Emergency general surgery in geriatric patients: How should we evaluate hospital experience?

INTRODUCTION As the aging American population poses unique challenges to acute care services, we determined if either hospital proportion or annual volume of geriatric patients undergoing emergency general surgery (EGS) procedures is associated with outcomes. METHODS Using criteria from the American Association of the Surgery of Trauma, we identified five EGS procedures in the 2012-2015 Nationwide Inpatient Sample common in geriatric patients (65+ years). We defined hospital proportion as the fraction of geriatric EGS patients divided by all EGS patients, where volume was the raw number of geriatric EGS patients. We then divided hospitals into quartiles both by proportion and then by volume of geriatric patients. Multivariable logistic regressions compared four outcomes between these quartiles: mortality, complications, failure to rescue (FTR; death after a complication), and extended length of stay (LOS; procedure-specific top decile of patients). RESULTS We identified 25,084 complex EGS procedures in geriatric patients at 3,528 hospitals (mortality, 10.6%; complications, 30.5%; FTR, 27.7%; extended LOS, 9.1%). The median hospital proportion of geriatric patients among EGS procedures was 42.8% (interquartile range, 33.3-52.2%), whereas the median hospital geriatric EGS volume after nationwide weighting was 40 per year (interquartile range, 20-70/year). After adjustment, the lowest hospital proportion quartile relative to the highest was associated with adverse outcomes: mortality (odds ratio, 1.21 [95% confidence interval, 1.03-1.44]), complications (1.16 [1.05-1.29]), FTR (1.32 [1.08-1.63]), and extended LOS (1.30 [1.12-1.50]). The lowest volume quartile relative to the highest was not associated with adverse outcomes. As the hospital proportion of geriatric patients increased by 10%, the odds of all adverse outcomes decreased: mortality by 7%, complications by 4%, FTR by 9%, and extended LOS by 8%. CONCLUSION When accounting for both, hospital proportion of geriatric EGS patients but not hospital volume is associated with postoperative outcomes, having important implications for quality improvement initiatives, benchmarking endeavors, and health services research. LEVEL OF EVIDENCE Care management, level IV; prognostic, level III.