Urinary tract infection after clean-contaminated pelvic surgery: a retrospective cohort study and prediction model.

Abstract

Postoperative urinary tract infection (UTI) leads to increased patient morbidity and health care costs. A prediction model may identify patients at highest risk for UTI development. Our primary objective was to determine the rate of UTI in the first 6weeks after benign gynecologic surgery. Our secondary objective was to identify risk factors and build a predictive model for postoperative UTI. We reviewed 310 patient records, which represent all patients who underwent clean-contaminated surgery at a tertiary center (2016-2017). UTI was defined as positive urine culture (> 100,000,000CFU/l) in a symptomatic patient. Pre-, intra- and postoperative variables were collected. The relation between these variables and UTI was assessed through logistic regression. A clinical prediction model was built. Patients' mean age was 58.5years and mean body mass index was 27.5kg/m2. Most were inpatients (65.8%) and 269 had urogynecologic procedures, with the remainder undergoing pelvic surgery for other indications. The most common operation was vaginal reconstruction for prolapse (59.7%), associated with concomitant synthetic midurethral sling in 1/3 cases. Forty patients (12.9%) developed UTI. Multivariate prediction modeling showed increasing age (OR 1.33, CI 1.01-1.75), increasing number of procedures (OR 1.42, CI 1.14-1.78) and prolonged voiding dysfunction (OR 3.78, CI 1.66-8.60) to be significant UTI predictors. Urinary tract infection in the first 6weeks after complex pelvic surgery is common. Our prediction model identifies that patients who are older women, have prolonged voiding dysfunction and have a greater number of concomitant pelvic floor surgeries have higher risk of postoperative UTI.