Abstract
IMPORTANCE: Surgical mortality remains a critical public health issue in resource-limited settings. Hospital ownership type may influence surgical outcomes, yet this relationship is not well understood in East, Central, and Southern Africa, where a diverse mix of public, private, and faith-based hospitals provides care. OBJECTIVE: To determine whether hospital ownership type (public, private, or faith-based) is associated with differences in surgical mortality rates in East, Central, and Southern Africa. DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study analyzed operative cases recorded by 214 general surgery trainees enrolled in surgical training programs at 85 public, private, and faith-based hospitals in East, Central, and Southern Africa from January 1, 2005, to December 31, 2020. Cases were documented in mandatory operative logbooks. Reported mortalities were analyzed using multilevel logistic regression to account for clustering by trainee while controlling for age category, emergency status, case complexity, specialty type, country Human Development Index, trainee postgraduate year, and self-reported autonomy. Data were analyzed in July 2024. EXPOSURES: Hospital ownership type categorized as public, private, or faith-based. MAIN OUTCOMES AND MEASURES: The primary outcome was perioperative mortality, defined as in-hospital, all-cause mortality before discharge. Mortality rates were compared across hospital types. RESULTS: Of the 106 106 operative cases analyzed, 48 474 (45.7%) were performed in public, 3507 (3.3%) in private, and 54 125 (51.0%) in faith-based hospitals. Patients' median (IQR) age was 34 (20-51) years, and 61.5% were male. The overall perioperative mortality rate was 1.6% (95% CI, 1.5%-1.6%). Among major cases (mortality, 1138 of 53 718 [2.1%; 95% CI, 2.0%-2.2%]), faith-based hospitals had 57% lower mortality (518 of 35 370 [1.5%; 95% CI, 1.3%-1.6%]) than public hospitals (589 of 17 223 [3.4%; 95% CI, 3.2%-3.7%]) and 47% lower mortality than private hospitals (31 of 1125 [2.8%; 95% CI, 1.9%-3.9%]). Mixed-effects logistic regression revealed that faith-based hospitals had lower odds of mortality compared with public hospitals (odds ratio, 0.67; 95% CI, 0.51-0.86; P = .002) and private hospitals (odds ratio, 0.57; 95% CI, 0.34-0.95; P = .03). CONCLUSIONS: In this study, faith-based hospitals in East, Central, and Southern Africa were associated with significantly lower surgical mortality rates compared with public and private hospitals. These findings suggest that practices and resources in faith-based hospitals contribute to improved surgical outcomes, warranting further investigation to inform health care policy and improve surgical care and outcomes in the region.