Route of Hysterectomy Influence and Teaching Hospital Status

Abstract

OBJECTIVE The purpose of this study was to evaluate whether teaching hospitals are more likely to perform abdominal (laparotomy) compared with nonabdominal (vaginal, laparoscopic) hysterectomies for benign indications. METHODS We conducted a retrospective, observational study comparing abdominal and nonabdominal inpatient hysterectomies performed in Illinois between 2000 and 2005 using the COMPdata database. Obvious cancer, prolapse, or indicated-abdominal surgeries (infection and pregnancy-related cases) were excluded. The final analysis included 94,599 cases. Diagnoses and patient demographics were analyzed from the database and hospitals' teaching status, as determined by telephone interviews. The relationship between route of hysterectomy and teaching hospital status was modeled using multivariable logistic regression with a P value cutoff of less than 0.05. RESULTS Eight-two percent of hysterectomies performed at teaching hospitals were performed abdominally compared with 77% at nonteaching hospitals. After adjusting for age and diagnoses, teaching hospitals were less likely to perform hysterectomy by abdominal approach (odds ratio 0.69, 95% confidence interval 0.49-0.97, P<.035). Clinical variables associated positively with vaginal hysterectomies included primary diagnoses of menstrual disorders, other female genital disorders, and menopausal disorders. Complication rate did not differ by teaching status. Laparoscopic hysterectomy, even after adjustment for confounders, was associated with fewer complications compared with both abdominal and vaginal routes. CONCLUSION The route of hysterectomy is only minimally influenced by teaching hospital status. These findings are important for clinician-educators responsible for teaching the nation's next generation of gynecologic surgeons. Strategies to overcome presumed physician-level factors are needed to optimize patient outcomes through appropriate use of nonlaparotomy surgery. LEVEL OF EVIDENCE III.