Transvaginal mesh implants independent review: interim report

Chapter 5: Review of the evidence from safety reviews and systematic reviews

5.1 Evidence availability

This section of the Independent Review (IR) was undertaken in line with a modified form of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline^[7].

This review considered systematic review evidence of two sorts. The first were the reviews of evidence undertaken by those agencies responsible for the safety of medical devices on an international and national basis. The second were the published, peer-reviewed Cochrane systematic reviews and health technology assessments undertaken in relation to mesh devices for SUI and POP.

Cochrane systematic reviews are produced by the Cochrane Collaboration. This is a global, independent network of researchers, professionals, patients, carers and people interested in health. It is formed as a not-for-profit organisation which spans contributors from more than 120 countries. Its work is always free from commercial sponsorship and other conflicts of interest. Cochrane Collaborators do this by producing reviews that summarise the best available evidence generated through research to inform decisions about health and health care. These Cochrane Reviews are systematic reviews of primary research in human health care and health policy. They are internationally accepted as providing evidence-based health care advice of the highest standard. Cochrane Reviews are updated as needed, ensuring that treatment decisions can be based on the most up-to-date and reliable evidence. The full text of every Cochrane systematic review and the Review Protocols for work in progress, are published online in the Cochrane Database of Systematic Reviews in the Cochrane Library^[8]. In the UK Cochrane Reviews are used to inform the National Institute for Health and Care Excellence (NICE) and the Scottish Intercollegiate Guidelines Network (SIGN) guidelines, as well as informing policy and decision making in health care commissioning and development.

The following databases were searched for relevant reviews: Cochrane Library (2004 to 2015) and Medline (2004 to 2015). The search strategies used a range of key words and subject headings to identify information focussed on mesh implants for the treatment of stress urinary incontinence (SUI) or pelvic organ prolapse (POP), either alone or in comparison with other, alternative approaches to treatment. These two limits were applied to the identification of systematic reviews and safety reviews for inclusion in the review. Selected material was also limited to data reviews published in English, or where an English translation existed.

National and international websites of medical device safety organisations were also searched.

In total ten safety reviews were included^[9]. These were:

• Australia - Therapeutic Goods Administration (2014) [AUS];
• Canada - Health Canada (2014) [CA];
• Denmark - Danish Health and Medicines Authority (2012) [DK];
• European Union (consultation draft) - Scientific Committee on Emerging and Newly Identified Health Risks (2015) [EU];
• Netherlands - Health Care Inspectorate (2013) [NL];
• New Zealand - Accident Compensation Corp. (2015) [NZ1];
• New Zealand - MedSafe (2014) [NZ2];
• UK - Medical Devices and Healthcare Regulatory Authority (2014) [UK1];
• UK - York Health Economics Consortium for the Medical Devices and Healthcare Regulatory Authority (2012) [UK2]; and
• USA - Food and Drug Administration (2011) [USA].

Of these, six were full, completed reviews [NL, NZ 1&2, UK 1&2 and USA], one [EU] was reviewed as a provisional draft report published for consultation and three were based on reported summaries on, or news alerts published on, official websites [AUS, CA and DK].

There were 12 pertinent Cochrane systematic reviews completed, of which nine related to mesh use and alternative management approaches for SUI:

• Glazener CMA & Cooper K. (2001). Anterior vaginal repair for urinary incontinence in women. [Glazener 1];
• Dean N et al (2006). Laparoscopic colposuspension for urinary incontinence in women. [Dean];
• Rehman H (2011). Traditional suburethral sling operations for urinary incontinence in women [Rehman];
• Lapitan MCM et al (2012). Open retropubic colposuspension for urinary incontinence in women. [Lapitan];
• Kirchin V (2012).Urethral injection therapy for urinary incontinence in women.[Kirchin];
• Bakali E et al (2013). Treatment of recurrent stress urinary incontinence after failed minimally invasive synthetic suburethral tape surgery in women [Bakali];
• Nambiar A (2014). Single-incision sling operations for urinary incontinence in women. [Nambiar];
• Glazener CMA & Cooper K (2014). Bladder neck needle suspension for urinary incontinence in women. [Glazener 2]; and
• Ford et al (2015). Minimally invasive synthetic suburethral sling operations for stress urinary incontinence in women. [Ford]

In addition there was one health technology assessment completed in relation to SUI:

• Cody J et al. (2003). Systematic review of the clinical effectiveness and cost-effectiveness of tension-free vaginal tape for treatment of urinary stress incontinence. ^[10] [Cody].

Finally, there were three Cochrane systematic reviews were completed in relation to POP;

• Maher C et al (2013). Surgical management of pelvic organ prolapse in women. [Maher];
• Bugge C et al (2013). Pessaries (mechanical devices) for pelvic organ prolapse in women [Bugge]; and
• Hagen S & Stark D. (2011). Conservative prevention and management of pelvic organ prolapse in women. [Hagen].

5.2 Methods

Following discussions with both patients and clinicians, a number of key outcome areas were identified to provide a data extraction framework. These were:

• Effectiveness of SUI or POP procedure(s):
  • effectiveness in terms of objective SUI / POP cure at one year or more;
  • effectiveness in terms of subjective SUI / POP cure at one year or more;
  • need for repeat SUI or POP surgery; or
  • further conservative treatment for SUI.
• Reported safety issues with SUI or POP procedure;
  • mesh technology; or
  • proprietary brand of mesh;
• Patient-focussed outcomes: Quality of Life (QoL):
  • measurable QoL at one year or more post procedure, specific to SUI or POP;
• Patient-focussed outcomes: adverse outcomes:
  • short term/postoperative complications;
  • long term disability due to adverse effects;
  • surgical treatment for adverse effects.
• Relative efficacy of alternative therapy to mesh.
• Systems efficacy;
  • Surgical capacity and competency issues;
  • Service capacity and feasibility;
  • Other factors.

Data were extracted and tabulated for further interpretation. The overall quality of the evidence reviewed was assessed using the Scottish Intercollegiate Guidelines Network (SIGN) on grades of evidence [SIGN 50 reference].

5.3 Safety reviews of mesh implants

5.3.1 Nature of the evidence

At its heart, any review of the safety of a medical device is seeking to determine if the device can continue to be used safely and how best to ensure that patient safety is maintained throughout the medical or surgical processes that implant or connect the device to the patient, without reducing the overall effectiveness of the device.

However, it is fair to say that the 10 safety reviews included in this IR differed in their specific focus, the content of the review, and in which actions were considered necessary. Some provided a comprehensive review of the evidence relating to adverse outcomes following mesh implantation; others considered the effectiveness of the original safety review process; whilst others simply provided health care systems with advice on how to proceed in the current set of circumstance.

In the context of the outcomes being considered, such safety reviews are most likely to focus mainly on the nature, severity and frequency of any surgical complications and adverse outcomes. They are also likely to consider aspects of efficiency and effectiveness in the delivery of care. Finally, they may consider whether there has been any failure in the regulatory system that was used to determine the original safety of the device as "safe" for health care use.

Different reviews may use varying methods. In most cases, the reviews can be classified as being "narrative reviews", reporting on available evidence. For the purposes of this review of safety reviews, the quality of this evidence has been assessed to be in a range from SIGN 1++ to SIGN 4 evidential levels. As such, they represent very good sources of evidence, within the context of the review's stated aim and focus. Four reviews specifically reviewed the safety of synthetic surgical mesh implants for SUI and POP [AUS, EU, NL, and UK 1&2], of which and one specifically considered whether a withdrawal of mesh for POP [NL]. Three reviews were undertaken to provide updated advice to patients, health care providers and clinicians [CA, DK and USA]. Finally, the two reviews from New Zealand [NZ 1&2] only considered data on adverse outcomes following mesh surgery. It is noteed that whilst seven of the reviews considered both SUI and POP procedures as being within scope [AUS, EU, NZ 1&2 & UK 1&2], three reviews only considered POP procedures [DK, NL and USA].

5.3.2 Results

The extracted data in relation to the safety reviews are summarised within two tables. Table 1a provides a detailed analysis of the data contained in each of the International Agency's Safety Reviews, whilst Table 1b gives general observations and findings form the International Agency Safety Reviews. In this section, the main findings are summarised across the reviews.

None of the safety reviews concludes that there is sufficient evidence to withdraw synthetic mesh from clinical use for either SUI or POP [AUS, CA, EU, DK, NL, NZ 1&2, UK 1&2, USA], though one review does recommend that women with mesh implants for POP are recalled to hospital for clinical assessment [DK].

In a similar way, none of the reviews under-estimate the reality that for some women, the use of mesh devices has been associated with long-term, adverse outcomes that have had severe effects which limit their everyday activities and reduce their quality of life. [AUS, CA, EU, DK, NL, NZ1&2, UK1&2, USA]

Mesh safety in treating SUI

The scientific rationale for the use of synthetic mesh is specifically considered in the safety review from the EU's Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR) [EU].

They conclude in relation to SUI procedures that there is a robust evidence-base to support the use of Mid Urethral Slings (MUS) which they described as being "the most extensively reviewed and evaluated procedure for female SUI now in use" [EU]. This is the case for both retropubic and transobturator MUS procedures [EU]. It should be noted that the safety review by the Australian Therapeutic Goods Agency, as reported on their website described the evidence base for SUI as "adequate". (AUS) The EU review notes that, as with any surgical procedure, there can be complications associated with MUS. Where these do occur, the surgical complication rates are low [AUS, EU, NZ1, UK 1&2]. Complications found to be associated with MUS procedures include bleeding, damage to the bladder and bowel, voiding difficulty, mesh tape exposure and pelvic pain. All these complications may require repeated surgery, but this is uncommon. [CA, EU, NZ1, UK 1&2] What evidence of longer term effectiveness that does exist suggests that treatment success decreases over five years for both retropubic and transobturator MUS procedures, though patient satisfaction with both types of slings remained high. [EU] New mesh erosions occurred in both types over time at a similar rate. [EU, NZ1, UK1&2]

Mesh safety in treating POP

For POP procedures the EU safety review concluded that there is convincing evidence in favour of the use of a synthetic mesh to repair a prolapsed anterior vaginal wall. The evidence suggests that mesh implants are both subjectively and objectively superior in terms of clinical outcomes to a native tissue repair, though the reported health-related QoL post-surgery is no different (EU). Complications are reported with the rate of new pelvic organ prolapse in the untreated vaginal compartment significantly higher when synthetic mesh is used, though there is no evidence that this leads to a need for subsequent operations for POP. A similar pattern is observed for post-procedure SUI where the use of mesh is associated with higher rates of reported SUI; again this is not reflected in the need for SUI surgery. [EU] Mesh exposure is reported frequently following anterior wall repair with mesh, though there are no differences in reported rates for new dyspareunia or sexual dysfunction. [EU]

For repairs to the posterior vaginal compartment, the review concluded that there is moderate evidence that the use of mesh results in higher rates of objective cure. This, however, is also associated with higher rates of new POP of the anterior vaginal compartment. Subjective cure or cases of new SUI are observed to be no different from native tissue repair. As with anterior vaginal wall repairs, mesh exposures are reported frequently. [EU, NZ1] This pattern is also broadly observed for mesh repairs in more than one vaginal compartment. Repairs using mesh were found to result in higher rates of subjective and objective cure, but also in significantly higher rates of new POP of the untreated vaginal compartments. No differences in other outcomes or surgical complications were found, though mesh exposures were frequently reported. [EU, NZ1]

Surgical approach to POP repair was considered in the Australian review. This concluded that whilst there was evidence of effective use of mesh in abdominal POP, there was insufficient evidence to support its use in transvaginal POP repairs. [AUS] This is consistent with the advice provided by both Health Canada and the Food and Drug Administration (FDA) in the United States which advised clinicians to note that transvaginal procedures may carry a higher risk of complications than abdominal POP or native tissue repairs. [CA, USA]

For all types of POP repair, reported mesh exposure, (symptomatic or asymptomatic), obviously differed from that in native tissue repair (observed complication rates 4-19%); however no differences in complication rates were observed for dyspareunia (post-surgical or new), post-surgical pain, haemorrhage, bowel and/or rectal injury, urinary infection and postoperative urinary retention between mesh and native tissue surgery. [EU] Overall, complication rates for POP were reported to be low. [AUS, EU] However, it was noted that the follow-ups in papers in these meta-analyses were mainly short-term (up to 12 months); a few were medium-term (1-5 years). Long-term results (5-10 years) of RCT's are not yet published but are needed for the full appreciation of outcomes. [EU, UK 1&2, USA]

Risk factors for adverse outcomes

The EU review undertook a very detailed analysis to identify possible sources of risk associated with adverse outcomes in the use of mesh implants. This review has been used as the basis for this section of the report, augmented by other evidence from safety reviews where appropriate.

An analysis of adverse outcomes associated with the type of synthetic surgical meshes for treating SUI and POP was undertaken. At present, four major types of mesh are produced commercially. The data suggested that two types of mesh are "most appropriate" for mesh implants:

• synthetic mesh type 1 (polypropylene monofilament, macroporous >75μm):
  • synthetic mesh for vaginal use; and
  • synthetic mesh for insertion via the abdominal route; and
• synthetic mesh type 3 (polyester ,multifilament , microporous <10μm):
  • synthetic mesh for insertion via the abdominal route.[EU]

Synthetic mesh type 2 (mono and multifilament, microporous, and synthetic mesh type 4 (monofilament, nanoporous <1μm) were considered to be "not appropriate" for this clinical use. For all other forms of synthetic mesh materials, the EU review concluded that there was insufficient evidence on which to base an opinion. [EU] When considering factors which may be associated with mesh design, the review highlighted a number of factors that may be possible potential sources of risk. These included: overall surface area of mesh used, (which is greater for POP than for SUI); the composition of the mesh weave and its porosity; the physical character of the mesh and its durability within the context of long-term indwelling of the device in human tissue on a long-term basis. [EU]

Whilst the available evidence only allows for a two year follow up, specific surgical techniques were noted to be associated with a higher risk of adverse outcomes. At the most fundamental level, it was noted that mesh exposure is only seen with a non-absorbable material such as synthetic mesh, this is true of all synthetic materials. Generally, it was concluded that vaginal surgery is associated with a higher risk of mesh-related complications and morbidity than abdominal mesh procedures. [AUS, CA, EU, NZ1, UK 1&2, USA] Overall, the EU review was of the opinion that the risk assessment of the use of mesh needs to differentiate between its use in SUI and POP in that the evidence:

• on efficacy and use of implanted meshes for SUI suggested that the associated risk of complications was low (albeit that follow data were limited and there was an absence of long-term -up (5-10 years) follow up data;
• on vaginal insertion of non-absorbable synthetic mesh with a large surface area for POP suggests it is associated with the highest incidence of complications; and
• that vaginally implanted mesh for POP is associated with increased risks compared to mesh implantation for SUI. [AUS, CA, EU, NZ1, UK 1&2. USA]

In the light of these considerations, the EU review concluded that the use vaginally implanted mesh for POP should be restricted [EU]. This is, however, the only review to reach such a conclusion.

Surgeons' experience of the procedures in question was considered a risk factor for adverse outcomes in a number of the safety reviews. These can be summarised as:

• surgical experience in SUI and POP MUS procedures - the evidence suggests only surgeons with experience should perform these procedures, though there is not clarity on the definition of an "experienced surgeon"; [AUS, EU, UK1&2]
• level of surgical training and maintaining competence - the evidence suggests that successful learning may vary from one trainee to another and may be affected by factors such as: the trainee's prior surgical experience; the difficulty of the procedures; and the level/quality of the clinical supervision; [CA, EU, UK1, USA] and
• adherence with clinical guidelines - the evidence suggests that there is a greater risk of adverse outcomes if surgeons do not follow appropriate clinical guidelines or the manufacturer's instructions. [EU, NZ1, UK 1&2]

The potential to identify patient groups that were at a higher risk of complication was noted in four safety reviews [AUS, EU, UK1&2]. However, there is at present very little robust evidence available to inform patient selection when synthetic mesh is proposed for use in POP or SUI procedures. More research needs to be done on this, at present it is recognised that: smoking is statistically associated with an increased risk of mesh exposure; and factors such as obesity and age may also be important. In this latter regard, the EU review concluded that it was prudent to be "more reluctant" to use mesh devices for POP in younger age groups. [EU]

Patient Consent

The need to ensure that patient decisions to undergo a mesh procedure must be based on appropriately informed patient consent is noted in six safety reviews [AUS, CA, EU, NL, UK1, USA]. They note that gaining effective patient consent should be the result of a wide-ranging discussion regarding the patient's specific situation and all the potential benefits and risks from the use of synthetic mesh for either SUI or POP procedures. Safety reviews providing specific guidance on what should be included in such patient consent discussions include those from Health Canada [CA], SCENIHR [EU], The Dutch Healthcare Inspectorate [NL], the UK Medicines and Healthcare products Regulatory Agency [UK1], and the US Food and Drug Administration. [USA]

Data gaps and long-term follow up

The need for more detailed data, or the relative lack of such data, was mentioned in some way by all the safety reviews. In general data identified as being lacking related to:

• research evidence on long-term follow up (greater than 5 years post-surgery) for patients receiving mesh procedures for SUI or POP; [EU, UK1&2, USA] ,
• the lack of traceability for individual mesh devices used in such procedures; [EU]
• the lack of data on the specific surgical approaches used in mesh procedures; [EU],

and

• the lack of comprehensive adverse outcome reporting. [CA, EU, UK1&2, NL, USA]

Whilst there were differences in proposed approaches to deal with the evidential gaps identified, it is notable that these focused on the need for effective data capture and reporting.

Medical device regulatory systems

Finally, six safety reviews commented on aspects of the processes by which medical devices are assessed for their safety and what could be done to improve this. [AUS, CA, EU, NL, UK1, USA] Whilst the Therapeutic Goods Administration in Australia have indicated that they are reviewing the safety compliance of all mesh devices [AUS] and the US Food and Drug Administration has taken action to bring mesh devices for POP into a level of regulatory requirement more in line with that in Europe, [USA] the most comprehensive consideration of the overall system of assessing medical devices and safety was provided by the Dutch Healthcare Inspectorate. [NL]. They concluded that whilst the processes by which mesh devices had been assessed as being safe were in line with the Dutch regulatory framework, there were areas for improvement. The first of these was that the requirements for the evaluation process of devices should be made stricter, with more time taken to assess safety and judgements taken in the light of clear criteria for effectiveness and safety. The second improvement was that the formal Vigilance and Post-Marketing Surveillance stage of the evaluation process be strengthened in European legislation. This improvement, which was recommended in three other safety reviews [CA, UK1, USA], would allow for longer-term assessment of complications and adverse outcomes, especially when novel procedures were being used with devices deemed to already be safe. [CA]. Finally, the agency recommended the creation of a central, independent registry for implants, recording product information and patient information as a minimum. It was noted, however, that this was not specific to mesh devices and that the registry should include all implants, across all specialties. [NL] Clearly these recommendations are specific to the Dutch circumstances, though the revised EU Medical Devices Directive will address several of these requirements [http://ec.europa.eu/growth/sectors/medical-devices/regulatory-framework/index_en.htm]

5.3.3 Interpretation

The key messages from this analysis are:

• No safety review by an international agency has called for mesh devices used in SUI and POP procedures to be withdrawn from use.
• For some women, there are long-term, adverse outcomes associated with the use of mesh devices that have had severe effects on their everyday activities and quality of life.
• The use of mesh for MUS in treating SUI is generally considered effective with low complication rates, accepting that long-term follow up data are not presently available.
• The use of mesh for some forms of POP repair is considered effective, provided that it is used in abdominal rather than vaginal procedures. The risk of complications/adverse outcomes is higher folollowing the use of mesh for POP than for SUI.
• Risk factors for adverse outcomes include: surgical approach, the experience of the surgeon and some patient characteristics. Factors which may influence risk include the physical characteristics of the mesh device used.
• Well-informed patient consent is essential.
• Data associated with long-term follow up of mesh procedures are not currently collected. This needs to be remedied.
• The systems used to assess the safety of medical devices could be improved, notably in the area of vigilance and post-marketing surveillance.

5.4 Systematic reviews of effectiveness of mesh in stress urinary incontinenece and pelvic organ prolapse

5.4.1 Nature of the evidence

The systematic reviews included in this section of the report are either Cochrane systematic reviews (12 reviews) or Health Technology Assessment s (one review). In each case the methods adopted in creating these reviews mean that they fulfil the highest level of SIGN evidence grades in that they are based on evidence ranging from SIGN 1++ (High quality meta-analysis, systematic reviews of Randomised Controlled Trials (RCTs), or RCTs with a very low risk of bias) to SIGN 1+ (Well conducted meta-analyses, systematic reviews, or RCTs with a low risk of bias).

Using such systematic reviews, allows the professional and patients within the Independent Review (IR), and those who will read this IR more widely, access to the most recent clinical knowledge in a more readily accessible form. In the context of this review, they can facilitate the process to evaluate clinical effectiveness data of treatments and services across a range of settings and circumstances.

However using such systematic reviews does have some drawbacks. For example, they are not good at summarising evidence "gaps" (you cannot summarise what is not there), at the same time the review system will still work with evidence which is of differing "robustness". Though in this latter regard, the methods used do include ways of assessing the underlying strength or weakness of data included in the systematic review. Another limitation that should be noted is that systematic reviews tend to focus on the experiences of patient groups and not on individual patients. As such it is better at summarising research, than personal experiences.

Whilst such systematic reviews still require interpretation of the reviewed evidence, not least in how it relates to clinical effectiveness and public / patient experience, they are the best available evidence on which to base such assessments.

5.4.2 Results

As noted above (section 5.2 above) 12 systematic reviews were included in this analysis. Nine of these related to SUI procedures and three to POP procedures. The systematic reviews cover not only mesh procedures, but also comparisons with conventional treatment. This was to allow consideration of the effectiveness of mesh against other treatment approaches. The results of the analysis in relation to SUI procedures is shown in Table 2a, whilst that for POP procedures is shown in Table 2b.

In this section, the major conclusions from each of the systematic reviews in relation to effectiveness (clinical and patient outcomes and QoL measures) are presented.

Effectiveness of SUI Procedures

Anterior vaginal repair (without mesh) for urinary incontinence in women [Glazener1]

Open abdominal surgery (retropubic suspension) was more effective than anterior vaginal repair for the treatment of primary urodynamic stress incontinence. The effect was longer lasting, whether or not the women had associated prolapse. Marginal differences in QoL recorded.

No differences in effectiveness or QoL were observed between anterior vaginal repair and needle suspension. No comparative trials were found between anterior vaginal repair and mock operation, laparoscopic colposuspension, and suburethral sling procedures.

Overall the evidence was assessed as being of poor overall quality. Long term follow up data in trials included were limited.

Effectiveness of tension-free vaginal tape [Cody]

Laparoscopic colposuspension and traditional slings have broadly similar cure rates to Tension-free Vaginal Tape (TVT) and open colposuspension. The QoL of patients treated with TVT were significantly better in relation to their emotional state, social functioning, and mental health. However generally all QoL was shown to be better post-operatively compared to pre-operation levels.

Overall, the authors considered that there were "limited" data on which to base the review. The lack of long-term follow up data was specifically noted.

Laparoscopic colposuspension for urinary incontinence [Dean]

Laparoscopic colposuspension is reported to provide a lower objective cure rate for SUI over colposuspension by open surgery (in the shorter-term, less than 18 months). However, no significant differences were found in an 18 month to five year period. There were significantly fewer perioperative complications with Laparoscopic colposuspension and some evidence for less pain. The QoL data did not suggest any differences between the two.

In comparison with vaginal mesh slings (self-fixing slings), laparoscopic colposuspension was found to be less effective than mesh in objective cure rate for SUI, though there was no statistically significant difference in subjective cure rate between them. The data on QoL was not pooled, meaning a single analysis was not possible. Of five studies, only one suggested that mesh was associated with improved QoL

When differing approaches to laparoscopic colposuspension were compared it was reported that objective and subjective cure rates were higher when it was performed using sutures for the repair than when it was performed using mesh fixed with surgical staples. In a further comparison, suture repairs that used two sutures were more effective than those using a single suture. QoL was not assessed in these trials.

No research trials which compared laparoscopic colposuspension to mock operation, conservative management, needle suspension, traditional sling procedures, anterior vaginal repairs or peri-urethral injections were found.

The authors considered that the data was adequate in 13 trials and of uncertain quality in a further eight included in the systematic review. One trial was described as inadequate. The lack of long-term follow up data was noted as a specific omission which limited the analysis.

Traditional suburethral sling operations for urinary incontinence [Rehman]

The systematic review concluded that in trials which compared traditional slings with minimally invasive slings using synthetic mesh, there were no differences in the effectiveness of either type of sling on objective or subjective cure rate for SUI. Whilst QoL data were collected in some studies, no differences were found between types of sling procedure.

In comparisons with other surgical techniques - open abdominal retropubic colposuspension, abdominal and vaginal needle suspension and trials of different types of traditional sling materials - suggested that traditional slings may be equally as effective as other surgical approaches. Whilst not suitable for detailed analysis QoL data suggested that patients undergoing open abdominal retropubic colposuspension had better post-operative QoL than those with a traditional sling procedure.

Traditional slings may be more effective than either drugs or injectable agents, synthetic material. However, in comparisons with conservative management, anterior vaginal repair and laparoscopic procedures were not found to be more effective.

The authors considered that caution was needed in interpreting the results of the systematic review. The quality of evidence in the studies was variable; with only short-term follow-up and the lack of focus on primary outcome data (e.g. complication rates).

Open retropubic colposuspension for urinary incontinence [Lapitan]

Open retropubic colposuspension was found to be an effective treatment for stress urinary incontinence compared with other forms of surgery (anterior colporrhaphy (repair), needle suspension, and laparoscopic colposuspension). Long term data were analysed suggesting that it was effective in the long term with approximately 80% patients undergoing open retropubic colposuspension still continent at 5 years. QoL data in these trials were sparse and whilst post-operative improvements were observed, little evidence for differences between procedures was found.

Sling procedures, both traditional and minimally invasive mesh sling procedures, were found to be not significantly different from open retropubic colposuspension in objective or subjective SUI cure. QoL data also showed no significant differences.

More limited evidence in relation to open retropubic colposuspension compared with conservative management, pharmaceutical drug treatment, and injectable synthetic material suggested that surgical intervention may be more successful. No trial data was found for other possible comparisons.

Overall the authors considered the data quality to be classed as "unclear", though the ability to analyse longer term data was welcomed.

Urethral injection therapy for urinary incontinence [Kirchin]

Injection therapy with synthetic particulate material shows a short-term advantage over home pelvic floor muscle training in reducing SUI and an increased QoL. However, as follow up was limited to three months, it is not clear if this advantage is maintained.

Injection therapy appears inferior to open surgery at 12 months, but has a better safety profile.

Out of 14 trials in the systematic review, risk of bias was assessed as either "low" or "unclear" in all but one trial. This one trial was assessed at "high" risk of bias. Only short-term follow data was considered.

Treatment of recurrent stress urinary incontinence after failed minimally invasive synthetic suburethral tape surgery [Bakali]

No trials were found suitable for inclusion.

Non RCT data suggest that repeat suburethral tape surgery is less effective than for primary surgery. There is some evidence that retropubic suburethral tapes are superior to transobturator tapes as secondary procedures.

Single-incision sling operations for urinary incontinence in women [Nambiar]

Single incision (mini) slings were found to be less effective than retropubic, minimally invasive slings in achieving objective or subjective cure. QoL was reported to be statistically significantly better in the retropubic group.

Twenty trials were included which comparison between single incision slings with obturator minimally invasive slings; either the medial-to-lateral 'inside out' surgical approach (TVT-O) or the lateral-to-medial 'outside-in' approach (TOT). Objective and subjective cure rates were found to be significantly better for both TVT-O and TVT-O/TOT combined than for single incision slings. No difference was found for TOT alone compared with the single incision sling. No QoL data was reported.

No trials were identified in which single-incision slings were compared with no treatment, conservative treatment, open colposuspension or laparoscopic procedures. Whilst trials comparing them to traditional sub-urethral slings were found, the authors did not consider the data of an appropriate quality on which to confidently identify any differences between any of the different types of single-incision sling.

The overall quality of the data included was assessed by the authors as variable. About half of the trials were considered to have used using adequate methods to reduce the risk of bias, while in the other half, the methods used were considered to be inadequate or were not described. Long term follow up data were noted as lacking.

Bladder neck needle suspension for urinary incontinence in women [Glazener2]

Needle suspension was compared with open abdominal retropubic suspension using different techniques. Subjective outcome at both under and at one year suggested that open abdominal retropubic suspension was the more effective treatment.

When compared with anterior vaginal repair, needle suspension was found to be similar in terms of subjective cure rates after 12 months and long-term problems with voiding dysfunction.

One small trial compared needle suspension with suburethral sling procedures. However it was too small to address differences in cure rates,

No statistically significant differences were found in the one trial that presented comparisons between types of NS. No trials were found which compared needle suspension with mock procedure, conservative management, laparoscopic colposuspension, periurethral injections, or pharmaceutical drug treatment.

Overall the authors considered the quality of the data low. Long term follow data was only available for three of the included trials.

Minimally invasive synthetic suburethral sling operations for stress urinary incontinence in women [Ford]

Mid-urethral slings using mesh implants were found to be a highly effective treatment for SUI. Robust short term data suggests no significant differences between the two insertion routes, transobturator; and retropubic, in subjective or objective cure of incontinence. . There is some evidence that the observed equivalence in subjective and objective cure rates in the medium (1 to 5 years) and longer term (over 5 years).

The trials that compared the retropubic, bottom-to-top approach with the retropubic, top-to-bottom one approach showed that inserting the mesh tape through the retropubic route from bottom to- top is the more effective than the top-to-bottom approaches.

Comparisons between the transobturator, medial-to-lateral approach with the transobturator, lateral-to-medial approach showed no evidence of any differences between the two approaches with respect to SUI outcomes. When a retropubic route is employed, a bottom-to-top approach is more effective in terms of subjective cure than a top-to-bottom approach. When traversing the transobturator route, the evidence suggested that medial-to-lateral ('inside-out') and lateral-to-medial ('outside-in') approaches have similar effects.

No significant differences in efficacy or surgical outcomes were observed in those studies which compared one method of mid-urethral tape insertion with another using the same insertion route.

Differences between monofilament tapes and multifilament tapes were assessed. No statistical differences in objective or subjective cure rates were found.

As a general finding, QoL was found to improve significantly post-operatively within treatment groups, whatever the comparison being undertaken. However, no statistically significant differences were found between groups in the comparisons.

The majority of the trials included in the systematic review were rated by the authors as being of unclear quality. Longer term data, particularly on the long-term effects of surgery, and how the different insertion routes affect long-term outcomes are lacking.

Effectiveness of POP Procedures

Surgical management of pelvic organ prolapse in women [Maher]

The findings in relation to POP surgery are complex. Overall, the authors conclude that the data they have reviewed does not provide sufficient evidence to guide clinical practice.

Abdominal sacral colpopexy was associated with a lower rate of recurrent vault prolapse and less dyspareunia than vaginal sacrospinous colpopexy.

The use of absorbable polyglactin mesh overlay, absorbable porcine dermis or polypropylene mesh at the time of anterior vaginal wall repair reduces the risk of recurrent cystocele on examination, however improved outcomes including patient satisfaction, quality of life and reduced operations for recurrences have not yet been demonstrated.

Posterior vaginal wall repair may have a better anatomical success rate than transanal repair in the management of posterior vaginal wall prolapse, but the clinical effects are uncertain. There is no evidence to support the use of graft materials in the posterior compartment.

The evidence at this stage does not support the use of transvaginal combined total, anterior or posterior mesh kits for multi-compartment prolapse. Whilst anatomical outcome may be improved (as compared to native tissue repair) no difference was found in symptoms or quality of life outcomes. The mesh exposure rate was nearly 1 in 5, with nearly 1 in 10 requiring surgical intervention.

Performing continence surgery at the time of prolapse surgery in women with stress urinary incontinence is likely to be beneficial. This benefit is also associated with women undergoing prolapse who have been found to have occult stress incontinence pre-operatively.

Generally, the quality of the trials was described as "variable" by the authors in almost all cases, they considered that it was "unclear" what risk of bias the included trials presented.

Long-term outcomes were noted as being absent. They should be reported at least at two and five years after surgery, preferably longer.

Pessaries (mechanical devices) for pelvic organ prolapse in women [Bugge]

Formal comparison between the use of a mechanical device and the use of any surgery, with or without any form of mesh, were not considered in this review.

Conservative prevention and management of pelvic organ prolapse in women [Hagen]

In relation to this systematic review, only data relating to the comparison of physical and/or lifestyle interventions supplementing surgery with surgery alone was included. In three areas: physical intervention versus surgery; lifestyle intervention versus surgery; and combined physical and lifestyle intervention against surgery, no trials were identified.

The two trials identified in this systematic review provide contradictory findings. Whilst both compared pelvic floor muscle training (PFMT) following surgery with a control group who underwent surgery alone. In one trial the results indicate that despite the tendency towards improvement in the PFMT group over time, there were no significant differences in manometry scores between the controls and those in the PFMT arm. Change from baseline in the other objective measures (vaginal resting pressure, peak maximum vaginal squeeze pressure, and area maximum vaginal squeeze pressure) did not differ between groups. The second trial, however, reported that improvement in mean maximum pelvic floor muscle squeeze was significantly greater in the PFMT group than the control group.

Both trials reported on urinary function. In one trial it was reported that there were no significant differences between the intervention and control groups in reported incontinence using validated instruments. The second trial reported a significant improvement in urine leakage for both the intervention and control groups, but no significant difference in improvement between the groups.

5.4.3 Interpretation

SUI Procedures

The key messages from these analyses are:

• Abdominal surgery (retropubic colposuspension) was more effective than anterior vaginal repair for the treatment of primary urodynamic stress incontinence.
• Mid-urethral sling procedures were found to be as effective as traditional surgical approaches for SUI. Marginal benefits in QoL were noted.
• Mid-urethral sling procedures were found to be objectively more effective than laparoscopic colposuspension, but not subjectively so. Laparoscopic colposuspension was found to be no more effective than open colposuspension in the medium term. The type of surgical approach to make repairs when undertaking laparoscopic colposuspension may be a factor in achieving successful outcomes. Findings from analysis of QoL data were limited.
• Traditional sling operations are as effective as either mid-urethral sling procedures for SUI or other surgical approaches.
• Whilst open retropubic colposuspension was found to be more effective than other surgical approaches, it was not found to be significantly different from mid-urethral tape procedures.
• What data there is suggests that treating SUI with injectable materials is better than conservative management, but less effective than open surgery.
• There is limited (non-RCT) evidence that retropubic suburethral mesh tapes are superior to transobturator mesh tapes when used in repeat procedures.
• Women were more likely to remain incontinent after surgery with single-incision (mini) slings than after use of inside-out transobturator (TVT-O) tapes.
• Open abdominal retropubic suspension was found to be more effective than needle suspension. Anterior vaginal repair and suburethral (mesh) slings were not significantly different.
• Mid-urethral slings using mesh implants were found to be a highly effective treatment for SUI.

POP Procedures

The key messages from these analyses are:

• The findings in relation to POP surgery are complex. Overall, the authors conclude that the data they have reviewed does not provide sufficient evidence to guide clinical practice.
• Whilst some trial findings to support the use of pelvic floor muscle training as a treatment for women with prolapse, the evidence remains complex and limited. There was insufficient evidence about other interventions or combinations of interventions to inform practice

5.5 Systematic Reviews of Adverse Outcomes in SUI and POP

5.5.1 Nature of the evidence

See section 5.4.1 above.

5.5.2 Results

As noted above (section 5.2 above) 12 systematic reviews were included in this analysis. Nine of these related to SUI procedures and three to POP procedures. The systematic reviews cover not only mesh procedures, but also comparisons with conventional treatment. This was to allow consideration of the effectiveness of mesh against other treatment approaches. The results of the analysis in relation to SUI procedures in shown in Table 2a, whilst that for POP procedures is shown in Table 2b.

In this section, the major conclusions from each of the systematic reviews in relation to adverse outcomes are presented.

Adverse Outcomes in SUI procedures

Anterior vaginal repair for urinary incontinence in women (Glazener1]

Clinically relevant post-operative complications were reported for both abdominal retropubic suspension and anterior vaginal repair, but the complication rates were not different. New or recurrent prolapse was found to be less likely after anterior vaginal repair, whilst repeat surgery for recurrent incontinence was higher after anterior vaginal repair. No other differences were found for any other adverse outcomes.

No long-term adverse outcomes were considered.

Adverse outcomes associated with tension-free vaginal tape [Cody]

There were no significant differences in adverse outcomes between TVT and comparative surgery reported.

Laparoscopic colposuspension for urinary incontinence [Dean]

Significantly fewer perioperative complications were reported for laparoscopic colposuspension compared with open colposuspension.

No robust, statistically significant differences were found in adverse outcome rates for any other comparisons included in the systematic review. This included the comparison between laparoscopic colposuspension and vaginal mid-urethral mesh tape procedures.

No long-term adverse outcomes were considered.

Traditional suburethral sling operations for urinary incontinence [Rehman]

When comparing traditional slings with mid-urethral mesh tape procedures, perioperative complication rates were found to be higher for the traditional sling. New cases of overactive bladder function were also noted, but no other adverse outcomes were found to be significantly different.

Significantly fewer perioperative complications were found amongst open colposuspension patients than traditional sling patients. Voiding dysfunction was also significantly higher after traditional sling surgery compared with open colposuspension. Significantly more women who had traditional slings had post-operative complications compared with those who had needle suspension.

No long term adverse outcomes were considered.

Open retropubic colposuspension for urinary incontinence [Lapitan]

Perioperative complication rates for open retropubic colposuspension were lower than those observed for both needle suspension or anterior colporrhaphy. The procedure may also be associated with lower rates of bladder perforation when compared with laparoscopic colposuspension.

The long-term profile of adverse outcomes for sling procedures, in particular with the use of mid-urethral mesh tape procedures, is still unclear.

Urethral injection therapy for urinary incontinence [Kirchin]

Overall, the complication rate for injections was lower than for open surgery. However, the follow up periods were very short and as the approach is not efficacious, the adverse events are not clinically relevant.

Treatment of recurrent stress urinary incontinence after failed minimally invasive synthetic suburethral tape surgery [Bakali]

No trials were included in the systematic review.

Single-incision sling operations for urinary incontinence in women [Nambiar]

Both repeat stress incontinence surgery and new cases of urinary urgency were found to be associated with single incision (mini) slings compared to minimally invasive retropublic slings.

In the comparison between single incision slings and transobturator mid-urtheral mesh tapes, a complex set of adverse outcomes were reported. These may be summarised as:

• Vaginal mesh exposure (erosion): TVT-O had a significantly lower risk of mesh erosion than single-incision slings. No statistically significant difference between the TOT and single incision sling treatment groups.
• Post-operative pain or discomfort: Both TVT-O and TOT patients had more post-operative pain and discomfort than single-incision sling patients.
• Long-term pain or discomfort: TOT was found to be associated with higher rates of long-term pain than single-incision. No differences were observed for TVT-O compared with single-incision slings. TOT and TVT-O are both obturator devices.
• Repeat stress incontinence surgery: women undergoing single-incision slings were nearly six times more likely to need further stress incontinence surgery after single-incision sling surgery than after TVT-O. There was no evidence of a difference between single-incision slings and TOT slings.
• Need for any other additional or new surgical procedure to treat complications: TVT-O was found to be associated with a statistically lower chance of needing surgery whilst no differences in risk was found for TOT compared with single-incision sling surgery.

No long term adverse outcome data were reported.

Bladder neck needle suspension for urinary incontinence in women [Glazener2]

No significant differences were found in complication rates for needle suspension when compared with either open abdominal retropubic suspension or anterior vaginal repair. Sling operations were associated with higher complication rates than needle suspension.

One small trial (n= 9 treatment v 10 controls) found that postoperative pain was significantly less at three months in women whose needle suspension used polytetrafluoroethylene sutures compared with polypropylene ones.

No long term data were considered.

Minimally invasive synthetic suburethral sling operations for stress urinary incontinence in women [Ford]

Overall perioperative complication rates for the transobturator route compared to the retropublic route were not statistically significant different.

Where individual adverse outcomes were reported, the transobturator route was associated with significantly fewer: major vascular injuries, bladder / urethral perforations, and post-operative voiding dysfunction. However the clinical importance of these adverse outcomes vary. Vascular and bladder perforation, for example, sound serious, but clinically they should always be detected by cystoscopy and remedied by re-positioning.

Pain rates were found to vary between groups as to which approach was associated with greater pain. Groin pain was higher for the transobturator route whilst suprapubic pain was lower. Most cases of pain resolved within six months. However, in at least one RCT (Teo et al (2011), the trial team decided to finish recruitment early, due to excess leg pain in the tension-free vaginal tape transobturator group^[11]. This highlights the clinical importance of adverse outcomes.

The need for repeat incontinence surgery was found to be not significantly different between groups under 12 months, though it was found to be more common for patients undergoing transobturator procedures over one year.

No statistically significant difference was seen in overall perioperative complications when comparing the retropubic bottom-to-top approach with the retropubic top-to-bottom approach. Significantly fewer women undergoing the retropubic bottom-to-top approach experienced bladder perforation, voiding dysfunction or vaginal tape erosion.

In the transobturator, medial-to-lateral approach compared with the transobturator, lateral-to-medial approach analysis, the former was found to be associated with fewer vaginal wall perforations, but higher levels of voiding dysfunction. There were no statistically significant differences between the two groups for: overall perioperative complication rate; major vascular / visceral injury; bladder perforation; de novo urgency symptoms; detrusor overactivity; vaginal tape erosions; and groin/thigh pain. No significant difference in the rates of repeat incontinence surgery in the medium term was found.

Whilst data greater than five years was included in these analyses, it was acknowledged that the long term effects of mesh sling insertion required further detailed research.

Adverse Outcomes in POP procedures

Surgical management of pelvic organ prolapse in women [Maher]

The data relating to adverse outcomes following POP surgery is of relatively low quality and few conclusions can be drawn from it.

Where adverse outcomes are reported, it is difficult to assess the clinical significance of these. In many cases the differences between groups within comparisons are reported as not significant. Of all possible complications, mesh erosion is one of the more commonly identified, adverse outcomes noted following the treatment of POP

Pessaries (mechanical devices) for pelvic organ prolapse in women [Bugge]

No adverse outcomes considered in the report.

Conservative prevention and management of pelvic organ prolapse in women [Hagen]

No data on adverse outcomes was reported in the systematic review.

5.5.3 Interpretation

SUI Procedures

The key messages from these analyses are:

• Anterior vaginal repair was found to increase likelihood of further SUI surgery, but reduce the risk of new or repeat prolapse when compared with abdominal retropublic surgery.
• No significant differences were found in the risk of adverse effects between retropubic and transobturator, mid-urethral mesh tape procedures.
• Mid-urethral mesh tape procedures were not found to be associated with greater risk of adverse outcomes than laparoscopic colposuspension, though long-term, data was not collected.
• Mid-urethral mesh tape procedures were associated with lower complication rates than traditional suburethral sling operations.
• The long-term profile of adverse outcomes associated with the use of TVT mesh, remains unclear due to the absence of adequate research.
• Adverse events are lower for treatment by injection than for open surgery, although efficacy is significantly lower.
• Treatment with single-incision slings is more likely to need further continence surgery and experience mesh exposure more often than those treated with transobturator (TVT-O) tapes.
• Minimally invasive sling procedures appear to have similar in their adverse outcome rates, though long term effects have not been adequately researched.
• The clinical importance of these adverse outcomes do differ: bladder perforation (more common in retropubic procedures) is of little or no clinical importance, whilst groin pain (more common for transobturator procedures) is of greater importance clinically.

POP Procedures

The key messages from these analyses are:

• The data relating to adverse outcomes following POP surgery is of relatively low quality and few conclusions can be drawn.
• Mesh erosion is one of the main adverse outcomes noted following the treatment of POP.
• No data on adverse outcomes were reported in the systematic review on conservative management of POP compared with surgery.

5.6 Conclusions

On the safety of mesh

The evidence from systematic reviews into the safety and effectiveness of SUI and POP mesh procedures and the adverse outcomes associated with them presents a complex picture.

Although the international safety reviews have differing emphasis and explore the issues in a variety of ways, all the international safety reviews recognise that, for some women, there are long-term, adverse outcomes associated with the use of mesh devices that have had severe effects on their everyday activities and quality of life.

At the same time, none of the international safety reviews conclude that this is a need for mesh devices used in SUI and POP procedures to be withdrawn from use.

In those safety reviews which directly addressed the effectiveness of mesh devices, the general conclusion is that mesh use in mid-urethral sling procedures to treat SUI is effective with low, short-term complication rates. The use of mesh for some forms of POP repair is considered effective, provided that it is used in abdominal rather than vaginal procedures. The risk of complications/adverse outcomes, notably mesh erosion is higher for POP than for SUI.

The reviews which have explored the possible risk factors associated with women who have had serious adverse outcomes have identified the following factors:

• the surgical approach adopted: transobturator versus retropublic in SUI procedures and vaginal versus abdominal in POP procedures;
• the experience of the surgeon undertaking the procedures;
• patient characteristics, including health risk behaviours; and
• the physical characteristics of the mesh device used.

It is accepted in several of the reviews that long-term follow up data are not presently available to capture late complications of mesh surgery. It is also recommended that this gap in the data needs to be addressed and routinely collected. More widely the systems used to assess the safety of medical devices were identified as areas for further improvement, notably in the area of vigilance and post-marketing surveillance.

All the reviews note the essential need for well-informed patient consent.

On the effectiveness of mesh

The effectiveness of mesh was considered in the systematic reviews. This is also a complicated picture. The key messages sections try and draw out the specific findings, but even they can be complex and difficult to interpret.

The main thrust of the findings in relation to SUI suggest that mid-urethral slings using mesh implants were found to be a highly effective treatment for SUI. There is limited (non-RCT) evidence that retropubic suburethral mesh tapes are superior to transobturator mesh tapes when used in repeat procedures. In comparison with other surgical techniques traditional sling procedures could be as effective as mesh procedures, though this was not the case for other non-mesh surgery.

The findings in relation to POP surgery are complex. Overall, the reviews conclude that the data does not provide sufficient evidence on effectiveness to guide clinical practice regarding the use of mesh implants over other surgical and non-surgical interventions.

For both SUI and POP the absence of appropriate long term data and on patient focussed outcomes was a persistent issue. Long-term data on surgical effectiveness outcomes were simply not being collected in many RCTs. Where patient important outcomes were assessed, these suggested on marginal benefits in terms of improvement in formal Quality of Life measurements, these may not be suitable measure for the types of adverse outcomes experienced by some women.

On adverse outcomes associated with mesh

The data from the reviews on adverse outcomes associated with the use of mesh procedures for SUI procedures suggests that mid-urethral mesh tape procedures were associated with fewer adverse outcomes than traditional suburethral sling or Laparoscopic colposuspension operations, though long-term, data was not collected

There were no statistically significant differences found in the risk of adverse effects between retropubic and transobturator, mid-urethral mesh tape procedures. Though once again, possible long term effects have not been adequately researched.

The clinical importance of these adverse outcomes to mesh procedures for SUI do differ. For example, bladder perforation (more common in retropubic procedures) is of limited or no clinical importance, whilst groin pain (more common for transobturator procedures) is of greater importance clinically.

The reviews were consistent in finding that the data relating to adverse outcomes following POP surgery with mesh is of relatively low quality and few conclusions can be drawn. That aid, even with the limited data that does exist, mesh erosion is one of the main adverse outcomes noted following surgical treatment with mesh

No data on adverse outcomes were reported in the systematic review on conservative management of POP compared with either mesh or traditional POP surgery.