Objective: To evaluate the impact of perioperative variables on length of postoperative catheterization.
Study design: A multicenter, prospective, cohort study of women undergoing pelvic reconstructive and/or incontinence surgery was performed. Perioperative variables associated with length of catheterization and prolonged catheterization (catheterization >30 days) were analyzed. Univariate logistic regression was used identify variables associated with urinary retention. Multivariate logistic regression analysis was performed on variables identified by univariate analysis to construct the best model predicting prolonged postoperative catheterization.
Results: Catheterization data were available for 408 patients. Mean catheterization length was 11 +11 days. Thirty-four patients required prolonged catheterization. Logistic regression analysis selected 3 variables as predictors of prolonged catheterization: performance of pubovaginal sling (OR 5.44), performance of vaginal apex suspension (OR 2.32) and preoperative presence of grade 3/4 vaginal apex descent (OR 2.75) (all p < or =0.05). The probability of prolonged catheterization occurring in the absence of any of the predictors was 2% and increased to 5-11% if 1 predictor was present. When all 3 were present, the probability of prolonged catheter use increased to 45%.
Conclusion: The performance of a pubovaginal sling and of a vaginal apex suspension or the preoperative presence of grade 3/4 vaginal apex descent are associated with prolonged postoperative catheterization.