Complications of Synthetic Midurethral Slings: Is There a Relevant Discrepancy Between Observational Data and Clinical Trials?

Context: The complications of synthetic midurethral slings (MUSs) in women with stress urinary incontinence (SUI) have become a globally debated issue. Objective: To systematically review the short- and long-term complications of mesh slings reported in observational data compared with clinical trial data, to determine whether the complication rates from clinical trials reflects “real-world” observational data. Evidence acquisition: PubMed and Cochrane Library were searched. Methods as detailed in Preferred Reporting Items for Systematic Reviews and Meta-analyses and Cochrane Handbook for Systematic Reviews of Interventions were followed. Evidence synthesis: Thirty registries/databases including 709 335 MUS procedures (1–22 yr of follow-up) were identified. MUS procedures were associated with intraoperative bladder perforation in 0.86–3.6%, urethral perforation in 0–0.1%, vascular injury in 0.04–0.1%, voiding lower urinary tract symptoms (LUTS) in 1.47–3.5%, vaginal exposure in 0.2–1.9%, and reoperation in up to 9% of cases. Forty-three randomised clinical trials were identified, including 6284 women who underwent MUS procedures and 2177 women who underwent other interventions (1–10 yr of follow-up). MUS procedures were associated with urinary tract perforation in 2.58%, vaginal injury in 1.43%, de novo voiding LUTS in 4.37%, de novo storage LUTS in 5.41%, mesh extrusion/exposure (vaginal/urinary tract) in 2.54%, dyspareunia in 2.26%, pain (pelvic/suprapubic/perineal) in 2.83%, and reoperation for complications required in 1.82% of cases. Meta-analyses of the randomised controlled trials revealed that retropubic MUSs were associated with more events of urinary tract perforation (risk ratio [RR] 9.81, 95% confidence interval [CI] 5.05–19.04, high certainty of evidence [COE]) and voiding LUTS (RR 1.57, 95% CI 1.19–2.07, high COE) than transobturator MUSs. MUSs were associated with more events of pain than mini-slings (RR 1.72, 95% CI 1.04–2.87, moderate COE). Conclusions: Short- and long-term data on complications of polypropylene mesh used for female SUI are fairly comparable when using outcome data from well-designed clinical trials or from less structured prospective or retrospective registries. Comparisons have to be made with caution since the two systems of data collection are inherently incomparable. This knowledge should be incorporated in the discussion on how to implement polypropylene mesh for female stress incontinence. Patient summary: In order to know whether mesh tapes used for treating stress incontinence work well and are safe, high-quality information is important. It appears that well-designed clinical studies give similar results to large registration databases. These data should be interpreted with caution in view of the different ways the information was collected. These results will help physicians and patients understand the risks of mesh tapes.