Does Site of Surgery Matter With High-Volume Surgeons? A Comparison of Short-Term Urogynecologic Surgery Adverse Events Across a Large Health Care System

INTRODUCTION: Surgical care is shifting to a model where surgeons are increasingly tasked with operating at satellite hospitals to expand access to specialty services. Little is known about the impact of site of surgery on perioperative outcomes. OBJECTIVE: To determine if composite perioperative adverse events (AEs) differ by hospital location among high-volume urogynecologists performing pelvic organ prolapse (POP) surgery within a large academic health system. We hypothesized that AEs did not differ by surgical site. METHODS: This is a retrospective cohort study of POP surgeries in patients aged ≥61 years performed at a large health system by 14 urogynecologists between January 2016 and May 2023. We compared surgeries performed at the flagship women’s hospital (WH) and nearby hospitals (grouped as “central” hospitals), to satellite hospitals (greater than 25 miles from WH). All surgeries were performed by the same group of urogynecologists. Our primary outcome was a composite of intraoperative and postoperative AEs within the first 8 weeks: bowel injury, bladder injury, ureteral injury, blood loss >500 cc, conversion to open surgery, unplanned return to OR, ileus or small bowel obstruction, hematoma, blood transfusion, wound dehiscence, wound infection, venous thromboembolism or pulmonary embolism, and urinary tract infection. Secondary outcomes included emergency room visits related to surgery, readmissions, postoperative urinary retention, and independent evaluation of AEs included in the composite variable. Demographic and clinical characteristics were compared using t-tests, chi-square, and Wilcoxon rank-sum tests as appropriate using Stata software (version 16, Statacorp, Texas). Clinically significant variables with p<0.05 were then used in a multivariable logistic regression to evaluate the association to AEs. RESULTS: The cohort consisted of 2336 patients: 1934 (82.8%) surgeries were performed at central hospitals and 402 (17.2%) at satellite hospitals. Average age was 71.2 (SD 6.7) with no difference between sites (Table 1). Patients at satellite hospitals had higher average body mass index, greater median parity, and were more likely to be White and have a history of urogynecologic surgery. There were no differences in hypertension, diabetes, smoking status, history of stroke or cancer, and Charlson Comorbidity Index score. More surgeries were performed via the vaginal route at satellite hospitals, which was largely influenced by higher numbers of vaginal repairs without apical suspension. More patients were discharged on the day of surgery at central hospitals. There were higher numbers of composite AEs at central hospitals compared to satellites (26.1% vs 19.1%, p 0.01; Table 2). There were no differences in individual postoperative or intraoperative AEs, emergency room visits or readmissions. There was a higher proportion of patients with postoperative urinary retention at central hospitals. On multivariable logistic regression, surgery at a satellite and same-day discharge had decreased odds of AEs (aOR 0.69, 95% CI 0.51–0.94; aOR 0.58, 95% CI 0.43–0.78). CONCLUSIONS: In this cohort, urogynecologic surgeries performed at distant satellite hospitals were less likely to have perioperative AEs. This may be due to the increased complexity of cases performed at central hospitals. These findings highlight the need to further evaluate the safety of evolving models of care intended to increase patient access to specialty surgical care.