Relationship between patient safety indicator events and hospital location for inpatient hysterectomy.

Abstract

Previous studies suggest surgical quality outcomes are similar between rural and urban hospitals, but data about gynecology in rural hospitals is sparse. This was a retrospective cohort study utilizing the National Inpatient Sample database from the Agency of Healthcare Research and Quality. Patients who underwent benign hysterectomy for non-prolapse indications between 2012-2016 were identified using ICD-9 and 10 codes. Patients were stratified into rural or urban non-teaching groups; urban teaching hospitals were the referent group. The primary outcome was the rate of patient safety indicator (PSI) events. PSI events were identified using ICD-9 and 10 codes. Statistical analysis was performed using analysis of variance and uni- and multivariate Poisson regressions. 154,810 patients met all inclusion criteria. The cumulative rate of PSI events was 11.9% at rural hospitals, 13.9% at urban non-teaching hospitals and 16.9% at urban teaching hospitals, P<0.001. The most common PSI events were postoperative metabolic derangement, hemorrhage, and accidental puncture. The rate of transfusion was highest in urban teaching hospitals (6.7%) and similar for rural (5.1%) and urban non-teaching hospitals (5.5%), P<0.001. The rate of genitourinary tract injury was between 1.4-1.6%, and similar across sites, P=0.89. After adjusting for confounders, the risk of PSI events was similar across locations. The risk of transfusion was lower at rural hospitals (aRR=0.84, 95% CI: 0.74-0.94). Hysterectomy performed at rural hospitals, typically thought of as having low surgical volume compared to urban hospitals, is associated with similar risk of PSI events and lower risk of transfusion.