Vaginal Hysterectomy in a Large Health Maintenance Organization: Retrospective Application of a Clinical Algorithm

Abstract

Study objective: The primary objective was to describe practice patterns of benign hysterectomy within a large community health maintenance organization (HMO) and evaluate clinical and surgeon characteristics associated with the performance of vaginal hysterectomy (VH). Secondary objectives were to retrospectively apply a VH algorithm to determine how our practice patterns conform, and compare rates of perioperative complications among patients who did and did not meet the algorithm. Patient and surgeon characteristics, and perioperative complications, were compared between patients who underwent VH and did or did not meet the VH algorithm.

Design: Retrospective cohort study.

Setting: Large community HMO.

Patients: Women undergoing benign hysterectomy.

Interventions: None.

Measurements and main results: Route of hysterectomy, patient and surgeon characteristics, perioperative complications. One hundred and thirty-one of 984 (13.3%) benign hysterectomies from January 1, 2013 to June 30, 2015 were vaginal. Patients who were vaginally parous, Hispanic, had normal preoperative uterine size and documentation of uterine descent were more likely to have VH (all p <.05). High-volume surgeons performed 18.8% of their hysterectomies vaginally, as compared to low-volume surgeons who performed 11.4% of their hysterectomies vaginally (p <.01). VH were more likely to be performed by surgeons with longer practice durations than non-vaginal hysterectomies (16.3 vs 12.2 years, p <.01). Seventy-five percent of patients who met the VH algorithm underwent non-vaginal hysterectomy and they had longer operative durations and higher rates of postoperative complications compared to patients who underwent VH. Conversely, patients who underwent VH despite not meeting the VH algorithm did not have significantly different rates of perioperative complications or blood loss than patients who met the VH algorithm.

Conclusion: Seventy-five percent of patients deemed appropriate for VH by our algorithm underwent non-vaginal hysterectomy and had more postoperative complications and longer operative durations. Our data suggest that surgeon characteristics, including surgical volume and duration of practice, may explain some of this performance gap. These findings contribute additional insight into current practice patterns and describe clinical factors that should be included in VH algorithms.

Keywords: Decision-directed algorithm; Hysterectomy algorithm; Minimally-invasive hysterectomy; Surgeon characteristics; Vaginal hysterectomy.