Intended for healthcare professionals


Surgical interventions for stress urinary incontinence

BMJ 2019; 365 doi: (Published 05 June 2019) Cite this as: BMJ 2019;365:l2350

Linked research

Surgical interventions for women with stress urinary incontinence

Linked opinion

What next for transvaginal mesh?

  1. Emily Carter, Cochrane fellow1,
  2. Rufus Cartwright, subspecialty trainee urogynaecology2
  1. 1Cochrane UK, Oxford, UK
  2. 2John Radcliffe Hospital, Oxford OX3 8HU, UK
  1. Correspondence to: R Cartwright rufus.cartwright{at}

Women and surgeons must navigate a challenging landscape of choices

Stress urinary incontinence—defined as leaking urine with coughing, sneezing, or exertion—affects more than one in 10 adult women,1 and represents a large unmet health need. The linked systematic review and network meta-analysis by Imamura and colleagues (doi:10.1136/bmj.l1842) includes and updates eight earlier Cochrane reviews, to provide a unified summary of comparative efficacy and safety for all contemporary surgical procedures for stress incontinence.2

From the late 1990s, there has been a rapid progression in surgical techniques for management of stress incontinence. The advent of the polypropylene midurethral sling seemed to many surgeons like a revolutionary advance, compared with earlier and more invasive options such as open colposuspension and native tissue fascial slings. The midurethral sling was a technically simple day-case procedure, with excellent efficacy rates and seemingly lower morbidity than previous procedures. It rapidly became the most widely offered procedure in many countries,3 and with its popularity, overall surgical treatment rates increased.4

However, increasing publicity about the harms of vaginal mesh implantation, including chronic pain and vaginal erosions, have caused a crisis of confidence among both the public and surgeons.3 Implantation of midurethral slings has declined, even in countries where they are not currently restricted or suspended,56 and with this decline, overall treatment rates for stress incontinence appear to have fallen.78

This systematic review comes at a crucial time in the debate about pelvic mesh, as clinicians have struggled to integrate older evidence about colposuspension and native tissue fascial slings with newer evidence from more than 150 randomised trials that were published during the time that retropubic, transobturator, and single incision polypropylene sling devices came to market. The work reported here has already been used to inform the recent National Institute of Health and Care Excellence (NICE) guidelines9 on surgical management of stress urinary incontinence, which despite controversy,10are likely to shape future practice in England and beyond.

The interpretation of results from such a complex network meta-analysis is itself complex,11 and the results reported here have to be considered in the light of a wider range of evidence, including from observational studies. Imamura and colleagues identify retropubic slings, transobturator slings, traditional fascial slings, and open colposuspension as the procedures that are most likely to be the most effective.

This focus obscures important differences in efficacy and safety between these procedures. Both the direct and indirect comparisons reported here favour retropubic slings over transobturator slings, with moderate confidence in this difference. In our view, correctly, NICE have therefore recommended specifically against the use of transobturator slings, except where a retropubic approach would be contraindicated.

Another concern is in the assessment of the evidence for laparoscopic colposuspension. No statistically significant difference exists in the cure or improvement rates between open and laparoscopic colposuspension; therefore, for many women, laparoscopic colposuspension might remain preferable in view of reduced morbidity.

As the mesh controversy has developed, women have more often chosen urethral bulking agents as first line surgical treatment. Although the efficacy of bulking agents is moderate at best, the safety profile is good.12 Here, because no randomised controlled trials were available to bring bulking agents into the network meta-analyses, they could not be included in the evidence synthesis, but remain a reasonable low risk option to discuss with patients.

While NICE have developed a decision aid13 to help women choose between so many surgical options, further work is clearly needed to understand how women should navigate what remains a complex choice, based on their individual priorities for cure and improvement.14 Imamura and colleagues’ work provides the best possible summary of current evidence of efficacy, but highlights the huge difficulties that will be faced in establishing adequate evidence of comparative long term safety. The existing randomised controlled trials are of short duration, and even in aggregate are underpowered for serious but rarer adverse events such as mesh complications, which can occur many years after implantation.15

These safety issues could be even more pronounced if, as seems likely, women will increasingly opt for colposuspension and traditional slings, which are both highly technical and operator dependent, but which have been neglected for almost two decades. This shift in practice will create huge challenges in the training or retraining of current and future gynaecologists and urologists. There remains a real risk that regulators, surgeons, and the public will reject the midurethral sling in favour of procedures that are eventually proven to have been both less safe and less effective.