PubMed] [Google Scholar]
- 40.Avula HR, Leong TK, Lee KK, Sung SH, Go AS. Long-term outcomes of adults with heart failure by left ventricular systolic function status. Am J Cardiol. 2018;122(6):1008-1016. doi: 10.1016/j.amjcard.2018.05.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Shah KS, Xu H, Matsouaka RA, et al. Heart failure with preserved, borderline, and reduced ejection fraction: 5-year outcomes. J Am Coll Cardiol. 2017;70(20):2476-2486. doi: 10.1016/j.jacc.2017.08.074 [DOI] [PubMed] [Google Scholar]
- 42.Cheng RK, Cox M, Neely ML, et al. Outcomes in patients with heart failure with preserved, borderline, and reduced ejection fraction in the Medicare population. Am Heart J. 2014;168(5):721-730. doi: 10.1016/j.ahj.2014.07.008 [DOI] [PubMed] [Google Scholar]
- 43.Steinberg BA, Fang JC. Long-term outcomes of acute heart failure: where are we now? J Am Coll Cardiol. 2017;70(20):2487-2489. doi: 10.1016/j.jacc.2017.08.075 [DOI] [PubMed] [Google Scholar]
- 44.Restrepo MI, Faverio P, Anzueto A. Long-term prognosis in community-acquired pneumonia. Curr Opin Infect Dis. 2013;26(2):151-158. doi: 10.1097/QCO.0b013e32835ebc6d [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Centers for Medicare & Medicaid Services BPCI Advanced. 2018. https://innovation.cms.gov/initiatives/bpci-advanced/. Updated March 15, 2019. Accessed March 21, 2019.
- 46.Centers for Medicare & Medicaid Services Value based programs. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/Value-Based-Programs.html. Updated July 25, 2018. Accessed March 5, 2019.
- 47.Khot UN, Johnson MJ, Wiggins NB, et al. Long-term time-varying risk of readmission after acute myocardial infarction. J Am Heart Assoc. 2018;7(21):e009650. doi: 10.1161/JAHA.118.009650 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.De Alba I, Amin A. Pneumonia readmissions: risk factors and implications. Ochsner J. 2014;14(4):649-654. [PMC free article] [PubMed] [Google Scholar]
- 49.Aggarwal G, Peden CJ, Mohammed MA, et al. ; Emergency Laparotomy Collaborative . Evaluation of the collaborative use of an evidence-based care bundle in emergency laparotomy. JAMA Surg. 2019;154(5):e190145. doi: 10.1001/jamasurg.2019.0145 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Huddart S, Peden CJ, Swart M, et al. ; ELPQuiC Collaborator Group . Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg. 2015;102(1):57-66. doi: 10.1002/bjs.9658 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Lee H, Shi SM, Kim DH. Home time as a patient-centered outcome in administrative claims data. J Am Geriatr Soc. 2019;67(2):347-351. doi: 10.1111/jgs.15705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Groff AC, Colla CH, Lee TH. Days spent at home: a patient-centered goal and outcome. N Engl J Med. 2016;375(17):1610-1612. doi: 10.1056/NEJMp1607206 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Myles PS, Shulman MA, Heritier S, et al. Validation of days at home as an outcome measure after surgery: a prospective cohort study in Australia. BMJ Open. 2017;7(8):e015828. doi: 10.1136/bmjopen-2017-015828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Porter ME. What is value in health care? N Engl J Med. 2010;363(26):2477-2481. doi: 10.1056/NEJMp1011024 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. International Classification Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) Codes Used to Identify the Study Population
eTable 2. Top Primary Admission Diagnoses in the Cohort
eTable 3. One-Year Mortality Rates by Primary Admission Procedure or Diagnosis
eTable 4. Proportion of Patients With at Least One Hospital Encounter in the Year After Discharge Among a Matched Cohort of Patients Admitted for Emergency General Surgery (EGS) or Acute Medical Admission
eFigure. Flow Diagram of Medicare Beneficiaries Admitted for Emergency General Surgery (EGS) and Acute Medical Conditions Included in Matched Analyses