Related Experiment Video
Updated: Apr 17, 2026
Vessel-sparing Excision and Primary Anastomosis
Published on: January 7, 2019
National estimates of predictors of outcomes for emergency general surgery
Adil Aijaz Shah1, Adil Hussain Haider, Cheryl K Zogg
1From the Department of Surgery (A.A.S., C.K.Z., D.A.S., E.R.H., E.B.S., C.G.V., D.T.E.), The Johns Hopkins University School of Medicine, Baltimore, Maryland; Center for Surgery and Public Health (A.H.H., C.K.Z.), Harvard Medical School and Harvard School of Public Health; and Department of Surgery (A.H.H., C.K.Z.), Brigham and Women's Hospital, Boston, Massachusetts; Department of Surgery (S.N.Z.), Howard University School of Medicine, Washington, District of Columbia; Institute for Health Care Research and Improvement (S.S.), Baylor Health Care System, Dallas, Texas; and Department of Surgery (H.Z.), Aga Khan University, Karachi, Pakistan.
Background:
Identifying predictors of mortality and surgical complications has led to outcome improvements for a variety of surgical conditions. However, similar work has yet to be done for factors affecting outcomes of emergency general surgery (EGS). The objective of this study was to determine the predictors of in-hospital complications and mortality among EGS patients.
Methods:
The Nationwide Inpatient Sample (2003-2011) was queried for patients with conditions encompassing EGS as determined by the American Association for Surgery of Trauma, categorizing them into predefined EGS groups using DRG International Classification of Diseases-9th Rev.-Clinical Modification codes. Primary outcomes considered included incidence of a major complication (pneumonia, pulmonary emboli, urinary tract infections, myocardial infarctions, sepsis, or septic shock) and in-hospital mortality. Separate multivariate logistic regression analyses for complications and mortality were performed to identify risk factors of either outcome from the following domains: patient demographics (age, sex, insurance type, race, and income quartile), comorbidities, and hospital characteristics (location, teaching status, and bed size).
Results:
This study included 6,712,151 discharge records, weighted to represent 32,910,446 visits for EGS conditions. Mean age was 58.50 (19.74) years; slightly more than half (54.66%) were female. Uninsured patients were more likely to die (odds ratio,1.25; 95% confidence interval, 1.20-1.30), whereas patients in the highest income quartile had the least likelihood of mortality (odds ratio, 0.86; 95% confidence interval, 0.84-0.87). Old age was an independent predictor of mortality for all EGS subdiagnoses. The overall mortality rate was 1.76%; the overall complication rate was 10.03%. Of the patients who died, 62% experienced at least one major complication. Patients requiring resuscitation had the highest likelihood of mortality followed by patients with vascular disease and hepatic disease.
Conclusion:
Death patterns of EGS patients were discerned using an administrative data set. Understanding patterns of mortality and complications derived from studies such as this could improve hospital benchmarking for EGS, akin to trauma surgery's previous success.
Level Of Evidence:
Prognostic and epidemiologic study, level III.
