An alternative statistical approach for predicting prolonged catheterization after Burch colposuspension during reconstructive pelvic surgery.

Our objective was to use an alternative statistical approach to identify clinical and urodynamic predictors of prolonged catheterization following Burch colposuspension. Seventy women with genuine stress incontinence underwent Burch colposuspension with suprapubic catheter placement at Rush Presbyterian-St. Luke's Medical Center from 1 July 1992 to 1 October 1993. Patient charts were retrospectively reviewed to extract pertinent variables from their history, examination and preoperative urodynamic evaluation. The day of suprapubic catheter removal was considered the end-point 'event' for the purposes of survival analysis. This statistical model allowed us to identify preoperative clinical parameters important in determining the percentage of patients requiring catheters as a function of time. The need for defining prolonged postoperative catheterization was eliminated. Aging (P=0.01), increasing maximal urethral pressures (P=0.02) and menopausal status (P=0.02) were important in determining the percentage of patients requiring catheters as a function of time. Data from our preoperative voiding studies were not predictive of prolonged catheterization following Burch colposuspension.