LAS VEGAS – A prediction model of seven combined risk factors provides a validated, individualized way to determine a patient’s risk for developing de novo stress urinary incontinence following pelvic prolapse surgery, a novel study showed.
"In women without stress urinary incontinence symptoms, prolapse surgery may cause de novo SUI in 16%-51% of patients," Dr. J. Eric Jelovsek said at the annual meeting of the American Urogynecologic Society.
Dr. J. Eric Jelovsek
"Recent studies have demonstrated effective prevention strategies, including prophylactic incontinence surgery such as concomitant midurethral sling or Burch urethropexy, and provided refined estimates of the average patient’s risk. While these studies have advanced our knowledge of the overall prevalence of de novo SUI, the risk prediction for a specific patient varies based on individual characteristics. A prediction model that more accurately predicts an individual’s risk may further help customize the shared decision-making process that occurs between a patient and her physician when planning for a concomitant continence operation," Dr. Jelovsek said.
In order to develop and validate a model to predict an individual’s risk of de novo SUI within 12 months of prolapse surgery, Dr. Jelovsek and his associates with the Eunice Kennedy Shriver National Institute of Child Health and Human Development’s Pelvic Floor Disorders Network used data from the OPUS (Outcomes Following Vaginal Prolapse Repair and Midurethral Sling) trial.
To externally validate the model, they also used data from the CARE (Colpopexy and Urinary Reduction Efforts) trial. The investigators identified 12 original preoperative patient and test characteristics commonly used to predict the risk of de novo SUI following surgery from the available data sets. They were increased age, white race, higher vaginal parity, higher body mass index (BMI), current smoker, current diagnosis of diabetes, strenuous physical activity, baseline urgency urinary incontinence symptoms, higher preoperative Pelvic Organ Prolapse Quantification (POP-Q) stage, higher POP-Q point Aa measure, positive preoperative prolapse reduction stress test, and performance of a concomitant retropubic midurethral sling.
"We hypothesized that all risk factors except for the performance of a concomitant [retropubic midurethral sling] would increase the risk of de novo SUI," said Dr. Jelovsek, who is director of the Cleveland Clinic Multidisciplinary Simulation Center. The outcome of the prediction model was defined as development of de novo SUI as determined by the response of "somewhat," "moderately," or "quite a bit" on the Pelvic Floor Distress Inventory (PFDI) questions 20-22, as these responses have been shown to be highly relevant outcomes to patients.
Of the 465 women in the OPUS trial, 457 had SUI data available 12 months after surgery. Of the 12 original risk factors hypothesized to predict de novo SUI, 7 final risk factors were identified that when combined together accurately predicted de novo SUI. They were decreased age, higher vaginal parity, higher BMI, current diagnosis of diabetes, baseline urgency urinary incontinence symptoms, positive preoperative prolapse reduction stress test, and performance of a concomitant retropubic midurethral sling.
"Contrary to our original hypothesis, as age increased, a participant’s risk of experiencing de novo SUI decreased when combined with other factors," Dr. Jelovsek said. The researchers created a calculator using the variables from the model for making predictions in a clinical setting.
He went on to report that the prediction model had useful discrimination between women who ultimately did or did not experience de novo SUI, with a concordance index of 0.73. The accuracy of the model on the entire CARE data set was also significantly better than that of random chance (a concordance index of 0.62) as well as that of using the preoperative prolapse reduction stress test alone (a concordance index of 0.54).
"Our model demonstrates good predictive accuracy, with a concordance index of 0.72 in women undergoing transvaginal prolapse surgery," Dr. Jelovsek said. "This compares favorably to the National Cancer Institute Gail Model for Breast Cancer (a concordance index of 0.59), and the Framingham Cardiovascular Risk Model (a concordance index of 0.72)."
He acknowledged certain limitations of the study, including the fact that although the model was better than random chance in predicting the probability of SUI after abdominal sacral colpopexy and Burch urethropexy in the CARE population, the accuracy was lower. "This may have been because women in the OPUS data set underwent sling with vaginal surgery for prolapse, while those in CARE underwent a Burch procedure and abdominal surgery," he explained.
"It is also possible that abdominal sacral colpopexy places a different amount of risk of de novo stress urinary incontinence on individuals than does vaginal surgery for pelvic organ prolapse. Despite this, a concordance index of 0.62 suggests that the model is also a valuable adjunct for shared decision-making between the clinician and patients prior to abdominal sacral colpopexy with Burch urethropexy surgery."