The Need to Consider Longer-Term Outcomes of Care

Racial/Ethnic Disparities Among Adult and Older Adult Emergency General Surgery Patients at 30, 90, and 180 Days

Cheryl K. Zogg, MSPH, MHS; Olubode A. Olufajo, MD, MPH; Wei Jiang, MS; Anna Bystricky, BA; John W. Scott, MD, MPH; Shahid Shafi, MD, MPH, FACS; Joaquim M. Havens, MD; Ali Salim, MD, FACS; Andrew J. Schoenfeld, MD, MSc; Adil H. Haider, MD, MPH, FACS


Annals of Surgery. 2017;266(1):66-40. 

In This Article

Abstract and Introduction


Objectives: Following calls from the National Institutes of Health and American College of Surgeons for "urgently needed" research, the objectives of the present study were to (1) ascertain whether differences in 30/90/180-day mortality, major morbidity, and unplanned readmissions exist among adult (18–64 yr) and older adult (≥65 yr) emergency general surgery (EGS) patients; (2) vary by diagnostic category; and (3) are explained by variations in insurance, income, teaching status, hospital EGS volume, and a hospital's proportion of minority patients.

Background: Racial/ethnic disparities have been described in in-hospital and 30-day settings. How longer-term outcomes compare—a critical consideration for the lived experience of patients—has, however, only been limitedly considered.

Methods: Survival analysis of 2007 to 2011 California State Inpatient Database using Cox proportional hazards models.

Results: A total of 737,092 adults and 552,845 older adults were included. In both cohorts, significant differences in 30/90/180-day mortality, major morbidity, and unplanned readmissions were found, pointing to persistently worse outcomes between non-Hispanic Black and White patients [180-d readmission hazard ratio (95% confidence interval):1.04 (1.03–1.06)] and paradoxically better outcomes among Hispanic adults [0.85 (0.84–0.86)] that were not encountered among Hispanic older adults [1.06 (1.04–1.07)]. Stratified results demonstrated robust morbidity and readmission trends between non-Hispanic Black and White patients for the majority of diagnostic categories, whereas variations in insurance/income/teaching status/EGS volume/proportion of minority patients all significantly altered the effect—combined accounting for up to 80% of risk-adjusted differences between racial/ethnic groups.

Conclusions: Racial/ethnic disparities exist in longer-term outcomes of EGS patients and are, in part, determined by differences in factors associated with emergency care. Efforts such as these are needed to understand the interplay of influences—both in-hospital and during the equally critical, postacute phase—that underlie disparities' occurrence among surgical patients.


In May 2015, the American College of Surgeons and National Institutes of Health established an agenda for surgical disparities research.[1] Among the priorities identified was an urgent need to evaluate disparities in longer-term outcomes of care.[1] Defined as "differential access, care, and outcomes due to factors such as race/ethnicity,"[1,2] healthcare disparities related to race/ethnicity have been widely described. They account for more than 83,000 deaths and more than $57 billion dollars in hospital costs each year,[3,4] yet despite recognition and calls for research, longer-term surgical disparities have been narrowly considered, particularly among emergency patients.[5,6]

Studies acknowledge the existence of racial/ethnic disparities within emergency general surgery (EGS).[7–11] Most have focused on outcomes experienced by patients during index hospital stays [7,8] or on subgroups of EGS patients with outcomes extending out to 30 days.[9,10] They have tended to employ complete-cohort assessments based on logistic regression. Although appropriate for shorter-term outcomes, such techniques lack the methodology needed to understand longitudinal follow-up of patients. There remains a need to consider how longer-term outcomes may differ and what factors affect their potential persistence among EGS patients for whom prolonged postacute recovery periods often extend well beyond patients' index hospital visits.[5,6,12,13] EGS patients represent an ideal population within which to examine questions related to race/ethnicity given the broad spectrum of disease burden, lack of significant lag in time between provider evaluation and operation, and disproportionate use of emergency services by racial/ethnic minority patients.[7–9,14–17]

Prior work demonstrated apparent mitigation of EGS racial/ethnic disparities within a universally insured military population.[14] The analysis of mortality, major acute care surgery–related (ACSR) morbidity, and readmission at 30/90/180 days[14] contrasted expectations based on available data[7–10] but lacked a direct comparator group. Such findings further underscore the need to consider to what extent longer-term racial/ethnic disparities occur among EGS patients in the general population. To the best of our knowledge, no study has previously examined these issues.

In light of calls by the National Institutes of Health and American College of Surgeons for research addressing disparities associated with longer-term outcomes of care[1] and to provide a meaningful comparison to the apparent mitigation of disparities within a universally insured population,[14] the objectives of the present study were to:

  1. Ascertain whether racial/ethnic disparities in 30-, 90-, and 180-day outcomes exist among a longitudinally followed cohort of adult (18–64 yr) and older adult (65 yr or older) EGS patients.

  2. Determine whether results vary among stratified subgroups of diagnostic categories.

  3. Consider possible explanations for differential outcomes related to patients' primary insurance, income, Council of Teaching Hospitals (COTH) teaching status, hospital EGS volume, and hospital's proportion of racial/ethnic minority patients managed.