Reduction in Rates of Symptomatic Urinary Tract Infection After Pelvic Reconstructive Surgery: A Quality Improvement Analysis.

Abstract

Urinary tract infections contribute to high health care costs. This study aimed to determine if a combination of interventions was successful at reducing the rate of postoperative symptomatic urinary tract infections (SUTIs) in a female pelvic medicine and reconstructive surgery (FPMRS) practice. Observational, retrospective quality improvement analysis looking at the rate of postoperative SUTI within 30 days of surgery in women who underwent gynecologic surgery performed by an FPMRS surgeon from October 2015 to October 2019. Symptomatic urinary tract infection was defined by symptoms and urinalysis, positive urine culture, or treatment for cystitis or urethritis within 30 days of surgery. Interventions were implemented between 2015 and 2016: perioperative cranberry use, intraoperative protocols for catheterization, and postoperative protocols for urinary retention management. In 2018, we added metronidazole to cefazolin for antibiotic prophylaxis. We developed a multivariable logistic regression to determine if postoperative SUTI rates decreased over the study period with adjustment for clustering by surgeons, patient factors, and surgery type. Of 2,389 procedures performed, 284 (11.8%) involved patients who had an SUTI within 30 days of surgery. The annual infection rate decreased 50% (year 1, 14.7%; year 4, 7.3%). Adjusting for age, race, body mass index, length of stay, surgery type, and surgeon, the odds of SUTI decreased 19% each year (odds ratio [OR], 0.81; 95% confidence interval [CI], 0.72-0.91; P < 0.001). Compared with women who had other gynecologic procedures, those who underwent vaginal prolapse surgery with or without incontinence procedures (OR, 2.75; 95% CI, 1.35-5.54; P = 0.01) or incontinence surgery alone (OR, 2.65; 95% CI, 1.25-5.62; P = 0.01) were more likely to have an SUTI. Combining interventions can be highly effective in reducing postoperative SUTI rates.