Systematic Review of the Efficacy and Safety of Using Mesh Or Grafts in Surgery for Uterine Or Vaginal Vault Prolapse

Systematic review of the efficacy and safety of using mesh or grafts in surgery for uterine or vaginal vault prolapse

Xueli Jia, Cathryn Glazener, Graham Mowatt, David Jenkinson, Cynthia Fraser, Jennifer Burr

June 2008 INTERVENTIONAL PROCEDURES PROGRAMME

NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE

REVIEW BODY REPORT

Title Systematic review of the efficacy and safety of using mesh or grafts in surgery for uterine or vaginal vault prolapse

Produced by Health Services Research Unit University of Aberdeen 3rd Floor, Health Sciences Building Foresterhill Aberdeen AB25 2ZD

Authors Xueli Jia, Cathryn Glazener, Graham Mowatt, David Jenkinson, Cynthia Fraser, Jennifer Burr

Correspondence to Xueli Jia Research Fellow Health Services Research Unit University of Aberdeen 3rd Floor, Health Sciences Building Foresterhill AB25 2ZD Tel: (01224) 559801 Fax: (01224) 554580 Email: [email protected]

Date completed June 2008

i ‘Home unit’ details

The Health Services Research Unit (HSRU) is a multidisciplinary research group of about 50 people based in the University of Aberdeen. The Unit is core-funded by the Chief Scientist Office of the Scottish Government Health Directorates, and has responsibility for the following general remit:

1. To study or evaluate clinical activities with a view to improving effectiveness and efficiency in health care; 2. To work for the implementation of proven changes in clinical activities; 3. To encourage and support similar work throughout Scotland; 4. To train NHS staff in Scotland, and others, in the principles and practice of health services research in general, and health care evaluation in particular.

The Unit has an established portfolio of health services research focusing on two main programmes – health care assessment and delivery of care. The Unit is one of the two research groups that make up the ‘Review Body’ for the National Institute for Health and Clinical Excellence Interventional Procedures Programme. The other is based at Health Services Research, University of Sheffield.

Contributions of authors

Xueli Jia screened the search results, contacted manufacturers, assessed studies for inclusion, undertook data abstraction and quality assessment, conducted data analysis, and drafted the review. Cathryn Glazener drafted the scope, determined outcome categories, drafted the background and methods sections of the review, and commented on drafts of the review. Graham Mowatt commented on the scope of the review, drafted letters for contacting mesh/graft manufacturers for additional information, supervised the conduct of the review, and commented on drafts of the review. David Jenkinson provided advice on statistical analysis, checked the analysis results, and commented on drafts of the review. Cynthia Fraser developed and ran the literature search strategies, obtained papers, formatted the references, and drafted sections concerning search strategies and search results. Jennifer Burr supervised the conduct of the review, and commented on drafts of the review.

ii Conflict of interest

There is no conflict of interest.

Acknowledgements

We thank Chris Bain (Consultant Obstetrician and Gynaecologist, Aberdeen Royal Infirmary) for providing specialist advice and commenting on the draft of the report. The Health Services Research Unit receives a core grant from the Chief Scientist Office of the Scottish Government Health Directorates. The views expressed are those of the authors and not necessarily those of the funding bodies.

Special thanks also go to James Browning (Mpathy Medical Devices Ltd.), Christine Clarke (Bard Ltd.), Hazel Edwards (American Medical Systems (UK), Ltd.), Adrian Griffin (Johnson & Johnson Medical Ltd.), William Hynes (WL Gore & Associates (UK) Ltd.), Nathalie Verin (Boston Scientific Ltd.), and Coloplast Ltd. for providing details of properties of mesh produced by their respective companies and of published studies reporting mesh efficacy and safety.

iii TABLE OF CONTENTS Executive summary x List of abbreviations xxiii

1 OBJECTIVE OF THE REVIEW 1

2 BACKGROUND 2 2.1 Description of the underlying health problem 2 2.1.1 Epidemiology 2 2.1.2 Burden of disease 2 2.1.3 Aetiology, pathology and prognosis 3 2.2 Current management and alternative procedures 4 2.3 The interventional procedure under review 4 2.3.1 Description of the technology (mesh) 4 2.3.2 Description of the interventional procedure 5 2.3.3 Suitable candidates and relevant subgroups 8 2.3.4 Proposed clinical indications/contraindications and putative impact of the 8 procedure

3 METHODS FOR REVIEWING SAFETY AND EFFICACY 9 3.1 Search strategy 9 3.2 Inclusion and exclusion criteria 10 3.2.1 Types of studies 10 3.2.2 Types of participants 11 3.2.3 Types of interventions 11 3.2.4 Types of outcomes 11 3.3 Quality assessment 13 3.4 Data extraction 13 3.5 Data analysis 13

4 RESULTS 16 4.1 Number and type of included studies 16 4.1.1 Included studies 17 4.1.2 Ongoing studies 20 4.2 Number and type of excluded studies; reason for exclusion 22 4.3 Quality of available evidence 22 4.4 Sacrocolpopexy (mesh) 24

iv 4.4.1 Sacrocolpopexy for uterine prolapse repair (mesh) 35 4.4.1.1 Efficacy 35 4.4.1.2 Safety 35 4.4.1.3 Operation time and hospital stay 38 4.4.2 Sacrocolpopexy for vault prolapse repair (mesh) 38 4.4.2.1 Efficacy 39 4.4.2.2 Safety 41 4.4.2.3 Operation time and hospital stay 46 4.4.3 Sacrocolpopexy for uterine and/or prolapse repair (not reported separately) 47 (mesh) 4.4.3.1 Efficacy 47 4.4.3.2 Safety 49 4.4.3.3 Operation time and hospital stay 56 4.4.4 Summary of the evidence for and against sacrocolpopexy 57 4.5 Infracoccygeal sacropexy (mesh) 61 4.5.1 Infracoccygeal sacropexy for uterine prolapse repair (mesh) 68 4.5.1.1 Efficacy 68 4.5.1.2 Safety 68 4.5.1.3 Operation time and hospital stay 69 4.5.2 Infracoccygeal sacropexy for vault prolapse repair (mesh) 70 4.5.2.1 Efficacy 70 4.5.2.2 Safety 70 4.5.2.3 Operation time and hospital stay 72 4.5.3 Infracoccygeal sacropexy for uterine and/or vault prolapse repair (not reported 72 separately) (mesh) 4.5.3.1 Efficacy 72 4.5.3.2 Safety 73 4.5.3.3 Operation time and hospital stay 75 4.5.4 Summary of the evidence for and against infracoccygeal sacropexy 75 4.6 Uterine suspension sling (mesh) 78 4.6.1 Uterine suspension sling for uterine prolapse repair (mesh) 82 4.6.1.1 Efficacy 82 4.6.1.2 Safety 83 4.6.1.3 Operation time and hospital stay 85 4.6.2 Summary of the evidence for and against uterine suspension sling 85 4.7 Other mesh techniques or a mixture of mesh techniques 87 4.7.1 Uterine and/or vault repair (not reported separately) (mesh) 91

v 4.7.1.1 Efficacy 91 4.7.1.2 Safety 91 4.7.1.3 Operation time and hospital stay 93 4.7.2 Summary of the evidence for and against other mesh techniques or a mixture of 93 techniques that use mesh

5 DISCUSSION 95 5.1 Summary of main findings 95 5.1.1 Sacrocolpopexy 95 5.1.2 Infracoccygeal sacropexy 95 5.1.3 Uterine suspension sling 96 5.1.4 Other mesh techniques or a mixture of mesh techniques 96 5.2 Assumptions, limitations, and uncertainties 96

6 CONCLUSIONS 99 6.1 Implications for the NHS 99 6.2 Implications for future research 99

7 REFERENCE LIST 101

vi LIST OF TABLES

Table 1 Classification of uterine or vault prolapse stages 3 Table 2 The different mesh types in pelvic organ prolapse repair 5 Table 3 Search results 16 Table 4 Papers selected for full assessment 17 Table 5 Number of studies and reports included 18 Table 6 Characteristics of women included and treatment details 19 Table 7 Characteristics of ongoing studies 21 Table 8 Sacrocolpopexy: summary of patient characteristics and surgical procedures 25 Table 9 Sacrocolpopexy, efficacy: summary of median (range) event rates 31 Table 10 Sacrocolpopexy, safety: summary of median (range) event rates 33 Table 11 Infracoccygeal sacropexy: summary of patient characteristics and surgical 62 procedures Table 12 Infracoccygeal sacropexy, efficacy: summary of median (range) event rates 64 Table 13 Infracoccygeal sacropexy, safety: summary of median (range) event rates 66 Table 14 Uterine suspension sling: summary of patient characteristics and surgical 79 procedures Table 15 Uterine suspension sling, efficacy: summary of median (range) event rates 80 Table 16 Uterine suspension sling, safety: summary of median rates (range) 81 Table 17 Other mesh techniques or a mixture of mesh techniques (not reported 88 separately): summary of patient characteristics and surgical procedures Table 18 Other mesh techniques or a mixture of mesh techniques (not reported 89 separately), efficacy: summary of median (range) event rates Table 19 Other mesh techniques or a mixture of mesh techniques, safety: summary of 90 median rates (range)

vii LIST OF FIGURES

Figure 1 Summary of quality assessment of the RCTs (full text, n=5) 22 Figure 2 Summary of quality assessment of the non-randomised comparative studies 23 (n=17) Figure 3 Summary of quality assessment of the case series (n=29) 23 Figure 4 Sacrocolpopexy, vault prolapse repair, efficacy, subjective failure 39 Figure 5 Sacrocolpopexy, vault prolapse repair, safety, blood loss requiring transfusion 41 Figure 6 Sacrocolpopexy, vault prolapse repair, safety, damage to surrounding organs 42 Figure 7 Sacrocolpopexy, vault prolapse repair, safety, infection 44 Figure 8 Sacrocolpopexy, uterine or vault prolapse repair, efficacy, subjective failure 47 Figure 9 Sacrocolpopexy, uterine or vault prolapse repair, efficacy, objective failure 48 Figure 10 Sacrocolpopexy, uterine or vault prolapse repair, safety, number of women 50 requiring blood transfusion Figure 11 Sacrocolpopexy, uterine or vault prolapse repair, safety, damage to 51 surrounding organs Figure 12 Sacrocolpopexy, uterine or vault prolapse repair, safety, de novo urinary 53 symptoms Figure 13 Sacrocolpopexy, uterine or vault prolapse repair, safety, infection 54 Figure 14 Sacrocolpopexy, uterine or vault prolapse repair, safety, other serious adverse 55 effects not otherwise specified Figure 15 Sacrocolpopexy, uterine or vault prolapse repair, safety, other adverse effects 56 not otherwise specified Figure 16 Ranges of event rates across studies: sacrocolpopexy, efficacy 58 Figure 17 Ranges of event rates across studies: sacrocolpopexy, safety 59 Figure 18 Ranges of event rates across studies: infracoccygeal sacropexy, efficacy 76 Figure 19 Ranges of event rates across studies: infracoccygeal sacropexy, safety 76 Figure 20 Ranges of event rates across studies: uterine suspension sling, efficacy 86 Figure 21 Ranges of event rates across studies: uterine suspension sling, safety 86 Figure 22 Ranges of event rates across studies: other mesh techniques or a mixture of 94 mesh techniques, efficacy , efficacy Figure 23 Ranges of event rates across studies: other mesh techniques or a mixture of 94 mesh techniques, efficacy , safety

viii LIST OF APPENDICES Appendix 1 Search strategies 108 Appendix 2 Checklist of quality assessment of randomised controlled trials 112 Appendix 3 Checklist of quality assessment of non-randomised studies 113 Appendix 4 Included studies 114 Appendix 5 Characteristics of the included studies by mesh techniques 119 Appendix 6 Excluded studies; reasons for exclusion 160 Appendix 7 Detailed quality assessment results for included studies: randomised 169 controlled trials (full-text), n=5 Appendix 8 Detailed quality assessment results for non-randomised comparative studies, 170 n=17 Appendix 9 Detailed quality assessment results for case series, n=29 171 Appendix 10 Summary of outcomes: sacrocolpopexy, efficacy 173 Appendix 11 Summary of outcomes: sacrocolpopexy, safety 177 Appendix 12 Summary of outcomes: sacrocolpopexy, operation time and hospital stay 191 Appendix 13 Summary of outcomes: infracoccygeal sacropexy, efficacy 193 Appendix 14 Summary of outcomes: infracoccygeal sacropexy, safety 195 Appendix 15 Summary of outcomes: infracoccygeal sacropexy, operation time and 199 hospital stay Appendix 16 Summary of outcomes: uterine suspension sling, efficacy 200 Appendix 17 Summary of outcomes: uterine suspension sling, safety 201 Appendix 18 Summary of outcomes: uterine suspension sling, operation time and hospital 203 stay Appendix 19 Summary of outcomes: other mesh techniques or a mixture of mesh 204 techniques, efficacy Appendix 20 Summary of outcomes: other mesh techniques or a mixture of mesh 205 techniques, safety Appendix 21 Summary of outcomes: other mesh techniques or a mixture of mesh 207 techniques, operation time and hospital stay

ix EXECUTIVE SUMMARY Background

Pelvic organ prolapse (POP) is a protrusion of one or more pelvic organs (bladder, rectum, uterus, vaginal vault, bowel) through vaginal fascia into the vagina and the downward displacement (‘prolapse’) of the associated vaginal wall from its normal location to or outside the vaginal opening. POP affects a woman’s quality of life by its local physical effects (pressure, bulging, heaviness or discomfort) or its effect on urinary, bowel or sexual function. POP can be classified according to the compartment affected as: anterior vaginal wall prolapse (urethrocele, cystocele); posterior vaginal wall prolapse (rectocele, enterocele); prolapse of the uterus; and prolapse of the vaginal vault (which can only occur after prior hysterectomy). A woman can present with prolapse of one or more of these sites. The present review focuses on uterine and vaginal vault pelvic organ prolapse.

Current treatment options for upper vaginal prolapse include pelvic floor muscle training (PFMT), use of pessaries (mechanical devices such as rings or shelves), and surgery. Surgical repairs which necessarily involve the use of mesh include: • Sacrocolpopexy; • Infracoccygeal sacropexy (also known as Posterior IntraVaginal Slingplasty, IVS); • Uterine suspension sling (including sacrohysteropexy); and • Other mesh techniques such as sacrocolpoperineopexy.

Surgeries which do not involve mesh include: • Hysterectomy; • Cervical amputation (often called Manchester repair); and • Uterine/vault suspension (without sling).

There are numerous types of mesh materials available, which vary according to type of material, structure, and physical properties such as absorbability. There are no existing classification systems for mesh. For the purposes of this review we have defined four classes of mesh: absorbable synthetic mesh, biological graft, combination of absorbable and non-absorbable mesh (termed ‘combined’ hereafter), and non-absorbable synthetic mesh.

Objective

To systematically review the evidence for the efficacy and safety of mesh in surgery for uterine or vaginal vault pelvic organ prolapse. The specific objectives were to:

x (a) Report efficacy and safety of procedures using mesh (each type); (b) Compare efficacy and safety between procedures using mesh and no mesh; (c) Compare efficacy and safety of different procedures that use mesh; and (d) Compare efficacy and safety of different types of mesh for the same procedure.

Methods

We searched electronic databases, conference proceedings and relevant websites, contacted experts in the field and scanned bibliographies of retrieved papers. Extensive highly sensitive electronic searches were conducted to identify reports (both full-text papers and conference abstracts) of published and ongoing studies on the safety and efficacy of mesh used in the repair of pelvic organ prolapse. Searches were restricted to publications from 1980 onwards and to those published in the English language. Conference proceedings from five major urogynaecological organisations were searched from 2005 onwards for additional reports of randomised controlled trials (RCTs). Reference lists of all included studies were scanned to identify additional potentially relevant studies. Eleven manufacturers were identified and contacted for information on properties of mesh produced and for any studies related to mesh.

We considered the following types of studies: • RCTs (full-text publication) • RCTs (conference abstracts from 2005 onwards) • Non-randomised comparative studies (full-text publication) • Case series (full-text publication). As the volume of literature for sacrocolpopexy was substantial, only case series with a sample size of at least 100 women were included. There was no sample size restriction placed on case series that reported uterine suspension sling or infracoccygeal sacropexy as the volume of literature for these techniques was very limited. • Population-based registry reports that prospectively collected audit data on the use of mesh for uterine or vault prolapse.

We did not search for studies that only reported procedures without mesh.

One reviewer screened titles/abstracts and undertook data extraction. The same reviewer assessed the quality of the included studies using one of two separate checklists according to study design. Both checklists were developed through the Review Body for Interventional Procedures.

xi For each mesh technique, data analysis was conducted for three subgroups of women according to the type of prolapse being repaired: • Uterine prolapse; • Vault prolapse; and • Uterine or vault prolapse (not reported separately).

Meta-analyses of full-text RCTs, RCTs available as conference abstracts, and non- randomised comparative studies, using Cochrane Collaboration Review Manager (RevMan 4.2) software, were conducted to directly compare the efficacy and safety of procedures using mesh versus procedures without mesh, and between different techniques that used mesh.

Crude event rates for each of the intervention categories were tabulated for each outcome by study design (RCTs, non-randomised comparative studies, case series/registry) to illustrate the potential heterogeneity of event rates across different study designs. The median and range of the event rates for each study design were also reported, to express the variability of the data within each study design.

Results

From the initial 655 publications identified by the search strategy, 54 studies (reported in 60 publications) were included, of which 5 were full-text RCTs, 3 were RCTs available as conference abstracts, 17 were non-randomised comparative studies, and 29 were case series. Seven manufacturers provided data on mesh properties or related studies, all of which had already been identified by our searches. The table on the following page shows the 54 included studies by type of prolapse repair and by study design.

The studies took place during the period 1991 – 2007 in 16 countries. Overall, 7054 women were treated in total, of whom 4456 were treated with sacrocolpopexy, 976 with infracoccygeal sacropexy, 159 with uterine suspension sling, 520 with other mesh techniques (sacrocolpoperineopexy) or a mixture of the above mesh techniques, and 943 with no-mesh techniques. Across studies, the average age of the women was 61 years. Women who were treated with uterine suspension sling were younger (average age 37 years) than those treated with other techniques. Only six studies provided information on whether the procedures were primary or secondary. The average follow up was about two years for sacrocolpopexy, one year for infracoccygeal sacropexy, and three years for uterine suspension sling.

xii

Number of studies and reports included Study design Studies Sacrocolpopexy Infracoccygeal Uterine Other mesh sacropexy suspension techniques or a sling mixture of mesh techniques Uterine repair RCTs (full text) 1 0 0 1 0 RCTs (conference abstracts) 1 1 0 0 0 Non-randomised comparative studies 4 2 1 1 0 Case series with sample size ≥100 1 1 0 0 0 Case series with sample size <100 5 0 1 4 0 Subtotal 12 4 2 6 0 Vault repair RCTs (full text) 2 2 0 0 0 RCTs (conference abstracts) 1 0 1 0 0 Non-randomised comparative studies 3 3 0 0 0 Case series with sample size ≥100 4 4 0 0 0 Case series with sample size <100 5 0 5 0 0 Subtotal 15 9 6 0 0 Uterine or vault repair RCTs (full text) 2 2 0 0 0 RCTs (conference abstracts) 1 0 1 0 0 Non-randomised comparative studies 12 10 0 0 2 Case series with sample size ≥100 13 9 2 0 2 Case series with sample size <100 4 0 4 0 0 Subtotal 32 21 7 0 4 Total 54* 32 14 6 4

*5 studies double counted: Three studies reported different types of prolapse repair separately. Two studies reported different mesh techniques separately.

The methodological quality of the full-text studies only was assessed. For the five RCTs, all studies reported methods of randomisation. Treatment allocation was adequately concealed in three studies. All follow-up periods were one year or more. All studies used intention-to- treat analysis in that women were analysed according to the group to which they were randomised. Among the 17 non-randomised comparative studies, only three studies collected data prospectively. Mean follow up was more than one year in 13 studies. The loss to follow up rates ranged from 4% to 55%. Most studies did not report the characteristics of women who were lost to follow up. For the 29 case series, 12 studies collected data

xiii prospectively. Mean follow up was one year or more in 25 studies. In 12 studies reporting loss to follow up, rates varied from 2% to 74% with follow up ranging from 2 to 66 months.

Nine ongoing studies were also identified. Seven of these were RCTs. For studies reporting a follow-up period, all follow-up periods were one year or more. These studies are due to be completed between 2008 and 2014.

Summary of evidence: sacrocolpopexy (mesh)

Thirty two studies (4 full-text RCTs, 1 RCT available as a conference abstract, 15 non- randomised comparative studies, and 12 case series with a sample size over 100) involving a total of 4456 women provided data on sacrocolpopexy. The average follow up was around two years across studies. Figures A and B show the ranges of event rates for efficacy and safety respectively for sacrocolpopexy, together with the number of women involved for each outcome.

Efficacy

Figure A Ranges of event rates across studies: sacrocolpopexy, efficacy

50 NA, 38/209, 94/429 NA, 4/9, 25/60

40 NA, 10/32, 3/193 NA, NA, 43/169 0/39, 15/60, 17/352 30

20 0/62, 8/306, 12/686 NA, NA, 7/54 Event rates (%) rates Event 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent urinary Persistent bowel Persistent sexual other sites prolapse symptoms symptoms problems

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair

Notes: 1. Where only one study reported the outcome, a single horizontal bar reports its event rate; where more than one study reported the outcome, the range of the event rates across all studies is shown by a vertical bar. 2. The numbers in the figures are the number of events which occurred across all studies divided by the total number of women who received the treatment. 3. NA: no studies reported data for that outcome.

xiv The objective failure rates ranged from 0% to 6.1% in 11 studies involving 1054 women, while the subjective failure rates ranged from 3.3% to 31% in 9 studies involving 638 women. In 4 studies involving 451 women, rates for those needing a further operation for recurrent or de novo prolapse ranged from 0% to 14.4%. Estimates for other outcomes were based on single studies only (shown as a single horizontal bar in the figure). Meta-analysis was carried out to compare subjective failure and objective failure in sacrocolpopexy versus sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in either outcome between the operations.

Safety

Figure B Ranges of event rates across studies: sacrocolpopexy, safety 3/39, 55/621, 158/731 70

40 2/39, 5/473, 57/1568 20/267, 45/969, 53/1686 NA, 2/22, 18/272 NA, NA, 12/87 6/62, 10/299, 68/1390 30 NA, 14/427, 32/1296 7/143, 18/610, 47/1321 NA, NA, 2/178 2/62, 30/309, 77/1688 Event rates (%) rates Event 20

0 Blood Organ Mesh Operation for New urinary New bowel New sexual Infection Other Other transfusion damage erosion erosion symptoms symptoms symptoms serious adverse adverse effects effects

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair (not separated)

The proportion of women who required a blood transfusion for sacrocolpopexy ranged from 0% to 17% (19 studies, n=2080). The range for women with organ damage varied from 0% to 7.9% (15 studies, n=1723). Mesh erosion occurred in 0% to 12% of women (27 studies, n=2922), of whom 0% to 11% required an operation for mesh erosion (17 studies, n=2074). New urinary symptoms in women who did not have these symptoms at baseline occurred in 3.8% to 9.2% (4 studies, n=294). The estimate for new bowel symptoms (1.1%, 2/178) was based on a single study, and the estimate for new sexual symptoms (range 9.1% to 15%,

xv n=87) was based on two small studies. The range of infection was wide (0.8% to 68%, 17 studies, n=1391).

The estimates for other serious adverse effects (not otherwise specified) ranged from 0% to 16% (21 studies, n=2059). The events that occurred included bowel obstruction, ureteral obstruction, wound dehiscence, incisional hernia, vesico-cutaneous fistula, deep vein thrombosis, pulmonary embolism, pneumonia, myocardial infarction, congestive heart failure, arrhythmia, and left ventricular failure. The proportion of women with other adverse effects (not otherwise specified) ranged from 0% to 16% (17 studies, n=1751). The events that occurred included voiding dysfunction, haematoma, persistent vaginal discharge, vaginal bleeding, phlebitis, wound complications, gastrointestinal complications, e.g. nausea, emesis, bloating or subacute intestinal obstruction, sciatica, and buttock pain.

Meta-analysis was carried out to compare blood transfusion, damage to surrounding organs, infection, new urinary symptoms and other adverse effects (not otherwise specified) between sacrocolpopexy (mesh) and sacrospinous colpopexy (no mesh). There was no evidence of a difference in any of these outcomes between the operations. Meta-analysis was also conducted for other serious adverse effects (not otherwise specified): in four non- randomised comparative studies, sacrocolpopexy (mesh) was associated with significantly more adverse events than sacrospinous colpopexy (no mesh) (pooled RR 6.94, 95% CI 1.02 to 26.38).

Operation time and hospital stay

The average operation time varied from 89 minutes to 267 minutes (based on 15 studies), whereas average hospital stay after sacrocolpopexy ranged from 1 to 7 days. The majority of the studies reported stays of 3 to 5 days (11/14 studies).

Summary of evidence: infracoccygeal sacropexy (mesh)

Fourteen studies (2 RCTs available as conference abstracts, 1 non-randomised comparative study, 2 case series with a sample size ≥100, and 9 case series with a sample size <100) involving 976 women provided data on infracoccygeal sacropexy. The average follow up was around one year across studies. The ranges of event rates for safety and efficacy, together with the number of women involved are shown in Figures C and D respectively. Meta-analysis was not possible because no studies used the same comparator.

xvi Figure C Ranges of event rates across studies: infracoccygeal sacropexy, efficacy

60 NA, NA, 3/7 50 0/10, 14/151, 10/127 40 2/89, 4/60, 18/253 NA, NA, 5/17 30 NA, 8/91, 28/171 20 Event rates (%) rates Event NA, NA, 3/63 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent Persistent bowel Persistent sexual other sites prolapse urinary symptoms (no problems symptoms data)

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair (not separated)

Figure D Ranges of event rates across studies: infracoccygeal sacropexy, safety

40 10/79, 24/265, 33/545 30 NA, 0/183, 11/492 0/10, 3/203, 1/492 10/79, 4/75, 32/524 20 0/10, 0/188, 1/185 0/55, NA, 0/27 NA, 6/183, 16/515 Event rates (%) rates Event 10

0 Blood Organ Mesh Operation for New urinary New bowel New sexual Infection Other Other transfusion damage erosion erosion symptoms symptoms symptoms serious adverse (no data) (no data) adverse effects effects (no data)

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair (not separated)

The proportion of women with subjective failure ranged from 2.3% to 21% (n=262, 3 studies) after infracoccygeal sacropexy, whereas the range for objective failure was from 0% to 25% (9 studies, n=402). In 3 studies (n=288), 0% to 30% of women required re-operation for prolapse. Estimates for other outcomes were based on single studies including small numbers of women.

xvii The need for blood transfusion ranged from 0% to 2.3% (7 studies, n=383). The proportion of women with organ damage ranged from 0% to 2.7% (9 studies, n=684). The occurrence of mesh erosion ranged from 0% to 21% (11 studies, n=889), and 1.5% to 17% women needed an operation for mesh erosion (6 studies, n=678). Little evidence was available for new urinary symptoms, bowel symptoms, and sexual symptoms in women who did not have these symptoms at baseline. Infection ranged from 0% to 9.0% (8 studies, n=698). Other adverse effects (not otherwise specified) ranged from 0% to 10% (8 studies, n=675). The events that occurred included haematomas, urine retention, pain at skin incision, and vaginal pressure.

Average operation times ranged from 10 minutes to 149 minutes with the majority of the studies (8/11) reporting operation time of less than one hour. Average hospital stays ranged from less than 24 hours to 5 days with the majority of the studies (7/11) reporting two days or less.

Summary of evidence: uterine suspension sling (mesh)

Six studies (one full-text RCT, one non-randomised comparative study, and four case series) involving 239 women reported uterine suspension sling, of which five studies involving 219 women reported sacrohysteropexy. The average follow up across the six studies was around three years across studies. Figure E (efficacy) and Figure F (safety) show the range of events across studies, together with the number of women involved. Meta-analysis was not possible because the studies used different comparators.

The range for women reporting subjective failure was wide (0% to 39%, 3 studies, n=91). Women judged to have objective failure ranged from 0% to 7.7% (5 studies, n=136). The proportion of women who required a re-operation for prolapse ranged from 0% to 22% (3 studies, n=107). Little evidence was available for estimating other efficacy outcomes or any of the safety outcomes. The average operation time for uterine suspension sling ranged from 97 to 115 minutes (based on 3 studies). In one study, the average hospital stay was 7 days.

xviii Figure E Ranges of event rates across studies: uterine suspension sling, efficacy

50 17/91 40 10/107 1/5 30 3/19 3/27 20 4/136

Event rates(%) 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent urinary Persistent bowel Persistent sexual other sites (no prolapse symptoms symptoms problems data)

Uterine prolapse repair

Figure F Ranges of event rates across studies: uterine suspension sling, safety

1/2

40 20/126 30 10/101 20 3/77 0/91 1/66 1/66 0/25 5/101

Event rates (%) rates Event 10

Summary of evidence: other mesh techniques or a mixture of mesh techniques

xix Figure G Ranges of event rates across studies: other mesh techniques or a mixture of mesh techniques, efficacy

20 2/169 1/169 6/297

Event rates(%) 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent urinary Persistent bowel Persistent sexual other sites (no prolapse symptoms (no symptoms (no problems (no data) data) data) data)

Uterine and/or vault prolapse repair (not separated)

Figure H Ranges of event rates across studies: other mesh techniques or a mixture of mesh techniques, safety

4/9 50 18/173 40

30 12/342 20 2/45 11/342 27/415 14/415 9/342

Event rates (%) rates Event 10

The proportion of women who needed a re-operation for prolapse ranged from 0% to 4.7% (2 studies, n=297). Organ damage, mesh erosion, and operation for mesh erosion all occurred in less than 10% of women. In one study, the average operation time was 192 minutes, and in two studies, the average hospital stay ranged from 3 to 7 days.

xx Conclusions

The review indicated that although sacrocolpopexy (mesh) is associated with an objective failure ranging from 0% to 6% at an average follow up of two years, the incidence of subjective failure (range 3% to 31%) and need for a further prolapse operation (range 2% to 14%) may be higher. There were risks from adverse effects such as blood transfusion (range 0% to 17%), infection (range 1% to 69%), mesh erosion (range 4% to 12%) and the need for a further operation for mesh erosion (range 1% to 11%). There was no evidence of a statistically significant difference between sacrocolpopexy (mesh) and sacrospinous colpopexy (no mesh) in subjective failure and objective failure but the numbers were small.

For infracoccygeal sacropexy (mesh), subjective failure rates ranged from 2.3% to 21% at an average follow up of one year, objective failure 0% to 25%, mesh erosion 0% to 21%, and operation for mesh erosion 1.5% to 17%. Operation times (the majority were less than one hour) and hospital stays (the majority were two days or less) were relatively short compared with other operations that use mesh.

For uterine suspension sling, the objective failure rates ranged from 0% to 7.7% at an average follow up of three years, whereas subjective failure ranged from 0% to 39%. However, little evidence was available for safety outcomes.

Sacrocolpoperineopexy was another mesh technique that had been used, but there was little evidence available for assessing the efficacy or safety of this technique.

It was not possible to compare the efficacy and safety between procedures using mesh (each type) and no mesh, between different procedures that use mesh, or between different types of mesh for the same procedure, because very few studies used the same comparator. It was also not possible to compare differences in efficacy and safety for women having primary versus recurrent prolapse repairs. Of the 54 included studies, only two exclusively reported on women having primary repairs, and the remainder did not report outcomes for these subgroups separately.

Implications for future research

The RCTs currently available only as conference abstracts may be available as full-text studies in the future. Full-text papers of the conference abstracts (RCTs published before 2005 and all non-randomised comparative studies and case series) that were excluded may

xxi also be available in the future. The expected completion times of the nine on-going studies (seven RCTs, one registry, and one case series) range from 2008 to 2014. A further review incorporating these studies may be informative.

Subjective symptoms (including quality of life related to prolapse) are important to women. Identifying the need for further operations for recurrent prolapse and adverse effects requires long term follow up over many years.

The RCTs included in the present review were relatively small. Therefore, high quality long- term RCTs, with adequate power to detect clinically meaningful differences, with validated patient-reported outcome measures, primarily comparing the failure rate between mesh and no-mesh techniques, and between different types of mesh techniques and between different types of mesh, taking account of the outcomes important to women, are required to determine the comparative efficacy of mesh techniques and their optimal place in clinical practice.

In addition, prospective data collection should be considered in which the operative and clinical details of women undergoing prolapse surgery with mesh can be recorded so that sufficient efficacy and safety data can be gathered to guide the use of mesh or grafts in the future. In particular, long term outcomes need to be considered as it may be many years before a prolapse recurrence becomes symptomatic.

xxii LIST OF ABBREVIATIONS

BMI Body Mass Index CENTRAL Cochrane Central Register of Controlled Trials CI Confidence interval GATE Generic Appraisal Tool for Epidemiology PFMT Pelvic Floor Muscle Training pIVS Posterior intravaginal slingplasty (= infracoccygeal sacropexy) POP Pelvic Organ Prolapse POP-Q Pelvic Organ Prolapse Quantification System RCOG Royal College of Obstetricians and Gynaecologists RCT Randomised Controlled Trial ReBIP Review Body for Interventional Procedures RR Relative risk SCI Science Citation Index SMD Standardised Mean Difference

xxiii 1 OBJECTIVE OF THE REVIEW

The aim of this study was to systematically review the evidence for the efficacy and safety of mesh in surgery for uterine or vaginal vault pelvic organ prolapse. The specific objectives were to: (a) Report efficacy and safety of procedures using mesh (each type); (b) Compare efficacy and safety between procedures using mesh and no mesh; (c) Compare efficacy and safety of different procedures that use mesh; and (d) Compare efficacy and safety of different types of mesh for the same procedure.

1 2 BACKGROUND

2.1 Description of the underlying health problem

2.1.1 Epidemiology

Pelvic organ prolapse (POP) is a protrusion of one or more pelvic organs (bladder, rectum, uterus, vaginal vault, bowel) through vaginal fascia into the vagina and the downward displacement (‘prolapse’) of the associated vaginal wall from its normal location to or outside the vaginal opening.1 POP can be classified, according to the compartment affected, as: 1. Anterior vaginal wall prolapse (urethrocele, cystocele), 2. Posterior vaginal wall prolapse (rectocele, enterocele), 3. Prolapse of the uterus (cervix if the remaining part of the uterus was removed previously), 4. Prolapse of the vaginal vault (which can only occur after hysterectomy).

A woman can present with prolapse of one or more of these sites. The present review focuses on uterine and vault prolapse. We focused on anterior and/or posterior vaginal wall prolapse in a previous review.2

In the UK, around 48,500 women undergo hysterectomy each year, of which about 15,000 are vaginal procedures, and 8000 of these are for vaginal prolapse.3 Approximately 2,800 women will require further surgery at a later time for vault prolapse (35% of the women who have a vaginal hysterectomy for prolapse).4 Uterine or vault prolapse has a major impact on quality of life and on the use of NHS services.5

2.1.2 Burden of disease

POP affects a woman’s quality of life by its local physical effects (pressure, bulging, heaviness or discomfort) or its effect on urinary, bowel or sexual function. Urinary symptoms include both symptoms related to incontinence or urinary retention (incomplete emptying), bowel symptoms include constipation or faecal incontinence, and symptoms of sexual dysfunction include dyspareunia (pain during intercourse) or avoiding intercourse due to embarrassment. There is no symptom which is specific to uterine or vault prolapse alone. Women may present with any or none of these symptoms.

2 2.1.3 Aetiology, pathology and prognosis

Vaginal delivery is the greatest risk factor for POP. Neuromuscular injury6,7 and mechanical injury during labour (for example damage to the fascial connective tissue or pelvic floor muscles)8 has been identified as the main reason for POP. Nulliparous women with connective tissue disorders (e.g., Ehlers-Danlos or Marfan’s syndrome) or women with spina bifida and poorly functioning pelvic floor muscles are susceptible to POP and can present with advanced POP at a young age.

A number of objective clinical classification or grading systems are in use. The simplest include the ordinal grading of prolapse as first, second and third degree such as the Baden- Walker9 and Beecham10 classification systems. In an attempt to improve the accuracy of prolapse assessment, the International Continence Society has devised the POP quantification (POP-Q) System with proven inter-observer and intra-observer reliability.11,12 It allows for a standardised, accurate objective description of physical findings, and provides a means of quantification of change in individual patients, and comparison between series of patients. The staging depends on the position of the leading edge of the prolapse: the lowest or worst level of maximum protrusion is used. The Baden-Walker and POP-Q classifications are summarised in Table 1.

Table 1 Classification of uterine or vault prolapse stages Baden-Walker classification system POP-Q system - - Stage 0 No descent of pelvic structures during straining

First The cervix or vaginal apex descends halfway to the Stage I The leading edge of the prolapse is >1 degree hymen. cm above the hymenal ring.

Second The cervix or vaginal apex extends to the hymen or Stage II The leading edge of the prolapse extends degree over the perineal body. from 1 cm above the hymen to 1 cm below the hymenal ring.

Third The cervix and corpus uteri extend beyond the hymen Stage III The leading edge of the prolapse extends degree or the vaginal vault is everted and protrudes beyond to between 2 and 3 cm below the the hymen. hymenal ring, but there is not complete vaginal eversion.

- - Stage IV The vagina is completely everted.

3 2.2 Current management and alternative procedures

Current treatment options for uterine or vault prolapse include pelvic floor muscle training (PFMT), use of pessaries (mechanical devices such as rings or shelves), and surgery.13 Surgery is usually considered when the prolapse is more severe and is causing symptoms or because conservative treatment or a pessary do not work or are unacceptable to patients. Recent developments to try to reduce the need for re-operation have been the development of new operations such as sacrocolpopexy and infracoccygeal sacropexy which involves the use of mesh.

2.3 The interventional procedure under review

2.3.1 Description of the technology (mesh)

A variety of types of mesh are used for POP repair2. Table 2 summarises the most common types of mesh according to their absorbability, material, and structure.

Absorbable synthetic mesh (polyglactin) loses 75% of its strength after 21 days. Its purpose is to act as a supportive material for the formation of granulation tissue, and therefore is not suitable for operations which depend on the mesh to form a permanent bridge between two anatomical structures e.g. ligament, fascia, and organs.

Biological grafts show a spectrum of absorbability, and are eventually absorbed by the host tissues to a greater or lesser degree. Some may be so long lasting that they are suitable for use as a bridge between two anatomical structures.

Combined mesh/graft materials use the permanent strength of a non-absorbable mesh to provide the bridge function but are covered with other materials which are thought to decrease the risk of erosion and promote integration with the host tissues (e.g. collagen, and polyglactin).

Non-absorbable synthetic mesh does not degrade and acts as a mechanical bridge between tissues. It can be made from knitted single-fibre filaments (monofilament materials) or they can be braided with monofilament yarns, or woven as multifilament fibres in different ways. Non-absorbable synthetic mesh is usually further classified according to its pore size, and can be divided into four types (Amid types):14-16 totally macroporous with a pore size of >75μm (type I), totally microporous (type II), a mixture of macro and microporous (type III),

4 and submicronic (type IV). It is thought that the various fibre structures and pore sizes result in different mesh attributes (e.g. risk of infection, rejection and erosion), but it is not clear how function is affected by these attributes.

The choice of mesh therefore depends on a balance between strength and degree of absorbability and the risk of infection, rejection or erosion.

Table 2 The different mesh types in pelvic organ prolapse repair Type of mesh and grafts Material Examples (commercial or trade name) Absorbable synthetic mesh Polyglactin Vicryl (multifilament)

Biological graft Autograft Fascia lata NA (patient’s own tissue used) Rectus sheath NA (patient’s own tissue used)

Allograft Fascia lata Tutoplast Dura mata Duraderm Dermis AlloDerm

Xenografts Porcine dermis Pelvicol Porcine small-intestine submucosa SIS

Combined mesh/graft Polyglactin and polypropylene Vypro II (mono-multifilament) Porcine collagen and Avaulta (monofilament) polypropylene Non-absorbable synthetic mesh Type I (totally macroporous) Polypropylene Marlex (monofilament) Prolene (monofilament) Type II (totally microporous) Polytetrafluoroethylene Goretex (multifilament) Type III (macro-microporous) Polyethylene tetraphalate Mersilene (multifilament) Type IV (submicronic pore size) Silicone or silicone-coated mesh Polyatex

2.3.2 Description of the interventional procedure

Types of surgical repair which necessarily involve the use of mesh include sacrocolpopexy, infracoccygeal sacropexy, uterine suspension sling, and other mesh techniques such as colpoperineopexy. Types of surgeries which do not involve mesh include hysterectomy, cervical amputation, and uterine/vault suspension.

5 The procedures, with or without mesh, are usually conducted as an inpatient procedure under a general or spinal anaesthetic. Some women may have other gynaecological problems which require surgical correction at the same time. These concomitant procedures include anterior or posterior vaginal wall prolapse repair or surgery for urinary incontinence. These are usually conducted under the same anaesthetic.

Sacrocolpopexy (mesh)

Sacrocolpopexy can be carried out for vault prolapse or at the same time as hysterectomy for uterine prolapse with the aim of preventing vault prolapse. It involves attachment of the vaginal vault to the sacrum with the use of mesh as a necessary bridge. The mesh is attached to both the vault and the sacrum with sutures. This procedure is carried out abdominally (open or by laparoscope).

Infracoccygeal sacropexy (mesh)

Infracoccygeal sacropexy (also known as Posterior IntraVaginal Slingplasty, IVS) can be carried out for either uterine or vault prolapse. In women who have a uterine prolapse, the uterus can be preserved. Infracoccygeal sacropexy is usually carried out under local or regional anaesthesia. An incision is made in the posterior wall of the vagina and a small puncture incision is made in each buttock. A tape is introduced through the buttock incision and a tunnelling device is used to take it up around the rectum, along one side of the vagina, across the top of the vagina, down along the other side of the vagina and out through the incision in the other buttock. The tunneller is guided by a finger in the vaginal incision. The tape is sutured to the top of the vagina and acts as a tension-free sling to suspend the vaginal vault in its natural position. Delayed absorbable sutures are used to close the vaginal incision.

Uterine suspension sling (including sacrohysteropexy) using mesh

'Uterine suspension sling' involves the attachment of the uterus, using mesh, either to the sacrum or ligaments. Two main variants of the procedure exist (both necessitating an abdominal approach, using either open or laparoscopic techniques): Firstly, sacrohysteropexy, which involves the attachment of the uterus to the sacrum using a mesh bridge. In sacrohysteropexy, the mesh can be attached to the uterus in two different ways: either in the midline, and usually at the level of the back of the uterine cervix; or bilaterally from the insertion points of each uterosacral ligament at the uterus. (In both cases the other

6 end of the mesh is attached to the sacrum). Secondly, another mesh suspension technique involves attaching the uterus at the level of the front of the uterine cervix to the lateral pectineal ligaments. Each of the above procedures can be described as a 'uterine suspension sling'.’

Other mesh techniques

A variation of sacrocolpopexy is where extra mesh is inserted between the anterior vaginal wall and the bladder, and/or between the posterior vaginal wall and the rectum/colon (sometimes down to the perineum, colpoperineopexy). The extra mesh may be absorbable (biological) aiming to reduce the risk of erosion. The extra mesh is then attached at the vault to the mesh bridge to improve support through broader coverage of the vagina. This can be done as an abdominal or a combined abdominal-vaginal procedure.

Hysterectomy (no mesh)

A hysterectomy can be carried out for uterine prolapse. It can be done vaginally or abdominally. If carried out via the abdomen, it is possible to do a total or sub-total hysterectomy. In a sub-total abdominal hysterectomy, the lower (cervical) half of the uterus is not removed. After total hysterectomy, it is normal to support the vault intra-operatively by suturing the vaginal vault to the uterosacral ligaments (uterosacral suspension) or the sacrospinous ligament(s) (sacrospinous suspension), or closing the area between the rectum (levatorplasty) and vagina (culdoplasty) to prevent a hernia of the bowel. An anterior and/or posterior repair can be carried out at the same time as a vaginal hysterectomy if there is coexisting anterior or posterior prolapse.

Cervical amputation (no mesh)

Vaginal cervical amputation (often called Manchester repair) can be carried out for uterine prolapse instead of a vaginal hysterectomy. It involves removal of the lower (cervical) half of the uterus from below. The top of the vagina is supported by plication to the cardinal (transverse cervical) ligaments. An anterior and/or posterior repair can be carried out at the same time if there is co-existing anterior or posterior prolapse.

7 Uterine/vault suspension (no mesh)

Uterine/vault suspension can be carried out for either uterine or vault prolapse. The uterus or vault is sutured to the pelvic ligaments, e.g. sacrospinous ligament (also known as sacrospinous suspension, fixation, or colpopexy [for vault prolapse]) or the uterosacral ligaments (also known as uterosacral suspension, fixation, or colpopexy [for vault prolapse]). The procedure can be carried out vaginally or abdominally.

2.3.3 Suitable candidates and relevant subgroups

Mesh repair is theoretically suitable for any degree of symptomatic uterine or vault prolapse. However, prolapse repair is generally not suitable for women who wish to have more children, even if uterine preservation is considered possible.

2.3.4 Proposed clinical indications/contraindications and putative impact of the procedure

Women with prolapse only require treatment if they have symptoms which affect their quality of life. Surgical treatment is only indicated if conservative treatments such as PFMT, pessaries or adjustment of lifestyle risk factors have failed or are not possible.

Surgical treatment is contraindicated in women who are not fit for surgery or anaesthesia, but it is a matter of clinical judgment when this applies to individual women. Surgical prolapse repair is generally not appropriate if the conditions which have caused or exacerbated it (such as obesity, heavy lifting, constipation, smoking or chest problems resulting in coughing) will still be present afterwards. Most gynaecologists would regard a BMI over 35 as a complete contraindication, and would be reluctant to operate if the BMI is between 30 and 35.

The aim of treatment is the cure or improvement of the prolapse symptoms which are affecting quality of life, without introducing new symptoms which may present further problems.

8 3 METHODS FOR REVIEWING SAFETY AND EFFICACY

3.1 Search strategy

The literature search previously undertaken, primarily for the previous review on anterior and posterior prolapse, had been designed to retrieve all relevant publications for the repair of pelvic organ prolapse and included uterine and vaginal vault prolapse repair as this was felt to be the most efficient approach. We previously screened these search results to identify potentially relevant studies for both reviews, and from this 198 reports were set aside for full assessment for this review. Subsequently, we designed more specific search strategies for this review, focusing on uterine and vaginal vault prolapse repair only.

We searched electronic databases, conference proceedings and relevant websites, contacted experts in the field and scrutinised bibliographies of retrieved papers to identify reports (both full-text papers and conference abstracts) of published and ongoing studies on the efficacy and safety of mesh used in the repair of pelvic organ prolapse. Searches were restricted to publications from 1980 onwards, to those published in the English language and, for conference proceedings, to randomised controlled trials.

The databases searched were: Medline (1980 – Nov Wk 2 2007), Medline In-Process (4th Jan 2008), EMBASE (1980 – 2008 Wk 1), Biosis (1985 – 3rd Jan 2008), Science Citation Index (1980 – 5th Jan 2008), Cochrane Controlled Trials Register (The Cochrane Library, Issue 4 2007), and ISI Conference Proceedings (1990 – 11th Feb 2008) as well as current research registers (National Research Register (Issue 2, 2007), Current Controlled Trials (Jan 2008) and Clinical Trials (Jan 2008). Additional databases searched for systematic reviews and other background information included the Cochrane Database of Systematic Reviews (The Cochrane Library, Issue 4, 2007), Database of Abstracts of Reviews of Effectiveness (Jan 2008) and the HTA Database (Jan 2008). Conference proceedings of major urogynaecological organisations (American Urogynecologic Society, American Urological Association, European Association of Urology, European Society of Gynecological Endoscopy, International Continence Society and International Urogynecological Association) for 2005 onwards were scrutinised for additional reports of randomised controlled trials (RCTs). We scanned reference lists of all included studies to identify additional potentially relevant studies. Appendix 1 documents full details of the search strategies used and websites consulted.

9 We identified eleven manufacturers and contacted them for properties of mesh/graft produced and for information on any studies related to mesh/graft. The eleven manufacturers were: American Medical Systems UK Ltd., Atrium Medical Corp., Bard Ltd., Boston Scientific Ltd., Johnson & Johnson Medical Ltd., WL Gore & Associations (UK) Ltd., Coloplast Ltd., Mpathy Medical Devices Ltd., Tyrx Pharma Inc., Cook Biotech Inc., and Organogenesis Inc.

3.2 Inclusion and exclusion criteria

3.2.1 Types of studies

We considered the following types of studies: • Randomised controlled trials (RCTs) (full text, i.e. as published in journals) • RCTs (conference abstracts from 2005 onwards) • Non-randomised comparative studies (full text) • Case series (full text). As the volume of literature for sacrocolpopexy was substantial, only case series with a sample size of at least 100 women were included. There was no such restriction placed on infracoccygeal sacropexy or uterine suspension sling due to the limited evidence available. • Population-based registry reports that prospectively collected audit data on the use of mesh for uterine or vault prolapse.

We did not search for studies that only reported procedures which do not use mesh, hence data related to such procedures are derived from control groups in RCTs and non- randomised comparative studies only. We excluded case reports as we considered that there were sufficient data to capture rare adverse events. For conference abstracts, only those published from 2005 onwards were screened and only RCTs were considered. Other conference abstracts (i.e. published before 2005 and non-RCTs published from 2005 onwards) were excluded in order to ensure that the number of included studies remained manageable within the timeframe for the review. We excluded earlier reports of studies reproduced in later publications, where the results related to the same outcomes. Animal studies were not considered.

For case series and registries, we considered one year a minimum adequate period of time in which to assess the efficacy of prolapse repair. Studies with a mean follow up of less than one year were included for safety outcomes only.

10 3.2.2 Types of participants

We considered women undergoing surgery for uterine or vault prolapse. We excluded studies of women with prolapse caused by pelvic trauma, congenital disease, or prolapse after creation of neovagina. Women could be undergoing other concomitant operations, such as anterior or posterior repair or a continence procedure as long as the main indication for surgery was uterine or vault prolapse.

3.2.3 Types of interventions

The interventions considered were: • sacrocolpopexy; • infracoccygeal sacropexy; • uterine suspension sling; and • other techniques (including sacrocolpoperineopexy) or a mixture of above mesh techniques but groups not reported separately.

For RCTs and non-randomised comparative studies, there were no restrictions on the comparator. The comparator could be another operation using mesh or graft, or a type of surgery which did not involve mesh.

There were no restrictions on type of mesh. The types of mesh or grafts were classified as follows: • Absorbable synthetic mesh (e.g. polyglactin); • Biological graft (e.g. porcine dermis, fascia lata); • Combined mesh/graft (e.g. polypropylene and polyglactin composite, polypropylene mesh coated with absorbable porcine collagen); and • Non-absorbable synthetic mesh (e.g. polypropylene).

3.2.4 Types of outcomes

Efficacy

Primary outcomes for efficacy included: • Persistent prolapse symptoms (patient-reported/subjective failure); and • Recurrent prolapse at original site (clinician-observed/objective failure). We included any stage of recurrent prolapse at the original site (upper vagina). In the studies reporting this outcome, if more than one stage was reported, the lowest (best) one was used, for

11 instance, if a study reported failure rate at stage II or more and failure rate at stage III or more, the result based on stage II was used.

Secondary outcomes for efficacy included: • De novo prolapse, i.e. new prolapse at another site which was free of prolapse at baseline; • Need for further surgery for prolapse (both recurrent and de novo); • Persistent urinary symptoms in women having these symptoms at baseline; • Persistent bowel symptoms in women having these symptoms at baseline; • Persistent dyspareunia in women having these symptoms at baseline; and • Disease specific quality of life (RCTs and non-randomised comparative studies only).

For objective failure, outcomes measured by different systems, such as the POP-Q system and Baden-Walker system, were combined. Persistent urinary symptoms were included as an efficacy outcome whether or not a procedure aimed at curing urinary symptoms had been conducted at the same time. For disease specific quality of life, only RCTs and non- randomised comparative studies were considered because we could not adjust for baseline quality of life in registry or case series.

Safety

Safety outcomes included: • Blood loss; • Damage to surrounding organs during operation; • Mesh erosion; • Further operation needed for mesh erosion; • De novo urinary incontinence in women without urinary incontinence at baseline; • De novo bowel symptoms in women without these symptoms at baseline; • De novo dyspareunia in women without these symptoms at baseline; • Infection; • Other serious adverse effects not otherwise specified; and • Other adverse effects not otherwise specified.

We were aware that some of the outcomes are not directly associated with prolapse repair, for example, blood loss, infection, other serious adverse effects and other adverse effects not otherwise specified. They are also affected by concomitant procedures such as procedures for urinary incontinence.

12 Other outcomes

We considered other outcomes including operation time and hospital stay. We were aware that the difference in operation time across studies is also affected by the conduct of concomitant procedures, such as surgery for incontinence and additional anterior or posterior repair. However, in RCTs, this would be expected to balance out across the arms.

3.3 Quality assessment

One reviewer assessed the quality of the studies, using one of two separate checklists according to study design. Both checklists were developed through the Review Body for Interventional Procedures (ReBIP): • A 13-question checklist was used to assess the quality of RCTs (see Appendix 2). This was adapted from Verhagen and colleagues.17 • An 18-question checklist was used to assess the quality of non-randomised comparative studies with the same checklist minus four questions used to assess the quality of case series (see Appendix 3). The checklist was adapted from several sources, including the NHS Centre for Reviews and Dissemination’s guidance for those carrying out or commissioning reviews,18 Verhagen and colleagues,17 Downs and Black19 and the Generic Appraisal Tool for Epidemiology (GATE).

Each question was worded so that a rating of ‘Yes’ was always optimal in terms of methodological quality.

3.4 Data extraction

One reviewer screened the titles (and abstracts where available) of all papers identified by the search strategy. The same reviewer obtained and assessed full-text copies of all reports deemed to be potentially relevant for inclusion. A data extraction form was developed and piloted.2 One reviewer extracted details of study design, methods, participants, interventions and outcomes.

3.5 Data analysis

We conducted data analysis for techniques that use mesh separately, i.e. sacrocolpopexy, infracoccygeal sacropexy, uterine suspension sling, and other mesh techniques and a mixture of the above mesh techniques. Within each technique, data analysis was conducted for three subgroups of women according to the type of prolapse being repaired: • Uterine prolapse;

13 • Vault prolapse; and • Uterine or vault prolapse but groups not reported separately.

Direct comparisons (RCTs and non-randomised comparative studies)

We conducted meta-analyses of comparative studies, using Cochrane Collaboration Review Manager (RevMan 4.2) software, to directly compare the efficacy and safety between procedures using mesh and procedures without mesh, and between different procedures that used mesh. Pooled estimates from full-text RCTs, RCTs available as conference abstracts, and non-randomised comparative studies are presented separately.

For categorical outcomes we related the numbers reporting an outcome to the numbers at risk in each group to derive a relative risk (RR). For continuous variables we used means and standard deviations to derive a standardised mean difference (SMD). We used a random effects model.20

Crude event rates and median (ranges) event rates by study design

We tabulated crude event rates for each of the intervention categories for each outcome by study design (RCTs, non-randomised comparative studies, case series/registry). The median and range of the event rates for each study design were reported to illustrate the potential heterogeneity of event rates across different study designs.

Given the broad categorisation of treatments that we have used, i.e. irrespective of the intervention routes (vaginal, abdominal (open or laparoscopic)) and mesh types, in some of the comparative studies, both arms fell into the same treatment category. We found that in these studies the results between the arms were similar. For each outcome we therefore added together the results from the two arms and considered that to be one data point when calculating the median and the range. We have indicated where this occurred by an asterisk next to the number of studies in the tables.

14 Subgroup analysis

Pre-specified subgroup analysis was planned for: • Different intervention routes including vaginal, abdominal open, and abdominal laparoscopic; • Different mesh types; and within non-absorbable mesh, different Amid types (I to IV); and • Primary versus secondary repairs.

However, these subgroup analyses could not be conducted due to the lack of data.

15 4 RESULTS

4.1 Number and type of included studies

The results of the updated searches are summarised in Table 3. The numbers retrieved from the searches in Science Citation Index, Biosis and CENTRAL include only the additional reports found after excluding those identified from the Medline/Embase multifile search; the results of the searches of ISI conference proceedings and of conference abstracts include additional reports identified after deduplication against SCI and Biosis; and the results of the search of CCT include additional trials identified after deduplication against Clinical Trials.gov. We identified a total of 655 publications. In addition, the details of 28 potentially relevant ongoing studies were noted. The titles and abstracts (where available) were screened and 83 reports were selected for full assessment. One hundred and ninety-eight reports had already been identified as potentially relevant from the screening of the search results from the previous review therefore a total of 281 reports were selected for further assessment.

Table 3 Search results Database Number retrieved Number selected for assessment Primary reports: Medline/Embase/Medline In Process multi file 354 56 search(after deduplication in Ovid) SCI 156 13 Biosis 88 4 CENTRAL 7 0 ISI Conference proceedings 20 0 Conference abstracts 12 10 Total 637 83 Background: CDSR 1 0 DARE 8 0 HTA database 9 0 Total 18 0 Reports identified from previous review 198 Total assessed for review 655 281 Ongoing studies: NRR 14 1 Clinical Trials.gov 12 7 CCT 2 0 Personal communication 1 Total 28 9

16 4.1.1 Included studies

Table 4 details the outcome of the assessment process with 60 publications21-80 (57 full-text papers and 3 conference abstracts) included in the review. Appendix 4 lists the bibliographic details for all included studies. Seven of the 11 contacted manufacturers responded (American Medical Systems (UK) Ltd., Bard Ltd., Boston Scientific Ltd., Johnson & Johnson Medical Ltd., WL Gore & Associations (UK) Ltd., Coloplast Ltd., Mpathy Medical Devices Ltd.) and provided data on mesh properties or related studies, all of which had already been identified.

Table 4 Papers selected for full assessment Full-text papers Conference abstracts Included in review 57 3 Retained for background information 0 0 Excluded 138 83 Total 195 86

The included publications reported 54 studies (5 full-text RCTs, 3 RCTs available as a conference abstracts, 17 non-randomised comparative studies, and 29 case series) meeting the inclusion criteria for this review (see Table 5).

Appendix 5 provides details of the characteristics of studies by study design.

The studies took place during the period 1991 – 2007. Nineteen studies took place in the USA, six studies in Australia, six in the UK, three in France, two each in Canada, India, Italy, the Netherlands, South Africa, Sweden, and Turkey, and one each in Chile, China, Iran, Israel, Singapore and South Korea.

Table 6 summarises the characteristics of patients treated with mesh and the treatment details for the 54 included studies. Overall, 7054 women were treated in total, of whom 4456 were treated with sacrocolpopexy, 976 with infracoccygeal sacropexy, 159 with uterine suspension sling, 520 with other mesh techniques (sacrocolpoperineopexy) or a mixture of the above mesh techniques, and 943 with no-mesh techniques. Across studies, the average age of the women was 61 years. Women who were treated with uterine suspension sling were younger (average age 37 years) than those treated with other techniques. Only six studies provided information on whether the procedures were primary or secondary. The average follow up was around two years for sacrocolpopexy, one year for infracoccygeal sacropexy, and three years for uterine suspension sling.

Table 5 Number of studies and reports included Study design Studies Sacrocolpopexy Infracoccygeal sacropexy Uterine suspension sling Other mesh techniques or a mixture of mesh techniques Uterine repair RCTs (full text) 1 0 0 167 0 RCTs (conference abstracts) 1 129 0 0 0 Non-randomised comparative studies 4 233,45 160 133 0 Case series with sample size ≥100 1 179 0 0 0 Case series with sample size <100 5 0 170 423,24,52,54 0 Subtotal 12 4 2 6 0 Vault repair RCTs (full text) 2 2 35,57 0 0 0 RCTs (conference abstracts) 1 0 159 0 0 Non-randomised comparative studies 3 3 43,58,64 0 0 0

18 Case series with sample size ≥100 4 4 40,45,50,79 0 0 0 Case series with sample size <100 5 0 539,41,42,63,70 0 0 Subtotal 15 9 6 0 0 Uterine or vault repair RCTs (full text) 2 2 27,56 0 0 0 RCTs (conference abstracts) 1 0 174 0 0 Non-randomised comparative studies 12 1021,22,25,44,47,51,61,73,77,80 0 0 237,77 Case series with sample size ≥100 13 926,28,30,34,36,49,55,71,75 248,76 0 238,72

Case series with sample size <100 4 0 453,62,65,69 0 0 Subtotal 32 21 7 0 4 Total 54* 32 14 6 4 *5 studies are double counted: three studies45,70,79 reported different types of prolapse repair separately and two studies33,77 reported different mesh techniques separately.

Table 6 Characteristics of women included and treatment details Other mesh Infracoccygeal Uterine suspension techniques or a Sacrocolpopexy No mesh Total sacropexy sling mixture of mesh techniques N N N N N N N women N women N women N women N women N women studies studies studies studies studies studies Total number of women 32 4456 14 976 6 159 4 520 14 943 54 7054

Average age (y), median (range) 26 61 (53-68) 11 64 (54-73) 6 37 (22-61) 4 61 (44-62) 12 63 (56-66) 45 61 (22-73)

Prolapse repaired Uterine 4 311 (7%) 2 124 (13%) 6 159 (100%) 0 0 2 65 (7%) 12 659 (9%) Vault 9 1128 (25%) 6 285 (29%) 0 0 0 0 3 129 (14%) 15 1542 (22%)

19 Uterine or vault 21 3017 (67%) 7 567 (58%) 0 0 4 520 (100%) 9 749 (79%) 32 976 (16%)

Primary/secondary repair Primary repair 4 527 (67%) 0 2 48 (96%) 0 0 6 575 (69%) Secondary repair 4 261 (33%) 0 2 2 (4%) 0 0 6 263 (31%)

Average follow up (months), median 30 23 (8-66) 13 13 (5-30) 6 33 (12-95) 4 11 (1.5-19) 13 24 (1.5-38) 54 17 (1.5-95) (range)

Notes: Three studies45,70,79 reported different types of prolapse repair separately; T20 studies33,77 reported different mesh techniques separately.

4.1.2 Ongoing studies

We identified nine ongoing studies (Table 7). Seven of these were RCTs, (personal communication: A Griffiths, Johnson & Johnson, 2007)81-86 one was a registry87 of efficacy and safety results of using a medical device (product of American Medical Systems, USA) and one was a case series.88 In studies reporting this information, four studies were of vault prolapse repair, one of uterine prolapse repair, and one of prolapse repair in any compartment. Only three studies reported sample size. The registry study aims to recruit 1000 women.

Four RCTs compared variations of sacrocolpopexy. The other three RCTs compared total mesh (using introducers) with no mesh or sacrocolpopexy.

For studies reporting a follow-up period, all follow-up periods were one year or more. These studies are due to be completed between 2008 and 2014.

Table 7 Characteristics of ongoing studies ID Uterine/vault N Surgical techniques Follow-up Time to be completed RCT Brown, USA NR 300 A: sacrocolpopexy without Burch suspension 10y May 2014 B: sacrocolpopexy with Burch suspension

Halaska, Czech Republic Vault NR A: total mesh (using introducers) 1y Dec. 2009 B: Amreich Procedure (NR details)

Iglesia, USA Uterine NR A: total mesh (using introducers) 3y Jan. 2010 B: no mesh

Maher, Australia (personal communication, NR NR A: total mesh (using introducers) NR NR, interim results Adrian Griffin, 2 Aug. 2007) B: sacrocolpopexy available in Dec. 2008

Chen, USA Vault NR A: robotic assisted laparoscopic sacrocolpopexy 1y Dec. 2008 21 B: conventional laparoscopic sacrocolpopexy

Freeman, UK (pilot study) Vault 24 A: abdominal sacrocolpopexy 2y Apr. 2008 B: laparoscopic sacrocolpopexy

Culligan, USA NR NR A: sacrocolpopexy (polypropylene mesh) 5y Dec. 2009 B: sacrocolpopexy (absorbable graft) Registry Zylstra, USA Uterine, vault 1000 Non-absorbable synthetic mesh (using 2y Sep. 2008 introducers)

Case series Gannon, USA Vault NR Vaginal sling (device) NR NR (started in Mar. 2004)

NR: not recorded

4.2 Number and type of excluded studies; reason for exclusion

We obtained one hundred and thirty-eight potentially relevant studies identified by the search strategy, for which full-text papers, failed to meet the inclusion criteria given in Section 3.2, and were therefore excluded. A list of these studies is given in Appendix 6.

4.3 Quality of available evidence

We assessed the methodological quality of the full-text studies only. Appendices 7 to 9 provide details of the quality assessment of the included RCTs, non-randomised comparative studies, and case series respectively. The results are summarised in Figures 1, 2 and 3 respectively.

In the five RCTs, all studies reported methods of randomisation. Treatment allocation was adequately concealed in three studies.27,35,67 All follow-up periods were one year or more. All studies used intention-to-treat analysis in that women were analysed according to the group to which they were randomised. All studies reported the numbers of women lost to follow up (all less than 16%; results were extracted as reported). Only one of these five studies reported information on surgeon experience.67 In this study, all surgeons had performed more than 50 mesh procedures prior to the study commencing.

Figure 1 Summary of quality assessment of the RCTs (full text, n=5)

Q1. Randomisation adequate

Q2. Concealment of allocation

Q3. Similar baseline characteristics betw een groups

Q4. Inclusion criteria clearly specified Q5. Intervention cleary defined

Q6. Treatments w ere similar betw een groups Yes Q7. Mean follow -up ≥1y No

Q8. Outcome assessor blinded Unclear

Q9. Patients blinded

Q10. For continuous data, variability reported

Q11. Withdraw als/drop-outs unlikely to cause bias

Q12. ITT or no w ithdraw s/drop-outs after recruitment

Q13. Experienced practitioner

0% 20% 40% 60% 80% 100%

Among the 17 non-randomised comparative studies, only three studies collected data prospectively.33,51,73 Mean follow up was more than one year in 13 studies and less than one

year in four studies.21,37,77,80 The loss to follow up rates ranged from 4% to 55%. Most studies did not report the characteristics of women who were lost to follow up.

Figure 2 Summary of quality assessment of the non-randomised comparative studies (n=17)

Q1. Participant representativeness Q2. Participant inclusion/exclusion criteria described Q3. Similar participant baseline characteristics Q4. Selection of patients w as consecutive Q5. Data collection w as undertaken prospectively Q6. Similar baseline characteristics betw een groups Q7. Intervention clearly defined Q8. Experienced practitioner Yes Q9. Appropriate treatment place No Q10. Important outcomes considered Unclear Q11.Objective outcome measures used Q12. Outcome assessor blinded Q13. Mean follow -up ≥1y Q14. Information provided on non-respondents/drop-outs Q15. Withdraw als/drop-outs unlikely to cause bias Q16. Follow -up similar betw een groups Q17. Important prognostic factors identified Q18. Analyses adjusted for confounding factors

0% 20% 40% 60% 80% 100%

Amongst the 29 case series, 12 studies collected data prospectively. No studies reported surgeon experience. Mean follow up was one year or more in 25 studies and less than one year in four studies.40,41,62,65 In the 12 studies reporting loss to follow up, rates varied from 2% to 74% with follow up between 2 and 66 months.

Figure 3 Summary of quality assessment of the case series (n=29)

Q1. Participant representativeness

Q2. Participant inclusion/exclusion criteria described Q3. Similar participant baseline characteristics Q4. Selection of patients w as consecutive Q5. Data collection w as undertaken prospectively Q7. Intervention clearly defined Yes Q8. Experienced practitioner No Q9. Appropriate treatment place Unclear Q10. Important outcomes considered Q11.Objective outcome measures used

Q13. Mean follow -up ≥1y Q14. Information provided on non-respondents/drop-outs

Q15. Withdraw als/drop-outs unlikely to cause bias Q17. Important prognostic factors identified

0% 20% 40% 60% 80% 100%

4.4 Sacrocolpopexy (mesh)

Thirty two studies (4 full-text RCTs,27,35,56,57 1 RCT available as conference abstract,29 15 non-randomised comparative studies,21,22,25,33,43-45,47,51,58,61,64,73,77,80 and 12 case series with sample sizes over 100) involving a total of 4456 women provided data on sacrocolpopexy. Two studies29,33 reported uterine repair, 7 studies reported vault repair,35,40,43,50,57,58,64 2 studies reported uterine and vault repair separately,45,79 and 21 studies21,22,25- 28,30,34,36,44,47,49,51,55,56,61,71,73,75,77,80 did not report uterine and/or vault repair separately. Table 8 summarises the patient characteristics and surgical procedures in these 32 studies by types of prolapse. Mesh types varied across studies. The median (range) of mean follow up across these 32 studies was 23 months (8 to 66 months).

Data from individual studies reporting efficacy outcomes are given by type of prolapse in Appendix 10, the safety outcomes are in Appendix 11, and the operation time and hospital stay are in Appendix 12. The median rates and ranges for efficacy outcomes, by type of prolapse, are summarised in Table 9, and safety outcomes are in Table 10. No RCTs or non-randomised comparative studies reported data on disease-specific quality of life.

Methodological quality of the included studies

In the four full-text RCTs,27,35,56,57 the allocation was adequately concealed in two studies.27,35, All mean follow-up times were greater than one year. The number of women lost to follow up ranged from 9 to 15%. All studies used intention-to-treat analysis in that women remained in the groups to which they were randomised (Appendix 7).

In the 15 non-randomised comparative studies,21,22,25,33,43-45,47, 51,58,61,64,73,77,80 only two 33,80 33,43,44,51,64,73 studies collected data prospectively. Six studies recruited patients consecutively. Eleven studies22,25,33,43-45,47,58,61,64,73 had a follow-up period longer than one year. The number of women lost to follow up ranged from 4 to 55% in the five studies22,33,44,61,73 reporting this information (Appendix 8).

In the 12 case series,26,28,30,34,36,40,49,50,55,71,75,79 only three28,40,75 collected data prospectively. Four studies36,49,50,55 recruited patients consecutively. All studies but one40 had a follow up of over one year. Loss to follow up ranged from 5 to 73% in six studies26,28,34,49,50,71 reporting this information (Appendix 9).

Table 8 Sacrocolpopexy: summary of patient characteristics and surgical procedures ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft, trade name Concomitant Follow-up, Outcomes vault secondary procedure mean UTERINE RCT (abs.) Braun 200729 A, 23 A, 57 A, 23/0 NR A, abdominal A, hysterectomy + sacrocolpopexy; A, combined mesh, NR 33m (20-41) Efficacy B, 24 B, 56 B, 24/0 B, vaginal B, hysterectomy + anteroposterior polyglactin and prolene 1 :1, Safety colporrhaphy + Mayo McCall stitch Vypro B, no mesh Non-randomised comparative studies Costantini A, 36 A, 61 (12) A, 36/0 NR A, abdominal A, sacrohysteropexy A, polypropylene, Marlex Anti-incontinence: 51m (12-115) Efficacy 200533 B, 39 B, 62 (8) B, 39/0 B, abdominal B, hysterectomy + sacrocolpopexy (Amid type I) A, 28/36; B, 30/39 Safety B, same as A Hysterectomy: A, 0/36; B, 39/39

Griffis 200645 A, 60 NR A, 60/0 NR A, abdominal A, total hysterectomy + sacrocolpopexy A, polypropylene, Prolene NR 13m (12-15) Safety

25 B, 28 B, 28/0 B, abdominal B, supracervical hysterectomy + soft, Prolene (Amid type I), or sacrocolpopexy Atrium (NR Amid type); polyethylene tetraphalate, Mersilene (Amid type III) B, same as A Case series Wu 200679 101 64 (12) 101/0 101/0 Abdominal (open) Hysterectomy + sacrocolpopexy polyethylene tetraphalate, NR 15m (0.2-120) Safety Mersilene (Amid type III); polypropylene (NR trade name); or Gore-Tex (Amid type II) VAULT RCT Culligan 200535 A, 54 A, 60 (10) A, 0/54 NR A, abdominal (open) A, sacrocolpopexy A, polypropylene, Trelex NR 1y Efficacy B, 46 B, 58 (11) B, 0/46 B, abdominal (open) B, sacrocolpopexy B, cadaveric fascia lata, Safety Tutoplast

ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft, trade name Concomitant Follow-up, Outcomes vault secondary procedure mean Maher 200457 A, 47 A, 63 (39-84) A, 0/47 NR A, abdominal A, sacrocolpopexy A, polypropylene, Prolene NR 24m (6-60) Efficacy B, 48 B, 63 (35-88) B, 0/48 B, vaginal B, unilateral vaginal sacrospinous (Amid type I) Safety colpopexy B, no mesh Non-randomised comparative studies Govier 200543 A, 24 A+B, 67 (51- A, 0/24 NR A, abdominal (open A, sacrocolpopexy A, polypropylene, Prolene A+B, 20/45 A, 12m (1-38) Efficacy B, 21 86) B, 0/21 or laparoscopic) B, sacrocolpopexy B, silicone-covered B, 23m (16-41) Safety B, abdominal (open polyethylene mesh, or laparoscopic) American Medical Systems

Marcickiewicz A, 60 A, 58 (30-83) A, 0/60 NR A, abdominal A, sacrocolpopexy A, polypropylene, Prolene A, 6/60; B, 8/51 A, 34m (13-60) Efficacy 200758 B, 51 B, 66 (43-88) B, 0/51 (laparoscopic) B, sacrospinous colpopexy B, no mesh B, 38m (7-108) Safety B, vaginal

Paraiso 200564 A, 56 A, 62 (39-85) A, 0/56 NR A, abdominal A, sacrocolpopexy A, cadaveric fascia lata, A, 24/56; B, 38/61 A, 14m (1-46) Safety

26 B, 61 B, 61 (35-81) B, 0/61 (laparoscopic) B, sacrocolpopexy Tutoplast, or polypropylene, B, 16m (1-73) B, abdominal (open) Prolene B, save as above Case series Fedorkow 149 58 (11) 0/149 NR Abdominal Sacrocolpopexy Polypropylene, Prolene NR NR Safety 199340 (Amid type I)

Griffis 200645 196 NR 0/196 NR Abdominal Sacrocolpopexy Polypropylene, Prolene-Soft, NR 13m Safety Prolene (Amid type I), or Atrium (NR Amid type); polyethylene tetraphalate, Mersilene (Amid type III).

Higgs 2005a50 103 58 (10) 0/103 NR Abdominal Sacrocolpopexy Polypropylene, Prolene 39/103 66m (37-124) Efficacy (laparoscopic) (Amid type I) Safety

ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft, trade name Concomitant Follow-up, Outcomes vault secondary procedure mean Wu 200679 212 66 (11) 0/212 212/0 Abdominal (open) Sacrocolpopexy Polyethylene tetraphalate, NR 15m (0.3-121) Safety Mersilene (Amid type III); polypropylene (NR trade name); or Gore-Tex UTERINE AND/OR VAULT RCT Benson 199627 A, 40 A, 66 (10) NR NR A, abdominal A, sacrocolpopexy A, NR Anti-incontinence: 2.5y (1-5.5) Efficacy B, 48 B, 64 (9) B, vaginal B, bilateral sacrospinous ligament B, no mesh A, 14/40; B, 20/48 Safety vault suspension Hysterectomy: A, 20/40; B, 24/48

Lo 199856 A, 52 A, 63 (9) NR NR A, abdominal (open) A, sacrocolpopexy A, polyethylene tetraphalate, Anti-incontinence: 2.1y (1-5.2) Efficacy B, 66 B, 60 (10) B, vaginal B, sacrospinous ligament suspension Mersilene (Amid type III) A, 0/52; B, 0/66 Safety B, no mesh Hysterectomy:

27 A, 33/52; B, 20/66 Non-randomised comparative studies Altman 200521 A, 25 A, 66 (54-83) NR NR A, abdominal A, sacrocolpopexy A, 12 polypropylene, 13 Hysterectomy: Efficacy Safety B, 27 B, 69 (55-84) B, abdominal B, sacrocopopexy polytetrafluoroethylene (NR A, 2/25; B, 5/27 A, 7.4m; B, trade name) 7.1m B, porcine dermis, Pelvicol Safety A, 4.3y; B, 2.5y

Bai 200622 A, 20 NR NR NR A, abdominal A, sacrocolpopexy A, polyethylene tetraphalate, Performed if needed 1y Safety B, 54 B, abdominal B, high uterosacral colpopexy Mersilene (Amid type III) but NR numbers. B, no mesh

Begley 200525 A, 24 A, 69 (49-86) NR A+B, 85/7 A, abdominal (19 A, sacrocolpopexy A, polypropylene, Marlex or Anti-incontinence: A, 10m Efficacy B, 33 B, 66 (37-84) open, 5 B, sacrocolpopexy Prolene (Amid type I) A, 9/24; B, 14/33; B, 29m Safety C, 21 C, 66 (40-85) laparoscopic) C, sacrocolpopexy B, polypropylene, Gore-Tex C, 9/21; D, 5/14 C, 16m D, 14 D, 63 (25-83) B, abdominal (open) D, sacrocolpopexy (Amid type II) Hysterectomy: D, 19m

ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft, trade name Concomitant Follow-up, Outcomes vault secondary procedure mean C, abdominal (18 C, silicon-covered polyester, A, 1/24; B, 8/33; C, open, 3 American Medical Systems 3/21; D, 7/14 laparoscopic) D, 1 autologous and 13 D, abdominal (open) cadaveric fascia lata, Tutoplast

Gregory 200544 A, 49 NR NR NR A, abdominal (open) A, sacrocolpopexy A, polypropylene, Marlex Performed but NR A, 26 (10-48) Efficacy B, 33 B, abdominal (open) B, sacrocolpopexy (Amid type I); polyethylene numbers B, 21 (11-34) Safety tetraphalate, Mersilene (Amid type II) B, fascia lata, Community Tissue Services

Hardiman A, 80 A, 61 NR NR A, abdominal A, sacrocolpopexy A, polypropylene, Marlex Anti-incontinence: A, 3.9y Efficacy 47 28 1996 B, 125 B, 64 B, vaginal B, sacrospinous vault suspension (Amid type I) A, 76/80; B, 45/125 B, 2.2y Safety B, no mesh Hysterectomy: A+B, 6m-5y A, 18/80; B, 45/125

Hsiao 200751 A, 25 A, 66 A, 2/23 NR A, abdominal A, sacrocolpopexy A, 19 polypropylene, Anti-incontinence: A, 6m Safety B, 22 B, 71 B, 3/19 (laparoscopic) B, sacrocolpopexy Prolene; 3 silicone mesh A, 9/25; B, 10/22 B, 10m B, abdominal (open) B, 2 fascia lata; 13 Hysterectomy: polypropylene, Prolene; 5 NR silicon mesh; 2 polypropylene, Gore-Tex

Ng 200461 A, 113 A, 60 A, 78/35 NR A, abdominal (open) A, sacrocolpopexy A, polytetrafluoroethylene, Anti-incontinence: A, 18m (1-48) Efficacy B, 64 B, 63 B, 37/27 B, vaginal B, sacrospinous ligament fixation Gore-Tex (Amid type II) A, 28/113; B, 12/64 B, 13m (1-29) Safety B, no mesh Hysterectomy: A, 78/113; B, 38/64

ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft, trade name Concomitant Follow-up, Outcomes vault secondary procedure mean Sze 199973 A, 56 A, 57 (34-74) A, 19/35 NR A, abdominal (open) A, sacrocolpopexy A, NR Anti-incontinence: A, 23m (4-51) Safety B, 61 B, 64 (43-76) B, 7/40 B, vaginal B, sacrospinous ligament fixation B, no mesh A, 56/56; B, 61/61 B, 24m (7-72) Hysterectomy: A, 7/56; B, 19/61

Visco 200177 A, 155 A+B+C+D, NR NR A, abdominal A, sacrocolpopexy A, polyethylene tetraphalate, Anti-incontinence: A, 7m (1-87) B, 88 61 (31-84) B, abdominal B, sacrocolpoperineopexy Mersilene (Amid type III) or NR B, 5 (1-45) C, 25 C, vaginal+ C, sacrocolpoperineopexy Gore-Tex (Amid type II) Hysterectomy: C, 6m (1-28) D, 5 abdominal D, sacrocolpoperineopexy with mesh B, C, D, same as above A+B+C+D: 45/273 D, 7m (2-11) D, vaginal+ placed to vaginal field abdominal

Young 200480 A, 92 A, 61 (28-84) NR NR A, abdominal A, sacrocolpopexy A, a mixture of biological Anti-incontinence: NR, a safety Safety (prospective B, 187 B, 37-87 (laparoscopic) B, sacrospinous fixation grafts and non-absorbable A, 44/92; B, 28/187; registry in a

29 registry) C, 70 C, 64 (34-84) B, vaginal C, high uterosacral suspension synthetic mesh C, 8/70 year (Oct. C, vaginal B, no mesh Hysterectomy: 1998-Oct. C, no mesh A, 6/92; B, 63/187; 1999) C, 17/70 Case series Bensinger 121 53 (10) 86/35 121/0 Abdominal Sacrocolpopexy Polypropylene (NR trade Anti-incontinence: 13m (0.3-63) Safety 200526 name) 82/121 Hysterectomy: NR

Brizzolara 124 65 (9) 60/64 NR Abdominal Sacrocolpopexy Polypropylene, Prolene, or Anti-incontinence: 36m (0-74) Efficacy 200330 allograft (NR trade name) 92/124 Safety Hysterectomy: NR

Bradley 200728 305 NR NR NR Abdominal (open) Sacrocolpopexy Non-absorbable synthetic, Anti-incontinence: 1y Efficacy xenograft, autologous or 148/305 Safety cadaveric fascia lata

ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft, trade name Concomitant Follow-up, Outcomes vault secondary procedure mean Culligan 200234 245 61 (32-83) NR NR Abdominal Sacrocolpopexy Synthetic mesh (NR trade Anti-incontinence: Objective Efficacy name) 171/245 failure: >=4y; Safety Hysterectomy: subjective 11/245 failure: 3.3y (0.8-6.9) De Vries 199536 101 59 (37-82) 15/83 NR Abdominal (open) Sacrocolpopexy Polyethylene tetraphalate, Anti-incontinence : 4y (1-13) Efficacy Mersilene (Amid type III) 20/101 Safety Hysterectomy : NR

Higgs 2005b49 148 58 (10) 24/123 NR Abdominal (open) Sacrocolpopexy Combined mesh, Anti-incontinence : 45m (15) Efficacy polyproglactine and prolene 76/148 Safety 1 :1, Vypro; fascia lata; Hysterectomy : polypropylene, Prolene; 25/148 polyethylene tetraphalate,

30 Mersilene.

Lindeque 262 28-79 4/258 8/254 Abdominal (open) Sacrocolpopexy 18 dura mater strips, then Anti-incontinence: All >=16m Efficacy 200255 changed to 106/262 Safety polytetrafluoroethylene, Hysterectomy: NR Gore-Tex (Amid type II)

Snyder 199171 147 62 (30-83) 3/144 NR Abdominal (open) Sacrocolpopexy Dacron graft (NR details); NR 43m (1m-17y) Efficacy polytetrafluoroethylene, Safety Gore-Tex (Amid type II), and other type of mesh (NR)

Timmons 163 58 (19-81) 163/3 NR Abdominal (open) Sacrocolpopexy Fascial lata (NR trade NR 33m (9m-18y) Safety 199275 name); polyethylene tetraphalate, Mersilene (Amid type III)

Table 9 Sacrocolpopexy, efficacy: summary of median (range) event rates Uterine Vault Uterine and/or vault No mesh (vaginal No mesh (sacrospinous Sacrocolpopexy Sacrocolpopexy Sacrocolpopexy No mesh (uterine/vault suspension) hysterectomy) colpopexy) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, study % (range) study % (range) study % (range) study % (range) study % (range) study % median (range) A. Subjective failure: persistent prolapse symptoms RCT 0 0 1 3/46 (6.5%) 6.5% 1 4/43 (9.3%) 9.3% 1 6/38 (16%) 16% 1 14/42 (33%) 33% RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 1 13/60 (22%) 22% 1 6/51 (12%) 12% 2* 11/97 (11%) 14% (3.3-24) 1 11/48 (23%) 23% Case series 0 0 1 22/103 (21%) 21% 0 3 77/294 (26%) 26% (23-31) 0 B. Objective failure: recurrent prolapse at original site RCT 0 0 2* 2/135 (1.5%) 2.2% (0-4.3%) 1 8/43 (19%) 19% 1 0/52 (0%) 0% 1 NR RCT (abs.) 1 0/23 (0%) 0% 1 1/24 (4.2%) 4.2% 0 0 0 0 Non-rand 1 0/39 (0%) 0% 0 2* 2/105 (1.9%) 2.2% (0-4.4%) 1 0/51 (0%) 0% 2 4/140 (2.9%) 3.1% (1.3-5.0) 2 16/173 (9.2%) 15% (2.4-27%) Case series 0 0 1 4/66 (6.1%) 6.1% 0 4 8/494 (1.6%) 1.8% (0-3.1) 0 C. De novo prolapse at other sites

31 RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 1 10/32 (31%) 31% 1 4/28 (14%) 14% 0 0 Case series 0 0 0 0 1 3/193 (1.6%) 1.6% 0 D. Further operation needed for prolapse (recurrent or de novo) RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 1 0/39 (0%) 0% 0 1 15/60 (25%) 25% 1 4/51 (8%) 8% 0 0 Case series 0 0 0 0 2 17/352 (4.8%) 8.0% (1.5-14.4) 0 E. Persistent urinary symptoms RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 0 0 0 0 1 43/169 (25%) 25% 0

Note: 32 1. * Results from the two arms of one study were combined.

Table 10 Sacrocolpopexy, safety: summary of median (range) event rates Uterine Vault Uterine and/or vault No mesh (vaginal No mesh (sacrospinous Sacrocolpopexy Sacrocolpopexy Sacrocolpopexy No mesh (uterine/vault suspension) hysterectomy) colpopexy) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, study % (range) study % (range) study % (range) study % (range) study % (range) study % median (range) A. Blood loss requiring blood transfusion RCT 0 0 2* 2/147 (1.4%) 1.6% (1.0-2.1) 1 1/48 (2.1%) 2.1% 1 2/38 (5.3%) 5.3% 1 0/42 (0%) 0% RCT (abs.) 0 0 0 0 0 0 Non-rand 1 2/39 (5.1%) 5.1% 0 2* 3/177 (1.7%) 1.7% (both 1.7) 1 2/51 (3.9%) 3.9% 5* 17/388 (4.4%) 1.3% (0-12) 4 6/437 (1.4%) 2.1% (0-3.3) Case series 0 0 1 0/149 (0%) 0% 0 7 38/1142 (3.3%) 1.2% (0.8-17) 0 B. Damage to surrounding organs during operation RCT 0 0 2* 2/147 (1.4%) 1.6% (1.0-2.1) 1 1/48 (2.1) 2.1% 2 4/90 (4.4%) 4.9% (1.9-7.9) 2 1/108 (0.9%) 1.2% (0-2.4) RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 2* 10/177 (5.6%) 5.9% (5.1-6.7) 1 2/51 (3.9%) 3.9% 2 3/205 (1.5%) 1.5% (0.9-2.2) 2 8/321 (2.5%) 2.2% (1.6-2.8) Case series 0 0 1 2/103 (1.9%) 1.9% 0 6 25/1001 (2.5%) 2.0% (0.4-7.9) 0 C. Mesh erosion

33 RCT 0 0 2* 3/147 (2.0%) 2.1% (2.0-2.1) 1 NA 0 0 RCT (abs.) 1 1/23 (4.3%) 4.3% 1 NA NA 0 0 0 0 Non-rand 2* 12/143 (8.4%) 8.2% (7.7-8.7) 0 2** 7/162 (4.3%) 5.7% (2.6-8.9) 0 8**** 22/577 (3.8%) 2.4% (0-12) 3 NA Case series 1 7/101 (6.9%) 6.9%) 0 4 35/660 (5.3%) 6.5% (0-8.7) 0 7 31/1109 (2.8%) 2.4% (1.0-6.5) 0 D. Further operation needed for mesh erosion RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 2* 7/143 (4.9%) 5.8% (3.8-7.7) 0 2** 8/162 (4.9%) 6.8% (2.6-11) 0 4** 19/348 (5.5%) 6.1% (3.2-11) 1 NA Case series 0 0 3 10/448 (2.2%) 2.6% (0-4.9) 0 6 28/973 (2.9%) 2.9% (0.8-6.5) 0 E. De novo urinary symptoms in women who did not have these symptoms at baseline RCT 0 0 1 2/22 (9.1%) 9.1% 1 8/24 (33%) 33% 1 2/52 (3.8%) 3.8% 1 1/66 (1.5%) 1.5% RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 1 4/89 (4.5%) 4.5% 1 0/52 (0%) 0% Case series 0 0 0 0 1 12/131 (9.2%) 9.2% 0 F. De novo bowel symptoms RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0

Uterine Vault Uterine and/or vault No mesh (vaginal No mesh (sacrospinous Sacrocolpopexy Sacrocolpopexy Sacrocolpopexy No mesh (uterine/vault suspension) hysterectomy) colpopexy) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, study % (range) study % (range) study % (range) study % (range) study % (range) study % median (range) Non-rand 0 0 0 0 0 0 Case series 0 0 0 0 1 2/178 (1.1%) 1.1% 0 G. De novo sexual symptoms RCT 0 0 0 0 1 1/11 (9.1%) 9.1% 1 7/18 (39%) 39% RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 0 0 0 0 1 11/76 (15%) 15% 0 H. infection RCT 0 0 2* 5/147 (3.4%) 3.1% (2.1-4.0) 1 1/48 (2.1%) 2.1% 2 14/90 (16%) 17% (5.8-29) 2 20/108 (19%) 21% (11-31%) RCT (abs.) 0 0 0 0 0 0 Non-rand 1 3/39 (7.7%) 7.7% 0 3** 12/222 (5.4%) 5.0% (2.2-6.8) 1 4/51 (7.8%) 7.8% 4* 114/301 (38%) 28% (8.8-68) 3 53/250 (21.2%) 18.4% (7.8-39.1) Case series 0 0 2 38/252 (15%) 13.1% (1.9-24) 0 3 30/340 (8.8%) 10% (0.8-17) 0

34 I. Other serious adverse effects not otherwise specified RCT 0 0 2* 18/147 (12.2%) 10.1% (4.3-16) 1 0/43 (0%) 0% 2 4/90 (4.4%) 4.2% (2.6-5.8) 2 2/108 (1.9%) 1.9 (1.5-2.4) RCT (abs.) 1 1/23 (4.3%) 4.3% 1 0/24 (0%) 0% 0 0 0 0 Non-rand 1 1/39 (2.6%) 2.6% 0 2 12/162 (7.4%) 5.1% (0-10) 0 6** 23/440 (5.2%) 6.0% (0-8.5) 4 2/507 (0.4%) 0.2% (0-1.6) Case series 0 0 0 0 7 50/1158 (4.3%) 4.3% (2.7-6.1) 0 J. Other adverse effects not otherwise specified RCT 0 0 1 0/47 (0%) 0% 1 2/48 (4.2%) 4.2% 2 3/90 (3.3%) 3.9% (0-7.9) 2 19/108 (18%) 14% (0-29) RCT (abs.) 1 1/23 (4.3%) 4.3% 1 0/24 (0%) 0% 0 0 0 0 Non-rand 1 5/39 (12.8%) 12.8% 0 0 0 5* 11/393 (2.8%) 1.8% (1.1-5.8) 4 14/507 (2.8%) 2.4% (1.6-8.2) Case series 0 0 2 10/252 (4.0%) 5.0% (0-10) 0 5 54/907 (6.0%) 3.5% (0.4-16) 0

Note: 1. */**/****Results from the two arms of one/two/four studies were combined. 2. For other serious adverse effects not otherwise specified, the occurred effects included: bowel obstruction, ureteral obstruction, wound dehiscence, incisional hernia, vesico-cutaneous fistula, deep vein thrombosis, pulmonary embolism, pneumonia, myocardial infarction, congestive heart failure, arrhythmia, and left ventricular failure. 3. For other adverse effects not otherwise specified, the occurred effects included: voiding dysfunction, haematoma, persistent vaginal discharge, vaginal bleeding, phlebitis, wound complications, gastrointestinal complications, e.g. nausea, emesis, bloating or subacute intestinal obstruction, sciatica, and buttock pain.

4.4.1 Sacrocolpopexy for uterine prolapse repair (mesh)

Four studies29,33,45,79 involving 311 women reported sacrocolpopexy for uterine prolapse. One was an RCT available as a conference abstract29 that compared hysterectomy followed by sacrocolpopexy with hysterectomy alone. Two were non-randomised comparative studies: one33 compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy and the other45 compared total hysterectomy followed by sacrocolpopexy with supracervical hysterectomy followed by sacrocolpopexy. One was a case series79 that reported hysterectomy followed by sacrocolpopexy.

4.4.1.1 Efficacy

No studies reported data on subjective failure, new prolapse at other sites, persistent urinary symptoms, bowel symptoms, or sexual problems after sacrocolpopexy for uterine prolapse.

Objective failure: recurrent prolapse at original site (Table 9B)

One RCT available as a conference abstract29 and one non-randomised comparative study33 reported data on objective failure. In the RCT,29 there was no evidence of a difference in objective failure between hysterectomy followed by sacrocolpopexy and hysterectomy alone (0/23 vs. 1/24; RR 0.35, 95% CI 0.01 to 8.11). The non-randomised comparative study33 compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy. No objective failures, defined as prolapse less than 6cm above the hymen, occurred at a mean follow up of 51 months in either group (0/36 vs. 0/39).

Need for further surgery for prolapse (Table 9D)

One non-randomised comparative study33 reported the need for further surgery for prolapse. Sacrohysteropexy was compared with hysterectomy followed by sacrocolpopexy. None of the 75 women required a further operation for recurrent or de novo prolapse at a mean follow up of 51 months.

4.4.1.2 Safety

No studies reported damage to surrounding organs during operation, or new urinary symptoms, bowel, or sexual problems after sacrocolpopexy for uterine prolapse repair.

Blood loss i: Blood loss, ml

One non-randomised comparative study33 and one case series79 reported blood loss as continuous variable. The non-randomised comparative study compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy.33 Women who had a sacrohysteropexy had a lower blood loss (n=36, mean 200ml, SD not reported) than those who had a hysterectomy followed by sacrocolpopexy, (n=39, mean 325ml, SD not reported). Women in the case series79 had hysterectomy followed by sacrocolpopexy. The blood loss was similar (n=101, mean 328ml, SD 163ml) to that in the non-randomised comparative study33 mentioned above in which the same technique was used. ii: Blood loss requiring blood transfusion (Table 10A)

One non-randomised comparative study33 reported this outcome. There was no evidence of a difference between sacrohysteropexy and hysterectomy followed by sacrocolpopexy in blood loss requiring transfusion (2/36, 5.6% vs. 2/39, 5.1%; RR, 1.08, 95% CI 0.16 to 7.29).

Mesh erosion (Table 10C)

One RCT available as a conference abstract,29 two non-randomised comparative studies,33,45 and one case series79 involving 324 women reported mesh erosion. In the RCT,29 hysterectomy followed by sacrocolpopexy using a combined mesh (absorbable and non- absorbable) was used. One (1/23, 4.3%) woman had a mesh erosion at a mean follow up of 33 months. One non-randomised comparative study33 compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy. Non-absorbable synthetic mesh (Amid type I) was used. Three erosions (3/39, 7.7%) occurred in the hysterectomy group at a mean follow up of 51 months and no erosions (0/36) occurred in the sacrohysteropexy group. The other non-randomised comparative study45 found erosion rates of 11% (8/76) in women having total hysterectomy followed by sacrocolpopexy and 3.6% (1/28) in women having supracervical hysterectomy followed by sacrocolpopexy. Non-absorbable synthetic mesh (Amid type I or III) was used in this study. In the case series,79 6.9% (7/101) of women had a mesh erosion after hysterectomy followed by sacrocolpopexy. Non-absorbable synthetic mesh (Amid type I or III) was used in this case series.

Two studies45,79 reported the time interval when erosions occurred. The non-randomised comparative study comparing sacrocolpopexy with total or supracervical hysterectomy45 reported that the erosions (9/104) occurred at a median of four months (range 0.4 to 53). Similarly, in a case series,79 seven (7/101) erosions occurred at a median of 8.4 months (range 1.4 to 13 months) after hysterectomy and sacrocolpopexy.

Further operation needed for mesh erosion (Table 10D)

Two non-randomised comparative studies33,45 reported that further surgery was needed for mesh erosion in either the outpatient clinic or in the operating theatre. Among a total of 12 erosions, at least 7 required further surgery (partial or complete mesh removal) while the others were managed conservatively using oestrogen cream, antibiotic cream, or silver nitrate.

Infection (Table 10H)

One non-randomised comparative study33 reported data on infection. The outcome was reported as wound infection and fever. Three infections (3/39, 7.7%) occurred in the sacrocolpopexy group, and one case (1/36, 2.8%) occurred in the sacrohysteropexy group.

Other serious adverse effects not otherwise specified (Table 10I)

The only complication considered in this section was incisional hernia. One RCT available as a conference abstract29 and one non-randomised comparative study33 reported data on incisional hernia. The RCT29 compared hysterectomy followed by sacrocolpopexy with hysterectomy alone. One incisional hernia occurred in the sacrocolpopexy group (1/23, 4.3%) and no cases occurred in the hysterectomy alone groups (0/24). In the non- randomised comparative study,33 5.6% (2/36) of women had an incisional hernia after sacrohysteropexy, compared with 2.6% (1/39) of women who had a hysterectomy followed by sacrocolpopexy.

Other adverse effects not otherwise specified (Table 10J)

The complications considered in this section included haematoma and voiding dysfunction. One RCT available as a conference abstract29 and one non-randomised comparative study33 reported on these complications. In the RCT,29 one subaponeurotic haematoma (1/23, 4.3%) occurred in the sacrocolpopexy (with hysterectomy) group and none in the

hysterectomy alone group (0/24). In the non-randomised comparative study,33 17% (6/36) of women had perivesical haematoma or voiding dysfunction after sacrohysteropexy, compared with 13% (5/39) of women who had hysterectomy followed by sacrocolpopexy.

4.4.1.3 Operation time and hospital stay

Operation time

One RCT available as a conference abstract29 and one non-randomised comparative study33 reported operation time. In the RCT,29 operation time for abdominal hysterectomy followed by sacrocolpopexy was about 50 minutes longer than (vaginal) hysterectomy alone (mean 140min, n=23 vs. 90min, n=24). In the non-randomised comparative study,33 operation time for abdominal hysterectomy followed by sacrocolpopexy was about 25 minutes longer than for abdominal sacrohysteropexy (mean 115min, n=39 vs. 89min, n=36).

Hospital stay

One RCT available as a conference abstract29 and one case series79 reported hospital stay. In the RCT,29 hospital stay for abdominal hysterectomy followed by sacrocolpopexy was about two days longer than (vaginal) hysterectomy alone (mean 3.8 days, n=23 vs. 2 days, n=24). In the case series,79 the mean number of days of hospital stay was 3.5 days in women who underwent abdominal hysterectomy followed by sacrocolpopexy.

4.4.2 Sacrocolpopexy for vault prolapse repair (mesh)

Nine studies35,40,43,45,50,57,58,64,79 involving 1128 women reported vault prolapse repair (Table 9). Two were full-text RCTs: one57 compared non-absorbable synthetic mesh and a biological graft for sacrocolpopexy; the other35 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). Three were non-randomised comparative studies: one43 compared two types of non-absorbable synthetic mesh for sacrocolpopexy; another58 compared sacrocolpopexy with sacrospinous colpopexy (no mesh); the third64 compared (abdominal) laparascopic with (abdominal) open techniques for sacrocolpopexy. Four were case series which all reported non-absorbable synthetic mesh for sacrocolpopexy.

4.4.2.1 Efficacy

No studies reported persistent urinary and bowel symptoms after sacrocolpopexy for vault prolapse repair.

Subjective failure: persistent prolapse symptoms (Table 9A)

One RCT,57 one non-randomised comparative study58 and one case series50 reported subjective failure.

Direct comparison: techniques that use mesh versus techniques that do not use mesh

The RCT by Maher and colleagues57 and the non-randomised comparative study by Marcickiewicz and colleagues58 compared sacrocolpopexy (mesh) and sacrospinous colpopexy (no mesh). The RCT used an open abdominal approach for sacrocolpopexy (7% failure rate, 3/46) whereas the non-randomised comparative study used laparoscopy (22% failure rate, 13/60). There was no evidence of a difference in subjective failure between groups (Figure 4). The mean follow ups were 24 months and 34 months respectively. Both studies used non-absorbable synthetic mesh.

Figure 4 Sacrocolpopexy, vault prolapse repair, efficacy, subjective failure

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

One case series50 reported a subjective failure rate of 21% (22/103) at a mean follow up of 66 months after laparoscopic sacrocolpopexy. This was similar to the 22% reported in the non-randomised comparative study58 (see above).

Objective failure: recurrent prolapse at original site (Table 9B)

Two RCTs,35,57 two non-randomised comparative studies43,58 and one case series50 reported data on objective failure.

One RCT57 and one non-randomised comparative study58 compared sacrocolpopexy and sacrospinous colpopexy (no mesh). In the RCT, there was no statistically significant evidence of a difference in failure rates between groups at a mean follow up of 24 months (2/46, 4.3% vs. 8/43, 19%; RR 0.23, 95% CI 0.05 to 1.04). In the non-randomised comparative study, no case of objective failure occurred in either group at a mean follow up of 34 months (0/60 vs. 0/51).58

One RCT35 compared non-absorbable synthetic mesh and a biological graft for sacrocolpopexy. No case of objective failure occurred at a follow up of one year in 101 women. One non-randomised comparative study43 compared two types of non-absorbable synthetic mesh, polypropylene versus silicone-covered polyethylene, for sacrocolpopexy. No failure occurred in the polypropylene group and two (2/21, 9.5%) occurred in the silicone- covered polyethylene group with one detected at less than 20 months and the other at 24 months.

The case series50 reported a failure rate of 6.1% (4/66) after laparoscopic sacrocolpopexy at a mean follow up of 66 months.

De novo prolapse (i.e. new prolapse at other sites) (Table 9C)

Only one study, a non-randomised comparative study,58 reported data on this outcome. Cystocele occurred in 31% (10/32) women in the sacrocolpopexy group and in 14% (4/28) women in the sacrospinous colpopexy group (no mesh).

Need for further surgery for prolapse (Table 9D)

One non-randomised comparative study58 reported the need for further surgery for prolapse, with 25% (15/60) of women requiring a further operation for recurrent or de novo prolapse in the sacrocolpopexy group at a mean follow up of 34 months, and 8% (4/51) in the sacrospinous colpopexy group (no mesh) at a mean follow up of 38 months.

Persistent sexual symptoms (Table 9G)

Only one study, an RCT,57 reported data on persistent sexual symptoms; 44% (4/9) of women had persistent dyspareunia after sacrocolpopexy compared with 57% (4/7) after sacrospinous colpopexy (no mesh).

4.4.2.2 Safety

No studies reported new bowel or sexual symptoms after sacrocolpopexy for vault prolapse repair.

Blood loss i: Blood loss, ml

Two RCTs,35,57 two non-randomised comparative studies,58,64 and one case series79 reported blood loss as a continuous variable.

Mean blood loss during laparoscopic sacrocolpopexy ranged from 64ml (0-1000, n=60)58 to 172ml (0-1000, n=166).64 Mean blood loss during abdominal (open) sacrocolpopexy was similar across studies: 234ml (50-700, n=61),64 247ml (SD 148ml, n=54)35 and 298ml (SD 193ml, n=212).79 In the RCT comparing sacrocolpopexy and a no-mesh technique,57 there was no statistically significant difference between groups (362ml, n=47 vs 306, n=48; SMD 0.25, 95% CI -0.15 to 0.66). ii: Blood loss requiring blood transfusion (Table 10A)

Two RCTs35,57 two non-randomised comparative studies,58,64 and one case series40 reported blood loss requiring transfusion.

One RCT by Maher and colleagues57 and one non-randomised comparative study by Marcickiewicz and colleagues58 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in the need for transfusion between the groups (Figure 5) and two women in total (2/107) had a transfusion during sacrocolpopexy.

Figure 5 Sacrocolpopexy, vault prolapse repair, safety, blood loss requiring transfusion

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

One RCT35 compared non-absorbable synthetic mesh with a biological graft (cadaveric fascia lata) for sacrocolpopexy. One woman required transfusion (1/54, 1.9%, in the synthetic mesh group). One non-randomised comparative study64 compared laparoscopic with abdominal (open) sacrocolpopexy. One woman in each group (1/56, 1.8% vs. 1/61, 1.6%) required blood transfusion. No women required blood transfusion in the case series40 (n=149).

Damage to surrounding organs during operation (Table 10B)

Outcomes reported as cystotomy, bladder injury, bowel injury, colon perforation, enterotomy, bowel perforation and puncture of the inferior epigastric artery were considered in this section. Two RCTs35,57 two non-randomised comparative studies,58,64 and one case series50 reported damage to surrounding organs during operation.

One RCT by Maher and colleagues57 and one non-randomised comparative study by Marcickiewicz and colleagues58 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in damage to surrounding organs between the groups (Figure 6), although Marcickiewicz and colleagues used laparoscopy.

Figure 6 Sacrocolpopexy, vault prolapse repair, safety, damage to surrounding organs

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

One RCT35 compared non-absorbable synthetic mesh with biological graft (cadaveric fascia lata) in sacrocolpopexy. One bladder injury (1/54, 1.9%) occurred in the non-absorbable mesh group and none in the biological mesh group (0/46). One non-randomised comparative study64 compared laparoscopic with abdominal (open) sacrocolpopexy in cystotomy or enterotomy. There was no evidence of a statistically significant difference between the groups (2/56, 3.6% vs. 4/61, 6.6%; RR 0.54, 95% CI 0.10 to 2.86). The case series50 reported laparoscopic sacrocolpopexy, with two (2/103) women suffering a bowel perforation.

Mesh erosion (Table 10C)

Two full-text RCTs,35,57 two non-randomised comparative studies43,64 and four case series40,45,50,79 reported mesh erosion.

One RCT57 compared sacrocolpopexy (using non-absorbable synthetic mesh) with sacrospinous colpopexy (no mesh). In this study, 2.1% (1/47) of women had mesh erosion after sacrocolpopexy at a mean follow up of 24 months. The other RCT35 compared non- absorbable synthetic mesh and a biological graft (cadaveric fascia lata) for sacrocolpopexy. Of 54 women who received a non-absorbable mesh, two (2/54, 3.7%) had mesh erosion at one-year follow up compared with none (0/46) of those who received a biological graft.

One non-randomised comparative study43 compared two types of non-absorbable synthetic mesh, polypropylene and silicone-covered polyethylene, for sacrocolpopexy. No (0/24) erosion occurred in the polypropylene group but four cases of erosion occurred (4/21, 19%) in the silicone group. The four cases were detected at 4, 7, 12, and 20 months respectively.

The other non-randomised comparative study64 compared laparoscopic and abdominal (open) sacrocolpopexy. A mixture of absorbable and non-absorbable mesh was used in both groups. Two cases of erosion (2/56, 3.6%) occurred in the laparoscopic group and one (1/61, 1.6%) in the abdominal (open) group.

All of the four case series40,45,50,79 used non-absorbable synthetic mesh. Thirty-five (35/660, 5.3%) erosions occurred in total. The median (range) was 6.5% (0% to 8.7%, Table 10 C). The erosions occurred at various times ranging from 0.4 to 54 months. Median times to occurrence were reported as four months in 16 erosions (16/196) in one study45 and 8.4 months for 10 erosions (10/212) in another.79

Further operation needed for mesh erosion (Table 10D)

Two non-randomised comparative studies43,64 and three case series40,45,50 reported the need for a further operation for mesh erosion.

One non-randomised comparative study compared a polypropylene mesh with a silicone- covered polyethylene mesh in sacrocolpopexy.43 No women (0/24) needed a further operation for mesh erosion in the polypropylene group but five did (5/21, 23.8%) in the silicone group. The other non-randomised comparative study compared laparoscopic and

abdominal (open) sacrocolpopexy using a mixture of synthetic mesh and biological graft.64 Two women (2/56, 3.6%) had a further operation for mesh erosion in the laparoscopic group and one (1/61, 1.6%) in the abdominal (open) group. In the two studies, the mesh was removed transvaginally (n=3), laparoscopically (n=1), or abdominally (open) (n=4).

All three case series40,45,50 used non-absorbable synthetic mesh. A total of 10 (10/448, 2.2%; Table 10 D) women needed a further operation for mesh erosion.

De novo urinary symptoms (Table 10E)

Only one study, an RCT57 involving 46 women who had no urinary symptoms at baseline, reported this outcome. In this study, 9.1% (2/22) of women developed stress urinary incontinence after sacrocolpopexy compared with 33.3% (8/24) in the sacrospinous colpopexy group (no mesh).

Infection (Table 10H)

Outcomes reported as wound infection, urinary tract infection, mesh infection, pelvic abscess, delayed sepsis, postoperative fever, febrile morbidity, pneumonia, and infection (not specified) were considered in this section. Two RCTs35,57 three non-randomised comparative studies,43,58,64 and two case series40,50 reported data on infection.

One RCT by Maher and colleagues57 and one non-randomised comparative study by Marcickiewicz and colleagues58 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). Both used non-absorbable synthetic mesh. There was no evidence of a statistically significant difference in infection between groups (Figure 7).

Figure 7 Sacrocolpopexy, vault prolapse repair, safety, infection

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

One RCT35 compared non-absorbable synthetic mesh with biological graft (cadaveric fascia lata) in sacrocolpopexy. Two cases of postoperative fever occurred in each group (2/54 vs.

2/46). One non-randomised comparative study43 compared polypropylene mesh and silicone-covered polyethylene mesh for sacrocolpopexy. No mesh infection (0/24) occurred in the polypropylene group but one (1/21) occurred in the silicone group. Another non- randomised comparative study64 compared laparoscopic with abdominal (open) sacrocolpopexy. Six (6/56, 11%) cases of wound infection occurred in the laparoscopic group and two (2/61, 3.3%) in the abdominal (open) group.

In one case series,40 15% (22/149) of women had a urinary tract infection, 1.3% (2/149) a wound infection, 5.4% (8/149) a febrile morbidity, none (0/149) pelvic abscess, 2.7% (4/149) pneumonia and none (0/149) had delayed sepsis. The total infection rate, using a simple total of all types of infection, was 24.2% (36/149). This study did not report whether any women had more than one symptom, so it is possible that 24.2% is an overestimation of the infection rate at the individual level. A second case series50 reported that 1.9% (2/103) of women had a urinary tract infection. This is lower than the urinary tract infection rate found in the first case series40 (15%, 22/149).

Other serious adverse effects not otherwise specified (Table 10I)

Outcomes reported as ileus, small bowel obstruction, wound breakdown, incisional hernia, ventral hernia, osteomyelitis, deep vein thrombosis, pulmonary embolism, and cardiac complications were considered in this section. Two RCTs35,57 and two non-randomised comparative studies43,64 reported these outcomes.

One RCT57 compared sacrocolpopexy with sacrospinous colpopexy. Two (2/47, 4.3%) incisional hernias occurred in the sacrocolpopexy group and none (0/43) in the sacrospinous colpopexy group (no mesh). The other RCT35 compared non-absorbable synthetic mesh with biological graft (cadaveric fascia lata) in sacrocolpopexy; 20% (11/54) of women in the non- absorbable mesh group had pulmonary embolism (n=1), ileus (n=2) or wound breakdown (n=8), and 11% (5/46) in the biological graft group had wound breakdown.

One non-randomised comparative study43 compared polypropylene mesh with silicon- covered polyethylene mesh for sacrocolpopexy. No case of osteomyelitis occurred (0/24 vs. 0/21). The other non-randomised comparative study64 compared laparoscopic with abdominal (open) sacrocolpopexy. In this study, 5.4% (3/56) of women in the laparoscopic group had deep vein thrombosis (n=1), ventral hernia (n=1), or small bowel obstruction (n=1), while 15% (9/61) of women in the abdominal (open) group had deep vein thrombosis (n=1), ventral hernia (n=2), ileus (n=2), small bowel obstruction (n=2), or cardiac

complications (n=1 atrial fibrillation and mitral valve disease, n=1 atrial fibrillation with heart failure).

Other adverse effects not otherwise specified (Table 10J)

Outcomes considered in this section included rectovaginal haematoma, pelvic haematoma, wound haematoma, vaginal pain of undetermined origin, hypercarboxaemia or hyponatraemia requiring overnight stay in an intensive care unit, and sciatica requiring physiotherapy. One RCT57 and two case series40,50 reported these outcomes.

The RCT compared sacrocolpexy with sacrospinous colpopexy (no mesh). One case of rectovaginal haematoma and one case of vaginal pain of undetermined origin occurred in the no mesh group and none in the sacrocolpopexy group. In the two case series,40,50 ten events (10/252, 4.0%) occurred in total. The median rate (range) was 5.0% (0% to 10%, Table 10 J).

4.4.2.3 Operation time and hospital stay

(In this section averages are means unless stated otherwise.)

Operation time

Two RCTs,35,57 two non-randomised comparative studies,58,64 and one case series50 reported operation time. For laparoscopic sacrocolpopexy, the average operation time in each of three studies was 107 minutes (median, n=103),50 129 minutes (median, n=60)58 and 269 minutes (n=56).64 For abdominal (open) sacrocolpopexy the average operation times (across two studies and three arms) were 218 minutes (n=61),64 227 minutes (n=54)35 and 233 minutes (n=46).35 Sacrospinous colpopexy (no mesh) had shorter average operating times of 62 minutes (median, n=51)58 and 76 minutes (n=48).57,58

Hospital stay

One RCT,57 two non-randomised comparative studies58,64 and two case series50,79 reported hospital stay. Average hospital stay was reported as two days (n=56)64 and four days in two studies (median, n=10350 and n=6358) after laparoscopic sacrocolpopexy; four days in each of two studies (n=6164 and n=21279) for abdominal (open) sacrocolpopexy; and four days58 (n=51) or five days (n=48)57 after sacrospinous colpopexy (no mesh).

4.4.3 Sacrocolpopexy for uterine and/or vault prolapse (not reported separately) (mesh)

Twenty one studies21,22,25-28,30,34,36,44,47,49,51,55,56,61,71,73,75,77,80 involving 3017 women did not report data for uterine and vault prolapse repair separately (Table 9). Two were full-text RCTs: both27,56 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). Ten were non-randomised comparative studies21,22,25,44,47,51,61,73,77,80 with various comparators. Nine were case series26,28,30,34,36,49,55,71,75 which all used non-absorbable synthetic mesh for sacrocolpopexy.

4.4.3.1 Efficacy

Subjective failure: persistent prolapse symptoms (Table 10A)

One RCT,27 two non-randomised comparative studies,44,61 and three case series36,49,71 reported subjective failure.

The RCT (by Benson and colleagues)27 and one non-randomised comparative study (by Ng and colleagues)61 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). Subjective failure (reported as symptomatic prolapse) occurred more often in the RCT after sacrocolpopexy than in the non-randomised comparative study (6/38, 16% vs. 2/60, 3.3%). Longer mean follow up in the RCT (2.5 years vs. 16 months) may explain this result. The methodological quality of the RCT27 was reasonably high (Appendix 7). While there was no evidence of a statistically significant difference in subjective failure between the operations in the RCT27(RR 0.47, 95% CI 0.20 to 1.11, Figure 8), there was a statistically significant difference in favour of sacrocolpopexy in the non-randomised comparative study61 (RR 0.15, 95% CI 0.03 to 0.63, Figure 8).

Figure 8 Sacrocolpopexy, uterine or vault prolapse repair, efficacy, subjective failure

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

One small non-randomised comparative study44 compared non-absorbable synthetic mesh with biological graft (fascia lata) for sacrocolpopexy. There were no significant differences between the groups for subjective failure after a mean follow up of 26 months and 21 months respectively (2/19, 11% vs. 7/18, 39%; RR 0.27, 95% CI 0.06 to 1.13).

All three case series used an abdominal (open) technique for sacrocolpopexy. One36 used non-absorbable synthetic mesh and two49,71 used a variety of absorbable and non- absorbable mesh. The outcomes reported were prolapse-related complaints, prolapse symptoms, and pelvic pressure respectively. Across studies, the median subjective failure rate was 26% (range 23% to 31%; Table 10 A). The mean follow-up time was similar across studies (48, 45, and 43 months).

Objective failure: recurrent prolapse at original site (Table 10B)

One full-text RCT,56 two non-randomised comparative studies47,61 and four case series30,34,49,55 reported objective failure. Most studies used uterine or vault prolapse stage II or greater as the definition of objective failure. No study reported when the failure was detected.

The full-text RCT56 and the two non-randomised comparative studies47,61 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). No vault failure (0/52) occurred in the sacrocolpopexy group in the RCT56 at a mean follow up of 2.1 years. The vault failure rate in the sacrospinous colpopexy group was not reported: however, the failure rate for prolapse at any site (recurrent and de novo combined) was 3/52 (5.7%) for sacrocolpopexy versus 13/66 (19.7%) for sacrospinous colpopexy. The failure rate after sacrospinous colpopexy was high (13/48, 27%) in one non-randomised comparative study61 (Figure 9).

Figure 9 Sacrocolpopexy, uterine or vault prolapse repair, efficacy, objective failure

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

The mean follow up was more than 16 months in all four case series.30,34,49,55 The objective failure rates were similar across studies (median 1.8%, range 0% to 3.1%; Table 10 B).

De novo prolapse (i.e. new prolapse at other sites) (Table 10C)

Only one study, a case series28 reported de novo prolapse. In this study, posterior vaginal wall prolapse occurred in 1.6% (3/193) of women at one year after sacrocolpopexy.

Further operation needed for prolapse (recurrent or de novo) (Table 10D)

Two case series49,55 reported the number of women who needed a further operation for prolapse. Both used an abdominal (open) technique and a variety of biological, combined and non-absorbable meshes. In one case series,49 14% (13/90) of women required a further operation for prolapse at a mean follow up of 45 months. In the other,55 where all women had more than 16 months of follow up, 1.5% (4/262) required a further operation.

Persistent urinary symptoms (Table 10E)

Only one study, a case series28 reported persistent urinary symptoms in women having these symptoms at baseline. In this study, 25% (43/169) of women had persistent bothersome irritative symptoms at one year after sacrocolpopexy.

Persistent bowel symptoms (Table 10F)

One case series28 reported persistent bowel symptoms in women having these symptoms at baseline. At one year after abdominal (open) sacrocolpopexy, 13% (7/54) of women still needed digital assistance to empty their bowels.

Persistent sexual symptoms (Table 10G)

One case series28 reported persistent sexual symptoms in women having these symptoms at baseline. In this study, 42% (25/60) of women still had dyspareunia one year after sacrocolpopexy.

4.4.3.2 Safety

Blood loss i: Blood loss, ml

One RCT,56 six non-randomised comparative studies,21,44,47,51,61,80 and two case series34,71 reported blood loss as a continuous variable. In the RCT,56 mean blood loss during sacrocolpopexy (150ml) was significantly lower than during sacrospinous colpopexy (no mesh, 448ml, SMD -1.39, 95% CI -1.79 to -0.98). None of the non-randomised comparative studies reported standard deviations, precluding meta-analysis.

Mean blood loss was reported as 40ml (n=92)80 and 83ml (n=25)51 for laparoscopic sacrocolpopexy; 195ml (n=22),51 276ml (n=49),44 298ml (n=33),44 302ml (n=132)71 and 557ml (n=113)61 for abdominal (open) sacrocolpopexy; and 182ml (n=187),80 280ml (n=70),80 239ml (n=64),61 and 567ml (n=125)47 for sacrospinous colpopexy (no mesh). ii: Blood loss requiring blood transfusion (Table 10A)

One RCT,27 five non-randomised comparative studies,47,51,61,73,80 and seven case series26,30,34,36,55,71,75 reported blood loss requiring blood transfusion.

The RCT (by Benson and colleagues)27 and four non-randomised comparative studies47,61,73,80 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in the need for blood transfusion between sacrocolpopexy and sacrospinous colpopexy (Figure 10).

Figure 10 Sacrocolpopexy, uterine or vault prolapse repair, safety, number of women requiring blood transfusion

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

Another non-randomised comparative study51 compared laparoscopic with abdominal (open) sacrocolpopexy. No transfusion was required during either operation (0/25 vs. 0/22). Across the seven case series, a median of 1.2% of women required a blood transfusion (range 0.8%

to 17%, Table 10A). Some case series did not report the intervention route, i.e. laparoscopic or abdominal (open).

Damage to surrounding organs during operation (Table 10B)

Outcomes considered in this section included bladder injury, urethral injury, rectal mucosa injury, rectal laceration, small bowel serosal tear, and obturator nerve injury. Two full-text RCTs,27,56 two non-randomised comparative studies61,80 and six case series26,30,34,36,49,55 reported damage to surrounding organs during operation.

All four comparative studies compared sacrocolpopexy with sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in the risk of damage to surrounding organs in any of the individual studies or when they were combined (Figure 11).

Figure 11 Sacrocolpopexy, uterine or vault prolapse repair, safety, damage to surrounding organs

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

The six case series26,30,34,36,49,55 involved a total of 1001 women. Across these studies, a median (range) of 2.0% (0.4% to 7.9%; Table 10 B) of women had organ damage during operations.

Mesh erosion (Table 10C)

Eight non-randomised comparative studies21,22,25,44,47,51,61,77 and seven case series26,30,34,36,49,55,71 reported mesh erosion. The mean follow up ranged from four months to over four years.

Three non-randomised comparative studies22,47,61 compared sacrocolpopexy (mesh) with sacrospinous colpopexy (no mesh). All used non-absorbable synthetic mesh but of different Amid types. In operations where an Amid type I mesh was used, the erosion rate was 12% (4/34) at one-year follow up.22 In operations where an Amid type II mesh was used, the erosion rate was 0% (0/80) at a mean follow up of four years.47 In operations where an Amid type III mesh was used, the erosion rate was 1.7% (1/60) at a mean follow up of 18 months.64

A non-randomised comparative study51 compared laparoscopic with abdominal (open) sacrocolpopexy. A variety of biological (fascia lata) and non-absorbable materials were used. Five (3/25 vs. 2/22) erosions occurred in total. Another non-randomised comparative study77 compared sacrocolpopexy (mesh) with sacrocolpoperineopexy (mesh) using a variety of non-absorbable synthetic meshes. Five erosions (5/155, 3.2%) occurred in the sacrocolpopexy group compared with 10 (10/118, 8.5%) in the sacrocolpoperineopexy group.

Three non-randomised comparative studies21,25,44 compared non-absorbable synthetic mesh with biological graft (porcine dermis or fascia lata) for sacrocolpopexy. No erosions occurred in two of these studies (0/42 and 0/37)21,44 and seven (7/74) erosions associated with non- absorbable mesh occurred in the other study.25

In the seven case series26,30,34,36,49,55,71 that reported sacrocolpopexy, three used non- absorbable synthetic mesh with erosion rates of 1.0% (1/101),36 2.4% (6/245),34 and 6.5% (4/62)26 and another four case series30,49,55,71 used a mixture of absorbable and non- absorbable mesh with erosion rates ranging from 1.6% to 3.8%.

Time to mesh erosion

In the studies which reported this information, the time when erosion was detected ranged from 2 to 36 months.25,26,30,49,55,77

Further operation needed for mesh erosion (Table 10D)

Four non-randomised comparative studies22,25,51,77 and six case series26,30,34,49,55,71 reported further operation needed for mesh erosion. In the non-randomised comparative studies, 5.5% (19/348; median 6.1%, range 3.2% to 11%) of women treated with sacrocolpopexy required an operation for mesh excision or removal (Table 10 D). In the six case series,

2.9% of women (28/973, median 2.9%, range 0.8% to 6.5%) treated with sacrocolpopexy required an operation for mesh erosion.

De novo urinary symptoms (Table 10E)

One RCT,56 one non-randomised comparative study,61 and one case series28 reported data on de novo urinary symptoms in women who had no urinary symptoms at baseline. The RCT by Lo and colleagues56 and the non-randomised comparative study by Ng and colleagues61 compared abdominal (open) sacrocolpopexy with sacrospinous colpopexy (no mesh). The mean follow up was 2.1 years in the RCT. In the non-randomised comparative study, the mean follow up was 18 months in the sacrocolpopexy group and 13 months in the sacrospinous colpopexy group. There was no evidence of a statistically significant difference in the rate of de novo urinary incontinence in either study (Figure 12). In the case series, 9.2% (12/131) of women had bothersome urinary symptoms at one year after sacrocolpopexy.28

Figure 12 Sacrocolpopexy, uterine or vault prolapse repair, safety, de novo urinary symptoms

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

De novo bowel symptoms (Table 10F)

Only one study, a case series28 reported data on de novo bowel symptoms in women who had no bowel symptoms at baseline. At one year after sacrocolpopexy, 1.1% (2/178) of women needed digital assistance to empty their bowels.

De novo sexual symptoms (Table 10G)

One RCT56 and one case series28 reported data on de novo sexual symptoms in women who had no sexual symptoms at baseline. In the RCT, one woman (1/11, 9.1%) in the sacrocolpopexy group developed dyspareunia at a mean follow up of 2.1 years compared with seven (7/18, 39%) in the sacrospinous colpopexy group. In the case series, 11 (11/76, 15%) women developed dyspareunia within one year after sacrocolpopexy.

Infection (Table 10H)

Outcomes reported as urinary tract infection, wound infection, retroperitoneal infection, perineal infection, cystitis, pyrexia, febrile morbidity, and infection (not specified), were considered in this section. For studies that reported more than one infection outcome, we added the numbers together. However, an individual patient could present with more than one symptom of infection; therefore, we might have overestimated the infection rate at the individual level.

Two full-text RCTs,27,56 four non-randomised comparative studies,21,47,61,73 and three case series26,30,36 reported these outcomes. The two full-text RCTs and three non-randomised comparative studies47,61,73 compared sacrocolpopexy with sacrospinous colpopexy (no mesh). An overall pooled estimate was not calculated due to the statistical heterogeneity, different reporting of numbers (of women or events) and different study designs. There were no statistically significant differences in infection rates between interventions except in one non-randomised study by Ng and colleagues61 (Figure 13). In the three case series,26,30,36 the infection rates ranged from 0.8 to 17% (Table11 H) for sacrocolpopexy.

Figure 13 Sacrocolpopexy, uterine or vault prolapse repair, safety, infection

SCP: sacrocolpopexy; SSC: sacrospinous colpopexy

Note: In the study by Ng and colleagues,61 for sacrocolpopexy group, 77=12 urinary tract infection +6 wound infection + 59 pyrexia; for sacrospinous colpopexy group, 25=7 urinary tract infection + 0 wound infection + 18 pyrexia.

Other serious adverse effects not otherwise specified (Table 10I)

Outcomes considered in this section included ileus, small bowel obstruction, ureter obstruction requiring stent placement, wound dehiscence, wound breakdown, wound separation, fascia dehiscence, development of vaginal band requiring excision, incisional hernia, abdominal hernia, cicatricial hernia, rectovaginal fistula, vesico-cutaneous fistula, thrombo-embolism, deep vein thrombosis, pulmonary embolism, pneumonia, myocardial infarction, congestive heart failure, re-admission (the causes were not reported), arrhythmia, and left ventricular failure. Two full-text RCTs,27,56 six non-randomised comparative studies21,47,51,61,73,80 and seven case series26,30,34,36,49,55,75 reported data on these outcomes.

Two RCTs27,56 and four non-randomised comparative studies47,61,73,80 compared sacrocolpopexy with sacrospinous colpopexy (no mesh) (Figure 14). There was no evidence of a statistically significant difference in other serious adverse effects (not otherwise specified) in the RCTs between these two techniques but sacrocolpopexy was associated with a statistically significant higher risk of other serious adverse effects (not otherwise specified) in non-randomised comparative studies (pooled RR 6.94, 95% CI 1.02 to 26.38; Figure 14).

Figure 14 Sacrocolpopexy, uterine or vault prolapse repair, safety, other serious adverse effects not otherwise specified

In all of the studies reporting this outcome, the occurrence of other serious adverse effects (not otherwise specified) ranged from 0% to 8.5% for sacrocolpopexy (with mesh) and 0% to 2.4% for techniques not using mesh (Table 10 I).

Other adverse effects not otherwise specified (Table 10J)

Outcomes considered in this section included prolonged catheterisation (≥ 4 days), urinary retention >5 days, haematoma, pelvic haematoma, vaginal vault haematoma, wound haematoma, extra-peritoneal haematoma, persistent vaginal discharge, vaginal bleeding, phlebitis, wound complications, vaginal vault strictures, gastrointestinal complications (e.g. nausea, emesis, bloating or ileus), subacute intestinal obstruction, sciatica, transient right gluteal pain, buttock pain, foot drop, and problems of motion. Two full-text RCTs,27,56 five non-randomised comparative studies21,47,61,73,80 and five case series26,28,30,34,36 reported data on these outcomes.

In the six studies comparing sacrocolpopexy with sacrospinous colpopexy (no mesh), one RCT by Lo and colleagues56 favoured the mesh technique, but there was no overall evidence of a statistically significant difference in other adverse effects between techniques (Figure 15).

Figure 15 Sacrocolpopexy, uterine or vault prolapse repair, safety, other adverse effects not otherwise specified

In all studies reporting this outcome, the occurrence of other adverse effects (not otherwise specified) ranged from 0% to 16% for sacrocolpopexy, and 0% to 29% for techniques not using mesh (Table 10 J).

4.4.3.3 Operation time and hospital stay

In this section averages are means unless stated otherwise.

Operation time

Two full-text RCTs,27,56 five non-randomised comparative studies21,44,51,61,80 and one case series71 reported operation time. For abdominal (open) sacrocolpopexy the average operation time was reported as 97min (n=132),71 158min (n=52),56 185min (median, n=22),51 210min (n=49)44 and 270min (n=33).44 Average operation time for laparoscopic sacrocolpopexy was reported as 218min (n=92)80 and 220min (median, n=25).51 The reported average operation time ranged from 78min to 196min27,56,61,80 for sacrospinous colpopexy (no mesh).

Hospital stay

Two RCTs,27,56 four non-randomised comparative studies21,51,61,73 and one case series71 reported hospital stay. Mean hospital stay was reported as 1.2 days for laparoscopic sacrocolpopexy,51 ranged from 3.4 to 7.2 days for abdominal (open) sacrocolpopexy,51,56,61,71,73 and ranged from 4.0 to 8.8 days for sacrospinous colpopexy (no mesh)27,56,61,73

4.4.4 Summary of the evidence for and against sacrocolpopexy

Thirty two studies (4 full-text RCTs, 1 RCT available as a conference abstract, 15 non- randomised comparative studies, and 12 case series with sample sizes over 100) involving a total of 4456 women provided data on sacrocolpopexy. Figures 16 and 17 show the range of event rates from all studies presented in Table 9 and Table 10 respectively, to give an overview of the likely efficacy and safety for sacrocolpopexy. The data should be interpreted bearing in mind the often small numbers of women involved.

Efficacy

The objective failure rates ranged from 0% to 6.1% in 11 studies involving 1054 women, while the subjective failure rates varied from 3.3% to 31% in 9 studies involving 638 women. In 4 studies involving 451 women, the proportions of those needing a further operation for recurrent or de novo prolapse ranged from 0% to 25%. Estimates for other outcomes were based on single studies only (shown as a single horizontal bar in the figure). Meta-analysis was carried out to compare subjective failure (Figure 4 & 8, both with two studies) and objective failure (Figure 9, with two studies) in sacrocolpopexy versus sacrospinous

colpopexy (no mesh). There was no evidence of a statistically significant difference in either outcome between the operations.

Figure 16 Ranges of event rates across studies: sacrocolpopexy, efficacy

50 NA, 38/209, 94/429 NA, 4/9, 25/60

40 NA, 10/32, 3/193 NA, NA, 43/169 0/39, 15/60, 17/352 30

20 0/62, 8/306, 12/686 NA, NA, 7/54 Event rates (%) 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent urinary Persistent bowel Persistent sexual other sites prolapse symptoms symptoms problems

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair

Safety

The estimates for other serious adverse effects (not otherwise specified) ranged from 0% to 16% (21 studies, 2059). The events that occurred included bowel obstruction, ureteral obstruction, wound dehiscence, incisional hernia, vesico-cutaneous fistula, deep vein

thrombosis, pulmonary embolism, pneumonia, myocardial infarction, congestive heart failure, arrhythmia, and left ventricular failure. The proportion of women with other adverse effects (not otherwise specified) ranged from 0% to 16% (17 studies, n=1751). The events that occurred included voiding dysfunction, haematoma, persistent vaginal discharge, vaginal bleeding, phlebitis, wound complications, gastrointestinal complications, e.g. nausea, emesis, bloating or subacute intestinal obstruction, sciatica, and buttock pain.

Meta-analysis was carried out to compare blood transfusion (Figure 5, 2 studies; Figure 10, 5 studies), damage to surrounding organs (Figure 6, 3 studies; Figure 11, 4 studies), infection (Figure 7, 2 studies; Figure 13, 5 studies), new urinary symptoms (Figure 12, 2 studies) and other adverse effects (not otherwise specified; Figure 15, 6 studies) between sacrocolpopexy and sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in any of these outcomes between the operations. Meta- analysis was also conducted for other serious adverse effects (not otherwise specified; Figure 14, 5 studies): in four non-randomised comparative studies, sacrocolpopexy was associated with statistically significantly more adverse events than sacrospinous colpopexy (no mesh) (pooled RR 6.94, 95% CI 1.02 to 26.38).

Figure 17 Ranges of event rates across studies: sacrocolpopexy, safety 3/39, 55/621, 158/731 70

0 Blood Organ Mesh Operation for New urinary New bowel New sexual Infection Other Other transfusion damage erosion erosion symptoms symptoms symptoms serious adverse adverse effects effects

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair (not separated)

Notes: 1. Where only one study reported the outcome, a single horizontal bar reports its event rate; where more than one study reported the outcome, the range of the event rates across all studies is shown by a vertical bar; 2. The numbers in the figures are the number of events which occurred across all studies divided by the total number of women who received the treatment; 3. NA: no studies reported data for that outcome

Operation time and hospital stay

4.5 Infracoccygeal sacropexy (mesh)

Fourteen studies (2 RCTs available as conference abstracts,59,74 1 non-randomised comparative study,60 2 case series with sample sizes ≥100,48,76 and 9 case series with sample sizes <10039,41,42,53,62,63,65,69,70 involving 976 women provided data on infracoccygeal sacropexy. One study60 reported uterine repair, 539,41,42,59,63 reported vault repair, one70 reported uterine and vault prolapse repair separately, and 748,53,62,65,69,74,76, studies did not report uterine and/or vault repair separately. Table 11 summarises the patient characteristics and surgical procedures in these 14 studies. The median follow up across these 14 studies was 13 months (range 5 to 30 months).

Data from individual studies reporting efficacy outcomes are given by type of prolapse in Appendix 13, the safety outcomes are in Appendix 14, and the operation time and hospital stay are in Appendix 15. The median rates and ranges for efficacy outcomes, by type of prolapse, are summarised in Table 12, and safety outcomes are in Table 13. No RCTs or non-randomised comparative studies reported data on disease-specific quality of life.

Methodological quality of the included studies

The one non-randomised comparative study60 collected data prospectively but did not report loss to follow up (Appendix 8).

Among the 11 case series, seven studies were prospective.39,39,42,48,63,65,69 Two studies recruited patients consecutively.39,48 Three studies reported loss to follow up, which ranged from 2% to 47%.

Table 11 Infracoccygeal sacropexy: summary of patient characteristics and surgical procedures ID N Age, y Uterine/ Primary/ Intervention Technique Mesh/graft Anti-incontinence Follow-up Outcomes vault secondary route procedures UTERINE Non-randomised comparative studies Neuman 200760 A, 44 A, 64 (13) A, 44/0 NR A, vaginal A, hysterectomy + infracoccygeal NR Anti-incontinence: 30m (12-44) Efficacy B, 35 B, 51 (10) B, 35/0 B, vaginal sacropexy A, 9/44; B, 6/35 Safety B, infracoccygeal sacropexy with Hysterectomy: uterus preservation A, 44/44; B, 0/35 Case series (n<100) Sivaslioglu 200570 10 60 (7) 10/0 NR Vaginal Infracoccygeal sacropexy NR NR 16m (6) Efficacy (presume uterus was preserved, Safety as mean operation time was only 45min.) VAULT RCT (abs.) Meschia 200559 A, 30 NR A, 0/30 NR A, vaginal A, infracoccygeal sacropexy A, NR NR 24m Safety 62 B, 30 B, 0/30 B, vaginal B, sacrospinous fixation B, no mesh Case series (n<100) Farnsworth 200239 93 65 (36-77) 0/93 NR Vaginal Infracoccygeal sacropexy Polypropylene tape (IVS Tunneller) NR 12m (2-24) Efficacy Safety Foote 200741 52 64 0/52 NR Vaginal Infracoccygeal sacropexy Multifilament polypropylene (NR 5/52 20wk Safety trade name)

Ghanbari 200642 15 67 (50- 0/15 NR Vaginal Infracoccygeal sacropexy Polypropylene tape (IVS Tunneller) NR NR Safety 81)00 Petros 200163 75 54 (40-74) 0/75 NR Vaginal Infracoccygeal sacropexy Nylon tape (NR trade name) NR 1-4.5y Efficacy Safety Sivaslioglu 200570 20 60 (7) 0/20 NR Vaginal Infracoccygeal sacropexy NR NR 16m (6) Efficacy Safety UTERINE AND/OR VAULT RCT (abs.)

ID N Age, y Uterine/ Primary/ Intervention Technique Mesh/graft Anti-incontinence Follow-up Outcomes vault secondary route procedures De Tayrac 200674 A, 21 NR NR NR A, vaginal A, infracoccygeal sacropexy A, Tyco Healthcare Hysterectomy was A, 11m (1.5-34) Efficacy B, 20 B, vaginal B, unilateral sacrospinous B, no mesh performed as B, 16m (1.5-32) Safety suspension needed Case series (n ≥100) Hefni 200748 127 59 (9) 83/44 NR Vaginal Infracoccygeal sacropexy Multifilament polypropylene tape, Anti-incontinence : 14m (2-26) Efficacy IVS Tunneller 8/127 Safety Hysterectomy : 22/127 Vardy 200776 286 61 (14) NR NR Vaginal Infracoccygeal sacropexy Multifilament polypropylene tape, Anti-incontinence: 1y Efficacy (53 for IVS Tunneller device (approved by 122/186 Safety efficacy) FDA) Hysterectomy : NR Case series (n <100) Jordaan 200653 8 NR NR NR Vaginal Infracoccygeal sacropexy Absorbable and non-absorbable NR 13m Efficacy combined mesh, polyproglactine and prolene 1 :1, Vypro 63

Oliver 200662 14 73 (18) 7/7 NR Vaginal Modified infracoccygeal Polypropylene tape, IVS Tunneller Anti-incontinence: Efficacy: 6w Safety sacropexy with uterine NR Safety: 5m (2- preservation Hysterectomy: 3/7 11)

Petros 200565 67 65 (35-87) 23/40 NR Vaginal Tissue Fixation System posterior Polypropylene, Tissue Fixation NR 9m (3-15) Safety sling (a direct evolution of the System device, approved by the infracoccygeal sacropexy) Australian and European Government regulatory bodies

Sentihes 200769 44 66 (50-84) 24/20 NR Vaginal Non-absorbable hammock Polyester, Parietex, then changed to Anti-incontinence: 29m (9-47) Efficacy placement using anterior trans- multifilament polypropylene, 0/44 Safety obturator and posterior Surgipro, then low-weight Hysterectomy: 24/44 infracoccygeal extensions monofilament polypropylene, Ugytex

Table 12 Infracoccygeal sacropexy, efficacy: summary of median (range) event rates Uterine Vault Uterine and/or vault No mesh (sacrospinous No mesh (sacrospinous Infracoccygeal sacropexy No mesh (no data) Infracoccygeal sacropexy Infracoccygeal sacropexy suspension) suspension) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, median study % (range) study % (range) study % (range) study % (range) study % (range) study % (range) A. Subjective failure: persistent prolapse symptoms RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 0 0 1 8/91 (8.8%) 8.8% 0 2 28/171 (16%) 12% (2.3-21) 0 B. Objective failure: recurrent prolapse at original site RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 1 1/21 (4.8%) 4.8% 0 0/20 (0%) 0% Non-rand 1* 1/79 (1.3%) 1.3% 0 0 0 0 0 Case series 1 1/10 (10%) 10% 0 2 4/60 (6.7%) 5.0% (0-10) 0 4 17/232 (7.3%) 6.5% (0-25) 0 C. De novo prolapse at other sites RCT 0 0 0 0 0 0 64 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 0 0 0 0 1 3/63 (4.8%) 4.8% 0 D. Further operation needed for prolapse (recurrent or de novo) RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 1 0/10 (0%) 0% 0 3 14/151 (9.3%) 1.1% (0-30) 0 1 10/127 (7.9%) 7.9% 0 E. Persistent urinary symptoms RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 0 0 0 0 1 5/17 (29%) 29% 0 F. Persistent bowel symptoms (no studies reported this outcome)

Uterine Vault Uterine and/or vault No mesh (sacrospinous No mesh (sacrospinous Infracoccygeal sacropexy No mesh (no data) Infracoccygeal sacropexy Infracoccygeal sacropexy suspension) suspension) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, median study % (range) study % (range) study % (range) study % (range) study % (range) study % (range) G. Persistent sexual symptoms RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 0 0 0 0 1 3/7 (43%) 43% 0

*Results from the two arms of one study were combined. 65

Table 13 Infracoccygeal sacropexy, safety: summary of median (range) event rates Uterine Vault Uterine and/or vault No mesh (sacrospinous No mesh (sacrospinous infracoccygeal sacropexy No mesh (no data) infracoccygeal sacropexy infracoccygeal sacropexy suspension) suspension) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, median study % (range) study % (range) study % (range) study % (range) study % (range) study % (range) A. Blood loss requiring blood transfusion RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 0 0 0 0 0 0 Case series 1 0/10 (0%) 0% 0 3 0/188 (0%) 0% 0 3 1/185 (0.5%) 0% (0-2.3) 0 B. Damage to surrounding organs during operation RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 1 0/21 (0%) 0% 1 0/20 (0%) 0% Non-rand 0 0 0 0 0 0 Case series 1 0/10 (0%) 0% 0 4 3/203 (1.5%) 0.5% (0-2.7) 0 4 1/471 (0.2%) 0% (0-2.3) 0 C. Mesh erosion RCT 0 0 0 0 0 0 66 RCT (abs.) 0 0 1 2/30 (6.7%) 6.7% 1 NA 1 0/21 (0%) 0% 1 NA Non-rand 1* 10/79 (13%) 13% 0 0 0 0 0 Case series 0 0 4 22/235 (9.4%) 6.6% (5.3-21%) 0 4 33/524 (6.3%) 5.6% (1.5-17) 0 D. Further operation needed for mesh erosion RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 1* 10/79 (13%) 13% 0 0 0 0 0 Case series 0 0 1 4/75 (5.3%) 5.3% 0 4 32/524 (6.1%) 5.6% (1.5-17) 0 E. De novo urinary symptoms in women who did not have these symptoms at baseline RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 0 0 Non-rand 1* 0/55 (0%) 0% 0 0 0 0 0 Case series 0 0 0 0 1 0/27 (0%) 0% 0 F. De novo bowel symptoms (no studies reported this outcome) G. De novo sexual symptoms (no studies reported this outcome)

Uterine Vault Uterine and/or vault No mesh (sacrospinous No mesh (sacrospinous infracoccygeal sacropexy No mesh (no data) infracoccygeal sacropexy infracoccygeal sacropexy suspension) suspension) No. n/N, Median No. n/N, median No. n/N, median No. n/N, median No. n/N, median No. n/N, median study % (range) study % (range) study % (range) study % (range) study % (range) study % (range) H. infection RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 1 0/21 (0%) 0% 1 0/20 (0%) 0% Non-rand 0 0 0 0 0 0 Case series 0 0 3 6/183 (3.3%) 1.3% (0-5.4) 0 4 16/494 (3.2%) 4.8% (2.1-9.0) 0 I. Other serious adverse effects not otherwise specified (no studies reported this outcome) J. Other adverse effects not otherwise specified RCT 0 0 0 0 0 0 RCT (abs.) 0 0 0 0 1 2/21 (10%) 10% 1 2/20 (10%) 10% Non-rand 0 0 0 0 0 0 Case series 0 0 3 0/183 (0%) 0 0 4 9/471 (1.9%) 2.7% (1.0-7.1) 0

Note: 67 1. */** Results from the two arms of one/two studies were combined. 2. Other adverse effects (not otherwise specified) that occurred included haematomas, urine retention, pain at skin incision, and vaginal pressure.

4.5.1 Infracoccygeal sacropexy for uterine prolapse repair (mesh)

Two studies60,70 involving 89 women reported uterine prolapse repair using infracoccygeal sacropexy. One was a non-randomised comparative study that compared hysterectomy followed by infracoccygeal sacropexy with infracoccygeal sacropexy with uterus preservation. The other study was a case series where the women were treated with infracoccygeal sacropexy with uterine preservation.

4.5.1.1 Efficacy

No studies reported subjective failure, new prolapse at other sites, persistent urinary symptoms, bowel symptoms, or sexual problems after infracoccygeal sacropexy for uterine prolapse repair.

Objective failure: recurrent prolapse at original site (Table 12B)

In the non-randomised comparative study,60 one (1/35, 3%) objective failure occurred in the uterine preservation group and none (0/44) occurred in the hysterectomy group at a mean follow up of 30 months. In the case series,70 one of the ten women (10%) had an objective failure (grade 3 prolapse) at a mean follow up of 16 months. Neither of the studies reported time to failure.

Need for further surgery for prolapse (Table 12C)

The case series70 reported the number of women who needed further surgery for prolapse. One of the ten women had an objective failure (Grade 3) but she declined having a further operation.

4.5.1.2 Safety

No studies reported de novo bowel symptoms or sexual problems in women who did not have these symptoms at baseline, infection, and any other adverse effects not otherwise specified.

Blood loss (Table 12A)

The case series70 reported blood loss. The average blood loss in ten women was 200ml (SD 20ml) after infracoccygeal sacropexy with uterine preservation, and none required transfusion. Damage to surrounding organs during operation (Table 12B)

The case series70 reported this outcome and no organ damage occurred in the ten women.

Mesh erosion and further operation needed for mesh erosion (Table 12 C&D)

The non-randomised comparative study60 reported mesh erosion as tape protrusion. The rates were 14% (6/44) in the hysterectomy followed by infracoccygeal sacropexy group and 11% (4/35) in the infracoccygeal sacropexy with uterine preservation group. Time to occurrence of tape protrusion was not reported. All patients had segmental tape resection in the out-patient clinic.

De novo urinary symptoms (Table 12E)

The non-randomised comparative study60 reported the occurrence of de novo urinary symptoms. At a mean follow up of 30 months (range 12 to 44 months), no women reported de novo urinary incontinence in either group (0/29 vs. 0/26).

4.5.1.3 Operation time and hospital stay

Operation time

The average operation time was 45 minutes in the case series70 where infracoccygeal sacropexy with uterine preservation was used in ten women.

Hospital stay

In the non-randomised comparative study,60 women stayed almost 3 days longer in hospital after hysterectomy followed by infracoccygeal sacropexy than infracoccygeal sacropexy with uterine preservation (mean 4.2 days, n=79 versus 1.5 days, n=35). In the case series, the mean hospital stay was 2 days for the ten women who received infracoccygeal sacropexy with uterine preservation.70

4.5.2 Infracoccygeal sacropexy for vault prolapse repair (mesh)

Six studies39,41,42,59,63,70 involving 315 women reported vault prolapse repair using infracoccygeal sacropexy. One was an RCT available as a conference abstract59 in which infracoccygeal sacropexy was compared with sacrospinous colpopexy (no mesh), while the other five were case series with sample sizes of less than 100.

4.5.2.1 Efficacy

No studies reported new prolapse at other sites, persistent urinary symptoms, bowel symptoms, or sexual problems after infracoccygeal sacropexy for vault prolapse repair.

Subjective failure: persistent prolapse symptoms (Table 12A)

One case series39 reported subjective failure as having a less than 50% reduction in the severity of symptoms. In this study, 8.8% (8/91) of women reported this outcome at a mean follow up of 12 months after infracoccygeal sacropexy.

Objective failure: recurrent prolapse at original site (Table 12B)

Two case series63,70 reported objective failure. One63 reported a failure rate of 10% (4/40) at a mean follow up of 55 months with all cases occurring within 24 months. No failure (0/20) occurred in the other case series70 at a mean follow up of 16 months.

Need for further surgery for prolapse (Table 12D)

Three case series39,63,70 reported the number of women who needed further surgery for prolapse. Across these study, a median of 1.1% (2/91, range 0%, 0/20 to 30%, 12/40) of women needed further surgery for prolapse at a follow up of 12 months to 4.5 years.

4.5.2.2 Safety

No studies reported de novo urinary symptoms, bowel symptoms, or sexual problems after infracoccygeal sacropexy for vault prolapse repair.

Blood loss (Table 13A)

Five case series39,41,42,63,70 reported blood loss as a continuous variable. Mean blood loss ranged from 110ml to 200ml.39,41,42,63,70 Three case series39,63,70 involving 188 women reported that no women required a blood transfusion.

Damage to surrounding organs during operation (Table 13B)

Four case series39,42,63,70 involving 203 women reported damage to surrounding organs. Three women had organ damage (rectal perforations) (median 0.5%, range 0% to 2.7%).

Mesh erosion (Table 13C)

One RCT available as a conference abstract59 and four case series39,41,42,63 reported mesh erosion. The RCT available as a conference abstract59 reported that 6.7% (2/30) of women had mesh erosion after infracoccygeal sacropexy at a mean follow up of 24 months (type of mesh not reported). The four case series reported a total of 22 (22/235, 9.4%) erosions and the median (range) rate was 6.6% (5.3% to 21%) with at least five cases occurring within 12 months.

Further operation needed for mesh erosion (Table 13D)

One case series63 reported that four women (4/75, 5.3%) needed a further operation for mesh erosion.

Infection (Table 13H)

Outcomes considered in this section included urinary tract infection or unspecified infection. Three case series39,42,63 involving 183 women reported data. The rates ranged from 0% to 5.4% with a median of 1.3% across studies.

Other adverse effects not otherwise specified (Table 13J)

Outcomes considered in this section included significant bruising or haematoma. Three case series39,42,63 reported data with no events occurring in the 183 women involved.

4.5.2.3 Operation time and hospital stay

Five case series39,41,42,63,70 reported operation time. The times ranged from 30 to 66 minutes (n=255). Five case series reported length of hospital stay: less than one day (n=9339,42,63, n=1542 and n=1563), two days (n=20)70 and four days (n=52).41

4.5.3 Infracoccygeal sacropexy for uterine and/or vault prolapse repair (not reported separately) (mesh)

Seven studies48,53,62,65,69,74,76 involving 587 women reported uterine and/or vault prolapse repair (not separately) using infracoccygeal sacropexy. One was an RCT available as a conference abstract74 in which infracoccygeal sacropexy was compared with sacrospinous colpopexy (no mesh), two were case series48,76 with a sample size of more than 100, and four were case series53,62,65,69 with a sample size of less than 100.

4.5.3.1 Efficacy

Subjective failure: persistent prolapse symptoms (Table 12A)

Two case series48,69 reported data on subjective failure. Outcomes were reported as prolapse symptoms48 and feeling of vaginal lump48,69 respectively. However, the rates were very different (27/127, 21% vs. 1/44, 2.3%) although the study with a higher rate48 had a much shorter mean follow up (14 vs. 29 months).

Objective failure: recurrent prolapse at original site (Table 12B)

One RCT available as a conference abstract74 and four case series48,53,69,76 reported objective failure. The RCT74 compared infracoccygeal sacropexy with sacrospinous colpopexy (no mesh). One (1/21, 4.8%) objective failure occurred in the infracoccygeal sacropexy group and none (0/20) occurred in the no-mesh group. In the four case series, the median objective failure rate was 6.5% (range 0% to 25%) at a mean follow up ranging from 12 to 14 months. No study reported when the failure was detected.

De novo prolapse (i.e. new prolapse at other sites) (Table 12C)

One case series48 reported de novo prolapse. Cystocele occurred at a rate of 4.8% (3/63) at a mean follow up of 14 months after infracoccygeal sacropexy.

Further operation needed for prolapse (Table 12D)

One case series48 reported further operation needed for prolapse. At a mean follow up of 14 months, 7.9% (10/127) of women required a further operation for prolapse.

Persistent urinary symptoms (Table 12E)

One case series69 reported persistent urinary symptoms in women having these symptoms at baseline. In this study, 29% (5/17) of women still had stress urinary incontinence at a follow up of 29 months after infracoccygeal sacropexy.

Persistent bowel symptoms (no studies reported this outcome)

Persistent sexual symptoms (Table 12G)

One case series69 reported data on persistent sexual symptoms. In this study, 43% (3/7) of women had persistent sexual symptoms at a mean follow up of 29 months.

4.5.3.2 Safety

Blood loss (Table 13A)

One case series62 reported blood loss as a continuous variable. Mean blood loss was 182ml (n=14) during infracoccygeal sacropexy. Three case series48,62,69 reported blood loss requiring transfusion. A median of 0% (range 0% to 2.3%) was reported across these studies.

Damage to surrounding organs during operation (Table 13B)

Outcomes considered in this section included urethral injury and rectal injury. One RCT available as a conference abstract74 and four case series48,62,69,76 reported damage to surrounding organs during operation. The RCT74 compared infracoccygeal sacropexy with sacrospinous colpopexy (no mesh). No rectal injury occurred in either group (0/21 vs. 0/20). Across the four case series, a median of 0% (range 0% to 2.3%) women had organ damage.

Mesh erosion (Table 13C)

One RCT available as a conference abstract74 and four case series48,65,69,76 reported mesh erosion. The RCT74 compared infracoccygeal sacropexy with sacrospinous colpopexy (no mesh). No mesh erosion (0/21) occurred at a mean follow up of 11 months in the mesh group. The four case series involved 524 women. Across these studies, mesh erosion occurred at a median rate of 5.6% (range 1.5% to 17%; Table 13 C). In the studies which reported time to erosion, the time when erosion was detected ranged from six weeks to 20 months.48,69,76

Further operation needed for mesh erosion (Table 13D)

Four case series48,65,69,76 reported the number of women who needed a further operation for mesh erosion: 6.1% (32/524; median 5.6%, range 1.5% to 17%) required an operation for erosion.

De novo urinary symptoms (Table 13E)

One case series69 involving 27 women who had no urinary symptoms at baseline reported de novo urinary symptoms. No women had stress urinary incontinence at a mean follow up of 29 months after infracoccygeal sacropexy.

De novo bowel symptoms (no studies reported this outcome)

De novo sexual symptoms (no studies reported this outcome)

Infection (Table 13H)

Outcomes reported as urinary tract infection, vaginal infection, infection requiring tape removal, vaginitis, and febrile morbidity, were considered in this section. For studies that reported more than one infection outcome, we added the numbers together. However, an individual patient could present with more than one symptom of infection; therefore, we may have overestimated the infection rate at the individual level. One RCT available as a conference abstract74 and four case series48,62,65,76 involving 536 women reported infection. In total, the rates ranged from 0% to 9.0%.

Other adverse effects not otherwise specified (Table 13J)

Outcomes considered in this section included urinary retention, haematoma, para-rectal haematoma, vaginal wall adhesion, vaginal pressure, and pain at skin incision. One RCT available as a conference abstract74 and four case series48,62,69,76 involving 511 women reported other adverse effects not otherwise specified. In total, the rates ranged from 1.0% to 10% for infracoccygeal sacropexy.

4.5.3.3 Operation time and hospital stay

One RCT available as a conference abstract74 and four case series48,62,65,69, involving 293 women reported operation time. The reported average operation time for infracoccygeal sacropexy varied considerably between studies: from less than 10 minutes to 2.5 hours. Four case series48,62,65,69 involving 252 women reported hospital stay. Mean hospital stay ranged from 1.5 to 5.0 days.

4.5.4 Summary of the evidence for and against infracoccygeal sacropexy

Fourteen studies (2 RCTs available as conference abstracts, 1 non-randomised comparative study, 2 case series with a sample size ≥100, and 9 case series with a sample size <100) involving 976 women provided data on infracoccygeal sacropexy. The ranges of event rates across studies in Tables 12 and 13, together with the number of women involved are shown in Figure 18 and Figure 19 respectively. Meta-analysis was not possible because no studies used the same comparator.

The proportion of women with subjective failure ranged from 2.3% to 21% (n=262, 3 studies) after infracoccygeal sacropexy, whereas the range for objective failure was 0% to 25% (9 studies, n=402). In 3 studies (n=288), 0% to 30% of women required re-operation for prolapse. Estimates for other outcomes were based on single studies including small numbers of women.

For safety, the need for blood transfusion ranged from 0% to 2.3% (7 studies, n=383). The proportion of women with organ damage ranged from 0% to 2.7% (9 studies, n=684). The occurrence of mesh erosion ranged from 0% to 21% (11 studies, n=889), and 1.5% to 17% of women needed an operation for mesh erosion (6 studies, n=678). Little evidence was available for new urinary symptoms, bowel symptoms, and sexual symptoms in women who did not have these symptoms at baseline. Infection ranged from 0% to 9.0% (8 studies,

n=698). Other adverse effects (not otherwise specified) ranged from 0% to 10% (8 studies, n=675). The events occurred included haematomas, urine retention, pain at skin incision, and vaginal pressure.

Figure 18 Ranges of event rates across studies: infracoccygeal sacropexy, efficacy

60 NA, NA, 3/7 50 0/10, 14/151, 10/127 40 2/89, 4/60, 18/253 NA, NA, 5/17 30 NA, 8/91, 28/171 20 Event rates (%) rates Event NA, NA, 3/63 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent Persistent bowel Persistent sexual other sites prolapse urinary symptoms (no problems symptoms data)

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair (not separated)

Figure 19 Ranges of event rates across studies: infracoccygeal sacropexy, safety

40 10/79, 24/265, 33/545 30 NA, 0/183, 11/492 0/10, 3/203, 1/492 10/79, 4/75, 32/524 20 0/10, 0/188, 1/185 0/55, NA, 0/27 NA, 6/183, 16/515 Event rates (%) rates Event 10

Uterine prolapse repair Vault prolapse repair Uterine and/or vault prolapse repair (not separated)

Notes: 1. Where only one study reported the outcome, a single horizontal bar reports its event rate; where more than one study reported the outcome, the range of the event rates across all studies is shown by a vertical bar; 2. The numbers in the figures are the number of events which occurred across all studies divided by the total number of women who received the treatment; 3. NA: no studies reported data for that outcome

Average operation times ranged from 10 minutes to 149 minutes with the majority of the studies (8/11) reporting less than one hour. Average hospital stays ranged from less than 24 hours to 5 days with the majority of studies (7/11) reporting two days or less.

4.6 Uterine suspension sling (mesh)

Uterine suspension sling is a technique which can only be used for women who wish to conserve their uterus, i.e. for uterine prolapse repair only. Six studies23,24,33,52,54,67 involving 239 women reported on uterine suspension sling operations (Table 14). One was a full-text RCT67 which compared sacrohysteropexy with hysterectomy, one was a non-randomised comparative study33 which compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy (also using mesh), and four were case series in which three23,24,54 reported sacrohysteropexy and one52 a different uterine suspension sling technique (suspending the uterus to the pectineal ligaments). All of the case series had a sample size of less than 100. The median follow up across the six studies was 33 months (12 to 95 months).

Individual studies reporting efficacy outcomes are given in Appendix 16, safety outcomes in Appendix 17, and operation time and hospital stay in Appendix 18. The median rates and ranges for efficacy outcomes are summarised in Table 15, and safety outcomes in Table 16. No RCTs or non-randomised comparative studies reported data on disease-specific quality of life.

Methodological quality of the included studies

In the full-text RCT,67 the allocation was adequately concealed. The mean follow up was one year. The loss to follow up was 5% (4/82). Intention-to-treat analysis was used (Appendix 7).

The non-randomised comparative study33 collected data prospectively and recruited patients consecutively. The loss to follow up was 4% (3/75) at a mean follow up of 51 months (Appendix 8).

Of the four case series, two studies were prospective.52,54 Two studies recruited patients consecutively.24,54 One study reported loss to follow up with a rate of 5% (1/20).52

Table 14 Uterine suspension sling: summary of patient characteristics and surgical procedures ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft Concomitant procedure Follow up (mean) Outcomes vault secondary RCT Roovers 200467 A, 41 A, 58 (9) A, 41/0 NR A, abdominal A, sacrohysteropexy A, polypropylene, Gore-Tex Anti-incontinence: 1y Efficacy B, 41 B, 56 (11) B, 41/0 B, vaginal B, hysterectomy + anterior (Amid type II) A, 16/41; B,11/41 Safety and/or posterior colporrhaphy B, no mesh Hysterectomy: A, 2/41 (B was performed); B, 41/41 Non-randomised comparative studies Costantini A, 36 A, 61 (12) A, 36/0 NR A, abdominal A, sacrohysteropexy A, polypropylene, Marlex Anti-incontinence: 51m (12-115) Efficacy 200533 B, 39 B, 62 (8) B, 39/0 B, abdominal B, hysterectomy + sacropexy (Amid type I) A, 28/36; B, 30/39 Safety B, same as A Hysterectomy: A, 0/36; B, 39/39 Case series (n <100) Banu 199723 19 17-27 19/0 NR Abdominal (open) Sacrohysteropexy Polyester, Mersilene (Amid Hysterectomy: 0/19 3-5y Efficacy 79 type III) Safety

Barranger 30 35.7 (29-43) 30/0 29/1 Abdominal (open) Sacrohysteropexy Polyester, Mersuture Anti-incontinence: 30/30 Efficacy: 44.5m (2- Efficacy 200324 Hysterectomy: 0/30 156); Safety Safety: 94.6m (8- 160)

Leron 200154 13 39 (27-60) 13/0 NR Abdominal (open) Sacrohysteropexy Non-absorbable synthetic Anti-incontinence: 4/13 16m (4-49) Efficacy mesh, Teflon Hysterectomy: 0/13 Safety

Joshi 199352 20 27.5 (17-32) 20/0 19/1 Abdominal (open) Uterine suspension sling Non-absorbable synthetic Anti-incontinence: 5/20 6-30m Efficacy (uterus is suspended to the mesh, Mersilene (Amid type Hysterectomy: 0/19 Safety pectineal ligaments) III)

NR: not reported

Table 15 Uterine suspension sling, efficacy: summary of median (range) event rates Uterine Uterine suspension sling No mesh (hysterectomy) No. study n/N, % Median (range) No. study n/N, % median (range) A. Subjective failure: persistent prolapse symptoms RCT 1 16/41 (39%) 39% 1 5/41 (12%) 12% RCT (abs.) 0 0 Non-rand 0 0 Case series (n<100) 2 1/50 (2.0%) 1.6% (0-3.3) 0 B. Objective failure: recurrent prolapse at original site RCT 1 2/38 (5.3%) 5.3% 1 2/40 (5.0%) 5.0% RCT (abs.) 0 0 Non-rand 1 0/36 (0%) 0% 0 Case series (n<100) 3 2/62 (3.2%) 3.3% (0-7.7) 0 C. De novo prolapse at other sites (no studies reported this outcome) D. Further operation needed for prolapse (recurrent or de novo) RCT 1 9/41 (22%) 22% 1 1/41 (2.4%) 2.4% RCT (abs.) 0 0 Non-rand 1 0/36 (0%) 0% 0 Case series (n<100) 1 1/30 (3.3%) 3.3% 0 E. Persistent urinary symptoms RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series (n<100) 1 3/19 (15.8%) 15.8% 0 F. Persistent bowel symptoms RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series (n<100) 1 1/5 (20%) 20% 0 G. Persistent sexual symptoms RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series (n<100) 1 3/27 (11.1%) 11.1% 0

NR: not reported

Table 16 Uterine suspension sling, safety: summary of median rates (range) Uterine Uterine suspension sling No mesh (hysterectomy) No. study n/N, % Median (range) No. study n/N, % Median (range) A. Blood loss requiring blood transfusion RCT 1 1/41 (2.4%) 2.4% 1 2/41 (4.9%) 4.9% RCT (abs.) 0 0 Non-rand 1 2/36 (5.6%) 5.6% Case series 0 0 B. Damage to surrounding organs during operation RCT 1 0/41 (0%) 0% 1 1/41 (2.4%) 2.4% RCT (abs.) 0 0 Non-rand 0 0 Case series 2 0/50 (0%) 0% 0 C. Mesh erosion RCT 0 0 RCT (abs.) 0 0 Non-rand 1 0/36 (0%) 0 0 Case series 1 1/30 (3.3%) 3.3% 0 D. Further operation needed for mesh erosion RCT 0 0 RCT (abs.) 0 0 Non-rand 1 0/36 (0%) 0 Case series 1 1/30 (3.3%) 3.3% 0 E. De novo urinary symptoms in women who did not have these symptoms at baseline F. De novo bowel symptoms in women who did not have these symptoms at baseline RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series 1 0/25 (0%) 0% 0 G. De novo sexual symptoms in women who did not have these symptoms at baseline RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series 1 1/2 (50%) 50% 0 H. infection RCT 1 7/41 (17%) 17% 1 2/41 (4.9%) 4.9% RCT (abs.) 0 0 Non-rand 1 1/36 (2.8%) 2.8% 0 Case series 1 2/30 (6.7%) 6.7% 0 I. Other serious adverse effects not otherwise specified RCT 1 1/41 (2.4%) 2.4% 1 0/41 (0%) 0% RCT (abs.) 0 Non-rand 1 2/36 (5.6%) 5.6% Case series 1 2/30 (6.7%) 6.7% 0 J. Other adverse effects not otherwise specified RCT 1 9/41 (22%) 22% 1 9/41 (22%) 22% RCT (abs.) 0 0 Non-rand 1 6/36 (16.7%) 16.7% 0 Case series 2 5/49 (10.2%) 6.9% (3.3-10.5) 0 Note: 1. Other serious adverse effects (not otherwise specified) that occurred included incisional hernia and intestinal occlusion; 2. Other adverse effects (not otherwise specified) that occurred included urinary tract symptoms, dysmenorrhoae during the 1st menstrual cycle, haematoma, haemorrhage of presacral vein (no transfusion needed), dullness in upper leg, and sciatic pain.

4.6.1 Uterine suspension sling for uterine prolapse (mesh)

4.6.1.1 Efficacy

Subjective failure: persistent prolapse symptoms (Table 15A)

One RCT67 and two case series24,52 reported subjective failure. In the RCT, sacrohysteropexy was compared with vaginal hysterectomy (no mesh). At one year, the proportion of women who had visited a clinician because of prolapse symptoms was 39% (16/41) in the sacrohysteropexy group. This was statistically significantly higher than that in the no mesh group (5/41, 12%; RR 3.20, 95% CI 1.29 to 7.92). One case series24 reported sacrohysteropexy. One woman (1/30, 3.3%) had a symptomatic prolapse recurrence at three years after the treatment. The other case series52 reported a different uterine suspension sling technique: suspension of the uterus to the pectineal ligaments. No woman had prolapse symptoms during a follow up of six to 30 months.

Objective failure: recurrent prolapse at original site (Table 15B)

One full-text RCT,67 one non-randomised comparative study33 and three case series23,24,54 reported objective failure. The mean follow-up time ranged from 12 to 51 months. In the RCT,67 there was no evidence of a statistically significant difference in objective failure between sacrohysteropexy and vaginal hysterectomy (2/38 vs. 2/40; RR 1.05, 95% CI 0.16 to 7.10). The non-randomised comparative study33 compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy. No women had objective failure, defined as prolapse less than 6cm above the hymen, at a mean follow up of 51 months in either group. All three case series reported sacrohysteropexy. The proportion of women with objective failure was 3.2% (2/62) with a median across studies of 3.3% (0 to 7.7%). No studies reported time to failure.

De novo prolapse (no studies reported this outcome)

Need for further surgery for prolapse (Table 15D)

One RCT,67 one non-randomised comparative study,33 and one case series24 reported the need for further surgery for prolapse. In the RCT,67 22% (9/41) of women needed a further operation for recurrent or de novo prolapse after sacrohysteropexy. This was statistically significantly higher than hysterectomy (1/41, 2%; RR 9.00, 95% CI 1.19 to 67.85). In the non-

randomised comparative study that compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy,33 none of the 75 women required a further operation for recurrent or de novo prolapse at a mean follow up of 51 months. In the case series,24 one woman (1/30, 3.3%) required further surgery three years after sacrohysteropexy.

4.6.1.2 Safety

Blood loss (Table 16A)

One RCT,67 one non-randomised comparative study,33 and one case series54 reported blood loss as a continuous variable. In the RCT,67 the average blood loss in 41 women treated with sacrohysteropexy was 244ml (SD 52ml). This was similar to the blood loss in the hysterectomy group (no mesh, n=41, mean 248ml, SD 34ml). The non-randomised comparative study compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy.33 Women who had a sacrohysteropexy had a lower blood loss (n=36, mean 200ml, SD not reported) than those who had a sacrocolpopexy (n=39, mean 325ml, SD not reported). In the case series, the average blood loss in 13 women was 271ml (range 50 to 800ml) during sacrohysteropexy.

One RCT,67 one non-randomised comparative study,33 and two case series24,52 reported on the need for blood transfusion. In the RCT,67 there was no evidence of a statistically significant difference between sacrohysteropexy and vaginal hysterectomy (no mesh) in the number of women whose blood loss required transfusion (1/41, 2.4% vs. 2/41, 4.9%; RR 0.50, 95% CI 0.05 to 5.30), nor in the non-randomised comparative study,33 in which sacrohysteropexy was compared with hysterectomy followed by sacrocolpopexy (2/36, 5.6% vs. 2/39, 5.1%; RR, 1.08, 95% CI 0.16 to 7.29). No women required blood transfusion in the two case series (0/50).

Damage to surrounding organs during operation (Table 16B)

One RCT67 reported damage to surrounding organs. No organ damage occurred in the sacrohysteropexy group (0/40), and one bowel lesion occurred in the hysterectomy group (no mesh, 1/41).

Mesh erosion (Table 16C)

One non-randomised comparative study33 and one case series24 reported mesh erosion. The non-randomised comparative study33 compared sacrohysteropexy with hysterectomy followed by sacrocolpopexy. Non-absorbable synthetic mesh (Amid type I) was used. At a mean follow up of 51 months, no erosions (0/36) occurred in the sacrohysteropexy group and three erosions (3/39, 7.7%) occurred in the sacrocolpopexy group. In the case series, one woman (1/30, 3.3%) had an erosion after sacrohysteropexy.

Further operation needed for mesh erosion (Table 16D)

One non-randomised comparative study33 and one case series24 reported on the number of women who needed a further operation for mesh erosion. All erosions (1/66, 1.5%, after sacrohysteropexy; 3/39, 7.7% after hysterectomy followed by sacrocolpopexy) required further surgery (partial or complete mesh removal).

De novo urinary symptoms (no studies reported this outcome)

Infection (Table 16H)

One RCT,67 one non-randomised comparative study,33 and one case series24 reported infection. In the RCT,67 infections were reported as wound infection, vault abscess during admission, infected implant requiring surgery, and fever of unknown origin. In total, 17% (7/41) of women had an infection after sacrohysteropexy compared with 4.9% (2/41) in the vaginal hysterectomy group (no mesh). The outcome was reported as wound infection and fever in the non-randomised comparative study.33 Three cases of infection (3/39, 7.7%) occurred in the hysterectomy followed by sacrocolpopexy group, and one case (1/36, 2.8%) occurred in the sacrohysteropexy group. In the case series, one case of urinary tract infection (1/30, 3.3%) and one case of wound infection (1/30, 3.3%) occurred after sacrohysteropexy.

Other serious adverse effects not otherwise specified (Table 16I)

One RCT,67 one non-randomised comparative study,33 and one case series24 reported other serious adverse effects not otherwise specified. In the RCT, one incisional hernia occurred in the sacrohysteropexy group (1/41, 2.4%). In the non-randomised comparative study,33 5.6% (2/36) of women had an incisional hernia after sacrohysteropexy, compared with 2.6% (1/39)

in women who had a hysterectomy followed by sacrocolpopexy (RR 2.17, 95% CI 0.21 to 22.89). One (1/30, 3.3%) case of incisional hernia and one (1/30, 3.3%) case of intestinal occlusion by the mesh occurred after sacrohysteropexy in the case series.

Other adverse effects not otherwise specified (Table 16J)

One full-text RCT,67 one non-randomised comparative study,33 and two case series23,24 reported other adverse effects not otherwise specified. In the full-text RCT,67 outcomes were reported as lower urinary tract symptoms during admission, upper leg dullness during admission, and vaginal stricture requiring excision. In the sacrohysteropexy group, 22% (9/41) of women had one of these problems compared with 22% (9/41) in the vaginal hysterectomy group (no mesh). In the non-randomised comparative study,33 17% (6/36) of women had perivesical haematoma or voiding dysfunction after sacrohysteropexy, compared with 13% (5/39) in women who had hysterectomy followed by sacrocolpopexy. In the two case series, two women (2/19, 10.5%) had dysmenorrhea during the first menstrual cycle after sacrohysteropexy; one women each (1/30, 3.3%) had retroperitoneal haematoma, haemorrhage of presacral vein without need for transfusion, and sciatic pain after sacrohysteropexy.

4.6.1.3 Operation time and hospital stay

One full-text RCT67 and one non-randomised comparative study33 reported operation time. In the full-text RCT,67 mean operation times for sacrohysteropexy and vaginal hysterectomy were 97 minutes (n=41) and 107 minutes (n=41) respectively. In the non-randomised comparative study,33 operation time for hysterectomy followed by sacrocolpopexy was about 25 minutes longer than for sacrohysteropexy (mean 115 minutes, n=39 vs. 89 minutes, n=36).

One full-text RCT67 reported hospital stay. Hospital stay was similar between sacrohysteropexy and vaginal hysterectomy (mean 7.7 days, n=41 vs. 7.6 days, n=41).

4.6.2 Summary of the evidence for and against uterine suspension sling

Six studies (one full-text RCT, one non-randomised comparative study, and four case series) involving 239 women reported uterine suspension sling. Figure 20 and Figure 21 show the range of event rates across studies in Table 15 and Table 16, together with the number of women involved. Meta-analysis was not possible because the studies used different comparators.

Figure 20 Ranges of event rates across studies: uterine suspension sling, efficacy

50 17/91 40 10/107 1/5 30 3/19 3/27 20 4/136

Event rates(%) 10

0 Subjective failure Objective failure New prolapse at Reoperation for Persistent urinary Persistent bowel Persistent sexual other sites (no prolapse symptoms symptoms problems data)

Uterine prolapse repair

Figure 21 Ranges of event rates across studies: uterine suspension sling, safety

1/2

40 20/126 30 10/101 20 3/77 0/91 1/66 1/66 0/25 5/101

Event rates (%) rates Event 10

Notes: 1. Where only one study reported the outcome, a single horizontal bar reports its event rate; where more than one study reported the outcome, the range of the event rates across all studies is shown by a vertical bar; 2. The numbers in the figures are the number of events which occurred across all studies divided by the total number of women who received the treatment;

4.7 Other mesh techniques or a mixture of mesh techniques

Four studies37,38,72,77 involving 520 women reported sacrocolpoperineopexy38,72,77 or a mixture of techniques that use mesh but did not report them separately (Table 17).37 No studies reported uterine repair and vault repair separately or reported whether the repair was for uterine prolapse or for vault prolapse. Two studies were non-randomised comparative studies: one37 compared a mixture of sacrohysteropexy and sacrocolpopexy with sacrospinous colpopexy (no mesh); the other77 compared sacrocolpopexy with sacrocolpoperineopexy (both with mesh). Two were case series with one38 reporting a mixture of sacrocolpopexy, sacrohysteropexy, and sacrocervicopexy, and the other72 reporting sacrocolpoperineopexy. The median follow up across the four studies was 10 months (range 1.5 to 19 months).

Individual studies reporting efficacy outcomes are given in Appendix 19, safety outcomes in Appendix 20, and operation time and hospital stay in Appendix 21. The median rates and ranges for efficacy outcomes are summarised in Table 18, and safety outcomes in Table 19. No RCTs or non-randomised comparative studies reported data on disease-specific quality of life.

Methodological quality of the included studies

Both non-randomised comparative studies37,77 collected data retrospectively and did not report whether patients were recruited consecutively. Neither study reported loss to follow up.

The two case series38,72 also collected data retrospectively but both39,50 recruited patients consecutively. The loss to follow up rates were 30% (39/128) at a mean follow up of 19 months and 74% (125/169) of a mean follow up of 14 months.

Table 17 Other mesh techniques or a mixture of mesh techniques (not reported separately): summary of patient characteristics and surgical procedures ID N Age, y Uterine/ Primary/ Intervention route Technique Mesh/graft Concomitant procedure Follow up (mean) Outcomes vault secondary Non-randomised comparative studies Demirci 200737 A, 45 A, 44 (14) A, 38/7 NR A, abdominal A, 20 sacrohysteropexy, 25 A, polypropylene, Prolene Anti-incontinence: 6w postoperatively Safety B, 60 B, 58 (16) B, 54/6 B, vaginal sacrocolpopexy (Amid type I) A, 34/45; B, 45/60 B, sacrospinous fixation B, no mesh Hysterectomy: A, 18/45; B, 54/60

Visco 200177 A, 155 A+B+C+D, 61 NR NR A, abdominal A, sacrocolpopexy A, polyethylene tetraphalate, Anti-incontinence: A, 7m (1-87) Safety B, 88 (31-84) B, abdominal B, sacrocolpoperineopexy Mersilene (Amid type III) or NR B, 5m (1-45) C, 25 C, vaginal+ abdominal C, sacrocolpoperineopexy Gore-Tex (Amid type II) Hysterectomy: C, 6m (1-28) D, 5 D, vaginal+ abdominal D, sacrocolpoperineopexy B, C, D, same as above A+B+C+D: 45/273 D, 7m (2-11) with mesh placed to vaginal field

88 Case series

Elneil 200538 128 62 (38-91) NR NR Abdominal (open or Sacrocolpopexy, Polypropylene, Prolene Anti-incontinence: 18/128 19m (2-65) Efficacy laparoscopic) sacrohysteropexy, or (Amid type I), porcine small Hysterectomy: 6/128 Safety sacrocervicopexy intestine submucosa (Surgisis) was used in some patients below the polypropylene mesh

Su 200772 169 62 (10) 5/164 NR Abdominal + vaginal Sacrocolpoperineopexy Porcine dermis, Pelvicol; Anti-incontinence: 155/169 14m (1.5-24) Efficacy cadaveric fascia, Tutoplast; Hysterectomy: 11/169 Safety polypropylene, Gynemesh, Prolite, or Prolene (all Amid type I)

NR: not reported

Table 18 Other mesh techniques or a mixture of mesh techniques (not reported separately), efficacy: summary of median (range) event rates Uterine and/or vault Other mesh techniques or a mixture of No mesh (sacrospinous suspension) mesh techniques No. study n/N, % Median (range) No. study n/N, % median (range) A. Subjective failure: persistent prolapse symptoms RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series (n<100) 1 2/169 (1.2%) 1.2% 0 B. Objective failure: recurrent prolapse at original site RCT 0 0 RCT (abs.) 0 0 Non-rand 0 Case series (n<100) 1 1/169 (0.6%) 0.6% 0 C. De novo prolapse at other sites (no studies reported this outcome) D. Further operation needed for prolapse (recurrent or de novo) RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series (n<100) 2 6/297 (2.0%) 2.3% (0-4.7) 0 E. Persistent urinary symptoms (no studies reported this outcome) F. Persistent bowel symptoms (no studies reported this outcome) G. Persistent sexual symptoms (no studies reported this outcome)

Table 19 Other mesh techniques or a mixture of mesh techniques, safety: summary of median rates (range) Uterine and/or vault Other mesh techniques or a mixture of mesh techniques No mesh (sacrospinous suspension) No. study n/N, % Median (range) No. study n/N, % Median (range) A. Blood loss requiring blood transfusion RCT 0 0 RCT (abs.) 0 0 Non-rand 1 2/45 (4.4%) 4.4% 1 0/60 (0%) 0% Case series 0 0 B. Damage to surrounding organs during operation RCT 0 0 RCT (abs.) 0 0 Non-rand 1 1/45 (2.2%) 2.2% 1 1/60 (1.7%) 1.7% Case series 2 10/297 (3.4%) 3.1% (1.6-4.7) 0 C. Mesh erosion RCT 0 0 RCT (abs.) 0 0 Non-rand 1 10/118 (8.5%) 8.5% 0 Case series 2 17/297 (5.7%) 5.3% (2.3-8.3) 0 D. Further operation needed for mesh erosion RCT 0 0 RCT (abs.) 0 0 Non-rand 1 10/118 (8.5%) 8.5% 0 Case series 2 4/297 (1.3%) 1.5% (0.6-2.3) 0 E. De novo urinary symptoms in women who did not have these symptoms at baseline RCT 0 0 RCT (abs.) 0 0 Non-rand 0 0 Case series 1 4/9 (44%) 44% 0 F. De novo bowel symptoms (no studies reported this outcome) G. De novo sexual symptoms (no studies reported this outcome) H. infection RCT 0 0 RCT (abs.) 0 0 Non-rand 1 16/45 (36%) 36% 1 10/60 (16.7%) 16.7% Case series 1 2/128 (1.6%) 1.6% 0 I. Other serious adverse effects not otherwise specified RCT 0 0 RCT (abs.) 0 0 Non-rand 1 3/45 (6.7%) 6.7 1 0/60 (0%) 0% Case series 2 6/297 (2.0%) 1.9% (0.8-3.0) 0 J. Other adverse effects not otherwise specified RCT 0 0 RCT (abs.) 0 0 Non-rand 1 7/45 (16%) 16% 1 5/60 (8.3%) 8.3% Case series 2 5/297 (1.7%) 1.6% (0.8-2.4) 0 Note: 1. Other serious adverse effects (not otherwise specified) that occurred included: wound dehiscence, pulmonary oedema, and ileus. 2. Other adverse effects (not otherwise specified) that occurred included: urinary retention, haematoma, and wound cellulitis.

4.7.1 Uterine and/or vault repair (not reported separately) (mesh)

4.7.1.1 Efficacy

No studies reported de novo prolapse, persistent urinary, bowel or sexual symptoms in women who had these symptoms at baseline.

Subjective failure: persistent prolapse symptoms (Table 18A)

One case series72 reported subjective failure after sacrocolpoperineopexy: 1.2% (2/169) of women had prolapse symptoms at a mean follow up of 14 months.

Objective failure: recurrent prolapse at original site (Table 18B)

One case series72 reported objective failure: apical prolapse stage II or more occurred in one woman (1/169) at a mean follow up of 14 months after sacrocolpoperineopexy.

Further operation needed for prolapse (Table 18D)

Two case series38,72 reported the number of women who needed a further operation for prolapse. One case series38 reported sacrocolpopexy, sacrohysteropexy, and sacrocervicopexy, but did not report data separately. In this study, 4.7% (6/128) of women required a further operation for prolapse at a mean follow up of 19 months. The other case series72 reported sacrocolpoperineopexy. No women (0/169) required a further operation for prolapse at a mean follow up of 14 months.

4.7.1.2 Safety

Blood loss requiring transfusion (Table 19A)

One non-randomised comparative study37 reported blood transfusion. A variety of techniques that used mesh (sacrohysteropexy or sacrocolpopexy) were compared with sacrospinous colpopexy (no mesh). Two (2/45) women who underwent techniques that used mesh required transfusion compared with none (0/60) in the no-mesh group.

Damage to surrounding organs during operation (Table 19B)

Outcomes considered in this section included bladder injury, rectal injury, and bowel perforation. One non-randomised comparative study37 and two case series38,72 reported damaging to surrounding organs during operation. The non-randomised comparative study37 compared a mixture of techniques that used mesh (sacrohysteropexy and sacrocolpopexy) with sacrospinous colpopexy (no mesh). One (1/45) bladder injury occurred in the mesh group and one (1/60) rectal injury occurred in the no-mesh group. The two case series reported sacrocolpoperineopexy72 or a mixture of techniques that used mesh.38 Across these studies, median of 3.1% (range 1.6% to 4.7%) women suffered organ damage.

Mesh erosion (Table 19C)

One non-randomised comparative study77 and two case series38,72 reported mesh erosion. The non-randomised comparative study77 compared sacrocolpopexy with sacrocolpoperineopexy (both with mesh) using a variety of non-absorbable synthetic meshes. Fifteen erosions (5/155 vs. 10/118) occurred in total. One case series72 reported sacrocolpoperineopexy. A variety of biological grafts (porcine dermis or cadaveric fascia lata) and non-absorbable meshes were used. The mesh erosion rate was 8.3% (14/169). Another case series38 reported a mixture of techniques which used mesh, i.e. sacrocolpopexy, sacrohysteropexy and sacrocervicopexy. In this study, 2.3% (3/128) of women had mesh erosions. The time that erosion was detected ranged from six weeks to 40 months.38,72,77

Further operation needed for mesh erosion (Table 19D)

One non-randomised comparative study77 and two case series38,72 reported the number of women who needed a further operation for mesh erosion. Of 27 erosions (27/297, 5.7%), 14 required excision.

De novo urinary symptoms (Table 19E)

One case series72 involving nine women who had no urinary symptoms at baseline reported de novo urinary symptoms. In this study, 44% (4/9) of women had incontinence at one year after sacrocolpoperineopexy.

De novo bowel symptoms (no studies reported this outcome)

De novo sexual symptoms (no studies reported this outcome)

Infection (Table 19H)

Outcomes reported as urinary tract infection, wound infection, vaginal vault infection, and febrile morbidity were considered in this section. One non-randomised comparative study37 and one case series38 reported these outcomes. The infection rates were reported as 1.6% (2/128)38 and 36% (16/45).37

Other serious adverse effects not otherwise specified (Table 19I)

Outcomes considered in this section included ileus, wound dehiscence, and pulmonary oedema. One non-randomised comparative study37 and two case series38,72 reported these outcomes. The occurrence of other serious effects not otherwise specified ranged from 0.8% (1/128) to 6.7% (3/45).

Other adverse effects not otherwise specified (Table 19J)

Outcomes considered in this section included urinary retention for longer than five days, wound haematoma, extra-peritoneal haematoma, and wound cellulitis. One non-randomised comparative study37 and two case series38,72 reported these outcomes. The number of women experiencing other adverse effects not otherwise specified ranged from 0.8% (1/128) to 16% (7/45).

4.7.1.3 Operation time and hospital stay

One non-randomised comparative study37 reported on operation time. Average operation time for sacrohysteropexy and sacrocolpopexy (not reported separately) was 192 minutes (n=45), compared with 141 minutes (n=60) for sacrospinous colpopexy (no mesh).

One non-randomised comparative study37 and one case series72 reported hospital stay. Average hospital stay was 7 days after sacrohysteropexy and sacrocolpopexy (not reported separately),37 6 days after sacrospinous colpopexy (no mesh),37 and 3.3 days after sacrocolpoperineopexy.72

4.7.2 Summary of the evidence for and against other mesh techniques or a mixture of techniques that use mesh

separately. Figure 22 and Figure 23 show the range of event rates across studies in Table 18 and Table 19, together with the number of women involved, respectively. The proportion of women who needed a re-operation for prolapse ranged from 0% to 4.7% (2 studies, n=297). Organ damage, mesh erosion, and operation for mesh erosion all occurred in less than 10% of women. In one study reporting this information, the average operation time was 192 minutes, and across two studies the average hospital stay ranged from 3 to 7 days.

Figure 22 Ranges of event rates across studies: other mesh techniques or a mixture of mesh techniques efficacy

20 2/169 1/169 6/297

Event rates(%) 10

Uterine and/or vault prolapse repair (not separated)

Figure 23 Ranges of event rates across studies: other mesh techniques or a mixture of mesh techniques, safety

4/9 50 18/173 40

30 12/342 20 2/45 11/342 27/415 14/415 9/342

Event rates (%) rates Event 10

Notes: 1. Where only one study reported the outcome, a single horizontal bar reports its event rate; where more than one study reported the outcome, the range of the event rates across all studies is shown by a vertical bar; 2. The numbers in the figures are the number of events which occurred across all studies divided by the total number of women who received the treatment.

5 DISCUSSION

This systematic review was prompted by concerns regarding the efficacy and safety of techniques that used mesh in surgery for uterine or vault prolapse repair. The techniques using mesh include sacrocolpopexy,89 infracoccygeal sacropexy,90 and uterine suspension sling.91

5.1 Summary of main findings

In this review, we identified 54 studies (5 full-text RCTs, 3 RCTs available as conference abstracts, 17 non-randomised comparative studies, and 29 case series) involving 7054 women treated for uterine or vault prolapse in total.

5.1.1 Sacrocolpopexy

Thirty-two studies (4 full-text RCTs, 1 RCT available as conference abstract, 15 non- randomised comparative studies, and 12 case series with a sample size over 100) involving a total of 4456 women provided data on sacrocolpopexy. The average follow up across studies was around two years. For efficacy, the objective failure rates ranged from 0% to 6.1% in 11 studies that involved 1054 women, while the subjective failure rates varied from 3.3% to 31% in 9 studies that involved 638 women. Meta-analysis was carried out to compare subjective failure and objective failure between sacrocolpopexy and sacrospinous colpopexy (no mesh). There was no evidence of a statistically significant difference in either outcome between the operations. For safety, mesh erosion occurred in 0% to 12% of women (27 studies, n=2922), of whom 0% to 11% required an operation for mesh erosion (17 studies, n=2074). The average operation time ranged from 89 to 267 minutes (15 studies), whereas the average hospital stay ranged from 1 to 7 days. The majority of the studies (11/14 studies) reported hospital stays of 3 to 5 days.

5.1.2 Infracoccygeal sacropexy

Fourteen studies (2 RCTs available as conference abstracts, 1 non-randomised comparative study, 2 case series with sample size ≥100, and 9 case series with sample size <100) involving 976 women provided data on infracoccygeal sacropexy. The average follow up across studies was around one year. For efficacy, the proportion of women with subjective failure ranged from 2.3% to 21% (n=262, 3 studies), whereas the range for objective failure was from 0% to 25% (9 studies, n=402). For safety, the occurrence of mesh erosion ranged

from 0% to 21% (11 studies, n=889), and 1.5% to 17% of women needed an operation for mesh erosion (6 studies, n=678). Meta-analysis was not possible because no studies used the same comparator. Average operation times ranged from 10 minutes to 149 minutes with the majority of the studies (8/11) reporting less than one hour. Average hospital stays ranged from less than 24 hours to 5 days with the majority of the studies (7/11) reporting two days or less.

5.1.3 Uterine suspension sling

Six studies (one full-text RCT, one non-randomised comparative study, and four case series) involving 239 women reported uterine suspension sling. The average follow up across studies was around three years. For efficacy, the range for women reporting subjective failure rates ranged from 0% to 39% (3 studies, n=91). Women judged to have objective failure ranged from 0% to 7.7% (5 studies, n=136). Meta-analysis was not possible because no studies used the same comparator. Little evidence was available for estimating safety outcomes. The average operation time for uterine suspension sling ranged from 97 to 115 minutes (based on 3 studies). In one study reporting this information, the average hospital stay was 7 days.

5.1.4 Other mesh techniques or a mixture of mesh techniques

Four studies (two non-randomised comparative studies and two case series) involving 520 women reported sacrocolpoperineopexy, or a mixture of techniques that used mesh but did not report them separately. None of the studies reported uterine and vault prolapse repair separately. For efficacy, there was little evidence to assess subjective failure and objective failure. The proportion of women who needed a re-operation for prolapse ranged from 0% to 4.7% (2 studies, n=297). Organ damage, mesh erosion, and operation for mesh erosion all occurred in less than 10% of women. In one study reporting this information, the average operation time was 192 minutes, and across two studies, the average hospital stay ranged from 3 to 7 days.

5.2 Assumptions, limitations, and uncertainties

For conference abstracts, only RCTs published from 2005 onwards were included. As the volume of literature for sacrocolpopexy was substantial, case series with a sample size of less than 100 were excluded. We placed no such restriction on sample size for studies reporting uterine suspension sling or infracoccygeal sacropexy due to the more limited evidence.

We did not search for studies that reported procedures without mesh because these procedures were not the main focus of this review. Initial searches of MEDLINE and EMBASE from 1996 onwards identified 2008 titles/abstracts of such studies. The data on procedures without mesh in this report were taken from the control groups of RCTs and non- randomised comparative studies only. Therefore, the results for ‘no mesh’ were not derived from a comprehensive literature search and hence should be interpreted with caution.

In terms of the methodological quality of the included studies, in three of the five full-text RCTs allocation was adequately concealed. Four of these studies reported loss to follow up (range 5% to 15%). Only three of the 17 non-randomised comparative studies collected data prospectively. Five of these studies reported loss to follow up rates (range 4% to 55%). Ten of the 29 case series collected data prospectively. Twelve case series reported rates of loss to follow up (range 2% to 74%). Data were extracted as reported.

Three studies (RCTs published in 2005 or later) were available only as conference abstracts. Their methodological quality was not assessed because of the limited data reported. We have presented the evidence from these studies separately from other types of study designs, and it should be interpreted with caution.

For each technique using mesh, uterine prolapse and vault prolapse were reviewed separately because we considered that efficacy and safety may be different. However, the majority of the studies (32/54) included in this review reported a mixture of uterine and vault prolapse repair and did not report their data separately. As there was limited evidence for uterine prolapse repair and for vault prolapse repair alone, evidence from studies that reported a mixture of uterine and vault prolapse may be regarded as complementary information relating to the safety and efficacy of mesh.

The mesh techniques were classified into sacrocolpopexy, infracoccygeal sacropexy, uterine suspension sling, and other mesh techniques (such as sacrocolpoperineopexy) or a mixture of the above mesh techniques. The no-mesh techniques reported included hysterectomy and uterine/vault suspension without sling. However, there were variations within each technique, including whether abdominal (open or laparoscopic, for sacrocolpopexy) or vaginal intervention route (for uterine suspension sling), type of mesh material (biological, non- absorbable synthetic (Amid type I – IV), combination of absorbable and non-absorbable material), and preservation or not of the uterus in uterine prolapse repair. It was not possible to compare differences amongst these variations as these were not reported adequately or in sufficient numbers.

Categorising some of the reported outcomes was problematic. For instance, cut-off points used to determine objective failure rates, in terms of the degree of prolapse when failure was deemed to have occurred, varied between studies. All types of infections such as urinary tract infection, wound infection and pelvic abscess were grouped together.

Results from full-text RCTs, RCTs available as conference abstracts, and non-randomised comparative studies were presented separately in the meta-analysis. We used median and range of event rates by study design. This allowed the results to reflect the variability of the data. We did not attempt to combine the evidence from different study designs. This is because non-randomised comparative studies and case series are more prone to systematic biases than RCTs. These include selection bias in the types of participants, not adjusting for confounding factors (e.g. severity of prolapse, age, co-morbidities), and withdrawal and loss to follow up rates, which may affect the reliability and size of the treatment effects.

For this review, we considered one year as an adequate minimum period of time to assess the efficacy of prolapse repair. However, one year may be too soon to judge whether prolapse surgery is successful in the long-term, and therefore failure at one year may represent a serious underestimation of efficacy. Prospective studies would require extended follow up to assess meaningful mesh failure rates.

The estimation of some adverse effect rates may be prone to random error, particularly those that occur infrequently. In addition, for some adverse effects, there were risks of overestimation because some studies reported more than one symptom for an outcome at the individual level. For example, some studies reported urinary tract infection and wound infection under the outcome of infection. We calculated event rates by simply adding the number of events regardless of the fact that an individual may have two different types of infection.

It is important to determine whether safety and efficacy of mesh differ between primary and recurrent prolapse repair. However, of the 54 included studies, only two case series26,79 reported exclusively on women having primary repairs. Another four studies24,25,52,55 reported the case mix of primary and secondary operations but did not report outcome data separately for the two groups.

The results were considered generalisable as the majority of studies recruited participants from routine practice without restriction on the severity of prolapse or other patient characteristics.

6 CONCLUSIONS

6.1 Implications for the NHS

The review indicated that although sacrocolpopexy is associated with an objective failure ranging from 0% to 6% at an average follow up of two years, the incidence of subjective failure (range 3% to 31%) and need for a further prolapse operation (range 2% to 14%) may be higher. There were risks from adverse effects such as blood transfusion (range 0% to 17%), infection (range 1% to 69%), mesh erosion (range 4% to 12%) and the need for a further operation for mesh erosion (range 1% to 11%). There was no evidence of a statistically significant difference between sacrocolpopexy and sacrospinous colpopexy (no mesh) in subjective failure or objective failure.

For infracoccygeal sacropexy, subjective failure rates ranged from 2.3% to 21% at an average follow up of one year, objective failure 0% to 25%, mesh erosion 0% to 21%, and operation for mesh erosion 1.5% to 17%. Operation times (the majority were less than one hour) and hospital stays (the majority were two days or less) were relatively short compared with other operations that use mesh.

Sacrocolpoperineopexy was another mesh technique that had been used, but there was little evidence for the efficacy or safety of this technique.

It was not possible to compare differences in efficacy and safety for women having primary versus recurrent prolapse repairs. Of the 54 included studies, only two exclusively reported on women having primary repairs, and the remainder did not report outcomes for these subgroups separately.

6.2 Implications for future research

(seven RCTs, one registry, and one case series) range from 2008 to 2014. A further review incorporating these studies may be informative.

The RCTs included in the present review were relatively small. Therefore, high quality long- term RCTs, with adequate power to detect clinically meaningful differences, with validated patient-reported outcome measures, primarily comparing the failure rate between mesh and no-mesh techniques, and between different types of mesh techniques and between different types of mesh, taking account the outcomes important to women, are required to determine the comparative efficacy of mesh techniques and their optimal place in clinical practice.

100

7 REFERENCE LIST

1 Kohli N, Goldstein,DP. An overview of the clinical manifestations, diagnosis, and classification of pelvic organ prolapse. [webpage on the Internet] Waltham, MA: UpToDate, 2007. [accessed March 2008] Available from: URL: http://patients.uptodate.com/topic.asp?file=pelvic_s/2164.

2 Jia X, Glazener,C, Mowatt,G, Maclennan,G, Fraser,C, Burr,J. Systematic review of the efficacy and safety of using mesh or grafts in surgery for anterior and/or posterior vaginal wall prolapse. Interventional Procedures IP 660. [document on the Internet] London: National Institute for Health and Clinical Excellence, 2008. [accessed March 2008] Available from: URL: http://www.nice.org.uk/nicemedia/pdf/660_systematic_review_250208.pdf.

3 Hospital Episode Statistics: Main operations 2005-2006. [database on the Internet] Hospital Episode Statistics, 2007. [accessed March 2008] Available from: URL: http://www.hesonline.org.uk/Ease/servlet/ContentServer?siteID=1937&categoryID=19 2.

4 Boyles SH, Weber AM, Meyn L. Procedures for pelvic organ prolapse in the United States, 1979-199. Am J Obstet Gynecol 2003;188:108-15.

5 Mayne CJ, Assassa RP. Epidemiology of incontinence and prolapse. BJOG 2004;111:2-4.

6 DeLancey JO, Kearney R, Chou Q, Speights S, Binno S. The appearance of levator ani muscle abnormalities in magnetic resonance images after vaginal delivery. Obstet Gynecol 2003;101:46-53.

7 Snooks SJ, Swash M, Mathers SE, Henry MM. Effect of vaginal delivery on the pelvic floor: a 5 year follow-up. Br J Surg 1990;77:1358-60.

8 Tulikangas PK. Defect theory of pelvic organ prolapse. Clin Obstet Gynecol 2005;48:662-7.

9 Baden WF, Walker TA. Genesis of the vaginal profile: a correlated classification of vaginal relaxation. Clin Obstet Gynecol 1972;15:1048-54.

10 Beecham CT. Classification of vaginal relaxation. Am J Obstet Gynecol 1980;136:957-8.

11 Bump RC, Mattiasson A, Bo K, Brubaker LP, DeLancey JO, Klarskov P et al. The standardization of terminology of female pelvic organ prolapse and pelvic floor dysfunction. Am J Obstet Gynecol 1996;175:10-7.

12 Hall AF, Theofrastous JP, Cundiff GW, Harris RL, Hamilton LF, Swift SE et al. Interobserver and intraobserver reliability of the proposed International Continence Society, Society of Gynecologic Surgeons, and American Urogynecologic Society pelvic organ prolapse classification system. Am J Obstet Gynecol 1996;175:1467-70.

13 Hagen S, Stark D, Maher C, Adams EJ. Conservative management of pelvic organ prolapse in women. Cochrane Database of Systematic Reviews 2006;Issue 4. Art. No.: CD003882. DOI: 10.1002/14651858.CD003882.pub3.

101

14 Amid PK. Classification of biomaterials and their related complication in abdominal wall hernia surgery. Hernia 1997;1:15-21.

15 Baessler K, Maher CF. Mesh augmentation during pelvic-floor reconstructive surgery: Risks and benefits. Curr Opin Obstet Gynecol 2006;18:560-6.

16 Deprest J, Zheng F, Konstantinovic M, Spelzini F, Claerhout F, Steensma A et al. The biology behind fascial defects and the use of implants in pelvic organ prolapse repair. Int Urogynecol J 2006;17:S16-S25.

17 Verhagen AP, de Vet HC, de Bie RA, Kessels AG, Boers M, Bouter LM et al. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998;51:1235-41.

18 Undertaking systematic reviews of research on effectiveness. CRD's guidance for those carrying out or commissioning reviews. CRD Report No 4. University of York: Centre for Reviews and Dissemination; 2001.

19 Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health 1998;52:377-84.

20 Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- analyses. Br Med J 2003;327:557-60.

21 Altman D, Anzen B, Brismar S, Lopez A, Zetterstrom J. Long-term outcome of abdominal sacrocolpopexy using xenograft compared with synthetic mesh. Urology 2006;67:719-24.

22 Bai SW, Kwon HS, Chung DJ. Abdominal high uterosacral colpopexy and abdominal sacral colpopexy with mesh for pelvic organ prolapse. Int J Gynecol Obstet 2006;92:147-8.

23 Banu LF. Synthetic sling for genital prolapse in young women. Int J Gynecol Obstet 1997;57:57-64.

24 Barranger E, Fritel X, Pigne A. Abdominal sacrohysteropexy in young women with uterovaginal prolapse: long-term follow-up. Am J Obstet Gynecol 2003;18:1245-50.

25 Begley JS, Kupferman SP, Kuznetsov DD, Kobashi KC, Govier FE, McGonigle KF et al. Incidence and management of abdominal sacrocolpopexy mesh erosions. Am J Obstet Gynecol 2005;192:1956-62.

26 Bensinger G, Lind L, Lesser M, Guess M, Winkler HA. Abdominal sacral suspensions: analysis of complications using permanent mesh. Am J Obstet Gynecol 2005;193:2094-8.

27 Benson JT, Lucente V, McClellan E. Vaginal versus abdominal reconstructive surgery for the treatment of pelvic support defects: a prospective randomized study with long- term outcome evaluation. Am J Obstet Gynecol 1996;175:1418-21.

28 Bradley CS. Bowel symptoms in women 1 year after sacrocolpopexy. Am J Obstet Gynecol 2007;197:642 e1-642 e8.

102

29 Braun H, Fernandez M, Delloro A, Gonzalez F, Cuevas R, Rojas I. Prospective randomised study to compare colposacropexy and Mayo McCall technique in the correction of severe genital central prolapse. Int Urogynecol J 2007;18:S1-S24.

30 Brizzolara S, Pillai-Allen A. Risk of mesh erosion with sacral colpopexy and concurrent hysterectomy. Obstet Gynecol 2003;102:306-10.

31 Brubaker L, Cundiff GW, Fine P, Nygaard I, Richter HE, Visco AG et al. Abdominal sacrocolpopexy with Burch colposuspension to reduce urinary stress incontinence. N Engl J Med 2006;354:1557-66.

32 Burgio KL. Bladder symptoms 1 year after abdominal sacrocolpopexy with and without Burch colposuspension in women without preoperative stress incontinence symptoms. Am J Obstet Gynecol 2007;197:647 e1-647 e6.

33 Costantini E, Mearini L, Bini V, Zucchi A, Mearini E, Porena M. Uterus preservation in surgical correction of urogenital prolapse. Eur Urol 2005;48:642-9.

34 Culligan PJ, Murphy M, Blackwell L, Hammons G, Graham C, Heit MH. Long-term success of abdominal sacral colpopexy using synthetic mesh. Am J Obstet Gynecol 2002;187:1473-80.

35 Culligan PJ, Blackwell L, Goldsmith LJ, Graham CA, Rogers A, Heit MH. A randomized controlled trial comparing fascia lata and synthetic mesh for sacral colpopexy. Obstet Gynecol 2005;106:29-37.

36 de Vries MJ, van Dessel TH, Drogendijk AC, de H, I, Huikeshoven FJ. Short-term results and long-term patients' appraisal of abdominal colposacropexy for treatment of genital and vaginal vault prolapse. Eur J Obstet Gynecol Reprod Biol 1995;59:35- 8.

37 Demirci F. Perioperative complications in abdominal sacrocolpopexy and vaginal sacrospinous ligament fixation procedures. Int Urogynecol J Pelvic Floor Dysfunct 2007;18:257-61.

38 Elneil S, Cutner AS, Remy M, Leather AT, Toozs-Hobson P, Wise B. Abdominal sacrocolpopexy for vault prolapse without burial of mesh: a case series. BJOG 2005;112:486-9.

39 Farnsworth BN. Posterior intravaginal slingplasty (infracoccygeal sacropexy) for severe posthysterectomy vaginal vault prolapse--a preliminary report on efficacy and safety. Int Urogynecol J 2002;13:4-8.

40 Fedorkow DM, Kalbfleisch RE. Total abdominal hysterectomy at abdominal sacrovaginopexy - a comparative-study. Am J Obstet Gynecol 1993;169:641-3.

41 Foote AJ. Infracoccygeal sacropexy. Aust N Z J Obstet Gynaecol 2007;47:250-1.

42 Ghanbari Z, Baratali BH, Mireshghi MS. Posterior intravaginal slingplasty (infracoccygeal sacropexy) in the treatment of vaginal vault prolapse. Int J Gynaecol Obstet 2006;94:147-8.

43 Govier FE, Kobashi KC, Kozlowski PM, Kuznetsov DD, Begley SJ, McGonigle KF et al. High complication rate identified in sacrocolpopexy patients attributed to silicone mesh. Urology 2005;65:1099-103.

103

44 Gregory WT, Otto LN, Bergstrom JO, Clark AL. Surgical outcome of abdominal sacrocolpopexy with synthetic mesh versus abdominal sacrocolpopexy with cadaveric fascia lata. Int Urogynecol J 2005;16:369-74.

45 Griffis K, Evers MD, Terry CL, Hale DS. Mesh erosion and abdominal sacrocolpopexy: A comparison of prior, total, and supracervical hysterectomy. J Pelvic Med Surg 2006;12:25-30.

46 Handa VL. Sexual function before and after sacrocolpopexy for pelvic organ prolapse. Am J Obstet Gynecol 2007;197:629 e1-629 e6.

47 Hardiman PJ, Drutz HP. Sacrospinous vault suspension and abdominal colposacropexy: success rates and complications. Am J Obstet Gynecol 1996;175:612-6.

48 Hefni M. Morbidity associated with posterior intravaginal slingplasty for uterovaginal and vault prolapse. Arch Gynecol Obstet 2007;276:499-504.

49 Higgs P, Goh J, Krause H, Sloane K, Carey M. Abdominal sacral colpopexy: an independent prospective long-term follow-up study. Aust N Z J Obstet Gynaecol 2005;45:430-4.

50 Higgs PJ, Chua HL, Smith AR. Long term review of laparoscopic sacrocolpopexy. BJOG 2005;112:1134-8.

51 Hsiao KC. Comparison of laparoscopic and abdominal sacrocolpopexy for the treatment of vaginal vault prolapse. J Endourol 2007;21:926-30.

52 Joshi VM. A new technique of uterine suspension to pectineal ligaments in the management of uterovaginal prolapse. Obstet Gynecol 1993;81:790-3.

53 Jordaan DJ, Prollius A, Cronje HS, Nel M. Posterior intravaginal slingplasty for vaginal prolapse. Int Urogynecol J 2006;17:326-9.

54 Leron E, Stanton SL. Sacrohysteropexy with synthetic mesh for the management of uterovaginal prolapse. BJOG 2001;108:629-33.

55 Lindeque BG, Nel WS. Sacrocolpopexy--a report on 262 consecutive operations. S Afr Med J 2002;92:982-5.

56 Lo T-S. Abdominal colposacropexy and sacrospinous ligament suspension for severe uterovaginal prolapse: A comparison. J Gynecol Surg 1998;14:59-64.

57 Maher CF, Qatawneh AM, Dwyer PL, Carey MP, Cornish A, Schluter PJ. Abdominal sacral colpopexy or vaginal sacrospinous colpopexy for vaginal vault prolapse: a prospective randomized study. Am J Obstet Gynecol 2004;190:20-6.

58 Marcickiewicz J, Kjollesdal M, Engh ME, Eklind S, Axen C, Brannstrom M et al. Vaginal sacrospinous colpopexy and laparoscopic sacral colpopexy for vaginal vault prolapse. Acta Obstet Gynecol Scand 2007;86:733-8.

59 Meschia M, Barbacini P, Longatti D, Gattei U, Pifarotti P. Randomized comparison between infracoccygeal sacropexy (posterior IVS) and sacrospinous ligament fixation in the management of vault prolpase. Int Urogynecol J 2005;16:S54.

104

60 Neuman M, Lavy Y. Conservation of the prolapsed uterus is a valid option: medium term results of a prospective comparative study with the posterior intravaginal slingoplasty operation. Int Urogynecol J 2007;18:889-93.

61 Ng CC, Han WH. Comparison of effectiveness of vaginal and abdominal routes in treating severe uterovaginal or vault prolapse. Singapore Med J 2004;45:475-81.

62 Oliver R, Dasgupta C, Coker A. Posterior intravaginal slingplasty for vault and uterovaginal prolapse: An initial experience. Gynecol Surg 2006;3:88-92.

63 Papa Petros PE. Vault prolapse II: Restoration of dynamic vaginal supports by infracoccygeal sacropexy, an axial day-case vaginal procedure. Int Urogynecol J Pelvic Floor Dysfunct 2001;12:296-303.

64 Paraiso MF, Walters MD, Rackley RR, Melek S, Hugney C. Laparoscopic and abdominal sacral colpopexies: a comparative cohort study. Am J Obstet Gynecol 2005;192:1752-8.

65 Petros PE, Richardson PA. Tissue Fixation System posterior sling for repair of uterine/vault prolapse -- a preliminary report. Aust N Z J Obstet Gynaecol 2005;45:376-9.

66 Richter HE, Goode PS, Kenton K, Brown MB, Burgio KL, Kreder K et al. The effect of age on short-term outcomes after abdominal surgery for pelvic organ prolapse. J Am Geriatr Soc 2007;55:857-63.

67 Roovers JP, van der Vaart C, van der Bom JG, van Leeuwen JH, Scholten PC, Heintz PM. A randomised controlled trial comparing abdominal and vaginal prolapse surgery: effect on urogenital function. BJOG 2004;111:50-6.

68 Roovers JP, van der Bom JG, van der Vaart CH, van Leeuwen JH, Scholten PC, Heintz AP. A randomized comparison of post-operative pain, quality of life, and physical performance during the first 6 weeks after abdominal or vaginal surgical correction of descensus uteri. Neurourol Urodyn 2005;24:334-40.

69 Sentilhes L, Sergent F, Resch B, Verspyck E, Descamps P, Marpeau L. Midterm Follow-up of High-Grade Genital Prolapse Repair by the Trans-obturator and Infracoccygeal Hammock Procedure after Hysterectomy. Eur Urol 2007;51:1065-72.

70 Sivaslioglu AA, Gelisen O, Dolen I, Dede H, Dilbaz S, Haberal A. Posterior sling (infracoccygeal sacropexy): An alternative procedure for vaginal vault prolapse. Aust N Z J Obstet Gynaecol 2005;45:159-60.

71 Snyder TE, Krantz KE. Abdominal-retroperitoneal sacral colpopexy for the correction of vaginal prolapse. Obstet Gynecol 1991;77:944-9.

72 Su KC, Mutone MF, Terry CL, Hale DS. Abdominovaginal sacral colpoperineopexy: patient perceptions, anatomical outcomes, and graft erosions. Int Urogynecol J 2007;18:503-11.

73 Sze EH, Kohli N, Miklos JR, Roat T, Karram MM. A retrospective comparison of abdominal sacrocolpopexy with Burch colposuspension versus sacrospinous fixation with transvaginal needle suspension for the management of vaginal vault prolapse and coexisting stress incontinence. Int Urogynecol J 1999;10:390-3.

105

74 Tayrac R, Bader G, Deffieux X, Fazel A, Mathe ML, Fernandez H. A prospective randomized study comparing posterior IVS and sacrospinous suspension for the surgical treatment of uterine or vaginal vault prolapse. Int Urogynecol J 2006;17:S234-S235.

75 Timmons MC, Addison WA, Addison SB, Cavenar MG. Abdominal sacral colpopexy in 163 women with posthysterectomy vaginal vault prolapse and enterocele. Evolution of operative techniques. J Reprod Med 1992;37:323-7.

76 Vardy MD, Brodman M, Olivera CK, Zhou HS, Flisser AJ, Bercik RS. Anterior intravaginal slingplasty tunneller device for stress incontinence and posterior intravaginal slingplasty for apical vault prolapse: a 2-year prospective multicenter study. Am J Obstet Gynecol 2007;197:104-8.

77 Visco AG, Weidner AC, Barber MD, Myers ER, Cundiff GW, Bump RC et al. Vaginal mesh erosion after abdominal sacral colpopexy. Am J Obstet Gynecol 2001;184:297- 302.

78 Whitehead WE, Bradley CS, Brown MB, Brubaker L, Gutman RE, Varner RE et al. Gastrointestinal complications following abdominal sacrocolpopexy for advanced pelvic organ prolapse. Am J Obstet Gynecol 2007;197:78-7.

79 Wu JM, Wells EC, Hundley AF, Connolly A, Williams KS, Visco AG. Mesh erosion in abdominal sacral colpopexy with and without concomitant hysterectomy. Am J Obstet Gynecol 2006;194:1418-22.

80 Young SB, Kohorn EI, Braz-Martin S, Baker SP. A survey of the complications of vaginal prolapse surgery performed by members of the Society Of Gynecologic Surgeons. Int Urogynecol J 2004;15:165-70.

81 Brown M. Long-term effectiveness of abdominal sacrocolpopexy for the treatment of pelvic organ prolapse. ClinicalTrials.gov Identifier NCT00099372. ClinicalTrials.gov, 2005.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00099372?term=nct00099372&rank=1.

82 Chen CC. Laparoscopic versus robotic assisted laparoscopic sacrocolpopexy for vaginal prolapse. ClinicalTrials.gov Identifier NCT00551993. ClinicalTrials.gov, 2007.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00551993?term=nct00551993&rank=1.

83 Culligan P. Randomized controlled trial comparing acellular collagen biomesh (Pelvisoft) to polypropylene mesh (Pelvitex) for sacral colpopexy. ClinicalTrials.gov Identifier NCT00564083. ClinicalTrials.gov, 2005.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00564083?term=nct00564083&rank=1.

84 Freeman R. LAS study - a randomised controlled trial of abdominal versus laparoscopic sacrocolpopexy for the treatment of post-hysterectomy vaginal vault prolapse. NRR ID N0185181850. National Research Register Archive, 2006.[accessed February 2008] Available from: URL: https://portal.nihr.ac.uk/Profiles/NRR.aspx?Publication_ID=N0185181850.

85 Halaska M. Randomized multicentric study to treat prolapse after hysterectomy with Amreich procedure or total Prolift procedure. ClinicalTrials.gov Identifier NCT00572702. ClinicalTrials.gov, 2007.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00572702?term=nct00572702&rank=1.

106

86 Iglesia C. Efficacy study of vaginal mesh for prolapse (VAMP). ClinicalTrials.gov Identifier NCT00475540. ClinicalTrials.gov, 2007.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00475540?term=nct00475540&rank=1.

87 Zylstra S. Observational data collection of surgical outcomes in the treatment of vaginal prolapse with AMS products. ClincialTrials.gov Identifier NCT00388947. ClinicalTrials.gov, 2006.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00388947?term=nct00388947&rank=1.

88 Gannon NA. Treatment of stress and mixed urinary incontinence and vaginal vault prolapse. ClinicalTrials.gov Identifier NCT00223106. ClinicalTrials.gov, 2005.[accessed February 2008] Available from: URL: http://clinicaltrials.gov/ct2/show/NCT00223106?term=nct00223106&rank=1.

89 Mesh sacrocolpopexy for vaginal vault prolapse. Interventional Procedure Guidance Number: IP215. [document on the Internet] London: National Institute for Health and Clinical Excellence, 2007.[accessed March 2008] Available from: URL: http://www.nice.org.uk/guidance/index.jsp?action=byID&o=11202.

90 Posterior infracoccygeal sacropexy for vaginal vault prolapse. Interventional Procedure Guidance Number: IPG125. [document on the Internet] London: National Institute for Health and Clinical Excellence, 2005.[accessed March 2008] Available from: URL: http://www.nice.org.uk/guidance/index.jsp?action=byID&o=11163.

91 Laparoscopic uterine suspension sling IP372. [webpage on the Internet] London: National Institute for Health and Clinical Excellence, 2008.[accessed March 2008] Available from: URL: http://www.nice.org.uk/guidance/index.jsp?action=byID&o=11311.

107 APPENDIX 1 Search strategies

MEDLINE (1966- Nov Wk 2 2007) EMBASE (1980-2008 Wk 1) Medline In Process (4th Jan 2008) Ovid Multifile Search URL: http://gateway.ovid.com/athens

1 vaginal vault prolapse/ use emez 2 uterine prolapse/ 3 (prolapse adj2 (uter$ or cervic$ or apical or vagin$)).tw. 4 or/1-3 5 urogenital surgical procedures/ or gynecologic surgical procedures/ or urologic surgical procedures/ use mesz 6 gynecologic surgery/ or urologic surgery/ use emez 7 (surgery or surgical or repair$).tw. 8 4 and (5 or 6 or 7) 9 4 and su.fs. 10 (sacro$ hysteropex$ or sacral hysteropex$ or sacrohysteropex$).tw. 11 (sacro$ colpopex$ or sacral colpopex$ or sacrocolpopex$).tw. 12 (sacro$ cervicopex$ or sacral cervicopex$ or sacrocervicopex$).tw. 13 (sacropex$ or intravaginal sling$).tw. 14 (suspension adj2 (uter$ or vault or sacro$ or sling$)).tw. 15 or/10-14 16 surgical mesh/ 17 exp fascia/ 18 (mesh or graft$ or allograft$ or fascia or tape$).tw. 19 polypropylenes/ 20 (prolene or trelex or avaulta or pelvitex or prolift or polyform or marlex or gynemesh or gore text).tw,dv. 21 polyglactin 910/ 22 vicryl.tw. 23 (tutoplast or faslata or fortagen).tw. 24 (porcine dermis or pelvicol or pelvisoft).tw. 25 or/16-24 26 (8 or 9) and 25 27 15 or 26 28 limit 27 to yr="1980 - 2007" 29 english.la. 30 28 and 29 31 animals/ not humans/ 32 (animal/ or nonhuman/) not human/ use emez 33 30 not (31 or 32) 34 clinical trial.pt. 35 exp controlled study/ use emez 36 randomi?ed.ab. 37 randomly.ab. 38 trial.ab. 39 groups.ab. 40 comparative study.pt. 41 follow-up studies/ use mesz 42 time factors/ use mesz 43 treatment outcome/ use emez 44 major clinical study/ use emez 45 clinical trial/ use emez 46 (preoperat$ or pre operat$).mp. use mesz 47 (chang$ or evaluat$ or reviewed or baseline).tw. 48 (prospective$ or retrospective$).tw. use mesz 49 (cohort$ or case series).tw. use mesz 50 (compare$ or compara$).tw. use emez 51 case report/ use emez 52 case reports.pt. 53 case?.tw.

108 54 or/34-53 55 33 and 54 56 remove duplicates from 55

Science Citation Index (1980- 5th Jan 2008) Biosis (1985-3rd Jan 2008) ISI Proceedings (1990- 6th Jan 2008) Web of Knowledge URL: http://wok.mimas.ac.uk/

#1 TS=(prolapse SAME (uter* OR cervic* OR apical OR vagin* OR vault)) #2 TS=(mesh OR fascia OR tape*) #3 TS=graft #4 TS=allograft #5 TS=(polypropylene* OR prolene OR trelex OR avaulta OR pelvitex OR prolift OR polyform OR marlex OR gynemesh OR gore tex) #6 TS=(polyglactin OR vicryl OR tutoplast OR faslata OR fortagen) #7 TS=(porcine dermis OR pelvicol OR pelvisoft) #8 #1 AND (#2 OR #3 OR #4 OR #5 OR #6 OR #7) #9 TS=(sacro* hysteropex* OR sacral hysteropex* OR sacrohysteropex*) #10 TS=(sacro* colpopex* OR sacral colpopex* OR sacrocolpopex*) #11 TS=(sacro* cervicopex* OR sacral cervicopex* OR sacrocervicopex*) #12 TS=(sacropex* OR intravaginal sling*) #13 TS=(suspension SAME (uter* OR vault OR sacro* OR sling*)) #14 #8 OR #9 OR #10 OR #11 OR #12 OR#13

Cochrane Library ( Issue 4, 2007) URL: http://www3.interscience.wiley.com/cgi-bin/mrwhome/106568753/HOME

#1 MeSH descriptor Uterine Prolapse explode all trees #2 (prolapse near/2 (cervic* or apical or vagin* or uter*)):ti,ab,kw #3 #1 or #2 #4 MeSH descriptor Urogenital Surgical Procedures explode all trees #5 MeSH descriptor Gynecologic Surgical Procedures explode all trees #6 MeSH descriptor Urologic Surgical Procedures explode all trees #7 (surgery or surgical or repair*):ti,ab,kw #8 (#4 OR #5 OR #6 OR #7 #9 SU subheading #10 #3 and (#8 or #9) #11 MeSH descriptor Surgical Mesh explode all trees #12 MeSH descriptor Fascia explode all trees #13 MeSH descriptor Polypropylenes, this term only #14 MeSH descriptor Polyglactin 910, this term only #15 (mesh or graft* or allograft* or fascia):ti,ab,kw #16 (prolene or trelex or avaulta) or (pelvitex or prolift or polyform) or (marlex or gynemesh or gore tex) i\#17 (vicryl or tutoplast or faslata or fortagen) #18 (porcine dermis or pelvicol or pelvisoft) #19 (#11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18) #20 #10 and #19 #21 (sacro* hysteropex*):ti,ab,kw or (sacral hysteropex*):ti,ab,kw or (sacrohysteropex*):ti,ab,kw #22 (sacro* colpopex*):ti,ab,kw or (sacral colpopex*):ti,ab,kw or (sacrocolpopex*):ti,ab,kw #23 (sacro* cervicopex*):ti,ab,kw or (sacral cervicopex*):ti,ab,kw or (sacrocervicopex*):ti,ab,kw #24 (sacropex* or intravaginal slingplast*):ti,ab,kw #25 (suspension near/2 (uter* or vault or sacro*)):ti,ab,kw #26 #20 or #21 or #22 or #23 or #24 or #25

109 DARE and HTA Databases (Jan 2008) NHS Centre for Reviews & Dissemination URL: http://nhscrd.york.ac.uk/welcome.htm

# 1 MeSH Uterine Prolapse # 2 ( prolapse AND ( uter* OR cervic* OR apical OR vagin* ) ) # 3 #1 or #2 # 4 ( sacro* AND hysteropex* ) OR ( sacral AND hysteropex* ) OR ( sacrohysteropex* ) # 5 ( sacro* AND colpopex* ) OR ( sacral AND colpopex* ) OR ( sacrocolpopex* ) # 6 ( sacro* AND cervicopex* ) OR ( sacral AND cervicopex* ) OR ( sacrocervicopex* ) # 7 sacropex* OR ( intravaginal AND sling* ) # 8 suspension AND ( uter* OR vault OR sacro* ) # 9 #3 OR #4 OR #5 OR #6 OR #7 OR #8

Clinical Trials (Jan 2008) URL: http://clinicaltrials.gov/ct/gui/c/r

Condition = prolapse

Current Controlled Trials (Jan 2008) URL: http://www.controlled-trials.com/

(prolapse and (vagina% or pelvic or uter% or cervic% or genital or apical)

Conference Proceedings Abstracts:

American Urogynecologic Society 2006 Annual Scientific Meeting, Palm Springs,CA., 19-21 October. 2005 Scientific Meeting, Atlanta, GA., October.

American Urological Association Annual Meetings 2005-7

Available from: URL: http://www.abstracts2view.com/aua/sessionindex.phpUTH

European Association of Urology 2007 Annual Congress, Berlin, Germany, 21-24 March. 2006 Annual Congress, Paris, France, 5-8 April. 2005 Annual Congress, Istanbul, Turkey, 16-19 March.

European Society of Gynecological Endoscopy 15th Annual Congress, Strasbourg, France, 5-7 October. 14th Annual Congress, Athens, Greece, 6-8 October.

International Continence Society 2007 Annual Meeting, Rotterdam, The Netherlands, 20-24 August. 2006 Annual Meeting,Christchurch, New Zealand, 27 November-1 December. 2005 Annual Meeting,Montreal, Canada, 22 August-2 September.

International Urogynecological Association 2007 Annual Meeting, Cacun, Mexico, 12-16 June. 2006 Annual Meeting, Athens, Greece, 6-9 September. 2005 Annual Meeting, Copenhagen, Denmark, 9-12 August. 2007 Annual Congress, Berlin, 21-24 March.

110 Websites searched for other evidence-based reports and background information: American Medical Systems URL: http://www.americanmedicalsystems.com/ American Urogynecologic Society URL: http://www.augs.org/ Americal Urological Association URL: http://www.auanet.org/ Atrium Medical URL: http://www.atriummed.com/atrium.asp Bard Urological URL: http://www.bardurological.com/ Boston Scientific URL: http://www.bostonscientific.co.uk/ Colplast URL: http://www.urology.coloplast.com/ Ethicon Women’s Health & Urology URL: http://www.womenshealthsolutions.co.uk/ European Society of Gynecological Endoscopy URL: http://www.esge.org/ European Association of Urology URL: http://www.uroweb.org/ Gore Medical URL: http://www.goremedical.com/ International Continence Society: URL https://www.icsoffice.org/ International Urogynecological Association URL: http://www.iuga.org/ Johnson & Johnson URL: http://www.jnjgateway.com/ Mpathy URL: http://www.mpathy.org/ Organogenesis URL: http://www.organogenesis.com/ Tutogen Medical URL: http://www.tutogen.com/

111 APPENDIX 2 Checklist of quality assessment of randomised controlled trials

Criteria Yes No Unclear Comment 1. Was the assignment to the treatment groups really random?

2. Was the treatment allocation concealed from those responsible for entering patients into trials, i.e. not knowing upcoming assignments in advance?

3. Were the groups similar at baseline in terms of prognostic factors, e.g. age, duration of disease, disease severity?1 4. Were the eligibility criteria specified? 5. Was the intervention (and comparison) clearly defined? 6. Were the groups treated in the same way apart from the intervention received? 7. Was there a follow-up period ≥ 1 year? 8. Was the outcome assessor blinded to the treatment allocation? 9. If patient blind is possible, were the patients blinded?2 10. If having primary outcome measures as continuous data, were the point estimates and measures of variability presented?3 11. Were the withdrawals/drop-outs having similar characteristics as those completed the study and therefore unlikely to cause bias?4 12. Did the analyses include all women according to randomised groups, i.e. intention-to-treat analysis?5 13. Was the operation undertaken by somebody experienced in performing the procedure?6

Note:

1. ‘Yes’ if two or more than two factors were similar. 2. If patient blinding is impossible, note ‘impossible’ in comment area and leave other cells blank. 3. If having no primary outcome measures as continuous data, note ‘no continuous data’ in comment area and leave other cells blank. 4. ‘Yes’ if no withdrawal/drop out; ‘No’ if drop-out rate ≥30% or differential drop-out. 5. ‘Yes’ if no withdrawals/drop out after enrol 6. ‘Yes’ if the practitioner received training on conducting the procedure before or conducted same kind of procedure before, i.e. no learning curve.

112 APPENDIX 3 Checklist of quality assessment of non-randomised studies

Criteria Yes No Unclear Comments 1. Were participants a representative sample selected from a relevant patient population, e.g. randomly selected from those seeking for treatment despite of age, duration of disease, primary or secondary disease, and severity of disease? 2. Were the inclusion/exclusion criteria of participants clearly described? 3. Were participants entering the study at a similar point in their disease progression, i.e. severity of disease? 4. Was selection of patients consecutive? 5. Was data collection undertaken prospectively? 6. Were the groups comparable on demographic characteristics and clinical features? 7. Was the intervention (and comparison) clearly defined? 8. Was the intervention undertaken by someone experienced at performing the procedure?1 9. Were the staff, place, and facilities where the patients were treated appropriate for performing the procedure? (E.g. access to back-up facilities in hospital or special clinic) 10. Were all the important outcomes considered? 11. Were objective (valid and reliable) outcome measures used, including satisfaction scale? 12. Was the assessment of main outcomes blind? 13. Was follow-up long enough (≥1y) to detect important effects on outcomes of interest? 14. Was information provided on non-respondents, dropouts?2 15. Were the withdrawals/drop-outs having similar characteristics as those completed the study and therefore unlikely to cause bias?3 16. Was length of follow-up similar between comparison groups 17. Were all the important prognostic factors identified, e.g. age, duration of disease, disease severity?4 18. Were the analyses adjusted for confounding factors?

The same form was adapted to assess the quality of case series after taking out question 6, 12, 16 and 18.

Note: 1. ‘Yes’ if the practitioner received training on conducting the procedure before or conducted same kind of procedure before, i.e. no learning curve. 2. ‘No’ if participants were from those whose follow up records were available (retrospective) 3. ‘Yes’ if no withdrawal/drop out; ‘No’ if drop-out rate ≥30% or differential drop-out, e.g. those having most severe disease died during follow up but the death was not due to treatment; no description of those lost. 4. ‘Yes’ if two or more than two factors were similar.

113 APPENDIX 4 Included studies

Altman 2006 Altman D, Anzen B, Brismar S, Lopez A, Zetterstrom J. Long-term outcome of abdominal sacrocolpopexy using xenograft compared with synthetic mesh. Urology 2006;67:719-24.

Bai 2006 Bai SW, Kwon HS, Chung DJ. Abdominal high uterosacral colpopexy and abdominal sacral colpopexy with mesh for pelvic organ prolapse. Int J Gynecol Obstet 2006;92:147-8.

Banu 1997 Banu LF. Synthetic sling for genital prolapse in young women. Int J Gynecol Obstet 1997;57:57-64.

Barranger 2003 Barranger E, Fritel X, Pigne A. Abdominal sacrohysteropexy in young women with uterovaginal prolapse: long-term follow-up. Am J Obstet Gynecol 2003;18:1245-50.

Begley 2005 Begley JS, Kupferman SP, Kuznetsov DD, Kobashi KC, Govier FE, McGonigle KF et al. Incidence and management of abdominal sacrocolpopexy mesh erosions. Am J Obstet Gynecol 2005;192:1956-62.

Bensinger 2005 Bensinger G, Lind L, Lesser M, Guess M, Winkler HA. Abdominal sacral suspensions: analysis of complications using permanent mesh. Am J Obstet Gynecol 2005;193:2094-8.

Benson 1996 Benson JT, Lucente V, McClellan E. Vaginal versus abdominal reconstructive surgery for the treatment of pelvic support defects: a prospective randomized study with long-term outcome evaluation. Am J Obstet Gynecol 1996;175:1418-21.

Bradley 2007 Primary reference Bradley CS. Bowel symptoms in women 1 year after sacrocolpopexy. Am J Obstet Gynecol 2007;197:642e1-e8.

Secondary references Brubaker L, Cundiff GW, Fine P, Nygaard I, Richter HE, Visco AG et al. Abdominal sacrocolpopexy with Burch colposuspension to reduce urinary stress incontinence. N Engl J Med 2006;354:1557-66.

Burgio KL. Bladder symptoms 1 year after abdominal sacrocolpopexy with and without Burch colposuspension in women without preoperative stress incontinence symptoms. Am J Obstet Gynecol 2007;197:647e1-e6.

Handa VL. Sexual function before and after sacrocolpopexy for pelvic organ prolapse. Am J Obstet Gynecol 2007;197:629e1-e6.

Richter HE, Goode PS, Kenton K, Brown MB, Burgio KL, Kreder K et al. The effect of age on short-term outcomes after abdominal surgery for pelvic organ prolapse. J Am Geriatr Soc 2007;55:857-63.

Whitehead WE, Bradley CS, Brown MB, Brubaker L, Gutman RE, Varner RE et al. Gastrointestinal complications following abdominal sacrocolpopexy for advanced pelvic organ prolapse. Am J Obstet Gynecol 2007;197:78-7.

114 Braun 2007 Braun H, Fernandez M, Delloro A, Gonzalez F, Cuevas R, Rojas I. Prospective randomised study to compare colposacropexy and Mayo McCall technique in the correction of severe genital central prolapse. Int Urogynecol J 2007;18:S1-S24.

Brizzolara 2003 Brizzolara S, Pillai-Allen A. Risk of mesh erosion with sacral colpopexy and concurrent hysterectomy. Obstet Gynecol 2003;102:306-10.

Constantini 2005 Costantini E, Mearini L, Bini V, Zucchi A, Mearini E, Porena M. Uterus preservation in surgical correction of urogenital prolapse. Eur Urol 2005;48:642-9.

Culligan 2002 Culligan PJ, Murphy M, Blackwell L, Hammons G, Graham C, Heit MH. Long-term success of abdominal sacral colpopexy using synthetic mesh. Am J Obstet Gynecol 2002;187:1473- 80.

Culligan 2005 Culligan PJ, Blackwell L, Goldsmith LJ, Graham CA, Rogers A, Heit MH. A randomized controlled trial comparing fascia lata and synthetic mesh for sacral colpopexy. Obstet Gynecol 2005;106:29-37.

De Vries 1995 de Vries MJ, van Dessel TH, Drogendijk AC, de H, I, Huikeshoven FJ. Short-term results and long-term patients' appraisal of abdominal colposacropexy for treatment of genital and vaginal vault prolapse. Eur J Obstet Gynecol Reprod Biol 1995;59:35-8.

Demirci 2007 Demirci F. Perioperative complications in abdominal sacrocolpopexy and vaginal sacrospinous ligament fixation procedures. Int Urogynecol J Pelvic Floor Dysfunct 2007;18:257-61.

Elneil 2005 Elneil S, Cutner AS, Remy M, Leather AT, Toozs-Hobson P, Wise B. Abdominal sacrocolpopexy for vault prolapse without burial of mesh: a case series. BJOG 2005;112:486-9.

Farnsworth 2002 Farnsworth BN. Posterior intravaginal slingplasty (infracoccygeal sacropexy) for severe posthysterectomy vaginal vault prolapse--a preliminary report on efficacy and safety. Int Urogynecol J 2002;13:4-8.

Fedorkow 1993 Fedorkow DM, Kalbfleisch RE. Total abdominal hysterectomy at abdominal sacrovaginopexy - a comparative-study. Am J Obstet Gynecol 1993;169:641-3.

Foote 2007 Foote AJ. Infracoccygeal sacropexy. Aust N Z J Obstet Gynaecol 2007;47:250-1.

Ghanbari 2006 Ghanbari Z, Baratali BH, Mireshghi MS. Posterior intravaginal slingplasty (infracoccygeal sacropexy) in the treatment of vaginal vault prolapse. Int J Gynaecol Obstet 2006;94:147-8.

115 Govier 2005 Govier FE, Kobashi KC, Kozlowski PM, Kuznetsov DD, Begley SJ, McGonigle KF et al. High complication rate identified in sacrocolpopexy patients attributed to silicone mesh. Urology 2005;65:1099-103.

Gregory 2005 Gregory WT, Otto LN, Bergstrom JO, Clark AL. Surgical outcome of abdominal sacrocolpopexy with synthetic mesh versus abdominal sacrocolpopexy with cadaveric fascia lata. Int Urogynecol J 2005;16:369-74.

Griffis 2006 Griffis K, Evers MD, Terry CL, Hale DS. Mesh erosion and abdominal sacrocolpopexy: A comparison of prior, total, and supracervical hysterectomy. J Pelvic Med Surg 2006;12:25- 30.

Hardiman 1996 Hardiman PJ, Drutz HP. Sacrospinous vault suspension and abdominal colposacropexy: success rates and complications. Am J Obstet Gynecol 1996;175:612-6.

Hefni 2007 Hefni M. Morbidity associated with posterior intravaginal slingplasty for uterovaginal and vault prolapse. Arch Gynecol Obstet 2007;276:499-504.

Higgs 2005a Higgs PJ, Chua HL, Smith AR. Long term review of laparoscopic sacrocolpopexy. BJOG 2005;112:1134-8.

Higgs 2005b Higgs P, Goh J, Krause H, Sloane K, Carey M. Abdominal sacral colpopexy: an independent prospective long-term follow-up study. Aust N Z J Obstet Gynaecol 2005;45:430-4.

Hsiao 2007 Hsiao KC. Comparison of laparoscopic and abdominal sacrocolpopexy for the treatment of vaginal vault prolapse. J Endourol 2007;21:926-30.

Jordaan 2006 Jordaan DJ, Prollius A, Cronje HS, Nel M. Posterior intravaginal slingplasty for vaginal prolapse. Int Urogynecol J 2006;17:326-9.

Joshi 1993 Joshi VM. A new technique of uterine suspension to pectineal ligaments in the management of uterovaginal prolapse. Obstet Gynecol 1993;81:790-3.

Leron 2001 Leron E, Stanton SL. Sacrohysteropexy with synthetic mesh for the management of uterovaginal prolapse. BJOG 2001;108:629-33.

Lindeque 2002 Lindeque BG, Nel WS. Sacrocolpopexy--a report on 262 consecutive operations. S Afr Med J 2002;92:982-5.

Lo 1998 Lo T-S. Abdominal colposacropexy and sacrospinous ligament suspension for severe uterovaginal prolapse: A comparison. J Gynecol Surg 1998;14:59-64.

116 Maher 2004 Maher CF, Qatawneh AM, Dwyer PL, Carey MP, Cornish A, Schluter PJ. Abdominal sacral colpopexy or vaginal sacrospinous colpopexy for vaginal vault prolapse: a prospective randomized study. Am J Obstet Gynecol 2004;190:20-6.

Marcickiewicz 2007 Marcickiewicz J, Kjollesdal M, Engh ME, Eklind S, Axen C, Brannstrom M et al. Vaginal sacrospinous colpopexy and laparoscopic sacral colpopexy for vaginal vault prolapse. Acta Obstet Gynecol Scand 2007;86:733-8.

Meschia 2005 Meschia M, Barbacini P, Longatti D, Gattei U, Pifarotti P. Randomized comparison between infracoccygeal sacropexy (posterior IVS) and sacrospinous ligament fixation in the management of vault prolapse. Int Urogynecol J 2005;16:S54.

Neuman 2007 Neuman M, Lavy Y. Conservation of the prolapsed uterus is a valid option: medium term results of a prospective comparative study with the posterior intravaginal slingoplasty operation. Int Urogynecol J 2007;18:889-93.

Ng 2004 Ng CC, Han WH. Comparison of effectiveness of vaginal and abdominal routes in treating severe uterovaginal or vault prolapse. Singapore Med J 2004;45:475-81.

Oliver 2006 Oliver R, Dasgupta C, Coker A. Posterior intravaginal slingplasty for vault and uterovaginal prolapse: An initial experience. Gynecol Surg 2006;3:88-92.

Papa Petros 2001 Papa Petros PE. Vault prolapse II: Restoration of dynamic vaginal supports by infracoccygeal sacropexy, an axial day-case vaginal procedure. Int Urogynecol J Pelvic Floor Dysfunct 2001;12:296-303.

Paraiso 2005 Paraiso MF, Walters MD, Rackley RR, Melek S, Hugney C. Laparoscopic and abdominal sacral colpopexies: a comparative cohort study. Am J Obstet Gynecol 2005;192:1752-8.

Petros 2005 Petros PE, Richardson PA. Tissue Fixation System posterior sling for repair of uterine/vault prolapse -- a preliminary report. Aust N Z J Obstet Gynaecol 2005;45:376-9.

Roovers 2004 Primary reference Roovers JP, van der Vaart C, van der Bom JG, van Leeuwen JH, Scholten PC, Heintz PM. A randomised controlled trial comparing abdominal and vaginal prolapse surgery: effect on urogenital function. BJOG 2004;111:50-6.

Secondary reference Roovers JP, van der Bom JG, van der Vaart CH, van Leeuwen JH, Scholten PC, Heintz AP. A randomized comparison of post-operative pain, quality of life, and physical performance during the first 6 weeks after abdominal or vaginal surgical correction of descensus uteri. Neurourol Urodyn 2005;24:334-40.

117 Sentilhes 2007 Sentilhes L, Sergent F, Resch B, Verspyck E, Descamps P, Marpeau L. Midterm Follow-up of High-Grade Genital Prolapse Repair by the Trans-obturator and Infracoccygeal Hammock Procedure after Hysterectomy. Eur Urol 2007;51:1065-72.

Sivaslioglu 2005 Sivaslioglu AA, Gelisen O, Dolen I, Dede H, Dilbaz S, Haberal A. Posterior sling (infracoccygeal sacropexy): An alternative procedure for vaginal vault prolapse. Aust N Z J Obstet Gynaecol 2005;45:159-60.

Snyder 1991 Snyder TE, Krantz KE. Abdominal-retroperitoneal sacral colpopexy for the correction of vaginal prolapse. Obstet Gynecol 1991;77:944-9.

Su 2007 Su KC, Mutone MF, Terry CL, Hale DS. Abdominovaginal sacral colpoperineopexy: patient perceptions, anatomical outcomes, and graft erosions. Int Urogynecol J 2007;18:503-11.

Sze 1999 Sze EH, Kohli N, Miklos JR, Roat T, Karram MM. A retrospective comparison of abdominal sacrocolpopexy with Burch colposuspension versus sacrospinous fixation with transvaginal needle suspension for the management of vaginal vault prolapse and coexisting stress incontinence. Int Urogynecol J 1999;10:390-3.

Tayrac 2006 Tayrac R, Bader G, Deffieux X, Fazel A, Mathe ML, Fernandez H. A prospective randomized study comparing posterior IVS and sacrospinous suspension for the surgical treatment of uterine or vaginal vault prolapse. Int Urogynecol J 2006;17:S234-S235.

Timmons 1992 Timmons MC, Addison WA, Addison SB, Cavenar MG. Abdominal sacral colpopexy in 163 women with posthysterectomy vaginal vault prolapse and enterocele. Evolution of operative techniques. J Reprod Med 1992;37:323-7.

Vardy 2007 Vardy MD, Brodman M, Olivera CK, Zhou HS, Flisser AJ, Bercik RS. Anterior intravaginal slingplasty tunneller device for stress incontinence and posterior intravaginal slingplasty for apical vault prolapse: a 2-year prospective multicenter study. Am J Obstet Gynecol 2007;197:104-8.

Visco 2001 Visco AG, Weidner AC, Barber MD, Myers ER, Cundiff GW, Bump RC et al. Vaginal mesh erosion after abdominal sacral colpopexy. Am J Obstet Gynecol 2001;184:297-302.

Wu 2006 Wu JM, Wells EC, Hundley AF, Connolly A, Williams KS, Visco AG. Mesh erosion in abdominal sacral colpopexy with and without concomitant hysterectomy. Am J Obstet Gynecol 2006;194:1418-22.

Young 2004 Young SB, Kohorn EI, Braz-Martin S, Baker SP. A survey of the complications of vaginal prolapse surgery performed by members of the Society Of Gynecologic Surgeons. Int Urogynecol J 2004;15:165-70.

118 APPENDIX 5 Characteristics of the included studies by mesh techniques

Abdominal sacrocolpopexy by study design and by types of prolapse

FULL-TEXT RCTs

Uterine prolapse repair (no full-text RCTs available)

Vault prolapse repair Author, year Inclusion criteria: women with INTERVENTION A (n=45): EFFICACY, 1y Culligan 200535 posthysterectomy vaginal vault prolapse abdominal (open) sacrocolpopexy Objective failure: POP-Q stage II or more at C point: A, 0/45; B, 0/44 scheduled for sacrolcolpopexy Type of prolapse: vault Type of mesh/graft: polypropylene Other efficacy outcomes: the authors reported 18 ‘failures’ (A, 4/45; B, Exclusion criteria: NR mesh, Trelex, Boston Scientific, 14/44), 15 of which were classified as such because POP-Q point Aa, i.e. Publication type: full-text Boston, MA anterior prolapse, reached the -1 position or beyond, 3 were classified as Number of patients lost to follow-up, such because POP-Q point Ap, i.e. posterior prolapse, reached the -1 Study design: RCT n/N: A, 9/54; B, 2/46; 4 patients point or beyond, but not reported these were recurrent prolapse or de randomised to receive fascia were INTERVENTION B (n=44): novo. Setting: University hospital actually given mesh because of a abdominal (open) sacrocolpopexy transient shortage of graft The authors also found significant differences between A and B groups Country: US Type of mesh/graft: solvent- with respect to the 1y postoperative comparisons of points Aa, C, and Age (mean, SD): A, 60.4 (10.1); B, 57.5 dehydrated cadaveric fascia lata, POP-Q stages, but did not report the actual data.

119 Recruitment/Treatment dates: (10.8) Tutoplast, Mentor Corporation, Santa Jul 2001 – Jun 2003 Barbara, CA SAFETY Parity (median): A, 3; B, 2 Blood loss, ml, mean (SD): A, 247.2 (148.4); B, 264.7 (261.4) Length of follow-up: 1y results available Preoperative prolapse stage (any Concomitant procedures: Blood loss requiring transfusion, n/N: A, 1/54; B, 0/46 prolapse), mean (SD): A, 2.5 (0.5); B, Paravaginal repairs: A, 36/54; B, Source of funding: this study was 2.4 (0.7) 42/46 Damage to surrounding tissue, e.g. urethra, bladder, bowel: bladder funded by the Mentor Corporation, Posterior repairs: A, 26/54; B, 20/46 injury, A, 1/54; B, 0/46 Santa Barbara, California Number of surgeons: NR Mesh/graft erosion, n/N: A, 2/54; B, 0/46

Other serious adverse events: Pulmonary embolism: A, 1/54; B, 0/46

Other safety outcomes: postoperative fever: A, 2/54; B, 2/46; ileus: A, 2/54; B, 0/46; wound breakdown: A, 8/54; B, 5/46

OTHER OUTCOMES Operating time, mean, SD, min: A, 227.3 (63.3); B, 233.4 (66.9)

Author, year Inclusion criteria: women with INTERVENTION A (n=47): EFFICACY, mean 24m (6-60) Maher 200457 symptomatic posthysterectomy vaginal abdominal sacral colpopexy Subjective failure: ‘no symptoms of prolapse’: A, 43/46; B, 39/43 vault prolapse that extended to or Type of prolapse: vault beyond the introitus Type of mesh/graft: Prolene mesh, Objective failure: vault prolapse at grade 2 or more: A, 2/46; B, 8/43 Ethicon Inc., Somerville, NJ

Publication type: full-text Exclusion criteria: women who had The authors also reported ‘no prolapse at grade 2 or more at any vaginal undergone a previous sacral colpopexy site’: A, 36/47; B, 33/48, but NR the prolapse were recurrent or de novo. Study design: RCT or had a significantly foreshortened INTERVENTION B (n=48): unilateral vagina vaginal sacrospinous colpopexy (no Persistent sexual problems, n/N: dyspareunia: A, 4/9; B, 4/7 Setting: hospital mesh) Number of patients lost to follow-up, SAFETY Country: Australia n/N: A, 1/47 (lost to contact); B, 5/48 (1 Blood loss, ml, mean (SD) range: A, 362 (239), 100-1100; B, 306 (201), died of pneumonia, 2 had dementia, 2 Concomitant procedures: 100-1000 Recruitment/Treatment dates: were lost to contact) Posterior colporrhaphy: A, 11/47; B, Sep 1997 – Dec 2000 44/48 Blood loss requiring transfusion, n/N: A, 1/47; B, 1/48 Age (mean, SD, range): A, 63.4 (10.3), Anterior colporrhaphy: A, 0/47; B, Length of follow-up (mean, 39-84; B, 62.9 (10.3), 35-88 21/48 Damaged to surrounding organs, n/N: range): 24m (6-60) Enterocele repair: A, 0/47; B, 28/48 Cystotomy: A, 1/47; B, 1/48 Parity (mean, SD, range): A, 3.0 (1.6), Paravaginal repair: A, 19/47; B, 0/48 Source of funding: NR 1-8; B, 3.2 (1.0), 2-6 Mesh/graft erosion, n/N: A, 1/47; B, NA

BMI (mean, SD, range): A, 26.7 (4.9), New urinary symptoms: SUI: A, 2/22; B, 8/24, the authors mentioned the 19.6-42.2; B, 26.6 (4.9), 18.1-39.9 difference may be explained by the differences in paravaginal repairs between groups. This procedure is likely to offer superior bladder neck Pre-operative prolapse: support than the anterior repair alone. Urethrocele: A, 22/47; B, 20/48 Cystocele: A, 33/47; B, 28/48 Infection: A, 1/47; B, 0/48 Vault (to introitus): A,24/47; B,25/48 Vault (+introitus): A, 19/47; B, 23/48 Other adverse events: 120 Enterocele: A, 36/47; B, 38/48 Incisional hernia: A, 2/47; B, 0/48 Rectocele: A, 35/47; B, 30/48 Rectovaginal hematoma: A, 0/47; B, 1/48 Perineum: A 32/47; B, 28/48 Vaginal pain of undetermined origin: A, 0/47; B, 1/48

Number of surgeons: NR, all surgical OTHER OUTCOMES procedures were performed under the Operating time, min, mean (SD), range: A, N=47, 106min (37), 45-100; direct supervision of a consultant B, N=48, 76min (42), 26-300 urogynaecologist Length of hospital stay, d: A, 5.4 (2.2), 3-16; B, 4.8 (1.4), 3-10

Uterine and/or vault prolapse repair

Author, year Inclusion criteria: women with cervical INTERVENTION A (n=38): EFFICACY Benson 199627 prolapse to or beyond the hymen or with abdominal approach colposacral Subjective failure: women had symptomatic vaginal apex descent >50% vaginal vault in version >50% of its suspension and an abdominal of its length or vaginal wall protrusion beyond the hymen: A, 6/38; B, 14/42 Type of prolapse: uterine/vault length and anterior vaginal descent to or paravaginal repair (the others were asymptomatic regardless of the prolapse stage) beyond the hymen. Publication type: full-text Type of mesh/graft: NR SAFETY Exclusion criteria: a uterus >12w size, Blood loss requiring transfusion, n/N: A, 2/38; B, 0/42 Study design: RCT adnexal mass, fore shortened vaginal Surgical route: abdominal (NR vault, cystocele solely of the central or open or laparoscopic) Damage to surrounding tissue, e.g. urethra, bladder, bowel: Setting: university hospital midline defect type, history of >=2 enterotomy, A 1/38; B, 0/42; abdominal pelvic surgeries, morbid cystotomy, A, 1/38; B, 1/42; Country: US obesity. INTERVENTION B (n=42): vaginal obturator nerve injury, A, 1/38; B, 0/42

approach bilateral sacrospinous Recruitment/Treatment dates: Number of patients lost to follow-up, ligament vault suspension and Mesh/graft erosion, n/N: development of vaginal band requiring excision 14 Dec. 1989 – 27 Feb 1992 n/N: A, 2/40; B, 6/48 vaginal paravaginal repair (no mesh) (NR detail): A, 0/38; B, 1/42

Length of follow-up: mean, range, Age (mean, SD): A, 66.2 (9.7); B, 63.6 Infection: urinary tract infection, A 0/38; B, 9/42; 2.5y (1-5.5) (9.3) Concomitant procedures: wound infection, A, 3/38; B, 0/42; Needle suspension urethropexy, A, febrile morbidity, A, 8/38; B, 4/42 Source of funding: NR Parity (mean, SD): A, 3.3 (1.5); B, 2.6 0, B, 20/48; Burch procedure, A, (1.4) 14/40, B, 0. Other adverse events: sciatica, A 2/38; B, 0/42; ileus, A, 1/38; B, 0/42; Pre-operation prolapse stage: Hysterectomy: A, 20/40; B, 24/48; phlebitis: A, 1/38; B, 0/42 Women with grade 4 prolapse, i.e. Anterior colporrhaphy, A, 12/40; B, structures being beyond the introitus at 16/48; OTHER OUTCOMES rest: A, 18/40; B, 19/48 Posterior colporrhaphy: A, 20/40; B, Operating time: mean (SD), min, A, 215 (47); B, 196 (38) 32/48 Number of surgeons: all surgeries Length of hospital stay: mean (SD, day, A, 5.4 (1.1); B, 5.1 (1.2) were performed under the direct supervision of the first author.

Author, year Inclusion criteria: women with uterine INTERVENTION A (n=52): EFFICACY Lo 199856 prolapse or vaginal inversion greater abdominal (open) colposacropexy Objective failure: vault prolapse >=stage II, A, 0/52; B, NR. than stage III (POP-Q) Type of prolapse: uterine/vault Type of mesh/graft: Mersilene Any prolapse >= stage II, A, 3/52; B, 13/66. Exclusion criteria: patients with voiding mesh 121 Publication type: full-text problems necessitating surgical SAFETY correction INTERVENTION B (n=66): vaginal Blood loss, ml, mean (SD): A, 150 (137); B, 448 (258) Study design: RCT sacrospinous ligament suspension Number of patients lost to follow-up, Damage to surrounding tissue, e.g. urethra, bladder, bowel: ureteral Setting: medical centre n/N: 14/66; 6/72. The reason for lost of injury, A, 1/52; B, 0/66 follow up was advanced age, disability, Concomitant procedures: Country: Taiwan, China or unwilling to participate. Continence procedures: 0 New urinary symptoms: urinary incontinence, A, 2/52; B, 1/66 Post colporrhaphy: A, 19/52; B, 0; Recruitment/Treatment dates: Age (mean, SD): A, 63 (9); B, 60 (10) Anterior and posterior colporrhaphy: New sexual problems: dyspareunia, A, 1/11; B, 7/18 Jan. 1991 – Jan. 1996 Parity (mean, SD): A, 5.6 (5.1); B, 4.2 A, 64/66; B, 0; Length of follow-up: mean, range, (1.9) Hysterectomy: A, 33/52; B, 20/66 Infection: urinary tract infection, A, 2/52; B, 4/66; 2.1y, 1-5.2y retroperitoneal infection, A, 1/52, required surgery; B, 0/66; Number of surgeons: NR perineal infection, A, 0; B, 3, required surgery Source of funding: NR Other serious adverse events: rectovaginal fistula, A, 0; B, 1 required surgery

Other adverse events: vaginal vault strictures, A, 0/52; B, 2/66; Intestinal ileus, A, 3/52, resolved spontaneously with conservative treatment; B, 0; Prolonged catheterization (>=4d): A, 0; B, 17.

OTHER OUTCOMES Operating time: mean (SD), h: A, 2.63 (0.59); B, 2.36 (0.61)

Length of hospital stay: mean (SD), d: A, 7.24 (2.07); B, 8.77 (3.78)

RCTS PUBLISHED AS CONFERENCE ABSTRACTS ONLY

Uterine prolapse repair Author, year Inclusion criteria: POP-Q grade III-IV INTERVENTION A (n=23): EFFICACY 29 Braun 2007 prolapse. abdominal hysterectomy + Objective failure: vault prolapse relapse, no definition, A, 0/23; B, 1/24 at sacrocolpopexy 8m after surgery. Type of prolapse: presume to be Age (mean, SD): A, 56.5; B, 56,2 uterine (NR whether all had Type of mesh/graft: Vypro mesh One severe cystocele occurred in B group 3m after surgery but NR hysterectomy or not.) recurrent or de novo. Surgical route: abdominal Publication type: conference abstract SAFETY INTERVENTION B (n=24): vaginal Mesh/graft erosion, n/N: A, 1/23 Study design: RCT hysterectomy + anteroposterior colporrhaphy +Mayo McCall stitch Other serious adverse events: incisional hernia, A, 1/23; B, 0/24 Setting: hospital Surgical route: vaginal Other adverse events: subaponeurotic hematoma, A, 1/23; B, 0/24 Country: Chile

Recruitment/Treatment dates: OTHER OUTCOMES Operating time: A, 140min (100-240); B, 90min (50-130) 122 Jun. 2003 – Oct. 2004 Length of follow-up: 33m (20-41) Length of hospital stay (mean): A, 3.8d; B, 2d Source of funding: no conflict of interest with industry

Vault prolapse repair (no RCTs (abstracts) available)

Uterine and/or vault prolapse repair (not reported separately) (no RCTs (abstracts) available)

NON-RANDOMISED COMPARATIVE STUDIES

Uterine prolapse repair Author, year Inclusion criteria: patients with INTERVENTION A (n=36): EFFICACY Costantini 200533 symptomatic uterovaginal prolapse sacrohysteropexy Objective failure: recurrent cervical/vault prolapse (<6cm above the grade III – IV. hymen): A, 0/36; B, 0/39 Type of prolapse: uterine Type of mesh/graft: non- Exclusion criteria: for uterine absorbable synthetic mesh, Marlex Further need for pelvic floor repair: A, 0; B, 0 Publication type: full-text conservation group, no post- menopausal bleeding, no previous Surgical route: vaginally Persistent bladder symptoms, n/N: incontinence, A, 4/22; B, 7/20; but all Study design: non-randomised cervical intraepithelial neoplasia, no had surgeries for incontinence. comparative study, prospective abnormal cervical smears or uterine disease including uterine enlargement or INTERVENTION B (n=39):

Setting: university hospital cervical ulceration. sacropexy + hysterectomy SAFETY Blood loss, ml, median: A, 200; B, 325 Country: Italy Number of patients lost to follow-up, Type of mesh/graft: Marlex n/N: A, 2/36; B, 1/39; all were included Blood loss requiring transfusion, n/N: A, 2; B, 2 Recruitment/Treatment dates: in the analysis. Surgical route: vaginally Jun. 1995 – Dec. 2003 Mesh/graft erosion, n/N: A, 0; B, 3 Age (mean, SD): A, 61 (12); B, 62 (8) Length of follow-up: mean, range, Parity (median, range): A, 2 (1 – 8); B, Concomitant procedures: for Further operation needed for mesh erosion, n/N: A, 0; B, 3 vaginal 51m (12 – 115) 2 (0 – 3) incontinence, A, 28; B, 30; revision BMI (median, range): A, 24.4 (18.6 – hysterectomy, A, 0; B, 39 Source of funding: NR 31.2); B, 25.4 (16 – 35.1) Wound infection: A, 0; B, 2 Incontinence: A, 22; B 20 Other serious adverse events: incisional hernia, A, 2; B, 1 Pre-operative prolapse stage: uterine prolapse, A, I – II, 20, III, 14; B, I – II, 27, Other adverse events: fever, A, 1; B, 1; III, 11. Perivesical haematoma, A, 2; B, 4; Voiding dysfunction, A, 4; B, 1 Vault/uterine repair: A, 0/36; B, 0/39 OTHER OUTCOMES Number of surgeons: NR, all surgeries Operating time: min, median (range): A, 89 (60 – 110); B, 115 (80 – 135) were performed by or under the supervision of the senior author.

Author, year Inclusion criteria: NR INTERVENTION A (n=76): total EFFICACY 123 Griffis 200645 hysterectomy + sacrocolpopexy NR. Exclusion criteria: NR Type of prolapse: uterine (another Type of mesh/graft: Prolene-soft SAFETY group was treated as case series Vault/uterine repair: A, 0/76; B, 0/28 (Ethicon, Somerville, NJ), Mersilene Mesh/graft erosion, n/N: A, 8/76; B, 1/28; [vault repair]) (Ethicon), Prolene (Ethicon), or Mean time to erosion, 10.4m, median (range): 4m (0.4-53.5) Number of surgeons: NR Atrium (Atrium Medical Co., Hudson, Publication type: full-text NH) Further operation needed for mesh erosion, n/N: surgical excision in the operation room, A, 4; B, 0; the others were managed with conservative Study design: non-randomised Surgical route: abdominal (NR treatment, which included one or all of the following: estrogen cream, comparative study, retrospective open or laparoscopic); a vaginal wall antibiotic cream, silver nitrate, and local excision in the office. incision is required. Setting: hospital INTERVENTION B (n=28): Country: US supracervical hysterectomy + sacrocolpopexy Recruitment/Treatment dates: 1 Aug. 1995 – 31 Jul. 2002 Type of mesh/graft: same as Length of follow-up: mean 13.1m, above. 95% CI, 11.7 – 14.5 Surgical route: abdominal (NR Source of funding: NR open or laparoscopic); vaginal wall incision is avoided.

Concomitant procedures: rectocele repair using biograft, A,

6/76; B, 1/28

Vault prolapse repair Author, year Inclusion criteria: patients underwent INTERVENTION A (n=24): EFFICACY Govier 200543 abdominal or laparoscopic open/laproscopic sacrocolpopexy Objective failure: recurrent vault prolapse, NR definition, A, 0/24; B, 2/21, sacrocolpopexy one was detected after mesh removal, NR timing, the other at 24m. Type of prolapse: vault Type of mesh/graft: polypropylene Exclusion criteria: immunosuppression mesh, Johnson & Johnson, New Publication type: full-text that may have placed the patient at risk Brunswick, NJ SAFETY of bone anchor complications or patients Mesh/graft erosion, n/N: A, 0/24; B, 4/21, detected at 4, 7, 12, 20m Study design: non-randomised who were not surgical candidates respectively. comparative study (retrospective) INTERVENTION B (n=21): Age (mean, range): A+B, 67y (51-86) open/laproscopic sacrocolpopexy Mesh infection: A, 0/24; B, 1/21 Setting: medical center Number of surgeons: NR Type of mesh/graft: silicone- Further operation needed for mesh erosion/infection, n/N: A, 0/24; B, Country: US covered polyethylene mesh, 5/21 of which 2 operated abdominally and 3 operated transvaginally, one American Medical Systems, mesh was removed transvaginally, four was abdominally. Recruitment/Treatment dates: Minnetonka, Minn Jan 2001 – Mar 2004 Other serious adverse events: osteomyelitis: A, 0/24; B, 0/21 Concomitant procedures: Length of follow-up (mean, Cystocele repair: A+B, 11/45 range): A, 12m (1-38); B, 23m (16- Rectocele repair: A+B, 7/45 41) Slings: A+B, 20/45 124 Source of funding: two of the authors were paid consultants to American Medical Systems

Author, year Inclusion criteria: patients with post- INTERVENTION A (n=60): EFFICACY Marcickiewicz 200758 hysterectomy vault prolapse laparoscopic sacral colpopexy Subjective failure: sensation of prolapse: A, 13/60; B, 6/51

Type of prolapse: vault Exclusion criteria: NR Type of mesh/graft: Prolene mesh, Objective failure: vault prolapse >= grade II (Baden-Walker): A, 0; B, 0; polypropylene, Ethicon, Johnson & there were 5 grade I in A, and 3 grade I in B. Publication type: full-text Age (median, range): A, 58 (30-83); B, Johnson 66 (43-88) New prolapse at other sites, n/N: New cystocele: A, 10/32; B, 4/28 Study design: non-randomised Parity (median, range): A, 2.1 (0-5); B, Recurrent cystocele: A, 5/28; B, 10/23 comparative study, retrospective 2.3 (0-6) INTERVENTION B (n=51): vaginal BMI (median, range): A, 24.1 (18-27); sacrospinous colpopexy Further operation needed for prolapse (recurrent or new at other Setting: university hospital B, 27.9 (21-36.4) sites), n/N: all types of recurrence and de novo: A, 15/60 (occurred in the initial period, in 6 of these patients, the net (mesh) was attached to the Country: Sweden Number of surgeons: 4 experienced Concomitant procedures: vault with staples. These patients underwent another surgery A within a gynaecologists, 2 of whom performed all Burch/TVT/stamey: A, 6/60; B, 8/51; short time from the initial operation); B, 4/51 Recruitment/Treatment dates: surgery A, the other 2 performed all Anterior wall repair: A, 4/60; B, Jan 1994 – Dec 2002 surgery B. The vaginal surgeons had 14/51; SAFETY Length of follow-up: mean, range, performed >20 surgery B before the Posterior repair: A, 19/60; B, 20/51 Blood loss, ml, median (range): A, 64 (0-1000); B, 61 (0-400) A, 33.6m (13-60); B, 38.4m (7-108) onset of the study. The leaning curve of the laproscopic surgeons performing A Blood loss requiring transfusion, n/N: A, 1/60; B, 2/51 Source of funding: NR was partly included in the study.

Damage to surrounding tissue, e.g. urethra, bladder, bowel: bowel injury, A, 2/60; B, 2/51

Urinary tract infection: A, 3/60; B, 4/51

Other serious adverse events: Laparotomies due to perioperative complications, A, 3 (1 puncture of the inferior epigastric artery, 1 colon perforation, 1 severe adhesion); B, 0

OTHER OUTCOMES Operating time: median, min: A, 129; B, 62

Length of hospital stay: mean (range), d: A, 4.0 (2-21); B, 3.7 (1-18)

Author, year Inclusion criteria: patients with post- INTERVENTION A (n=56): EFFICACY Paraiso 200564 hysterectomy vaginal prolapse laparoscopic sacral colpopexy No preoperative and postoperative POP-Q examinations.

Type of prolapse: vault Exclusion criteria: NR Type of mesh/graft: solvent SAFETY dehydrated cadaveric fascia lata Blood loss, ml, mean (SD), range: A, 172 (166), 0-1100; B, 234 (149), Publication type: full-text Number of patients lost to follow-up, (Tutoplast, Mentor Corp., Santa 50-700 n/N: only mentioned many patients did Barbara, California), n=1; and Study design: non-randomised not return after their first postoperative polypropylene (Ethicon, Inc., Blood loss requiring transfusion, n/N: A, 1/56; B, 1/61 comparative study, retrospective visit. Cincinnati, Ohio) Damage to surrounding tissue, e.g. urethra, bladder, bowel: 125 Setting: hospital Age (mean, SD, range): A, 62 (11), 39- INTERVENTION B (n=61): cystotomy/bladder sutures (may be due to aggressive dissection of the 85; B, 61 (9.5), 35-81 abdominal (open) sacral colpopexy bladder for lower placement of the anterior mesh): A, 6/56; B, 2/61; Country: US Parity (median, range): A, 3 (0-10); B, Enterotomy: A, 1/56; B, 2/61 3 (0-9) Type of mesh/graft: save as above/ Recruitment/Treatment dates: BMI (mean, SD, range): A, 26.3 (6.6), Mesh/graft erosion, n/N: A, 2/56; B, 1/61 Late 1998 – 2003 13.4-55.3; B, 26.2 (4.5), 17.7-35.5 Length of follow-up: mean, SD, Concomitant procedures: Further operation needed for mesh erosion, n/N: A, 2 (1 vaginally, 1 range, A, 13.5 (12.1), 1-46; B, 15.7 Number of surgeons: >=1. All Burch or suburethral sling: A, 24/56; laparoscopically) B, 1 (vaginally) (18.1), 1-73 surgeons were in their learning curve of B, 38/61 laparoscopic sacral colpopexy at the Anterior repair, A, 5; B, 4; Would infection: A, 6/56; B, 2/61 Source of funding: NR time the patients in the laparoscopic Posterior repair: A, 23; B, 19; cohort underwent surgery. Rectopexy, A, 4; B, 8 Other serious adverse events: DVT, A, 1/56; B, 1/61; cardiac complications, A, 0/56; B, 2/61 (1 atrial fibrillation and mitral valve disease, 1 atrial fibrillation with heart failure); Ventral hernia, A, 1/56; B, 2/61.

Other adverse events: conversion to open technique, A, 1/56; B, NA; Ileus, A, 0/56; B, 2/61; Small bowel obstruction, A, 1/56; B, 2/61;

OTHER OUTCOMES Operating time: mean, SD, range, min, A, 269 (65), 150-467; B, 218 (60), 45-372.

Length of hospital stay: mean, SD, range, d, A, 1.8 (1.0), 1-6; 4.0 (1.8),

2-12.

Uterine and/or vault prolapse repair Author, year Inclusion criteria: patients with vaginal INTERVENTION A (n= 25): EFFICACY Altman 200521 vault prolapse stage II or worse (Baden- abdominal sacrocolpopexy with a Objective failure: vaginal vault prolapse stage II (Baden-Walker, i.e. Walker staging) synthetic mesh prolapse extending from the midplane to the hymenal ring): A, 6/25 in a Type of prolapse: uterine/vault mean of 7.4m follow up; B, 8/27 in a mean of 7.1m follow up. Exclusion criteria: NR Type of mesh/graft: 12 Publication type: full-text polypropylene and 13 SAFETY Age (mean, range): A, 65.9 (54-83); B, polytetrafluoroethylene (NR trade Blood loss, ml, mean: A, 162; B, 150 Study design: non-randomised 68.6 (55-84) name). The use of material was comparative study (retrospective) random, depending on the supply. Mesh/graft erosion, n/N: A, 0/25; B, 0/27 Parity (mean, range): A, 2.3 (0-4); B, Setting: hospital 2.4 (0-6) Surgical route: abdominal Infection: wound infection: A, 0/25; B, 3/27; cystitis: A, 2/25; B, 5/27; fever >3d after surgery: A, 1/25; B, 8/27; fever 4d to 1w after surgery: A, 0/25; B, Country: Sweden Number of surgeons: NR INTERVENTION B (n=27): 4/27. abdominal sacrocolpopexy with a Recruitment/Treatment dates: porcine collagen graft Other serious adverse events: re-admission: A, 0/25; B, 4/27 (NR 1998-2002 reasons) Type of mesh/graft: Pelvicol, CR Length of follow-up: mean, Bard, Marray Hill, NJ Other safety outcomes: vaginal vault hematoma: A, 1/25; B, 1/27; wound clinical follow-up: A, 7.4m; B, 7.1m; hematoma: A, 0/25; B, 1/27. questionnaire follow-up: A, 4.3y; B, Surgical route: abdominal 126 2.5y OTHER OUTCOMES Operating time, h: A, 1.48h; B, 1.74h Source of funding: hospital Concomitant procedures: administered research funds Posterior colporraphy: A, 6/25; B, Length of hospital stay, d: A, 4.1d; B, 3.9d 5/27; Anterior colporraphy: A, 1/25; B, 0/27; Hysterectomy: A, 2/25; B, 5/27

Author, year Inclusion criteria: women diagnosed as INTERVENTION A (n=20): EFFICACY, 1y Bai 200622 pelvic organ prolapse and managed by abdominal sacral colpopexy Objective failure: POP-Q >= grade II, A, 3/20; B, 6/54, but NR apical, either abdominal high uterosacral anterior, and posterior prolapse separately. Type of prolapse: uterine/vault colpopexy or abdominal sacral Type of mesh/graft: non- colpopexy with Mersilene mesh. absorbable synthetic, Mersilene Publication type: brief (Ethicon, Somerville, NJ) SAFETY, 1y communication Exclusion criteria: NR Mesh/graft erosion, n/N: A, 4/34; B, NA Surgical route: abdominal (NR Study design: non-randomised Number of patients lost to follow-up, open or laparoscopic) Conservative treatment for mesh/graft erosion, n/N: 2 were treated comparative study, retrospective n/N: A, 14/34; B, 25/79, no data with estrogen available at 1y after surgery. ‘Mesh inlay’ or ‘total mesh’: Setting: university hospital Further operation needed for mesh erosion, n/N: 2 pieces of mesh Pre-operation prolapse stage: were removed transvaginally Country: South Korea INTERVENTION B (n=54): high A, n=20 B, n=54 uterosacral colpopexy (no mesh) Recruitment/Treatment dates: Stage II 0 14

Apr. 2000 – Mar. 2003 Stage III 6 40 Surgical route: abdominal (NR Stage IV 14 0 open or laparoscopic) Length of follow-up: 1y results available Number of surgeons: NR Additional information: this technique enables the vaginal apex Source of funding: NR to be fixed proximal to the sacrum, restores normal vaginal length.

Concomitant procedures: total abdominal hysterectomy, hysterectomy, culdoplasty, paravaginal repair, anterior colporrhaphy, posterior colporrhaphy was performed if needed.

Author, year Inclusion criteria: patients underwent INTERVENTION A (n=24): 19 EFFICACY Begley 200525 abdominal sacrocolpopexy laparotomy/5 laparoscopic Objective failure: vaginal apex prolapse, NR criteria: A, 0/24; B, 1/33; C, sacrocolpopexy 2/21; D, 0/14; failure were detected at 5, 8, 12, and 16.2m respectively. Type of prolapse: uterine/vault Exclusion criteria: NR Type of mesh/graft: monofilament SAFETY Publication type: full-text Age (mean, SD, range): A, 68.5 (11), polypropylene mesh, pores >75μm, Mesh/graft erosion, n/N: A, 0/24; B, 3/33, average interval to detection, 49-86; B, 66.1 (11), 37-84; C, 66.0 (13), Marlex, Bard or Prolene, Ethicon 5m, range 2.7-8.4; C, 4/21; average interval to detection, 11m, range 4.1- Study design: non-randomised 40-85; D, 63.0 (15), 25-83 21.6; D, 0/14. 127 comparative study (retrospective) INTERVENTION B (n=33): Primary/secondary repair: A+B: 85/7 laparotomy sacrocolpopexy Further operation needed for mesh erosion, n/N: A, 0/24; B, 3/33, all Setting: a medical centre partial graft resection and all were successful via transvaginally; C, 4/21, Number of surgeons: more than 1. NR Type of mesh/graft: multifilament partial or complete removal, more complicated, 3 via laparotomy and 1 Country: US experience. polypropylene, interstices <10 μm, transvaginally; D, 0/14. Gore-Tex, WL Gore Recruitment/Treatment dates: 1997-2003 INTERVENTION C (n=21): 18 laparotomy/3 laparoscopic Length of follow-up (mean): A, sacrocolpopexy 9.8m; B, 29.3m; C, 15.5m; D, 18.6m Type of mesh/graft: polyester backbone covered with a thin layer Source of funding: NR of solid silicone elasomer (American System Triangle silicone-coated mesh, American system)

INTERVENTION D (n=14): laparotomy sacrocolpopexy

Type of mesh/graft: 1 autologous and 13 cadaveric fascia lata, Tutoplast

Concomitant procedures:

Anterior and/or posterior colporrhaphy: A, 2/24; B, 8/33; C, 7/21; D, 0/14; Pubovaginal sling (fascia or mesh): A, 5/24; B, 6/33; C, 2/21; D, 1/14; CaPs: A, 1/14; B, 1/33; C, 2/21; D, 1/14; Burch suspension: A, 3/24; B, 7/33; C, 5/21; D, 3/14; Abdominal paravaginal repair: A, 0/24; B, 0/33; C, 5/21; D, 1/14; Total abdominal hysterectomy: A, 1/24; B, 8/33; C, 3/21; D, 7/14.

Author, year Inclusion criteria: patients underwent INTERVENTION A (n=19): EFFICACY Gregory 200544 abdominal sacrocolpopexy abdominal (laparotomy) Subjective failure: women with ‘optimal’ surgical outcome (NR definition Type of prolapse: uterine/vault sacrocolpopexy for ‘optimal’): A, 17/19; B, 11/18 Exclusion criteria: NR Publication type: full-text Type of mesh/graft: synthetic Objective failure: total vaginal length changes, i.e. relative vaginal Number of patients lost to follow-up, mesh, 94% Marlex (CR Bard USA) descent, mean (SD): A, 1.1 (1.2) cm (n=19); B, 3.0 (3.8) cm (n=18), there Study design: non-randomised n/N: A, 30/49; B, 15/33; reasons include and 6% Mersilene (Ethicon, US) were two patients in this group with changes >10cm. comparative study (patients were driving distance, unrelated death, lack of recruited retrospectively, some transportation, lack of time, and in one INTERVENTION B (n=18): SAFETY outcome data were collected patient, disappointed with outcome of abdominal (laparotomically) Blood loss, ml, mean (SD): A, 276 (153), n=49; B, 298 (162), n=33 128 prospectively) surgery sacrocolpopexy Mesh/graft erosion, n/N: A, 0/19; B, 0/18 Setting: University hospital Age (mean): A, 66.5 (n=49); B, 66.2 Type of mesh/graft: fascia lata, (n=33) lyophilized, gamma-irradiated, from Country: US Community Tissue Services OTHER OUTCOMES Number of surgeons: 2 senior (Portland, Ore, US) Operating time, h, mean (SD): A, 3.5h (1.2), n=49; B, 4.5h (1.5), n=33; Recruitment/Treatment dates: surgeons, NR experience authors admitted that the longer operative time in group B was due to May 1995 – May 1999 Concomitant procedures: concomitant procedures. Burch: A, 28/49; B, 22/33 Length of follow-up (mean, Hysterectomy: A, 5/49; B, 8/33 range): A, 26.3m (10-48); B, 21.2m Ant. colporrhaphy: A, 3/49; B, 3/33 (11-34) Post. Colporrhaphy: A, 22/49; B, 11/3 Other information: the choice of graft material was determined by material availability: synthetic mesh was used before Jul 1997, fascia lata became available after Jul 1997 until May 1999.

Author, year Inclusion criteria: NR INTERVENTION A (n=80): EFFICACY Hardiman 199647 abdominal sacrocolpopexy Objective failure: recurrent vault prolapse (NR criteria): A, 1/80; B, 3/125 Exclusion criteria: NR Type of prolapse: uterine/vault Type of mesh/graft: polypropylene, SAFETY Age (mean): A, 60.8; B, 63.7 Marlex, Bard/Davol, Cranston, RI Blood loss, ml, mean, range: A, 745 (100-5000); B, 567 (100-2700)

Publication type: full-text Parity (mean, range): A, 2.62 (0.7); B, 2.69 (0-12) Blood loss requiring transfusion, n/N: A, 1/80 (hemorrhage from the Study design: non-randomised INTERVENTION B (n=125): vaginal presacral vein, 5000ml): B, 0/125 comparative study (NR prospective Number of surgeons: 1 sacrospinous vault suspension or retrospective) Mesh/graft erosion, n/N: A, 0/80; B, 0/125 Additional information: it was Setting: University hospital performed with the Miya hook: the Infection: urinary tract infection: A, 2/80; B, 10/125; febrile morbidity: A, left and right vaginal angles were 5/80; B, 13/125 Country: Canada suspended to the sacrospinous ligament using hook and sutures. Other serious adverse events: thrombo-embolism: A, 0/80; B, 0/125 Recruitment/Treatment dates: Jan 1988 – Oct 1993 Concomitant procedures: Other events: would complications: A, 4/80; B, 4/125 Continence procedures (Burch Length of follow-up: mean: A, suspension/suburethral 3.9y; B, 2.2y; A+B, range 6m to 5y sling/modified suburethral sling/combined abdominal-vaginal Source of funding: NR sling/Repstein procedure): A, 76/80; B, 116/125; Abdominal/vaginal hysterectomy: A, 18/80; B, 45/125; Posterior colporrhaphy: A, NR; B, 125/125; Anterior colporrhaphy: A, 6/80; B, 118/125 129 Author, year Inclusion criteria: NR INTERVENTION A (n=25): EFFICACY Hsiao 200751 laparoscopic sacrocolpopexy Objective failure: recurrent apical prolapse (no definition), A, 0/25; B, Exclusion criteria: NR 1/22 (allograft cadaveric fascia was used) Type of prolapse: uterine/vault Type of mesh/graft: allograft fascia Age (mean, SD): A, 66.0; B, 71.1 and autologous fascia lata, 0 in A, 2 SAFETY Publication type: full-text Parity (mean): A, 3.3; B, 3.6 in B; polypropylene (Ethicon, Blood loss, ml, mean (range): A, 82.7 (20-200); B, 195 (100-500) Pre-operative prolapse grade, Baden- Sommerville, NJ, 19 in A, 13 in B; Study design: non-randomised Walker (mean): A, 3.4; B, 3.4 silicon mesh, 3 in A, 5 in B; Gore- Blood loss requiring transfusion, n/N: A, 0/25; B, 0/22 comparative study, retrospective Tex, 0 in A, 2 in B. Vault/uterine repair: A, 23/2; B, 19/3 Mesh/graft erosion, n/N: mesh extrusion, A, 3/25 (all silicone mesh); B, Setting: university hospital 2/22 (1 silicone mesh, 1 Gore-TeX) Number of surgeons: NR INTERVENTION B (n=22): Country: US abdominal (open) sacrocolpopexy Further operation needed for mesh erosion, n/N: mesh removal: A, 3; B, 2 Recruitment/Treatment dates: Type of mesh/graft: see above. Aug. 1999 - Oct. 2004 Other adverse events: prolonged postoperative ileus, A, 0; B, 3/22; Length of follow-up: mean, A, pneumonia, A, 1; B, 0 5.9m; B, 10.0m Concomitant procedures: Cystocele repair, A, 2/25; B, 4/22; OTHER OUTCOMES Source of funding: NR Rectocele repair: A, 1/25; B, 4/22; Operating time: median (range), A, 219.9min (142-331); B, 185.2min Transvaginal sling: A, 9/25; B, 10/22 (195-278)

Length of hospital stay: mean (range), A, 1.2 (1-5); B, 3.3 (2-8).

Others: learning curve of laparoscopic technique, mean time, the 1st five patients, 291.8min; the 2nd five patients, 256min; the 3rd five patients, 174min; the last 5 patients, 173.4min.

Author, year Inclusion criteria: women with at least INTERVENTION A (n=113): EFFICACY Ng 200461 grade 4 uterovaginal prolapse or grade 3 abdominal (open) sacrocolpopexy Subjective failure: symptomatic: A, 2/60; B, 11/48 vault prolapse Type of prolapse: uterine/vault Type of mesh/graft: Objective failure: vault prolapse >= grade II: A, 3/60; B, 13/48 Exclusion criteria: NR polytetrafluoroethylene, Gore-tex Publication type: full-text (Gore Creative Technologies Number of patients lost to follow-up, Worldwide, Flagstaff, Arizona, US) SAFETY Study design: non-randomised n/N: 53/113; B, 16/64 Blood loss, ml, mean (range): A, 557 (50-2700); B, 239 (5-700). comparative study, retrospective Mentioned the differences were due to contaminant procedures. Age (mean): A, 60; B, 63.1 INTERVENTION B (n=64): vaginal Setting: hospital Parity (mean): A, 4; B, 3 sacrospinous ligament fixation Blood loss requiring transfusion, n/N: A, 13/113; B, 2/64 BMI (mean): A, 24.6; B, 24.1 Country: Singapore Pre-operative prolapse stage: grade Damage to surrounding tissue, e.g. urethra, bladder, bowel: tissue IV uterovaginal prolapse, A, 78; B, 37; Concomitant procedures: injury: A, 1/113 to bladder; B, 1/64 to rectal mucosa. Recruitment/Treatment dates: grade III or IV vault prolapse, A, 35; B, Burch suspension/TVT: A, 27/1; B, Mar. 1998 – Dec. 2001 27. 0/12; Mesh/graft erosion, n/N: mesh/suture erosion: A, 1/60; B, 5/48 Length of follow-up: mean, range, Anterior repair: A, 0, B, 44; A, 18.1m (0.9-48.1); B, 13.2 (1.1- Vault/uterine repair: A, 35/78; B, 27/37 Posterior repair: A, 8; B, 52; New urinary symptoms: genuine stress incontinence, A, 4/89; B, 0/52 29.1) Total abdominal hysterectomy: A, Number of surgeons: >=1. The surgery 78; B, 0; Infection: urinary tract infection, A, 12/113; B, 7/64; 130 Source of funding: NR was performed or supervised by the Vaginal hysterectomy: A, 0; B, 38. Wound infection, A, 6/113; B, 0/64; same urogynaecologist to reduce inter- Pyrexia (>37.5C), A, 59/113; B, 18/64 operator errors. Other serious adverse events: DVT, A, 1/113 in the leg; B, 0/64; Myocardial infarct, A, 1; B, 0; Congestive heart failure: A, 1; B, 0; Incisional hernia: A, 4; B, 0;

Other adverse events: pelvic haematoma: A, 1; B, 1 Wound breakdown: A, 2; B, 1; Haematuria: A, 29; B, 3. The difference was a result of the bladder being pushed medially in order to perform a Burch colposuspension or paravaginal cystocele repair. Subacute intestinal obstruction: A, 1; B, 0.

OTHER OUTCOMES Operating time: mean (range), min: A, 133 (32-227); B, 78 (28-56); the difference is because of the technically more difficult nature of abdominal sacrocolpopexy compared to vaginal sacrospinous fixation.

Length of hospital stay: mean (range), d: A, 6 (2-28); B, 4 (2-14)

Author, year Inclusion criteria: women with vaginal INTERVENTION A (n=56): EFFICACY Sze 199973 vault prolapse and coexisting or abdominal (open) sacrocolpopexy + Objective failure: NR. potential stress incontinence Burch suspension

Type of prolapse: uterine/vault Mentioned prolapse to or beyond the hymen: A, 9/47 (5 were anterior and Exclusion criteria: NR Type of mesh/graft: NR 4 were posterior defects); B, 18/54 (13 of which were anterior defects, but Publication type: full-text NR the other 5). Number of patients lost to follow-up, Study design: non-randomised n/N: A, 9/56; B, 7/61 INTERVENTION B (n=61): vaginal comparative study, retrospective sacrospinous ligament fixation + SAFETY Age (mean, SD, range): A, 57.2 (9.2), needle suspension Blood loss requiring transfusion, n/N: A, 0/56; B, 2/61 Setting: university hospital 34-74; B, 63.7 (8.9), 43-76 Parity (mean, SD, range): A, 3.3 (1.8), Concomitant procedures: Febrile morbidity: A, 7/56; B, 5/61 Country: US 0-7; B, 3.3 (1.9), 0-9 Hysterectomy: A, 7; B, 19; Anterior and/or posterior repair: A, Other serious adverse events: ileus: A, 2/56; B, 0/61. Recruitment/Treatment dates: Vault/uterine repair: A, 35/19; B, 40/7 49; B, 61 1989-1996 Other adverse events: transient right gluteal pain: A, 0/56; B, 5/61; Length of follow-up: mean (SD), Number of surgeons: 1 Phlebitis: A, 1/56; B, 0/61; range: A, 23.1 (12.6), 4-51; B, 24.0 (15), 7-72. Women with objective OTHER OUTCOMES follow up for less than 3m Length of hospital stay: mean, d: A, 3.4 (0.9), 2-7; B, 4.7 (1.1), 3-8 postoperatively were excluded from the study.

Source of funding: NR

Author, year Inclusion criteria: NR INTERVENTION A (n=155): EFFICACY Visco 200177 abdominal sacral colpopexies NR 131 Exclusion criteria: NR Type of prolapse: uterine/vault INTERVENTION B (n=88): SAFETY Age (mean, SD, range): in total: 60.6 abdominal sacral colpoperinopexy Blood loss, ml, mean (SD): in total, N=273, 460ml (390) Publication type: full-text (10.6), 31-84 Additional information: the Mesh/graft erosion, n/N: overall mean time to erosion: 12.3 (16.7)m Study design: non-randomised Parity (median, range): in total: 2 (0- procedure was defined as: a vault comparative study (retrospective) 13) suspension in which the synthetic A, 5/155, median time to erosion: 15.6m (2-33); mesh is attached to the upper third B, 4/88, 12.4m (3-40); Setting: University medical centre Number of surgeons: more than 1 of the vagina and suspended to the C, 4/25, 9.0m (5-15); (member of the Division of Gynecologic sacrum. D, 2/5, 4.1m (2-7) Country: US specialties) INTERVENTION C (n=25): Further operation needed for mesh erosion, n/N: in total, 12/15 were Recruitment/Treatment dates: combined vaginal and abdominal treated successfully with a single transvaginal outpatient excision, 2/15 1 Mar 1992 – 28 Feb 1999 colpoperineopexy with vaginal suture required two vaginal procedures, 1/15 required two transvaginal excisions, passage a transabdominal resection, and anterior transvaginal excision. Length of follow-up, median (SD): in total, 5.8m (1-87); A, 6.5m INTERVENTION D (n=5): (1-87); B, 4.9m (1-45); C, 6.1m (1- combined vaginal and abdominal 28); D, 6.6m (2-11) colpoperineopexy with vaginal mesh placement Source of funding: NR Additional information: the procedure was defined as: combined abdominal-vaginal colpoperineopexy procedure included a vaginal

dissection of the posterior vaginal wall. In intervention C, permanent sutures were attached to the perineal body and brought into the abdominal operative field through the posterior cul-de-sac with a long needle or ring carrier. The mesh was then attached to these sutures transabdominally without the mesh entering the vaginal operative field; in intervention D, the mesh was placed and attached vaginally directly to the perineal body with permanent sutures and brought into the abdominal operative field for attachment to the sacrum.

Type of mesh/graft, for intervention A, B, C, D: non- absorbable synthetic mesh, Mersilene (Ethicon, Somerville, NJ), N=269, or Gore tex, N=4

Concomitant procedures: in total: 132 Hysterectomy: 45/273 Posterior repair: 84/273

Author, year Inclusion criteria: NR INTERVENTION A (n=92): EFFICACY Young 200480 colposacropexies (95 began NR Exclusion criteria: NR laparoscopically but 2 converted to Type of prolapse: uterine/vault an open procedure, NR the 3rd). Age (mean, range): A, 61 (28-84); B, 65 SAFETY Publication type: full-text (37-87); C, 64 (34-84) Type of mesh/graft: autologous Blood loss, ml, mean (range): A, 40 (10-300); B, 182 (10-300); C, 280 fascia, n=8; cadaver/donor fascia (40-1000) Study design: non-randomised Number of surgeons: >=1 lata, n=7; Mersilene mesh, n=39; comparative study, prospective Prolene, n=24; Gore-Tex, n=2; Blood loss requiring transfusion, n/N: excessive hemorrhage, A, 3/92; registry Prolene with donor fascia, n=1; no B, 2/187; C, NA synthetic material was used n=3; Z- Setting: university hospital plasty was performed to reach the Damage to surrounding tissue, e.g. urethra, bladder, bowel: sacrum, n=2; the vagina reached the Enterotomy: A, 1/92; B, 0/187; C, 1/70; Country: US sacrum, n=1; material not specified, Proctotomies, A, 0/92; B, 4/187; C, NR; n=8. Cystotomy: A, 0/92; B, 1/187; C, NR; Recruitment/Treatment dates: Small bowel serosal tear: A, 1/92; B, 0/187; C, NR; 1 Oct. 1998 – 30 Sep 1999 Surgical route: abdominal (open) Rectal laceration: A, 0/92; B, 0/187; C, 1/70 Length of follow-up: NR, safety n=85; laparoscopic n=5. registry Other serious adverse events: incisional hernia, A, 1/92

Source of funding: NR INTERVENTION B (n=187): Other adverse events: sacrospinous fixation Hematoma, A, 1/92;

Additional information: a survey Small bowel obstruction, A, 1/92; to 147 members of the Society of Surgical route: vaginal Ureteral obstruction/bilateral kinked ureters that required stent placement, Gynecologic Surgeons in the US. C, 1/70; 21 members participated, 349 Additional information: right Buttock pain, B, 2/187; completed survey forms were sacrospinous ligament was used 169 Sciatic pain, B, 1/187; received. times, left 6 times, bilateral 5 times, Foot drop, B, 1/187; unspecified 7 times. Small bowel obstruction requiring surgery, A, 2/92.

INTERVENTION C (n=70): high OTHER OUTCOMES uterosacrosuspensions (no mesh) Operating time: mean, range, min For individual colpopexy procedures: A, 35 (30-180); B, 40 (30-60); C, NA Surgical route: vaginal For entire cases including other procedures: A, 218 (120-460); B, 182 (50- 502); C, 118 (45-240) Concomitant procedures: Needle suspension/sling/Burch suspension: A, 44/92; B, 28/187; C, 8/70; Hysterectomy: A, 6/92; B, 63/187; C, 17/70; Anterior colporrhaphy: A, 7/92; B, 146/187; C, 51/70; Posterior colporrhaphy: A, 48/92; B, 63/187; C, 64/70; Paravaginal repair: A, 48/92; B, 133 63/187; C, 1/70.

Additional information: the apex is suspended very high on the uterosacral ligaments CASE SERIES (n≥100)

Uterine prolapse repair Author, year Inclusion criteria: women underwent INTERVENTION A (n=101): EFFICACY Wu 200679 abdominal sacral colpopexy hysterectomy + abdominal (open) NR Type of prolapse: uterine sacrocolpopexy Publication type: full-text Exclusion criteria: patients who Study design: case series, underwent a repeat or laparoscopic Type of mesh/graft: Mersilene or SAFETY retrospective (treated as 2 case sacral colpopexy polypropylene mesh, or Gore-Tex Blood loss, ml, mean (SD): 328 (163) series) mesh Setting: university hospital Age (mean, SD): 64.4 (11.8) Mesh/graft erosion, n/N: 7/101, median time to occur was 8.4m (range Country: US Parity (mean, SD): 2.7 (1.5) 1.4 – 13.1) Recruitment/Treatment dates: BMI (mean, SD): 27.2 (5.1) Jun 1994 – Dec 2004 OTHER OUTCOMES Length of follow-up: mean, SD, Primary/secondary repair: 101/0 Length of hospital stay: mean (SD), 3.5 (1.9) range, 14.5 (23.6), 0.2-119.5 Source of funding: NR Number of surgeons: NR

Vault prolapse repair Author, year Inclusion criteria: patients with prior INTERVENTION A (n=149): EFFICACY Fedorkow 199340 hysterectomy and undergone abdominal abdominal sacrovaginopexy NR sacrovaginopexy Type of prolapse: vault Type of mesh/graft: Prolene Exclusion criteria: NR SAFETY Publication type: full-text Surgical route: NR open or Intraoperative hemorrhage: 0/149 Age (mean, SD): 58.4 (11.4) laparoscopic Study design: case series, Parity (mean, SD): 3 (2.0) Mesh/graft erosion, n/N: 0/149 prospective (was a non-randomised comparative study, only the vault Vault/uterine: 149/0 Further operation needed for mesh erosion, n/N: 0/149 repair group was data abstracted) Number of surgeons: >=1 Infection: urinary tract infection, 22/149; Setting: tertiary care gynecologic Wound infection, 2/149; urology clinic Febrile morbidity: 8/149, occurred in 5d; Pelvic abscess, 0/149; Country: Canada Pneumonia, 4/149; Delayed sepsis: 0/149; Recruitment/Treatment dates: 1 Jan. 1986 – 30 Jul. 1991 Other adverse events: pelvic hematoma, 0/149. Length of follow-up: NR. The article was received for publication in Oct. 1992. 134 Source of funding: NR

Author, year Inclusion criteria: all had prior INTERVENTION A (n=196): EFFICACY Griffis 200645 hysterectomy sacrocolpopexy NR.

Type of prolapse: vault (another Exclusion criteria: NR Type of mesh/graft: Prolene-soft SAFETY two groups were treated as non- (Ethicon, Somerville, NJ), Mersilene Mesh/graft erosion, n/N: 16/196 randomised comparative studies Vault/uterine repair: 196/0 (Ethicon), Prolene (Ethicon), or Mean time to erosion, 10.4m, median (range): 4m (0.4-53.5) [uterine repair]) Atrium (Atrium Medical Co., Hudson, Number of surgeons: NR NH) Further operation needed for mesh erosion, n/N: surgical excision in Publication type: full-text the operation room, 5; the others were managed with conservative Surgical route: abdominal (NR treatment, which included one or all of the following: estrogen cream, Study design: non-randomised open or laparoscopic); a vaginal wall antibiotic cream, silver nitrate, and local excision in the office. comparative study, retrospective incision is required.

Setting: hospital Concomitant procedures: rectocele repair using biograft, Country: US 14/196

Recruitment/Treatment dates: 1 Aug. 1995 – 31 Jul. 2002 Length of follow-up: mean 13.1m, 95% CI, 11.7 – 14.5

Source of funding: NR

Author, year Inclusion criteria: women who had INTERVENTION A (n=103): EFFICACY Higgs 2005a50 laparoscopic sacrocolpopexy at St. laparoscopic sacrocolpopexy Subjective failure: presence of prolapse symptoms (vaginal lump, vaginal Mary’s hospital, Manchester for vault pain, or discomfort), measured by a questionnaire which was not Type of prolapse: vault prolapse. Type of mesh/graft: Prolene validated, ‘’no change or worse’’, 22/103. (polypropylene, Ethicon, Publication type: full-text Exclusion criteria: NR Sommerville, New Jersey, US) Objective failure: recurrent vault prolapse, POP-Q, 4/66 (3 stage II, 1 stage III). Study design: case series, Number of patients lost to follow-up, Concomitant procedures: retrospective n/N: 37/140 not contactable; in the 103 laparoscopic colposuspension, who were followed up, 66 had physical 39/103; SAFETY Setting: a tertiary urogynaecology examination, a further of 37 filled in Cystocele repair and rectocele Damage to surrounding tissue, e.g. urethra, bladder, bowel: bladder unit in the North West of England questionnaires only because they could repair, with or without mesh, 20/103; perforation, 2/103, both occurred during colposuspension; bowel not come to clinic. Paravaginal repair, 20/103; perforation, 2/103 Country: UK rectopexy, 4/103. Age (mean, SD): 58 (10.2) Mesh/graft erosion, n/N: 9/103, in all cases, Prolene (polypropylene, Recruitment/Treatment dates: Parity (mean, SD): 2.7 (1.3) Additional information: the surgical Ethicon, Sommerville, New Jersey, US) mesh was used; all presented 1993-1999 technique used for laparoscopic >=6m after surgery, with some occurring over 3y later. Length of follow-up: median, Vault/uterine repair: 103/0 sacrocolpopexy evolved over time. range, 66m (37-124) When the data from the early Further operation needed for mesh erosion, n/N: 5/103 (1 complete Number of surgeons: >=1. Surgery technique group and more recent excision laparoscopically, 2 vaginal excisions under general anaesthesia, Source of funding: NR was performed by the senior author or a technique group are compared, the 2 vaginal excisions); 3 asymptomatic women declined to have trimming of senior urogynaecology fellow with the analysis showed little difference in the mesh; NR the 9th one who had mesh erosion. 135 senior author in attendance. the outcomes. Infection: urinary tract infection, 2/103

Other adverse events: wound haematoma, immediate post-operative 5/103; Hypercarboxaemia (overnight ICU), 1/103; Hyponatremia (overnight ICU), 1/103; Sciatica requiring physiotherapy, 3/103.

OTHER OUTCOMES Operating time: median, range, min: 145 (98-265); when laparoscopic sacrocolpopexy was performed alone, the duration was 107 (62-170)

Length of hospital stay: median, range, 3.8d (1-12) Uterine and/or vault prolapse repair Author, year Inclusion criteria: women underwent INTERVENTION A (n=121): EFFICACY Bensinger 200526 abdominal sacral suspension with sacrocolpopexy NR polypropylene mesh Type of prolapse: uterine/vault Type of mesh/graft: non- SAFETY Exclusion criteria: patients with a absorbable synthetic mesh, Blood loss requiring transfusion, n/N: 2/115 Publication type: full-text previous suspension procedure or a polypropylene sacral colpoperineopexy Damage to surrounding tissue, e.g. urethra, bladder, bowel: Study design: treated as case Surgical route: abdominal (NR Cystotomy, 2/121; small bowel laceration, 1/121 series, retrospective (contain 3 Number of patients lost to follow-up, open or laparoscopic)

groups: A, sacrocolpopexy + n/N: 57/121 in a year. Mesh/graft erosion, n/N: 4/62, 3 of which occurred between 5-10m, and supracervical hysterectomy; B, Concomitant procedures: the 4th one occurred 2y postoperatively. sacrocolpopexy + total Age (mean, SD): 53.3 (9.6) Burch urethropexy: 72; hysterectomy, C, sacrocolpopexy in Parity (mean, SD): A, 2.6 (1.3); B, 2.8 Paravaginal repair: 47; Further operation needed for mesh erosion, n/N: 3 had simple women had prior hysterectomy) (1.3); C, 2.6 (1.3) Enterocele repair: 98; excision, 1 had abdominal resection of the mesh. Posterior repair: 40; Setting: university hospital Vault/uterine repair: 35/86 Anterior repair: 7; Infection: 1/115; Sling (NR details): 10 Febrile morbidity: 11/115. Country: US Primary/secondary repair: 121/0 Other adverse events: Recruitment/Treatment dates: Pre-operation prolapse stage: Partial SBO/ileus: 4/115; Mar. 1997 – Jul. 2004 Stage I or II: 3; stage III or IV: 118 Persistent vaginal discharge, 6m: 3/64; Vaginal bleeding, 6m: 1/64 Length of follow-up: mean (SD), Number of surgeons: NR range, 12.5m (15.2), 0.3 – 63.3

Source of funding: NR

Author, year Inclusion criteria: all cases treated by INTERVENTION A (n=124): EFFICACY Brizzolara 200330 the 2nd author. sacrocolpopexy Objective failure: vault prolapse (no definition): 3/124 (A, 1, B, 2)

Type of prolapse: uterine/vault Exclusion criteria: NR Type of mesh/graft: 80% Prolene mesh (A, 47; B, 52), 20% allograft SAFETY Publication type: full-text Age (mean, SD): 65.1 (9.4) material (A, 13; B, 12) Blood loss requiring transfusion, n/N: 1/124 (A, 1; B, 0) 136 Parity (mean, SD): A, 2.9 (1.8); B, 3.2 Study design: treated as case (1.5) Surgical route: abdominal (NR Damage to surrounding tissue, e.g. urethra, bladder, bowel: ureteral series, retrospective (contains 2 BMI (mean, SD): 25.8 (4.2) open or laparoscopic) transaction stent placement, 1/124 (A, 1; B, 0) groups, A, sacrocolpopexy + hysterectomy, n=60; B, Vault/uterine: 64 (B)/60 (A) ‘Mesh inlay’ or ‘total mesh’: Mesh/graft erosion, n/N: mesh erosion, 3y postoperative, 1/124 (A, 0; B, sacrocolpopexy in women had prior 1, Prolene was used); cuff granulation, 3m postoperative, 1/124 (A, 0; B, 1, hysterectomy, n=64) Pre-operation prolapse stage: A, 3.9 Concomitant procedures: Prolene was used). (0.4); B, 3.6 (0.5) Burch urethropexy: 91/124; Setting: university hospital Enterocele repair: 111/124; Further operation needed for mesh erosion, n/N: 1/124, managed by Number of surgeons: 1, NR Paravaginal repair: 63/124; office resection Country: US experience Posterior repair: 106/124; Sling (NR detail): 1/124 Wound infection: 1/124 (A, 1; B, 0) Recruitment/Treatment dates: 1 Jan. 1996 – 31 Dec. 1998 Other serious adverse events: heart block in 24h, 1/124 (A, 1; B, 0); arrhythmia, 1/124 (A, 0; B, 1); Length of follow-up: median, small bowel obstruction, late clinic complications, 3/124 (A, 1, Prolene was range, 35.5m (0-74) used; B, 2, one of which used Prolene, the other used allograft); Incision hernia, 18m, 2/124 (A, 1 Prolene was used; B, 1, Prolene was Source of funding: supported by used); a Research Centre in Minority Abdominal hernia mesh infection, late clinic complication, 1/124 (A, 0; B, Institutions award, from the 1, Prolene was used) National Centre for Research Resources, National Institute of Other adverse events: Health. Delayed vaginal healing, resolved in 3m, 1/124 (A, 1, allograft was used; B, 0);

Author, year Inclusion criteria: women with stage II INTERVENTION A (n=305): EFFICACY, 1y Bradley 200728,31,32,46,66,78 – IV prolapse, but without stress urinary abdominal (open) sacrocolpopexy New prolapse at other sites, n/N: new rectal prolapse in those had not incontinence. received posterior procedures, 3/193 Type of prolapse: uterine/vault Type of mesh/graft: permanent Exclusion criteria: NR synthetic mesh was used > 80%; Persistent bladder symptoms, n/N: persistent bothersome irritative Publication type: full-text xenograft, autologous or cadeveric symptom, 43/169 (89 of whom had Burch suspension, but there was no Number of patients lost to follow-up, fascia lata was also used. significant differences in the symptoms between having or not having Study design: treated as case n/N: 17/322 in a year Burch suspension). series, prospective (was a RCT, Concomitant procedures: Burch comparing sacrocolpopexy with and Pre-operative prolapse stage: II, suspension, 148; Persistent bowel symptoms, n/N: need of digital assistance, 7/54 without Burch suspension) 39/305; III 206/306, IV, 60/305 Posterior colporrhaphy, perineorrhaphy, and Persistent sexual problems, n/N: dyspareunia, 25/60 Setting: 14 hospitals Number of surgeons: >=1 sacrocolpoperineopexy, 87

Country: US SAFETY New urinary symptoms: bothersome irritative symptom, 12/131 (61 had Recruitment/Treatment dates: Burch suspension, but the results were not significantly different between Mar. 2002 – Feb. 2005 having and not having Burch suspension).

Length of follow-up: 1y results New bowel symptoms: need of digital assistance, 2/178; available New sexual problems: dyspareunia in those who had not received Source of funding: National posterior procedures, 11/76 Institute of Child Health and Human 137 Development Other adverse events: gastrointestinal complications, 58/322 reported nausea, emesis, bloating or ileus during hospitalisation and 32/322 at 6w.

Author, year Inclusion criteria: patients underwent INTERVENTION A (n=245): EFFICACY Culligan 200234 sacral colpopexy with synthetic mesh abdominal sacral colpopexy Objective failure: apical failures >= stage II, 0/44; Any postoperative POP-Q point >= stage II: 37/245, of these 26 occurred Type of prolapse: uterine/vault Exclusion criteria: NR Type of mesh/graft: synthetic within 6m, 30 within 1y, another 5 were discovered 1-2y. Of the objective mesh, NR details. failures, 22 were of the anterior segment, 14 were of the posterior Publication type: full-text Number of patients lost to follow-up, compartment, and 2 were combination of the two (NR recurrence or de n/N: 6m, 0; 1y, 17; 2y, 63; 3y, 123; Surgical route: abdominal (NR novo). No apical failures were observed. Study design: case series, >=4y, 169 open or laparoscopic) retrospective Quality of life: measured by a validated prolapse specific QoL instrument, Age (mean, SD, range): 61.2 (10.9), 32 Concomitant procedures: administered by a telephone survey. The first 11 Qs make up a prolapse Setting: university hospital – 83 Suburethral sling using donor fascia specific symptom inventory (PSI), Q12-15 determine QoL score related to BMI (mean, SD, range): 26.7 (4.7), 18 – lata, 17; those symptoms: Country: US 45 Tension-free vaginal tape, 48; Burch procedures, 106; Number with data available PSI QoL Recruitment/Treatment dates: Number of surgeons: NR Hysterectomy, 11; 1y 106 5.3 (3.9) 0.55 (1.8) Sep. 1994 – Jan. 2001 Paravaginal repairs, 145; 2y 42 4.5 (3.3) 0.60 (2.3) Posterior repairs, 39; 3y 20 4.5 (3.5) 0.2 (0.89) Length of follow-up: objective Perineorraphies, 22; 4y 14 5.3 (3.9) 0.6 (2.1) failure, >=4y results available; Halban procedures, 47; subjective failure, mean 3.3y, range Anterior colporrhaphies, 6. 0.8 – 6.9. SAFETY

Blood loss, ml, mean (SD), range: 328.5 (203.4), 50 – 1300 Source of funding: NR Blood loss requiring transfusion, n/N: 2/245

Damage to surrounding tissue: cystotomy, 1/245

Mesh/graft erosion, n/N: 6/245 (of the 11 patients who underwent a hysterectomy at the time of colpopexy, 3 had erosion, but of the 234 patients who had not have a hysterectomy at the time of colpopexy, 3 erosions occurred, this difference was significant).

Further operation needed for mesh erosion, n/N: 5/245, removal of eroded graft

Other serious adverse events: immediate postoperative: 2 DVT, 1 PE, Small bowel obstruction, 2/245, required re-operation.

Other adverse events: ureteral obstruction, intra-operative, 2/245 relieved during surgery; Femoral neuropathy, 1/245, spontaneously resolved over 3m; Fascia dehenscence, 1/245.

Author, year Inclusion criteria: women underwent INTERVENTION A (n=101): EFFICACY De Vries 199536 abdominal colposacropexy. In 38% of abdominal colposacropexy Subjective failure: had prolapse-related complaints, 1y, 22/72 138 the patients the indication was isolated Type of prolapse: uterine/vault genital or vaginal vault prolapse. In the Type of mesh/graft: non- SAFETY remaining cases there was also the absorbable synthetic mesh, Blood loss requiring transfusion, n/N: hemorrhage >=750ml, 5/101 Publication type: full-text presence of cystocele, rectocele, or Mersilene enterocele. Damage to surrounding tissue, e.g. urethra, bladder, bowel: bladder Study design: case series, Surgical route: abdominal (open) lesion, 6/101; lesions of intestines, 2/101 retrospective Exclusion criteria: NR Concomitant procedures: Mesh/graft erosion, n/N: coming off of the mesh, 1/101 Setting: university hospital Age (mean, range): 59 (37 – 82) Burch suspension, 20/101; Parity (median, range): 3 (0 – 14) Anterior vaginal repair, 9/101; Infection: wound infection, 17/101 Country: the Netherlands Vault/uterine prolapse: 83/15/3 Posterior repair: 9/101; cervical stump prolapse Both anterior and posterior repair: Other serious adverse events: ileus, 4/101 of which 2 were caused by Recruitment/Treatment dates: 30/101 intra-peritoneal adhesion and 2 by internal herniation underneath the 1979 – 1992 Number of surgeons: NR mesh; Length of follow-up: median, Cicatricial hernia, 2/101. range, 4y (1 – 13) Other adverse events: Source of funding: NR Extra-peritoneal haematoma, during surgery, 1/101; Wound haematoma, postoperative, 3/101; Urinary retention, 3/101; Platzbanch (NR details), 1/101; Problems of motion (NR details), 8/101

Author, year Inclusion criteria: women who had INTERVENTION A (n=148): EFFICACY Higgs 2005b49 undergone sacral colpopexy for vault abdominal (open) sacrocolpopexy Subjective failure: persistent prolapse symptoms, 21/90

prolapse Type of prolapse: uterine/vault Type of mesh/graft: Vypro mesh Objective failure: vault prolapse, Baden-Walker >=stage 2, 2/64 (one Exclusion criteria: NR (absorbable synthetic), Johnson & grade II and one grade III) Publication type: full-text Johnson, Ethicon, Somerville, NJ, Number of patients lost to follow-up, US; autologous fascia lata (6/148), Further operation needed for prolapse (recurrent or new at other Study design: case series, n/N: 55/148 not contactable; in those doner fascia lata (14/148); in most sites), n/N: 13/93 of whom 10 was a posterior repair prospective who were followed up, 64 were available cases, polypropylene mesh (Prolene for clinical examination, a further of 29 Atrium mesh, Johnson & Johnson, Patients satisfaction (visual analogue score, 0-100): subjective Setting: a tertiary unit were available for telephone interview Ethicon, Somerville, NJ, US) was success rate >=80%, 57/87; 50-79%, 17/87; <50%, 13/87. because they were not able to come to used or Mesilene (Johnson & Country: Australia the clinic for physical examination. Johnson, Ethicon, Somerville, NJ, US), n=3/148. SAFETY Recruitment/Treatment dates: Age (mean, SD): 58 (10.5) Damage to surrounding tissue, e.g. urethra, bladder, bowel: bladder 1998-2001 Parity (mean, SD): 2.9 (1.4) Concomitant procedures: injuries, 5/148; vascular injuries, 1/148; bowel injuries, 2/148. Length of follow-up: mean (SD), TVT, 2/148; 45m (14.6) Vault/uterine repair: 123/24 Burch colposuspension: 74/148; Mesh/graft erosion, n/N: 4/148 (3 were non-absorbable mesh, 1 was Total abdominal hysterectomy, absorbable), 2 of which occurred in a year, 1 between one year and 18m. Source of funding: NR Number of surgeons: >=1. Surgery 25/148; was performed by an urogynaecology Rectopexy, 11/148; Further operation needed for mesh erosion, n/N: 3 was able to be consultant or senior urogynaecology Posterior vaginal repair, 54/148 excised vaginally, the 4th one was removed abdominally. fellows under supervision by an urogynaecologist. Other adverse events: vesico-cutaneous fistula, 1/148; Incisional hernia, 8/148

139 Author, year Inclusion criteria: women underwent INTERVENTION A (n=262): EFFICACY Lindeque 200255 sacrocolpopexy for the surgical repair of abdominal (open) sacrocolpopexy Objective failure: recurrent vault prolapse, no definition, 3/262 vaginal vault prolapse. Indication: Type of prolapse: uterine/vault patients with good vaginal wall support, Type of mesh/graft: initially used Further operation needed for prolapse (recurrent or new at other typically after recent anterior and dura mater strips, n=18; then Gore- sites), n/N: 4/262 Publication type: full-text posterior colporrhaphy, who were tex soft tissue patch was used, n>19, diagnosed with vaginal vault prolapse Gore and Associates, Inc., Flagstaff, Study design: case series, NR (stage II) and enterocele. Patients with Ariz., US. SAFETY prospective or retrospective massive enterocele with the uterus in Blood loss requiring transfusion, n/N: 3/262 situ were rarely seen but were also Concomitant procedures: Setting: university hospital included in the analysis. Burch vesicosuspension, 106/262; Damage to surrounding tissue, e.g. urethra, bladder, bowel: Posterior colporrhaphy, 60/262; Minor bladder injury, 3/162; minor small-bowel injury, 1/262; vessel injury, Country: South Africa Exclusion criteria: NR Anterior repair, 1/262. 0.

Recruitment/Treatment dates: Number of patients lost to follow-up, Mesh/graft erosion, n/N: 10/262, all occurred in a year. Mar. 1994 – Feb. 2001 n/N: NR Length of follow-up: all >=16m Further operation needed for mesh erosion, n/N: in 7 cases mesh Age (range): 28-79 removal was done through laparotomy while in 3 cases the patch could be Source of funding: NR removed transvaginally by gentle pulling. Primary/secondary repair: 8/254 Other serious adverse events: medical and other complications, 7/262 Vault/uterine: 258/4 (1 pneumonia, 3 prolonged ileus, 2 arrhythmia, 1 left ventricular failure).

Number of surgeons: >=1. One of the Other adverse events: intestinal obstruction, 3/262. authors was present at all operations

and performed approximately three- quarters of the procedures, the remainder being performed by the second author.

Author, year Inclusion criteria: patients with INTERVENTION A (n=147): EFFICACY Snyder 199171 diagnosis of complete vaginal prolapse abdominal (open) sacral colpopexy Subjective failure: pelvic pressure, 34/132

Type of prolapse: uterine/vault Exclusion criteria: NR Type of mesh/graft: Dacron graft, NR producer, n=65; expanded SAFETY Publication type: full-text Number of patients lost to follow-up, polytetrafluoroethylene (Gore-Tex, Blood loss, ml, mean (range): n=147, 302ml (30-2600) n/N: 15/147 because of death or inability soft tissue patch; WL Gore and Study design: case series, to assess the status adequately Associates Inc. Medical Products, Blood loss requiring transfusion, n/N: >500ml, 23/147 retrospective Division Flagstaff, AZ), n=78; other Age (mean, range): 62 (30-83) graft, NR, n=4. Mesh/graft erosion, n/N: 4/132 Setting: university Kansas medical Parity (mean): 3 center Vault/uterine repair: 144/3 Further operation needed for mesh erosion, n/N: 4 surgical removal

Country: US Number of surgeons: >=1, surgery was OTHER OUTCOMES performed by the authors or by one of Operating time: mean, range, 97min (50-210) Recruitment/Treatment dates: the senior professors in the department. 1972 – 1989 Length of hospital stay: mean, range, 5d (2-13) Length of follow-up: mean, range, 43m (1m – 17y), 116 had at least 140 6m follow up.

Source of funding: NR

Author, year Inclusion criteria: women with INTERVENTION A (n=163): EFFICACY Timmons 199275 symptomatic vaginal vault prolapse and abdominal (open) sacral colpopexy Objective failure: ‘bad vaginal vault support’, NR details, 2/163 at 6w enterocele underwent abdominal sacral Type of prolapse: uterine/vault colpopexy Type of mesh/graft: fascial lata, n=1; Mersilene n=162. SAFETY Publication type: full-text Exclusion criteria: NR Blood loss requiring transfusion, n/N: severe hemorrhage, 2/163 Concomitant procedures: Study design: case series, Age (mean, range): 58.3 (19-81) Retropubic cystourethropexy, Other adverse events: wound dehiscence, 1/163; wound separation, prospective Parity (mean, range): 2.9 (0-9) 29/163; 6/163 Posterior colporrhaphy, 28/163; Setting: university medical centre Vault/uterine repair: 3/163 did not have Anterior colporrhaphy, 3/163 a posthysterectomy vault prolapse, 2 of Country: US which had congenital absence of the Additional information: Recruitment/Treatment dates: uterus, 1 underwent vaginal modifications of the surgical 1972-1989 hysterectomy at the time of techniques have evolved. Length of follow-up: median, sacrocolpopexy, she previously had range, 33m (9m-18y); 20 patients significant component of the levator sling were followed >=10y; 78 >=5y. removed.

Source of funding: NR Number of surgeons: NR

Author, year Inclusion criteria: women with prior INTERVENTION A (n=212): EFFICACY Wu 200679 hysterectomy and underwent abdominal abdominal (open) sacrocolpopexy NR Type of prolapse: vault sacral colpopexy Publication type: full-text Type of mesh/graft: Mersilene or Study design: case series, Exclusion criteria: patients who polypropylene mesh, or Gore-Tex SAFETY retrospective (treated as 2 case underwent a repeat or laparoscopic mesh Blood loss, ml, mean (SD): 298 (193) series) sacral colpopexy Setting: university hospital Mesh/graft erosion, n/N: 10/212, median time to occur was 22.9m (range Country: US Age (mean, SD): 65.5 (11.4) 2.9 – 70.3) Recruitment/Treatment dates: Parity (mean, SD): 2.9 (1.5) Jun 1994 – Dec 2004 BMI (mean, SD): 26.9 (5.0) OTHER OUTCOMES Length of follow-up: mean, SD, Length of hospital stay: mean (SD), 3.5 (2.2) range, 15.0 (23.8), 0.3-121.1 Primary/secondary repair: 212/0 Source of funding: NR Number of surgeons: NR

141

Vaginal infracoccygeal sacropexy

No full-text RCTs reported infracoccygeal sacropexy

RCTs PUBLISHED AS CONFERENCE ABSTRACTS ONLY

Uterine prolapse repair (no RCTs (abstract) available)

Vault prolapse repair Author, year Inclusion criteria: women with >= INTERVENTION A (n=30): EFFICACY 59 Meschia 2005 stage II vault prolapse (point C >= -1), infracoccygeal sacropexy Objective failure: reported only one woman had a stage II vault prolapse, Type of prolapse: vault prolapse required surgical treatment but NR which group she was from. Publication type: conference Type of mesh/graft: NR abstract Study design: RCT Surgical rout: vaginal SAFETY Setting: university hospital Mesh/graft erosion, n/N: B, 2/30, one of which further developed a Country: Italy pararectal abscess. Recruitment/Treatment dates: INTERVENTION B (n=30): Feb. 2002 – Dec. 2003 sacrospinous fixation Length of follow-up: median, 24m Source of funding: not supported Surgical route: vaginal

142 by industry

Uterine and/or vault prolapse repair Author, year Inclusion criteria: NR INTERVENTION A (n=21): posterior EFFICACY 74 De Tayrac 2006 IVS Objective failure: reoccurrence of uterine prolapse, no definition, A, 1/21 Type of prolapse: uterine/vault in a patient had uterus preservation Publication type: conference Type of mesh/graft: Tyco abstract Healthcare Study design: RCT SAFETY Setting: NR Surgical route: vaginal Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal Country: NR injury, A, 0/21; B, 0/20 Recruitment/Treatment dates: Mar 2003 – Dec 2005 INTERVENTION B (n=20): unilateral Mesh/graft erosion, n/N: A, 0/21 Length of follow-up: A, 10.5m sacrospinous suspension (1.5-34); B, 15.5 (1.5-32) Infection, n/N: Postoperative vaginal infection, A, 0/21; B, 0/20 Source of funding: not supported Surgical rout: vaginal by industry Other adverse events: intra-operative or post-operative haematomas, A, 10%; B, 10.5%. Concomitant procedures: hysterectomy was performed as OTHER OUTCOMES needed. Operating time: A, 87.9 (25); B, 87.1 (18)

NON-RANDOMISED COMPARATIVE STUDIES

Uterine prolapse repair Author, year Inclusion criteria: patients suffering INTERVENTION A (n=44): EFFICACY Neuman 200760 from advanced uterine prolapse (stage hysterectomy + infracoccygeal Objective failure: uterine prolapse (no definition): A, 0/44; B, 1/35 3 or 4) resulting from a vaginal apical sacropexy Type of prolapse: uterine support defect, diagnosed clinically Overall satisfaction with the operative therapeutic results: good, A, according to the POP-Q system. Type of mesh/graft: NR 39/44; B, 32/35 Publication type: full-text Exclusion criteria: NR Surgical route: vaginal Study design: non-randomised SAFETY comparative study, prospective Age (mean, SD): A, 63.5 (13.2); B, 51.0 Mesh/graft erosion, n/N: tape protrusion, A, 6/44; B, 4/35 (10.0) INTERVENTION B (n=35): Setting: university hospital Parity (mean, SD): A, 4 (2): B, 4 (2) infracoccygeal sacropexy with uterus Further operation needed for mesh erosion, n/N: all 10 patients preservation underwent segmental tape resection at the out-patient clinic. Country: Israel Vault/uterine repair: A, 0/44; B, 0/35 Type of mesh/graft: NR New urinary symptoms: incontinence, A, 0/29; B, 0/26 Recruitment/Treatment dates: Number of surgeons: 2 Jan. 2003 – Jun. 2005 Surgical route: vaginal OTHER OUTCOMES Length of follow-up: mean, range, Length of hospital stay: mean, d: A, 4.2; B, 1.5 29.8m (12-44) Concomitant procedures: Source of funding: NR TVT: A, 9; B, 6 143 Cervical amputation, A, 0; B, 6; Anterior or posterior colporrhaphy: A, 33; B, 24

Vault prolapse repair (no non-randomised comparative studies available)

Uterine and/or vault prolapse repair (not reported separately) (no non-randomised comparative studies available)

CASE SERIES

Uterine prolapse repair Author, year Inclusion criteria: NR INTERVENTION A (n=10): EFFICACY Sivaslioglu 200570 infracoccygeal sacropexy Objective failure: uterine prolapse >=grade 2, 1/10 (grade 3); grade I Exclusion criteria: NR uterine prolapse, 3 Type of prolapse: uterine Type of mesh/graft: NR Age (mean, SD): 60 (7), n=30 Further operation needed for prolapse (recurrent or new at other Publication type: short- Pre-operative prolapse stage: POP-Q Surgical route: vaginal sites), n/N: 0/10 (the grade 3 one refused further operation) communication uterine prolapse grade 4, 10 SAFETY Study design: case series, NR Vault/uterine repair: 0/10 Blood loss, ml, mean (SD): 200 (20), n=30 prospective or retrospective (this study contain 2 case series, one on Number of surgeons: NR Blood loss requiring transfusion, n/N: 0/10 vault prolapse, the other on uterine prolapse, see the study above) Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal

perforation, 0/10 Setting: teaching hospital OTHER OUTCOMES Country: Turkey Operating time: mean, SD, n=30, 45min (8)

Recruitment/Treatment dates: Length of hospital stay: mean, SD, n=30, 2d (1) Oct 2002 – Jun 2004 Length of follow-up: mean, SD, 16 (6), n=30

Source of funding: NR

Vault prolapse repair Author, year Inclusion criteria: women who had INTERVENTION A (n=93): posterior EFFICACY Farnsworth 200239 undergone a hysterectomy and had intravaginal slingplasty Subjective failure: symptomatic (‘’<50% reduction in the severity of clinical evident vault prolapse (grade 2 (infracoccygeal sacropexy) symptoms’’), 8/91. Type of prolapse: vault or 3) Type of mesh/graft: polypropylene Further operation needed for prolapse (recurrent or new at other Publication type: full-text Exclusion criteria: NR tape (IVS Tunneller, Tyco Healthcare, sites), n/N: 2/91 USA) or nylon tape (Medhealth Study design: case series, Number of patients lost to follow-up, Supplies, WA) SAFETY prospective n/N: 2/93 Blood loss, ml, mean, range: 110 (20 – 400) Surgical route: vaginal 144 Setting: hospital Age (mean, range): 65 (36 – 77) Blood loss requiring transfusion, n/N: 0/93

Country: Australia Number of surgeons: 1 Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal perforation, 1/93 Recruitment/Treatment dates: 1998 – 2000 Mesh/graft erosion, n/N: tape rejection, 5/49 nylon tape, and 0/44 Length of follow-up: mean, range, polypropylene tape; 12m (2-24) Rectal tape erosion, 1/93

Source of funding: NR Infection: urinary tract infection, 5/93, all occurred in <1w; Pyrexin, 0/93.

Other adverse events: significant bruising or hematoma, 0/93.

OTHER OUTCOMES Operating time: min, mean, range, 60min (45-90)

Length of hospital stay: all patients were discharged home <24h.

Author, year Inclusion criteria: women with vault INTERVENTION A (n=52): EFFICACY Foote 200741 prolapse underwent infracoccygeal infracoccygeal sacropexy Objective failure: vault was not well elevated (no definition): 9/52 sacropexy Type of prolapse: vault Type of mesh/graft: multifilament Exclusion criteria: NR mesh, NR details. SAFETY Publication type: short- Blood loss, ml, mean: 190

communication Age (mean): 63.6 Surgical rout: vaginal Parity (mean): 2.9 Mesh/graft erosion, n/N: 11/52 in a mean of 20.2 weeks follow-up Study design: case series, Vault/uterine repair: 52/0 Concomitant procedures: retrospective suburethral sling, 5/52 Further operation needed for mesh erosion, n/N: mesh was excised Number of surgeons: NR and oversewn as a day procedure, but NR number of patients had this. Setting: university hospital OTHER OUTCOMES Country: Australia Operating time: mean, min, 65.8

Recruitment/Treatment dates: Length of hospital stay: mean, day, 4.1 2002-2004 Length of follow-up: mean 20.2 week results available

Source of funding: NR

Author, year Inclusion criteria: vault prolapse INTERVENTION A (n=15): EFFICACY Ghanbari 200642 grades 3 and 4 (POP-Q) intravaginal slingplasty Subjective failure: reduction in the frequency and severity of a symptom was found to be <50%, 1/15 Type of prolapse: vault Exclusion criteria: NR Type of mesh/graft: Tyco Health Care, US, polypropylene tape Publication type: short- Age (mean, range): 67 (50-81) SAFETY communication Surgical route: vaginal Blood loss, ml: all less than 130ml Vault/uterine repair: 15/0 145 Study design: case series, Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal prospective Number of surgeons: 1 perforations, 0/15

Setting: University health research Mesh/graft erosion, n/N: tape rejection, 1/15, occurred 6m centre postoperatively.

Country: Iran Infection: urinary tract infection, 0/15; postoperative pyrexia, 0/15

Recruitment/Treatment dates: Other adverse events: haematoma, 0/15; significant bruising, 0/15 2003-2004 Length of follow-up: NR OTHER OUTCOMES Operating time: mean (range), 30 (20-40) Source of funding: NR Length of hospital stay: all patients were discharged within 24h of surgery

Author, year Inclusion criteria: vaginal prolapse of INTERVENTION A (n=75): EFFICACY Petros 200163 at least 2nd degree magnitude following infracoccygeal sacropexy Objective failure: vault prolapse >= II degree, mean follow up 4.5y (55m), abdominal or vaginal hysterectomy range 48-60m, 4/40, all occurred within 24m Type of prolapse: vault Type of mesh/graft: tension-free Exclusion criteria: no exclusion vaginal tape (nylon) Further operation needed for prolapse (recurrent or new at other Publication type: full-text criteria. The minimally invasive nature of sites), n/N: 12/40 (8 for de novo anterior prolapse, 3 for vault prolapse, the technique allowed even very ill Surgical route: vaginal and 1 for rectocele) Study design: case series, patients to have surgery.

prospective Number of patients lost to follow-up, SAFETY Setting: hospital n/N: 4 in a year, 11 in 2 years, 19 in 3 Blood loss, ml, mean (range): 120 (20-800) years, and 35 in 4 years. Country: Australia Blood loss requiring transfusion, n/N: 0/75 Age (mean, range): 54 (40-74) Recruitment/Treatment dates: Parity (median, range): 3 (0-5) Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectum 1992-1993 Pre-operative prolapse: 2nd degree perforation: 2/75 Length of follow-up: 1-4.5y prolapse, 22; 3rd degree, 45; 4th degree, 8; Mesh/graft erosion, n/N: tape erosion, 4/75, occurred between 6-12m Source of funding: NR Rectocele: 68 Further operation needed for mesh erosion, n/N: all 4 were removed as an office procedure Vault/uterine repair: 75/0 Infection: not specified, 1/75, occurred at day 5; postoperative pyrexia, Number of surgeons: NR 0/75

Other adverse events: hematomas: 0/75

OTHER OUTCOMES Operating time: 30-60 min.

Length of hospital stay: all patients were discharged within 24h.

146 Author, year Inclusion criteria: NR INTERVENTION A (n=20): EFFICACY Sivaslioglu 200570 infracoccygeal sacropexy Objective failure: vault prolapse >=grade 2, 0/20; vault prolapse at grade Exclusion criteria: NR 1, 5/20 Type of prolapse: vault Type of mesh/graft: NR Age (mean, SD): 60 (7), n=30 Further operation needed for prolapse (recurrent or new at other Publication type: short- Pre-operative prolapse stage: POP-Q Surgical rout: vaginal sites), n/N: 0/20 communication vault prolapse grade 2, 5; grade 3, 12; grade 4, 3. Presumed uterus was reserved, SAFETY Study design: case series, NR because the mean operation time was Blood loss, ml, mean (SD): 200 (20), n=30 prospective or retrospective (this Vault/uterine repair: 20/0 only 45min. study contain 2 case series, one on Blood loss requiring transfusion, n/N: 0/20 vault prolapse, the other on uterine Number of surgeons: NR prolapse, see the study below) Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal perforation, 0/20 Setting: teaching hospital OTHER OUTCOMES Country: Turkey Operating time: mean, SD, n=30, 45min (8)

Recruitment/Treatment dates: Length of hospital stay: mean, SD, n=30, 2d (1) Oct 2002 – Jun 2004 Length of follow-up: mean, SD, 16 (6), n=30

Source of funding: NR

Uterine and/or vault prolapse repair Author, year Inclusion criteria: uterovaginal INTERVENTION A (n=127): posterior EFFICACY Hefni 200748 prolapse (65%) and vault prolapse intravaginal slingplasty/infracoccygeal Subjective failure: persistent prolapse symptoms, 27/127 (35%) sacropexy Type of prolapse: uterine/vault Objective failure: upper genital support failure, stage II – IV, 14/127 (6 Exclusion criteria: NR Type of mesh/graft: IVS Tunneller uterine prolapse, 5 vault prolapse, 3 enterocele recurrence) Publication type: full-text (Tyco Healthcare, USS, Norwalk, CT, Number of patients lost to follow-up, US), polypropylene multifilament tape. New prolapse at other sites, n/N: cystocele, 3/63 Study design: case series, n/N: NR prospective Surgical route: vaginal Further operation needed for prolapse (recurrent or new at other Age (mean, SD): 59 (9) sites), n/N: 10/127 had sacrospinous colpopexy Setting: hospital Parity (mean, SD): 2.4 (1) Concomitant procedures: BMI (mean, SD): 27.9 (4.6) transobturator mid-urethral tape Country: UK Cystocele: 64 insertion, 8; SAFETY Rectocele: 56 Hysterectomy, 22; Blood loss requiring transfusion, n/N: 0/127 Recruitment/Treatment dates: Anterior colporrhaphy, 64. Apr. 2003 - Aug. 2005 Vault/uterine repair: 44/83 Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal, Length of follow-up: mean, range, visceral or ureteric injuries, 0/127 14m (2-26) Number of surgeons: NR Mesh/graft erosion, n/N: tape exposure through the posterior vaginal wall, Source of funding: no financial 21/127, mean time occurred 8m (3.5), range 1.5-20m after surgery. support Further operation needed for mesh erosion, n/N: 21, required surgical 147 tape excision and approximation of vaginal edges under anaesthesia and antibiotic cover.

Infection: urinary tract infection, in first 6w after surgery, 2/127; infection along the course of the tape in the ischio-rectal (para-rectal) space requiring administration of antibiotics, surgical drainage, and tape removal, 1/127.

Other adverse events: para-rectal haematoma (approximately 300ml), 1/127; Anterior-posterior vaginal wall adhesions, 3/127; Formation of posterior vaginal wall granulation tissue overlying the area of tape attachment requiring surgical excision, 1/127.

OTHER OUTCOMES Operating time: mean (SD), 46min (18.5); for posterior IVS alone, n=24, 27.4m (10).

Length of hospital stay: mean (SD), 3d (0.98)

Author, year Inclusion criteria: women had posterior INTERVENTION A (n=8): posterior EFFICACY Jordaan 200653 intravaginal slingplasty. Only 8 of these intravaginal slingplasty Objective failure: grade 3 or 4 vault/uterine prolapse (Baden-Walker), 2/8 patients had apical (uterine/vault) grade Type of prolapse: uterine/vault 3 or 4 prolapse. Type of mesh/graft: Vypro mesh (Johnson & Johnson, Brussels,

Publication type: full-text Exclusion criteria: NR Belgium). It is a semi-absorbable mesh consisting of polyproglactine Study design: case series, Age (median, range): n=43, 65 (40-93) and prolene 1:1. Tunneler (US retrospective Parity (median, range): n=43, 3 (0-9) Surgical, Tyco Health Care Group) was used. Setting: university hospital Number of surgeons: NR Surgical route: vaginal Country: South Africa Concomitant procedures: Recruitment/Treatment dates: Not reported separately for the 8 NR women who had uterine/vault Length of follow-up: mean, range: prolapse 13m (1021) for the 42 women recruited, but only 8 women had uterine or vault prolapse

Source of funding: NR

Author, year Inclusion criteria: NR INTERVENTION A (n=14): a EFFICACY Oliver 200662 modified IVS technique which does Subjective failure: >100% cure of prolapse symptoms, 1/14 who had Exclusion criteria: NR not need to remove the uterus prior to uterine prolapse, occurred at 6w. Type of prolapse: uterine/vault the repair Age (mean, SD): 72.5 (18) SAFETY Publication type: full-text Type of mesh/graft: IVS Tunneller Blood loss, ml, mean (SD): 182 (12) 148 Vault/uterine repair: 7/7 (Tyco Healthcare US) is used to place Study design: case series, a polypropylene tape. Blood loss requiring transfusion, n/N: 0/14 retrospective Pre-operation prolapse: 7 vault prolapse (5 III degree, 2 II degree); Surgical route: vaginal Damage to surrounding tissue, e.g. urethra, bladder, bowel: rectal Setting: hospital 7 uterine prolapse (5 III degree, 2 II perforation, 0/14 degree) Concomitant procedures: Country: UK hysterectomy, 3/7 (it was felt that the Urinary tract infection: 1/14, occurred 1w postoperatively. Number of surgeons: 1 uterine prolapse in these patients was Recruitment/Treatment dates: of such a degree as to potentially Other adverse events: haematomas, 0/14; urine retention, 1/14 Jun 2004 – May 2005 compromise the repair success); (cystoscopy confirmed no problems with the tape insertion). Length of follow-up: the mean Cystocele repair: NR number. time interval of the first Rectocele repair: NR number. OTHER OUTCOMES postoperative follow up was 6w, Operating time: mean (SD), 50.5 (8) min and that of the questionnaire was 5.1m (range 2-11) Length of hospital stay: mean, 2d.

Source of funding: NR Author, year Inclusion criteria: patients with at least INTERVENTION A (n=67): Tissue EFFICACY Petros 200565 2nd degree prolapse complaining of a Fixation System posterior sling (is a Objective failure: vault prolapse (no definition): 1/66 ‘lump’ in the vagina, or prolapse to the direct evolution of the posterior IVS Type of prolapse: uterine/vault middle part of the posterior vaginal wall operation) Further operation needed for prolapse (recurrent or new at other (1st degree) who complained of dragging sites), n/N: 3/66 (1 vault prolapse, 1 enterocele, and 1 rectocele) Publication type: full-text pain, or other symptoms (such as Type of mesh/graft: the Tissue emptying difficulty) related to prolapse of Fixation System (TFS Manufacturing) Study design: case series, the posterior fornix. comprises of two small polypropylene SAFETY

prospective soft tissue anchors connected to an Mesh/graft erosion, n/N: large granuloma: 1/67 Exclusion criteria: NR adjustable polypropylene tape. The Setting: university hospital device has been approved by the Further operation needed for mesh erosion, n/N: 1, the sling was Number of patients lost to follow-up, Australian and European Government excised. Country: Australia n/N: 1/67 regulatory bodies. Urinary tract infection: 6/67 Recruitment/Treatment dates: Age (mean, range): 65 (35-87) Surgical route: vaginal NR Parity (mean, range): 2.7 (0-8) OTHER OUTCOMES Length of follow-up: Mean Additional information: the posterior Operating time: range, min, 5-10. (range): 9m (3-15) Vault/uterine repair: 40/23 TFS sling works much like a McCall procedure. The anchors are inserted Length of hospital stay: mean (range), d: 1.5 (1-3). Source of funding: NR. The first Pre-operative prolapse stage: 4th to the tissue just lateral to the author has a financial interest in the degree uterine/vault prolapse: 2/67; 3rd uterosacral ligaments. Tightening the Tissue Fixation System instrument. degree, 17/67; 2nd degree, 20/67; sling elevates the prolapse symptomatic 1st degree, 28/67. uterus/vaginal vault.

Number of surgeons: NR

Author, year Inclusion criteria: women with stage III INTERVENTION A (n=44): non- EFFICACY Sentihes 200769 or IV prolapse with at least two pelvic absorbable hammock placement Subjective failure: feeling of vaginal lump, 1/44 compartments using anterior trans-obturator and Type of prolapse: uterine/vault posterior infracoccygeal extensions. Objective failure: apical prolapse >=stage II, 0/44 Exclusion criteria: NR Publication type: full-text Type of mesh/graft: initially the 7 had persistent asymptomatic stage I prolapse (4 anterior, 3 posterior); 1 149 Age (mean, SD, range): 66 (8.5) (50- prosthetic meshes used were made of had symptomatic recurrence of a posterior vaginal wall stage III prolapse, Study design: case series, 84) polyester (Parietex, Sofradim, but NR these were recurrence prolapse or de novo. prospective Parity (mean, SD, range): 3.3 (2.7) (0- Trevoux, France), n=8; then 11) multifilament polypropylene meshes Persistent bladder symptoms, n/N: SUI, 5/17, none of the 17 women Setting: university hospital BMI (mean, SD, range): 27 (5.4) (17- (Surgipro mesh, Tyco Healthcare, had treatment for SUI. 42) Plaisir, France), n=16; and Country: France Pre-operative prolapse stage: apical subsequently low-weight Persistent sexual problems, n/N: dyspareunia, 3/7 prolapse (POP-Q), stage 0, 2; stage I, 1; monofilament polypropylene (Ugytex, Recruitment/Treatment dates: stage II, 16; stage III, 13; stage IV, 12. Sofradim, Trevoux, France), n=20 Other outcomes: subjective satisfaction assessed by a visual analogue Jan 2002 – Jun 2005 scale (0-10 very satisfied), 44/44 expressed high or very high satisfaction Length of follow-up: mean, range, Vault/uterine repair: 20/24 Surgical route: vaginal with the surgery outcome. 29.3 (9-47) Number of surgeons: NR ‘Mesh inlay’ or ‘total mesh’: total Source of funding: NR mesh SAFETY Blood loss requiring transfusion, n/N: 1/44 Concomitant procedures: hysterectomy, 24/44; no anterior or Damage to surrounding tissue, e.g. urethra, bladder, bowel: uneventful posterior colporrhaphy, excision of rectal injury, 1/44; vaginal skin, or other concomitant procedures were performed. Mesh/graft erosion, n/N: vaginal erosions, 6/44 (3 were diagnosed at 6w, 1 at 6m, 1 at 12m, and 1 at 16m; 2 used polyester, 2 used multifilament polypropylene, and 2 used monofilament polypropylene); no mesh erosions through the bladder or the rectum.

Further operation needed for mesh erosion, n/N: 2 mucosal closures

under local or general anaesthesia, 2 partial removal under anaesthesia.

New urinary symptoms: SUI, 0/27

Erosion with mesh infection: 2/44

Other adverse events: haematoma, 1/44

OTHER OUTCOMES Operating time: mean (range), 149min (90-240)

Length of hospital stay: mean (range), 4.8d (3-22)

150

Vaginal or abdominal uterine suspension sling (a technique for uterine prolapse repair)

FULL-TEXT RCTS Author, year: Roovers 200467,68 Inclusion criteria: patients with intact INTERVENTION A (n= 41): EFFICACY, 1y uterine and undergoing surgical Abdominal sacrohysteropexy Subjective failure: measured by Urogenital Distress Inventory (UDI), Type of prolapse: uterine correction of uterine prolapse stages II- range 0-100, higher score more bothersome symptoms IV (ICS). Type of mesh/graft: non- Publication type: full-text absorbable synthetic mesh, A B Exclusion criteria: presence of an polypropylene, Gore-Tex, Amid Baseline, N, mean (SD) N=41, 68,3 (4.3) N=41, 58.0 (4.7) Study design: RCT adnexal mass, a history of >2 abdominal classification type II (microporous). 1y, N, mean (SD) N=38, 9.2 (3.8) N=40, 5.1 (3.0) pelvic surgical procedures, extreme Mean difference (95% CI): 4.1 (-5.4 to 13.6) 2 Setting: three teaching hospitals obesity (BMI > 35 kg/m ), prior Surgical route: abdominal inflammatory bowel or pelvic disease Visited a doctor because of prolapse symptoms: A, 16/41; B, 5/41 Country: the Netherlands and faecal incontinence because of an ‘Mesh inlay’ or ‘total mesh’: neither internal or external anal sphincter defect. Objective failure: Recruitment/Treatment dates: Concomitant procedures: ICS Jan 1998 – Jul 2000 Number of patients lost to follow-up, Colposuspension: 16/41; A B n/N: A, 3/41; B, 1/41 (no motivation to Hysterectomy: 2/41 (patients Uterine or vault prolapse ≥ stage II, 1y 2/38 2/40 Length of follow-up: 1y undergo pelvic examines) preferred intervention B) Cystocele ≥ stage II Baseline 35/41 36/41 Source of funding: NR Age (mean, SD): A, 57.9 (8.8); B, 56.4 1y 14/38 16/40 (10.9) INTERVENTION B (n= 41): Rectocele ≥ stage II No mesh/graft: vaginal hysterectomy

151 Baseline 13/41 15/41 Parity (median, range): A, 2.9 (1.1); B, combined with anterior and/or 1y 2/38 6/40 2.5 (1.2) posterior colporrraphy Further operation needed for prolapse (recurrent or new at other BMI (mean, SD): A, 25.1 (3.0); B, 26.0 Surgical route: vaginal sites), n/N: A, 9/41 (5 cystocele, NR new at other sites or not, 4 recurrent (3.6) uterine prolapse); B, 1/41 (recurrent vault prolapse) Concomitant procedures: Incontinence: evidence of SUI, A, 9/41; Colposuspension: 11/41; SAFETY B, 8/41 Hysterectomy: 41/41 Blood loss, mean (SD): A, N=41, 244ml (51.5); B, N=41, 248ml (34.1)

Primary/secondary repair: NR Blood loss requiring transfusion, n/N: A, 1/41; B, 2/41

Number of surgeons: ≥ 1. All surgeries Damage to surrounding tissue, e.g. urethra, bladder, bowel: bowel were performed by experienced lesion, A, 0/41; B, 1/41 gynaecologists who were familiar with both techniques. They had all performed Infection: ≥ 50 of each described surgical Would infection: A, 0/41; B, 1/41; procedure prior onset of the study. Fever of unknown origin during admission: A, 3/41; B, 1/41; Vault abscess during admission: A, 2/41; B, 0/41 Infected implant requiring surgery, NR timing, NR: A, 2/41, B, NA.

Other serious adverse events: Incisional peritoneal hernia requiring surgery: A, 1/41; B, 0/41.

Other less serious adverse events:

Lower urinary tract symptoms during admission: A, 8/41; B, 8/41; Dullness upper leg during admission: A, 1/41; B, 0/41; Vaginal stricture requiring readmission for excision: A, 0/41; B, 1/41;

OTHER OUTCOMES Operating time, mean (SD): A, N=41, 97min (3.6); B, N=41, 107min (4.7); mean difference, 10min (-2 to 22)

Length of hospital stay, mean (SD): A, N=41, 7.7d (0.2); B, N=41, 7.6d (0.3); mean difference, 0.1d (-0.6 to 0.7).

Others: Visited a doctor because of: symptoms related to surgery: A, 25/41; B 13/41; defecation symptoms (NR persistent or new): A, 12/41; B 5/41; micturition symptoms (NR persistent or new): A, 8/41; B, 3/41; other symptoms: A, 11/41; B, 10/41. NON-RANDOMISED COMPARATIVE STUDIES Author, year Inclusion criteria: patients with INTERVENTION A (n=36): EFFICACY Costantini 200533 symptomatic uterovaginal prolapse sacrohysteropexy Objective failure: recurrent cervical/vault prolapse (<6cm above the grade III – IV. hymen): A, 0/36; B, 0/39 Type of prolapse: uterine Type of mesh/graft: non- Exclusion criteria: for uterine absorbable synthetic mesh, Marlex Further need for pelvic floor repair: A, 0; B, 0 Publication type: full-text conservation group, no post- 152 menopausal bleeding, no previous Surgical rout: vaginally Persistent bladder symptoms, n/N: incontinence, A, 4/22; B, 7/20; but all Study design: non-randomised cervical intraepithelial neoplasia, no had surgeries for incontinence. comparative study, prospective abnormal cervical smears or uterine disease including uterine enlargement or INTERVENTION B (n=39): Setting: university hospital cervical ulceration. sacropexy + hysterectomy SAFETY Blood loss, ml, median: A, 200; B, 325 Country: Italy Number of patients lost to follow-up, Type of mesh/graft: Marlex n/N: A, 2/36; B, 1/39; all were included Blood loss requiring transfusion, n/N: A, 2; B, 2 Recruitment/Treatment dates: in the analysis. Surgical rout: vaginally Jun. 1995 – Dec. 2003 Mesh/graft erosion, n/N: A, 0; B, 3 Age (mean, SD): A, 61 (12); B, 62 (8) Length of follow-up: mean, range, Parity (median, range): A, 2 (1 – 8); B, Concomitant procedures: for Further operation needed for mesh erosion, n/N: A, 0; B, 3 vaginal 51m (12 – 115) 2 (0 – 3) incontinence, A, 28; B, 30; revision BMI (median, range): A, 24.4 (18.6 – Hysterectomy, A, 0; B, 39 Source of funding: NR 31.2); B, 25.4 (16 – 35.1) Wound infection: A, 0; B, 2 Incontinence: A, 22; B 20 Other serious adverse events: incisional hernia, A, 2; B, 1 Pre-operative prolapse stage: uterine prolapse, A, I – II, 20, III, 14; B, I – II, 27, Other adverse events: fever, A, 1; B, 1; III, 11. Perivesical haematoma, A, 2; B, 4; Voiding dysfunction, A, 4; B, 1 Vault/uterine repair: A, 0/36; B, 0/39 OTHER OUTCOMES Number of surgeons: NR, all surgeries Operating time: min, median (range): A, 89 (60 – 110); B, 115 (80 – 135) were performed by or under the

supervision of the senior author.

CASE SERIES Author, year Inclusion criteria: young women, with INTERVENTION (n=19): EFFICACY Banu 199723 or without children, with genital prolapse sacrohysteropexy (the tape was Objective failure: not defined, 0/19 Type of prolapse: uterine fixed on the posterior surface of the Exclusion criteria: women with uterus and anchored to the anterior SAFETY Publication type: full-text abnormal anatomical findings longitudinal ligament, over the sacral Other adverse events: dysmenorrhea during the 1st menstrual cycle, promontory. 2/19; Study design: case series (NR Number of patients lost to follow-up, prospective or retrospective) n/N: NR Type of mesh/graft: Merselene tape No significant intra-operative or post-operative complications. (polyester), Ethicon, tensile strength Setting: teaching hospital Age (range): 17-27y 39kg. Parity (range): 0-3 Country: India Surgical route: abdominal (open) Vault/uterine repair: 0/19 Recruitment/Treatment dates: Concomitant procedures: NR Primary/secondary repair: NR Cervical amputation, 2/19 Length of follow-up: 3-5y Pre-operation prolapse stage: NR Source of funding: a doctor supplied the required Mersilene tape Number of surgeons: NR

153 Author, year Inclusion criteria: women of INTERVENTION (n=30): EFFICACY Barranger 200324 childbearing age with uterovaginal Sacrohysteropexy (anterior vaginal Subjective failure: symptomatic prolapse, 1/30 (3.3%), occurred 3y after Type of prolapse: uterine prolapse who wanted uterine wall and posterior vaginal wall was procedure. preservation attached to the sacral promontory Publication type: full-text (ligament) using mesh as a bridge) Objective failure: modified Bade and Walker classification, recurrent Exclusion criteria: NR uterovaginal prolapse (no details on grade), 1/30 (3.3%), occurred 3y after Study design: case series, Type of mesh/graft: polyester (no. procedure. retrospective, patients were Number of patients lost to follow-up, 2-0 Mersuture, Ethicon Inc., Johnson consecutively recruited. n/N: NR and Johnson com., Belgium) Further operation needed for prolapse (recurrent or de novo): 1/30 (3.3%) Setting: tertiary referral Age (mean, SD): 37.5 (29-43) Surgical route: abdominal (open) Persistent bladder symptoms, n/N: incontinence, 3/19 urogynaecology unit Parity (mean, SD): 2.3 (1-5) Persistent bowel symptoms, n/N: 1/5 Concomitant procedures: Persistent sexual problems, n/N: 3/27 not continue to sexual activities Country: France Vault/uterine repair: 0/30 Burch suspension, 30/30 Posterior colporrhaphy, 30/30 SAFETY Recruitment/Treatment dates: Primary/secondary repair: 29/1 Cervical amputation, 5/30 Blood loss requiring transfusion, n/N: 0/30 1987-1999 Length of follow-up: Pre-operation prolapse stage: Mesh/graft erosion, n/N: vaginal mesh erosion, 1/30 (3.3%), occurred 2y Objective measurements, mean Grade 2 uterovaginal prolapse: 13 after procedure 44.5m (2-156); Grade 3 uterovaginal prolapse: 17 Subjective measurements, mean Further operation needed for mesh erosion, n/N: 1/30 (3.3%), treated 94.6m (8-160) Number of surgeons: 1 (senior successfully with a single transvaginal excision. Source of funding: NR surgeon) New bowel symptoms, n/N: 0/25 New sexual problems, n/N: 1/2 dyspareunia

Infection: urinary tract infection, 1/30 (3.3%); wound infection, 1/30 (3.3%)

Other serious adverse events: Post incision abdominal hernia, 1/30 (3.3%); Intestinal occlusion by the mesh 4y after treatment: 1/30 (3.3%) Other adverse events: Retroperitoneal hematoma, 1/30 (3.3%); Hemorrhage of presacral veins, but not required transfusion, 1/30 (3.3%); Sciatic pain, 1/30 (3.3%)

LENGTH OF HOSPITAL STAY Average (range), 7d (6-9) Author, year Inclusion criteria: women with INTERVENTION (n=13): EFFICACY Leron 200154 uterovaginal prolapse wishing to retain sacrohysteropexy (mesh was Objective failure: 1st degree uterine prolapse, i.e. prolapse reached about Type of prolapse: uterine their uterus. attached to the uterine isthmus and 1cm above the introitus, 1/13 (7.7%) to the anterior longitudinal ligament Publication type: full-text Exclusion criteria: NR of the first or second sacral vertebra SAFETY in a tension free fashion. Mesh was Blood loss, ml: mean (range), 271ml (50-800), n=13 Study design: prospective case Number of patients lost to follow-up, extended to correct a cystocele in series, patients consecutively n/N: NR one women and a rectocele in 3 There were no intra- and postoperative complications. recruited. women) Age (mean, SD): 39(27-60) HOSPITAL STAY Setting: tertiary referral Parity (mean, SD): NR Type of mesh/graft: non- Mean (range), 4.6d (4-6) 154 urogynaecology unit absorbable synthetic mesh, Teflon Vault/uterine repair: 0/13 Country: UK Surgical route: abdominal (open) Primary/secondary repair: NR Recruitment/Treatment dates: Concomitant procedures: Feb 1993 – Mar 1999 Pre-operation prolapse stage: Colposuspension for urethral Length of follow-up: mean (range), Uterine prolapse 2nd degree (prolapse sphincter incompetence, 4/13; 16m (4-49) was at the introitus), 12; Posterior repair, 2/13 3rd degree (prolapse beyond the Source of funding: NR introitus), 1

Number of surgeons: 1

Author, year Inclusion criteria: INTERVENTION (n=20): suspend EFFICACY Joshi 199352 the uterus to the pectineal ligament Subjective failure: prolapse symptoms, 0/20 Type of prolapse: uterine Exclusion criteria: on both sides with Mersilene tape SAFETY Publication type: full-text Number of patients lost to follow-up, Type of mesh/graft: non- Blood loss requiring transfusion, n/N: 0/20 n/N: absorbable synthetic mesh, Study design: case series, Mersilene, Ethicon, Johnson & No early or late morbidity. prospective Age (mean, SD): Johnson Ltd., India Parity (mean, SD): Setting: hospital Surgical route: abdominal (open) Vault/uterine repair: Country: India Concomitant procedures:

Primary/secondary repair: Burch colposuspension, 5/20; Recruitment/Treatment dates: Anterior colporrhaphy, 1/20; NR Pre-operation prolapse stage: Posterior colpoperineorraphy, 1/20 Length of follow-up: range, 6-30m Number of surgeons: Source of funding: NR

155

Other mesh techniques or a mixture of mesh techniques (no studies reported uterine and/or vault prolapse repair separately)

NON-RANDOMISED COMPARATIVE STUDIES Author, year Inclusion criteria: patients with INTERVENTION A (n=45): EFFICACY Demirci 200737 uterovaginal or vaginal vault prolapse. abdominal sacrohysteropexy n=20 or NR. The main indications for surgery were sacrocolpopexy n=25 Type of prolapse: uterine/vault stage 2 prolapse in the Baden-Walker or SAFETY POP-Q system. Type of mesh/graft: non- Blood loss requiring transfusion, n/N: A, 4/45 (2 paravaginal vessels Publication type: full-text absorbable synthetic mesh, during paravaginal repair; 2 presacral vessels during sacrocolpopexy); B, Exclusion criteria: NR polypropylene, Prolene, Ethicon, 0/60 Study design: non-randomised Somverville, US comparative study, retrospective Age (mean, SD): A, 43.7 (14.4); B, 58.2 Damage to surrounding tissue, e.g. urethra, bladder, bowel: bladder (15.6) Surgical route: abdominal (NR injury, A, 1; B, 0; rectal injury, A, 0; B, 1 (during dissection of the pararectal Setting: hospitals Parity (median, range): A, 3.9 (2.7); B, open or laparoscopic) area) 4.3 (2.5) Country: Turkey BMI (mean, SD): A, 28.6 (5.8); B, 29.1 Infection: urinary tract infection, A, 7; B, 6; (5.9) INTERVENTION B (n=60): vaginal Wound infection, A, 5; B, 1; Recruitment/Treatment dates: sacrospinous fixation Febrile morbidity, A, 4; B, 2; Jan. 1999 – Sep. 2005 Vault/uterine repair: A, 7/38; B, 6/54 Vaginal vault infection, A, 0; B, 1. Length of follow-up: 6w postoperative Number of surgeons: 1 Concomitant procedures: Other adverse events: wound dehiscence, A, 3; B, 0; Burch colposuspension, A, 34; B, 45; Urinary retention >5d, A, 7; B, 5. Source of funding: NR Hysterectomy, A, 18; B, 54; 156 Posterior colporrhaphy: A, 0; B, 38 (6 OTHER OUTCOMES used polypropylene mesh); Operating time: min, mean, SD: A, n=18, 191.7 (38.2); B, 140.9 (28.3) Anterior colporrhaphy: A, 34; B, 55 (5 used polypropylene mesh) Length of hospital stay: d, mean, SD: A, n=18, 7.3 (1.8); B, 5.5 (1.9)

Author, year Inclusion criteria: NR INTERVENTION A (n=155): EFFICACY Visco 200177 abdominal sacral colpopexies NR Exclusion criteria: NR Type of prolapse: uterine/vault INTERVENTION B (n=88): SAFETY Age (mean, SD, range): in total: 60.6 abdominal sacral colpoperineopexy Blood loss, ml, mean (SD): in total, N=273, 460ml (390) Publication type: full-text (10.6), 31-84 Additional information: the Mesh/graft erosion, n/N: overall mean time to erosion: 12.3 (16.7)m Study design: non-randomised Parity (median, range): in total: 2 (0- procedure was defined as: a vault comparative study (retrospective) 13) suspension in which the synthetic A, 5/155, median time to erosion: 15.6m (2-33); mesh is attached to the upper third B, 4/88, 12.4m (3-40); Setting: University medical centre Number of surgeons: more than 1 of the vagina and suspended to the C, 4/25, 9.0m (5-15); (member of the Division of Gynecologic sacrum. D, 2/5, 4.1m (2-7) Country: US specialties) INTERVENTION C (n=25): Further operation needed for mesh erosion, n/N: in total, 12/15 were Recruitment/Treatment dates: combined vaginal and abdominal treated successfully with a single transvaginal outpatient excision, 2/15 1 Mar 1992 – 28 Feb 1999 colpoperineopexy with vaginal suture required two vaginal procedures, 1/15 required two transvaginal excisions, passage a transabdominal resection, and anterior transvaginal excision. Length of follow-up, median (SD): in total, 5.8m (1-87); A, 6.5m (1-87); INTERVENTION D (n=5):

B, 4.9m (1-45); C, 6.1m (1-28); D, combined vaginal and abdominal 6.6m (2-11) colpoperineopexy with vaginal mesh placement Source of funding: NR Additional information: the procedure was defined as: combined abdominal-vaginal colpoperineopexy procedure included a vaginal dissection of the posterior vaginal wall. In intervention C, permanent sutures were attached to the perineal body and brought into the abdominal operative field through the posterior cul-de-sac with a long needle or ring carrier. The mesh was then attached to these sutures transabdominally without the mesh entering the vaginal operative field; in intervention D, the mesh was placed and attached vaginally directly to the perineal body with permanent sutures and brought into the abdominal operative field for attachment to the sacrum. 157 Type of mesh/graft, for intervention A, B, C, D: non- absorbable synthetic mesh, Mersilene (Ethicon, Somerville, NJ), N=269, or Gore tex, N=4

Concomitant procedures: in total: Hysterectomy: 45/273 Posterior repair: 84/273 CASE SERIES Author, year Inclusion criteria: patients having INTERVENTION A (n=128): either EFFICACY Elneil 200538 surgery following a diagnosis of vaginal abdominal or laparoscopic Further operation needed for prolapse (recurrent or new at other vault prolapse or uterine prolapse, and sacrocolpopexy, sacrohysteropexy sites), n/N: 6/128, of which 1 had a repeat sacrocolpopexy, 1 had repeat Type of prolapse: uterine/vault having open or laparoscopic or sacrocervicopexy cervicopexy, 3 had an anterior repair and 1 had a rectocele repair. sacrocolpopexy (n=121), hysteropexy Publication type: full-text (n=6), or cervicopexy (n=1) using non- Type of mesh/graft: monofilament absorbable mesh for vault prolapse. polypropylene (Prolene, Ethicon, SAFETY Study design: case series, Johnson & Johnson, Somerville, Damage to surrounding tissue, e.g. urethra, bladder, bowel: retrospective Exclusion criteria: NR New Jersey) with a macroporous Bowel perforations during division of dense adhesions around the vagina, structure. In some patients, porcine 2/128. Setting: 4 hospitals Number of patients lost to follow-up, small intestine submucosa Mesh/graft erosion, n/N: asymptomatic vaginal mesh erosion, 6w, 3/128 n/N: 2y, 39; 3y, 89; 4-5y, 117 absorbable mesh (Surgisis, soft Country: UK tissue graft, Cook, Spencer, Indiana) Further operation needed for mesh erosion, n/N: 3, trimming the mesh

Age (median, range): 62 (38 – 91) was used to reinforce the posterior and suturing the vaginal edges in the clinic. Recruitment/Treatment dates: vaginal wall below the polypropylene NR Number of surgeons: >=1 (from 4 mesh. Wound infection: early postoperative, 2/128 Length of follow-up: mean, range, hospitals) 19m (1.5 – 65) Concomitant procedures: Other adverse events: wound haematoma, early postoperative, 1/128; Colposuspension, 18; Wound dehiscence, 1/128. Source of funding: NR Hysterectomy with or without salpingo-oophorectomy, 6; Anterior and posterior colporrhaphy, 16; Paravaginal repair: 1

Additional information: the pelvic peritoneum was not closed over the mesh.

Author, year Inclusion criteria: patients with INTERVENTION A (n=169): EFFICACY Su 200772 advanced prolapse that included abdominovaginal sacral Subjective failure: symptomatic, 2/169 symptomatic rectocele and perineal colpoperinopexy Type of prolapse: uterine/vault descent who underwent the abdomino Objective failure: apical prolapse, POP-Q >=stage II: 1/169 vaginal sacrocolpoperineopexy. Type of mesh/graft: cellular porcine Publication type: full-text collagen matrix, n=94 (Pelvicol, Any postoperative POP-Q point >=stage II, 38/169, all were stage II (26 Exclusion criteria: NR Bard, Murray, Hill, NJ); or cadaveric anterior prolapse, 10 posterior prolapse, 1 combined anterior and Study design: case series, fascia (Tutoplast, IOP, Costa Mesa, posterior, and 1 anterior, posterior and apical combined.), 18 occurred in 158 retrospective Number of patients lost to follow-up, CA), n=75; polypropylene 6m, 33 between 1m to 1y. n/N: 47/169 in a year, and 125/169 in (Gynemesh, Ethicon, Somerville, NJ) Setting: university hospital two years. n=134; Prolite (Atrium Medical Further operation needed for prolapse (recurrent or new at other Corporation, Hudson, NH), n=18; sites), n/N: 0/169 Country: US Age (mean, SD): 62.4 (9.7) Prolene (Ethicon), n=17. All synthetic Parity (mean, SD): 2.8 (1.4) mesh were macroporous and Recruitment/Treatment dates: Vault/uterine repair: 164/5 monofilament. SAFETY 1 Jul. 2001 – 1 Jul. 2003 Damage to surrounding tissue, e.g. urethra, bladder, bowel: Length of follow-up: mean (SD), Number of surgeons: 5 of whom 3 Surgical route: abdominal (open) Enterotomy, 1/169; range, 14.2 (8.2), 1.5 - 24 were follows and 2 were attending combined with vaginal Cystotomies, 7/169, occurred either during dissection in the vesicovaginal physicians space or space of Retzius. Source of funding: NR Concomitant procedures: Burch procedures/TVT:124/31; Mesh/graft erosion, n/N: synthetic mesh minor erosion (respond to Hysterectomy: 11/169; conservative management [antibiotic cream and vaginal estrogen]), Paravaginal defect repair: 101/169; 10/169; Synthetic mesh major erosion (need further operation), 1/169; Additional information: the Biological graft minor erosion, 3/169; procedure provides continuous graft Median time of occurrence was 3m. support that extends from the perineal body to the sacrum. Mesh New urinary symptoms: urodynamic incontinence, 4/9 at 1y placement overlapped the vaginally placed biograft for the proximal two- Other serious adverse events: ileus, 4/169; thirds of the posterior vaginal wall. This allowed the biograft to be in Other adverse events: pulmonary edema, 1/169; contact with the posterior vaginal Wound cellulitis, 4/169

wall incision and not the synthetic mesh. OTHER OUTCOMES Length of hospital stay: mean (SD), range, 3.34d (0.8), 3-8

159

APPENDIX 6 Excluded studies; reasons for exclusion

Less than 100 patients (N=100)

Addison WA, Livengood CH, III, Sutton GP, Parker RT. Abdominal sacral colpopexy with Mersilene mesh in the retroperitoneal position in the management of posthysterectomy vaginal vault prolapse and enterocele. Am J Obstet Gynecol 1985;153:140-6.

Agarwala N. Laparoscopic sacral colpopexy with Gynemesh as graft material-Experience and results. J Min Invasive Gynecol 2007;14:577-83.

Amrute KV, Eisenberg ER, Rastinehad AR, Kushner L, Badlani GH. Analysis of outcomes of single polypropylene mesh in total pelvic floor reconstruction. Neurourol Urodyn 2007;26:53- 8.

Atlas I, Hallum III AV, Hatch K. Laparoscopic repair of vaginal Gore-Tex erosion after sacral colpopexy. J Gynecol Surg 1995;11:177-80.

Ayav A, Bresler L, Brunaud L, Zarnegar R, Boissel P. Surgical management of combined rectal and genital prolapse in young patients: transabdominal approach. Int J Colorect Dis 2005;20:173-9.

Baessler K, Schuessler B. Abdominal sacrocolpopexy and anatomy and function of the posterior compartment. Obstet Gynecol 2001; 97:678-84.

Baessler K, Stanton SL. Sacrocolpopexy for vault prolapse and rectocele: do concomitant Burch colposuspension and perineal mesh detachment affect the outcome? Am J Obstet Gynecol 2005;192:1067-72.

Bai SW, Kim EH, Shin JS, Kim SK, Park KH, Lee DH. A comparison of different pelvic reconstruction surgeries using mesh for pelvic organ prolapse patients. Yonsei Med J 2005; 46:112-8.

Baker KR, Beresford JM, Campbell C. Colposacropexy with Prolene mesh. Surg Gynecol Obstet 1990;171:51-4.

Barrington JW, Calvert JP. Vaginal vault suspension for prolapse after hysterectomy using an autologous fascial sling of rectus sheath. Br J Obstet Gynaecol 1998; 105:83-6.

Biertho I, Dallemagne B, Dewandre JM, Markiewicz S, Monami B, Wahlen C, Weerts J, Jehaes C. Intravaginal slingplasty: short term results. Acta Chirurg Belg 2004;104:700-4.

Blanchard KA, Vanlangendonck R, Winters JC. Recurrent pelvic floor defects after abdominal sacral colpopexy. J Urol 2006;175:1010-3.

Brieger GM, Korda AR, Houghton CR. Abdomino perineal repair of pulsion enterocele. J Obstet Gynaecol Res 1996;22:151-6.

Brubaker L. Sacrocolpopexy and the anterior compartment: Support and function. Am J Obstet Gynecol 1995;173:1690-6.

Buchsbaum GM, Duecy EE. Local anesthesia with sedation for transvaginal correction of advanced genital prolapse. Am J Obstet Gynecol 2005;193:2173-6.

Carey MP, Slack MC. Transvaginal sacrospinous colpopexy for vault and marked uterovaginal prolapse. BJOG 1994;101:536-40.

160 Clavero PA, Guerrero JA, Salamanca A. Gore-Tex mesh pelvic occlusion and secondary colpopexy: A new surgical technique for posthysterectomy vaginal vault prolapse. Eur J Obstet Gynecol Reprod Biol 2006;126:113-5.

Cosson M, Boukerrou M, Narducci F, Occelli B, Querleu D, Crepin G. Long-term results of the Burch procedure combined with abdominal sacrocolpopexy for treatment of vault prolapse. Int Urogynecol J 2003;14:104-7.

Costantini E, Lombi R, Micheli C, Parziani S, Porena M. Colposacropexy with Gore-tex mesh in marked vaginal and uterovaginal prolapse. Eur Urol 1998;34:111-7.

Cranney A, Feibel R, Toye BW, Karsh J. Osteomyelitis subsequent to abdominal-vaginal sacropexy. J Rheumatol 1994;21:1769-70.

Creighton SM, Stanton SL. The surgical management of vaginal vault prolapse. Br J Obstet Gynaecol 1991;98:1150-4.

Cronje HS, de Beer JA. Abdominal hysterectomy and Burch colposuspension for uterovaginal prolapse. Int Urogynecol J 2004;15:257-60.

Cundiff GW, Harris RL, Coates K, Low VH, Bump RC, Addison WA. Abdominal sacral colpoperineopexy: a new approach for correction of posterior compartment defects and perineal descent associated with vaginal vault prolapse. Am J Obstet Gynecol 1997; 177:1345-53.

David-Montefiore E, Barranger E, Dubernard G, Detchev R, Nizard V, Darai E. Treatment of genital prolapse by hammock using porcine skin collagen implant (Pelvicol). Urology 2005;66:1314-8. de Lima OA, Valente CA. Rectus fascia colpopexy for complete prolapse of the vaginal vault after hysterectomy. Int Surg 1987; 72:119-23.

Delima OA, Valente CA. Rectus Fascia Colpopexy for Complete Prolapse of the Vaginal Vault After Hysterectomy. Int Surg 1987;72:119-23.

Deval B, Rafii A, Azria E, Darai E, Levardon M. Vaginal mesh erosion 7 years after a sacral colpopexy. Acta Obstet Gynecol Scand 2003;82:674-5.

Diana M, Schettini M. Treatment of vaginal vault prolapse with abdominal sacral colpopexy using prolene mesh. Minerva Ginecol 1999;51:349-53.

Diana M, Schettini M, Gallucci M. Treatment of vaginal vault prolapse with abdominal sacral colpopexy using a prolene net. Urogynaecol Int J 1999;13:25-33.

Diana M, Zoppe C, Mastrangeli B. Treatment of vaginal vault prolapse with abdominal sacral colpopexy using prolene mesh. Am J Surg 2000;179:126-8.

Donato DM, Waller-Smith S. The vaginal-psoas suspension repair of uterovaginal prolapse. J Am Coll Surg 2002;195:51-4.

Draca P, Miljkovi S, Jakovljevi B. Abdominal sling surgery--artificial sacro-uterine ligament. Med Pregl 2002; 55:279-85.

Drutz HP, Cha LS. Massive Genital and Vaginal Vault Prolapse Treated by Abdominal- Vaginal Sacropexy with Use of Marlex Mesh - Review of the Literature. Am J Obstet Gynecol 1987;156:387-92.

161 Elliott DS, Chow GK, Gettman M. Current status of robotics in female urology and gynecology. World J Urol 2006;24:188-192.

Elliott DS, Krambeck AE, Chow GK. Long-term results of robotic assisted laparoscopic sacrocolpopexy for the treatment of high grade vaginal vault prolapse. J Urol 2006; 176:655- 9.

FitzGerald MP, Edwards SR, Fenner D. Medium-term follow-up on use of freeze-dried, irradiated donor fascia for sacrocolpopexy and sling procedures. Int Urogynecol J 2004;15:238-42.

Flam F. Sedation and local anaesthesia for vaginal pelvic floor repair of genital prolapse using mesh. Int Urogynecol J 2007;18:1471-5.

Flynn MK, Webster GD, Amundsen CL. Abdominal sacral colpopexy with allograft fascia lata: one-year outcomes. Am J Obstet Gynecol 2005;192:1496-1500.

Fox SD, Stanton SL. Vault prolapse and rectocele: assessment of repair using sacrocolpopexy with mesh interposition. BJOG 2000;107:1371-5.

Gadonneix P, Ercoli A, Salet-Lizee D, Cotelle O, Bolner B, Van Den AM, Villet R. Laparoscopic sacrocolpopexy with two separate meshes along the anterior and posterior vaginal walls for multicompartment pelvic organ prolapse. J Am Assoc Gynecol Laparosc 2004;11:29-35.

Geomini PM, Brolmann HA, van Binsbergen NJ, Mol BW. Vaginal vault suspension by abdominal sacral colpopexy for prolapse: a follow up study of 40 patients. Eur J Obstet Gynecol Reprod Biol 2001;94:234-8.

Godin PA, Nisolle M, Smets M, Squifflet J, Donnez J. Combined vaginal and laparoscopic sacrofixation for genital prolapse using a tacking technique: A series of 45 cases. Gynaecol Endoscop 1999;8:277-85.

Goldberg J, Weinstein M, Fagan M, Nagy M, Nyirjesy P. Gluteal necrotizing myofascitis: an unusual delayed complication of abdominal sacrocolpopexy. Am J Obstet Gynecol 2001;185:1273-4.

Grundsell H, Larsson G. Operative management of vaginal vault prolapse following hysterectomy. Br J Obstet Gynaecol 1984;91:808-11.

Hampton BS, Smilen SW, Kwon C. Recurrent vaginal and concurrent bladder mesh erosion after abdominal sacral colpopexy. J Pelvic Med Surg 2005;11:261-3.

Harer WB, Jr. Round ligament synthetic graft colpopexy. Obstet Gynecol 1994;83:1064-6.

Hart SR, Weiser EB. Abdominal sacral colpopexy mesh erosion resulting in a sinus tract formation and sacral abscess. Obstet Gynecol 2004;103:1037-40.

Hilger WS, Poulson M, Norton PA. Long-term results of abdominal sacrocolpopexy. Am J Obstet Gynecol 2003;189:1606-10.

Hoffman MS. Scarospinous colpopexy in the management of uterovaginal prolapse. J Reprod Med 1996;41:299-303.

Hopkins MP, Rooney C. Entero mesh vaginal fistula secondary to abdominal sacral colpopexy. Obstet Gynecol 2004; 103:1035-6.

162 Imparato E, Aspesi G, Rovetta E, Presti M. Surgical-Management and Prevention of Vaginal Vault Prolapse. Surg Gynecol Obstet 1992;175:233-7.

Iosif CS. Abdominal sacral colpopexy with use of synthetic mesh. Acta Obstet Gynecol Scand 1993;72:214-7.

Jean F, Tanneau Y, Blanc-Louvry I, Leroi AM, Denis P, Michot F. Treatment of enterocele by abdominal colporectosacropexy - Efficacy on pelvic pressure. Colorect Dis 2002;4:321-5.

Kammerer-Doak DN, Rogers RG, Bellar B. Vaginal erosion of cadaveric fascia lata following abdominal sacrocolpopexy and suburethral sling urethropexy. Int Urogynecol J 2002;13:106- 9.

Kauppila O, Punnonen R, Teisala K. Prolapse of the vagina after hysterectomy. Surg Gynecol Obstet 1985;161:9-11.

Kauppila O, Punnonen R, Teisala K. Operative technique for the repair of posthysterectomy vaginal prolapse. Ann Chir Gynaecol 1986;75:242-4.

Kenton KS, Woods MP, Brubaker L. Uncomplicated erosion of polytetrafluoroethylene grafts into the rectum. Am J Obstet Gynecol 2002;187:233-4.

Kerkhof MH, Geomini PJAG, Brolmann HAM. Organic grafts in sacral colpopexy; are they introduced prematurely? Long term results for 29 patients. Gynecol Surg 2005;2:279-282.

Klapper AS, Langer O, Richter A, Zakashanskiy K, Friedman AJ. Abdominal sacral colpopexy using a porcine dermal graft and bone anchors in the elderly overweight patient. J Pelvic Med Surg 2004;10:231-8.

Kohli N, Walsh PM, Roat TW, Karram MM. Mesh erosion after abdominal sacrocolpopexy. Obstet Gynecol 1998;92:999-1004.

Latini JM, Brown JA, Kreder KJ. Abdominal sacral colpopexy using autologous fascia lata. J Urol 2004; 171:1176-9.

Lefranc JP, Atallah D, Camatte S, Blondon J. Longterm followup of posthysterectomy vaginal vault prolapse abdominal repair: a report of 85 cases. J Am Coll Surg 2002;195:352-8.

Leonardo C, Gentili G, Leonardo F. Abdominal sacral colpopexy with Mersilene mesh. Urol Int 2002; 68:6-9.

Lim M, Gonsalves S, Thekkinkattil D, Landon C, Sagar P. Functional outcome of open mesh sacrocolporectopexy surgery for patients with combined vaginal and rectal prolapse. Dis Colon Rectum 2006;49:766-7.

Limb J, Wood K, Weinberger M, Miyazaki F, Aboseif S. Sacral colpopexy using mersilene mesh in the treatment of vaginal vault prolapse. World J Urol 2005;23:55-60.

Lo TS, Horng SG, Huang HJ, Lee SJ, Liang CC. Repair of recurrent vaginal vault prolapse using sacrospinous ligament fixation with mesh interposition and reinforcement. Acta Obstet Gynecol Scand 2005; 84:992-5.

Maloney JC, Dunton CJ, Smith K. Repair of vaginal vault prolapse with abdominal sacropexy. J Reprod Med 1990;35:6-10.

Marinkovic SP, Stanton SL. Triple compartment prolapse: sacrocolpopexy with anterior and posterior mesh extensions. BJOG 2003;110:323-6.

163 Mattox TF, Moore S, Stanford EJ, Mills BB. Posterior vaginal sting experience in elderly patients yields poor results. Am J Obstet Gynecol 2006;194:1462-6.

Molsted-Pedersen L, Rudnicki M, Lose G. Transvaginal repair of enterocele and vaginal vault prolapse using autologous fascia lata graft. Acta Obstet Gynecol Scand 2006;85:874-8.

Nicita G. A new operation for genitourinary prolapse. J Urol 1998;160:741-5.

Nicita G, Li M, V, Filocamo MT, Dattolo E, Marzocco M, Paoletti MC, Villari D. Uterus-sparing vaginal surgery of genitourinary prolapse employing biocompatible material. Urol Int 2005;75:314-8.

Patsner B. Abdominal sacral colpopexy in patients with gynecologic cancer: report of 25 cases with long-term follow-up and literature review. Gynecol Oncol 1999;75:504-8.

Paulson JD. Rectal perforation by a posterior intervaginal slingplasty. J Pelvic Med Surg 2006;12:169-170.

Pilsgaard K, Mouritsen L. Follow-up after repair of vaginal vault prolapse with abdominal colposacropexy. Acta Obstet Gynecol Scand 1999;78:66-70.

Powell JL, Joseph DB. Abdominal sacral colpopexy for massive genital prolapse and posthysterectomy vaginal vault prolapse. J Gynecol Tech 1999;5:45-50.

Rae D, Hawthorn R. Sacrocolpopexy for vaginal vault prolapse: A combined vaginal and laparoscopic approach. Gynaecol Endosc 2002;11:75-79.

Reddy K, Malik TG. Short-term and long-term follow-up of abdominal sacrocolpopexy for vaginal vault prolapse: initial experience in a district general hospital. J Obstet Gynaecol 2002; 22:532-6.

Romero AA, Amundsen CL, Weidner AC, Webster GD. Transvaginal endoscopic removal of eroded mesh after abdominal sacral colpopexy. Obstet Gynecol 2004; 103:1040-2.

Rose SL, Bunten CE, Geisler JP, Sood AK, Sorosky JI. Polypropylene mesh erosion into the bowel and vagina after abdominal sacral colpopexy. J Pelvic Med Surg 2006;12:45-7.

Ross JW, Preston M. Laparoscopic sacrocolpopexy for severe vaginal vault prolapse: Five- year outcome. J Minim Invasive Gynecol 2005;12:221-226.

Rutman MP, Deng DY, Rodriguez LV, Raz S. Repair of vaginal vault prolapse and pelvic floor relaxation using polypropylene mesh. Neurourol Urodyn 2005;24:654-8.

Salman MM, Hancock AL, Hussein AA, Hartwell R. Lumbosacral spondylodiscitis: an unreported complication of sacrocolpopexy using mesh. BJOG 2003;110:537-8.

Scarpero HM, Cespedes RD, Winters JC. Transabdominal approach to repair of vaginal vault prolapse. Tech Urol 2001;7:139-45.

Schettini M, Fortunato P, Gallucci M. Abdominal sacral colpopexy with prolene mesh. Int Urogynecol J 1999;10:295-9.

Sentilhes L, Sergent F, Resch B, Verspyck E, Descamps P, Marpeau L. Midterm Follow-up of High-Grade Genital Prolapse Repair by the Trans-obturator and Infracoccygeal Hammock Procedure after Hysterectomy. Eur Urol 2007;51:1065-1072.

164 Shah DK, Paul EM, Rastinehad AR, Eisenberg ER, Badlani GH. Short-term outcome analysis of total pelvic reconstruction with mesh: the vaginal approach. J Urol 2004;171:261- 3.

Shrotri MS, Hershman MJ, Farquharson RG. Laparoscopic sacral colpopexy for posthysterectomy recurrent vaginal vault prolapse. Minim Invasive Ther Allied Technol 1997;6:357-9.

Silva-Filho AL, Santos-Filho AS, Figueiredo-Netto O, Triginelli SA. Uncommon complications of sacrospinous fixation for treatment of vaginal vault prolapse. Arch Gynecol Obstet 2005;271:358-62.

Spennacchio M, Buonaguidi A, Bertola E, Guareschi BM, Vignali M. Abdominal sacral colpopexy for vaginal vault prolapse: A retrospective study. J Gynecol Surg 1997;13:77-81.

Surico N, Ruspa G, Longo D, Salini A, Arnulfo A, Baj G. Abdominal sacrocolpopexy with collagen biosynthetic mesh: Analysis of 21 cases. J Gynecol Surg 2002;18:45-48.

Tahseen S. E. Long-term patient satisfaction and function after sacrocolpopexy for vault prolapse repair. J Pelvic Med Surg 2007;13:133-9.

Taylor GB, Moore RD, Miklos JR. Osteomyelitis secondary to sacral colpopexy mesh erosion requiring laminectomy. Obstet Gynecol 2006;107:475-7.

Unger JB. A persistent sinus tract from the vagina to the sacrum after treatment of mesh erosion by partial removal of a GORE-TEX soft tissue patch. Am J Obstet Gynecol 1999;181:762-3.

Valaitis SR, Stanton SL. Sacrocolpopexy: a retrospective study of a clinician's experience. Br J Obstet Gynaecol 1994;101:518-22.

Van Der Weiden RMF, Bergkamp ABM. Colposacropexy with mesh or collagen implant and titanium bone anchors placed in sacral segments 3 and 4. J Pelvic Med Surg 2003;9:9-14.

Virtanen H, Hirvonen T, Makinen J, Kiilholma P. Outcome of thirty patients who underwent repair of posthysterectomy prolapse of the vaginal vault with abdominal sacral colpopexy. J Am Coll Surg 1994;178:283-7.

Wasserman GA. Correction of a vault suspension graft complication. J Am Assoc Gynecol Laparoscop 1995;2:359-61.

Wille S, Braun M, Heidenreich A, Hofmann R, Engelmann U. Sacral colpopexy with concurrent Burch colposuspension in patients with vaginal vault prolapse. Urol Int 2006;76:339-44.

Yan A, Anne M, Karine A, Vanessa F, Christophe P, Anne T, Patrick M. Cystocele repair by a synthetic vaginal mesh secured anteriorly through the obturator foramen. Eur J Obstet Gynecol Reprod Biol 2004;115:90-4.

165 No Mesh Used (N=18)

Andreas KA, Kohler C, Diebolder H, Vercellino F, Krause N, Schneider A. Repair of prolapse with vaginal sacrocolporectopexy: Technique and results. Eur J Obstet Gynecol Reprod Biol 2005;122:237-42.

Barker GK, Viggers JC, Mason T, Smith GO. Transvaginal sacrospinous colpopexy--a new and easier way. Aust N Z J Obstet Gynaecol 1997;37:355-7.

Brown WE, Hoffman MS, Bouis PJ, Ingram JM, Hopes SL. Management of vaginal vault prolapse: retrospective comparison of abdominal versus vaginal approach. J Florida Med Assoc 1989;76:249-52.

Chang WC, Huang SC, Sheu BC, Hsu WC, Torng PL, Chow SN et al. Using Veronikis ligature carrier to simplify transvaginal sacrospinous colpopexy. Acta Obstet Gynecol Scand 2006;85:721-5.

Chapin DS. Teaching sacrospinous colpopexy. Am J Obstet Gynecol 1997;177:1330-6.

Dietz V. The effectiveness of the sacrospinous hysteropexy for the primary treatment of uterovaginal prolapse. Int Urogynecol J Pelvic Floor Dysfunct 2007;18:1271-6.

Hefni M, El Toukhy T, Bhaumik J, Katsimanis E. Sacrospinous cervicocolpopexy with uterine conservation for uterovaginal prolapse in elderly women: an evolving concept. Am J Obstet Gynecol 2003;188:645-50.

Hefni MA, El Toukhy TA. Long-term outcome of vaginal sacrospinous colpopexy for marked uterovaginal and vault prolapse. Eur J Obstet Gynecol Reprod Biol 2006;127:257-63.

Hefni MA. Sacrospinous colpopexy at vaginal hysterectomy: Method, results and follow up in 75 patients. J Obstet Gynaecol 2000;20:58-62.

Hewson AD. Transvaginal sacrospinous colpopexy for posthysterectomy vault prolapse. Aust N Z J Obstet Gynaecol 1998;38:318-24.

Krause HG, Goh JTW, Sloane K, Higgs P, Carey MP. Laparoscopic sacral suture hysteropexy for uterine prolapse. Int Urogynecol J 2006;17:378-81.

Maher CF, Murray CJ, Carey MP, Dwyer PL, Ugoni AM. Iliococcygeus or sacrospinous fixation for vaginal vault prolapse. Obstet Gynecol 2001;98:40-4.

Maher CF, Cary MP, Slack MC, Murray CJ, Milligan M, Schluter P. Uterine preservation or hysterectomy at sacrospinous colpopexy for uterovaginal prolapse? Int Urogynecol J 2001;12:381-4.

Natale F, Mako A, Panei M, Weir J, Antomarchi F, Cervigni M. Prospective randomized controlled study between two different procedures to suspend the vaginal vault: High levator myorraphy and uterosacral vaginal vault suspension. Neurourol Urodyn 2007;26:608-10.

Ostrzenski A. Laparoscopic retroperitoneal hysteropexy. A randomized trial. J Reprod Med 1998;43:361-6.

Pohl JF, Frattarelli JL. Bilateral transvaginal sacrospinous colpopexy: preliminary experience. Am J Obstet Gynecol 1997;177:1356-61.

166 Rogers A, Barker G, Viggers J, Mason T, Swan J, Mayall P. A review of 165 cases of transvaginal sacrospinous colpopexy performed by the Endo Stitch technique. Aust N Z J Obstet Gynaecol 2001;41:61-4. van Brummen HJ, van de PG, Aalders CI, Heintz AP, van der Vaart CH. Sacrospinous hysteropexy compared to vaginal hysterectomy as primary surgical treatment for a descensus uteri: effects on urinary symptoms. Int Urogynecol J 2003;14:350-5

Data not presented separately for uterine or vault repair or for mesh techniques (N=14)

Antiphon P, Elard S, Benyoussef A, Fofana M, Yiou R, Gettman M et al. Laparoscopic promontory sacral colpopexy: is the posterior, recto-vaginal, mesh mandatory? Eur Urol 2004; 45:655-61.

Baessler K, Hewson AD, Tunn R, Schuessler B, Maher CF. Severe mesh complications following intravaginal slingplasty. Obstet Gynecol 2005; 106:713-6.

Brieger GM, MacGibbon AL, Atkinson KH. Sacrospinous colpopexy. Aust N Z J Obstet Gynaecol 1995;35:86-7.

Brubaker L, Cundiff G, Fine P, Nygaard I, Richter H, Visco A et al. A randomized trial of colpopexy and urinary reduction efforts (CARE): Design and methods. Control Clin Trials 2003;24:629-42.

Cosson M, Occelli B, Narducci F, Ego A, Querleu D, Crepin G. Causes of failure of abdominal colposacropexy for the treatment of genital and vaginal vault prolapse. J Gynecol Surg 2000;16:141-8.

Cronje HS. Colposacrosuspension for severe genital prolapse. Int J Gynaecol Obstet 2004;85:30-5.

Fatton B, Amblard J, Debodinance P, Cosson M, Jacquetin B. Transvaginal repair of genital prolapse: preliminary results of a new tension-free vaginal mesh (Prolift technique)--a case series multicentric study. Int Urogynecol J 2007;18:743-52.

Pajoncini C, Costantini E, Bini V, Tascini M, Rociola W, Guercini F et al. VLPP and surgical outcome: Preliminary data. Neurourol Urodyn 2003;22:373-4.

Seracchioli R, Hourcabie JA, Vianello F, Govoni F, Pollastri P, Venturoli S. Laparoscopic treatment of pelvic floor defects in women of reproductive age. J Am Assoc Gynecol Lap 2004;11:332-5.

Su KCH, Mutone MF, Terry CL, Hale DS. Abdominovaginal sacral colpoperineopexy: patient perceptions, anatomical outcomes, and graft erosions. Int Urogynecol J 2007;18:503-11.

Sullivan ES, Longaker CJ, Lee PYH. Total pelvic mesh repair: A ten-year experience. Dis Colon Rectum 2001;44:857-63.

Thompson PK, Pugmire JE, Sangi-Haghpeykar H. Abdominal sacrocolpopexy utilizing Gore- Tex in genital prolapse: Unresolved issues. J Pelvic Med Surg 2004;10:311-7.

Vancaillie TG. MycroMesh is not a suitable soft tissue prosthesis for repair of the defective vaginal wall. J Am Assoc Gynecol Lap 2003; 10:424-5.

167 Vanlindert ACM, Groenendijk AG, Scholten PC, Heintz APM. Surgical Support and Suspension of Genital Prolapse, Including Preservation of the Uterus, Using the Gore-Tex(R) Soft-Tissue Patch (A Preliminary-Report). Eur J Obstet Gynecol Reproduct Biol 1993;50:133- 9.

No data reported (N=4)

Costantini E, Zucchi A, Giannantoni A, Mearini L, Bini V, Porena M. Must colposuspension be associated with sacropexy to prevent postoperative urinary incontinence? Eur Urol 2007;51:788-94.

Cronje HS, de Beer JA, Bam R. The pathophysiology of an enterocele and its management. J Obstet Gynaecol 2004; 24:408-13.

Rane A, Lim YN, Withey G, Muller R. Magnetic resonance imaging findings following three different vaginal vault prolapse repair procedures: a randomised study. Aust N Z J Obstet Gynaecol 2004;44:135-9.

Yau JL, Rahn DD, McIntire DD, Schaffer JI, Wai CY. The natural history of posterior vaginal wall support after abdominal sacrocolpopexy with and without posterior colporrhaphy. Am J Obstet Gynecol 2007;196:e45-e47.

Other reasons (N=2)

Weidner AC, Cundiff GW, Harris RL, Addison WA. Sacral osteomyelitis: an unusual complication of abdominal sacral colpopexy. Obstet Gynecol 1997;90:689-91. [case report]

Woodruff AJ. Abdominal sacral colpopexy: Surgical pearls and outcomes. Curr Urol Rep 2007;8:399-404. [review]

168 APPENDIX 7 Detailed quality assessment results for included studies: randomised controlled trials (full-text), n=5

Study ID Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Abdominal sacrocolpopexy Benson 199627 + + + + - + + ? - + + (9%, 8/88) + ? Culligan 200535 + + + + + + + + + + + (11%, 11/100) + ? Lo 199856 + ? + + + + + ? ? + + (15%, 21/138) + ? Maher 200457 + ? + + + + + ? ? + + (6%, 6/95) + ? Vaginal infracoccygeal sacropexy (no RCTs reported this technique) Vaginal or abdominal uterine suspension sling Roovers 200467 + + + + + + + ? ? + + (5%, 4/82) + +

Note: 169

1. See Appendix 2 for checklist for quality assessment of randomised controlled trials 2. + Yes; - No; ? Unclear

APPENDIX 8 Detailed quality assessment results for non-randomised comparative studies, n=17 Study ID Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Abdominal sacrocolpopexy Altman 200521 + + + ? - + + ? + + + ? - - ? + + - Bai 200622 + - + ? - - - ? + + + ? + + (35%, 39/113) - + + - Begley 200525 + + + ? - + + ? + + + ? + - ? + + - Costantini 200533 + + + + + + + ? + + + ? + + (4%, 3/75) + + + - Govier 200543 + + + + - ? + ? + + + ? + - ? - + - Gregory 200544 + + + + - ? + ? + + + + + + (55%, 45/82) - + + + Griffis 200645 + - + - - ? + ? + + + ? + - ? + + + Hardiman 199647 + - + ? ? + + ? + + + ? + - ? + + - Hsiao 200751 + - + + - + + ? + + + ? + - ? - + - Marcickiewicz 200758 + - + ? - + + - + + + + + - ? + + - Ng 200461 + + + ? - + + ? + + + ? + + (39%, 69/177) - + + -

170 Paraiso 200564 + + + + - + + - + + + ? + - ? + + - Sze 199973 + + + + - + - ? + + + ? + + (14%, 16/117) + + + - Visco 200177 + - + ? - ? + ? + + + ? - - ? + + + Young 200480 + - ? - + + + ? + - + - - - ? + + - Vault infracoccygeal sacropexy Neuman 200760 + + + ? + - + ? + + + ? + - ? + + - Vaginal or abdominal uterine suspension sling Costantini 200533 + + + + + + + ? + + + ? + + (4%, 3/75) + + + - Other mesh techniques or a mixture of above mesh techniques Demirci 200737 + + + ? - - + ? + - + ? - - ? + + - Visco 200177 + - + ? - ? + ? + + + ? - - ? + + + Note: 1. See Appendix 3 for checklist for quality assessment of non-randomised comparative studies 2. + Yes; - No; ? Unclear 3. Costantini 200533 and Visco 200177 compared different mesh techniques.

APPENDIX 9 Detailed quality assessment results for case series, n=29 Study ID Q1 Q2 Q3 Q4 Q5 Q7 Q8 Q9 Q10 Q11 Q13 Q14 Q15 Q17 Abdominal sacroclopopexy 26 Bensinger 2005 + + + ? - + ? + + + + + (47%, 57/121) - + 28 Bradley 2007 + + + ? + + ? + + + + + (5%, 17/322) + + 30 Brizzolara 2003 + - + ? - + ? + + + + - ? + 34 Culligan 2002 + - + ? - + ? + + + + + (73%, 123/169) - + 36 De Vries 1995 + + + + - + ? + + + + - ? + 40 Fedorkow 1993 + - + ? + + ? + - + ? - ? + 50 Higgs 2005a + - + + - + ? + + + + + (26%, 37/140) + + 49 Higgs 2005b + - + + - + ? + + + + + (37%, 55/148) - + 55 Lindeque 2002 + + + + ? + ? + + + + - ? + 71 Snyder 1991 + - + ? - + ? + - + + + (10%, 15/147) + + 75 Timmons 1992 + + + ? + + ? + + + + - ? +

171 79 Wu 2006 + + + ? - + ? + + + + - ? + Vaginal infracoccygeal sacropexy 39 Farnsworth 2002 + + + + + + ? + + + + + (2%, 2/93) + + 41 Foote 2007 + - + ? - - ? + + + - - ? + 42 Ghanbari 2006 + + + ? + - ? + + + ? - ? + 48 Hefni 2007 + + + + + + ? + + + + - ? + 53 Jordaan 2006 + + + ? - + ? + + + + - ? + 62 Oliver 2006 + - + ? - + ? + - + - - ? + 63 Petros 2001 + + + ? + + ? + + + + + (47%, 35/75) - + 65 Petros 2005 + + + ? + + ? + + + - + (2%, 1/67) + + 69 Sentihes 2007 + + + ? + + ? + + + + - ? + 70 Sivaslioglu 2005 + - + ? ? - ? + + + + - ? + 76 Vardy 2007 + + + ? + + ? + + + + - ? + Vaginal or abdominal uterine suspension sling

23 Banu 1997 + + + ? ? + ? + + ? + - ? + 24 Barranger 2003 + + + + - + ? + + + + - ? + 54 Leron 2001 + + + + + + ? + + + + - ? + 52 Joshi 1993 + - + ? + + ? + - - + + (5%, 1/20) ? + Other mesh techniques or a mixture of above mesh techniques 38 Elneil 2005 + - + + - + ? + + + + + (30%, 39/128) - + 72 Su 2007 + - + + - + ? + + + + + (74%, 125/169) - + 1. See Appendix 3 for checklist for quality assessment of non-randomised comparative studies; the same checklist was adapted for case series by taking out question 6, 12, 16 and 18; 2. + Yes; - No; ? Unclear; 172

APPENDIX 10 Summary of outcomes: sacrocolpopexy, efficacy

Table Uterine prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that occurred reported as (follow up) Subjective failure (no studies reported this outcome) Objective failure RCTs (abs.) Braun 200729 Hysterectomy+sacrocolpopexy 0/23 (0%) Vaginal 1/24 (4.2%) Vault prolapse, Occurred 8m after hysterectomy no definition surgery (33m (20-41)) Non-rand. comp. Costantini 200533 A: sacrohysteropexy A: 0/36 (0%) - - Recurrent 51m (12-115) B: hysterectomy+sacrocolpopexy B: 0/39 (0%) cervical/vault prolapse (<6cm above the hymen) De novo prolapse at other sites (no studies reported this outcome) Further operation needed for prolapse (recurrent or de novo) Non-rand. comp. Costantini 200533 A: sacrohysteropexy A: 0/36 (0%) - - Operation 51m (12-115) B: hysterectomy+sacrocolpopexy B: 0/39 (0%) needed persistent urinary incontinence in women having this symptom at baseline persistent bowel symptoms in women having these symptoms at baseline persistent sexual problems in women having these problems at baseline

173 Table Vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that occurred reported as (follow up) Subjective failure RCTs Maher 200457 Sacrocolpopexy, non-absorbable 3/46 (6.5%) Sacrospinous 4/43 (9.3%) Prolapse 24m (6-60) synthetic mesh colpopexy symptoms Non-rand. comp. Marcickiewicz Laproscopic sacrocolpopexy, 13/60 (21.7%) Sacrospinous 6/51 (11.8%) Sensation of A, 34 (13-60) 200758 non-absorbable synthetic mesh colpopexy prolapse B, 38 (7-108) Case series Higgs 2005a50 Laproscopic sacrocolpopexy, 22/103 (21.4%) - - Prolapse 66m (37-124) non-absorbable synthetic mesh symptoms no change or worse Objective failure RCTs Culligan 200535 A: abdominal (open) A, 0/45 - - POP-Q stage II 1y Sacrocolpopexy, non-absorbable B, 0/44 or more at C synthetic mesh point B: abdominal (open) Sacrocolpopexy, absorbable biological graft Maher 200457 Sacrocolpopexy, non-absorbable 2/46 (4.3%) Sacrospinous 8/43 (18.6%) Vault prolapse 24m (6-60) synthetic mesh colpopexy >= grade II Non-rand. comp. Govier 200543 A, polypropylene A, 0/24 - - Recurrent vault One was detected B, silicone-covered polyethylene B, 2/21(9.5%) prolapse, no after mesh removal, definition <20m, the other at 24m. Marcickiewicz Laproscopic sacrocolpopexy, 0/60 (0%) Sacrospinous 0/51 (0%) Vault prolapse A, 34m (13-60) 200758 non-absorbable synthetic mesh colpopexy >=grade II B, 38m (7-108) Case series Higgs 2005a50 Laproscopic sacrocolpopexy, 4/66 (6.1%) - - Recurrent vault 66m (37-124) non-absorbable synthetic mesh prolapse >= stage II De novo prolapse at other sites Non-rand. comp. Marcickiewicz Laproscopic sacrocolpopexy, 10/32 (31.3%) Sacrospinous 4/28 (14.3%) Cystocele A, 34m (13-60) 200758 non-absorbable synthetic mesh colpopexy B, 38m (7-108) Further operation needed for prolapse (recurrent or de novo) Non-rand. comp. Marcickiewicz Laproscopic sacrocolpopexy, 15/60 (25%) Sacrospinous 4/51 (8%) Recurrent Occurred in the initial 200758 non-absorbable synthetic mesh colpopexy period, in 6 of them, the mesh was attached to the vault with staples. A, 34m (13-60) B, 38m (7-108) persistent urinary incontinence in women having this symptom at baseline persistent bowel symptoms in women having these symptoms at baseline persistent sexual problems in women having these problems at baseline RCTs Maher 200457 Sacrocolpopexy, non-absorbable 4/9 (44.4%) Sacrospinous 4/7 (57.1%) Dyspareunia 24m (6-60) synthetic mesh colpopexy

174 Table Uterine and/or vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Subjective failure RCTs Benson 199627 Sacrocolpopexy 6/38 (15.8%) Bilateral 14/42 Symptomatic 2.5y (1 – 5.5) sacrospinous (33.3%) prolapse suspension Non-rand. comp. Ng 200461 Abdominal (open) 2/60 (3.3%) Sacrospinous 11/48 Symptomatic A, 18m (1-48) sacrocolpopexy, non-absorbable fixation (22.9%) prolapse B, 13m (1-29) synthetic mesh (Gore-Tex, Amid type II) Gregory A, abdominal (open) A, 2/19 (10.5%) - - Non-optimal A, 26m (10-48) 200544 sacrocolpopexy, non-absorbable B, 7/18 (38.9%) surgical B, 21m (11-34) synthetic mesh (Marlex, Amid outcome, no type I) definition B, abdominal (open) sacrocolpopexy, biological graft (fascia lata) Case series De Vries Abdominal (open) 22/72 (30.6%) - - Prolapse- 4y (1-13) 199536 sacrocolpopexy, non-absorbable related synthetic mesh (Mersilene, Amid complaints type III) Higgs 2005b49 Abdominal (open) 21/90 (23.3%) - - Prolapse 45m (15) sacrocolpopexy, a mixture of symptoms absorbable and non-absorbable mesh/graft Synder 199171 Abdominal (open), a mixture of 34/132 (25.8%) - - Pelvic 43m (1m-17y) absorbable and non-absorbable pressure mesh/graft Objective failure RCTs Lo 199856 Abdominal (open), non- 0/52 (0%) Sacrospinous NR Vault prolapse 2.1y (1-5.2y) absorbable synthetic mesh suspension >= stage II (Mersilene mesh, Amid type III) Non-rand. comp. Hardiman Sacrocolppexy, non-absorbable 1/80 (1.3%) Sacrospinous 3/125 Vault prolapse, A, 4y 199647 synthetic mesh (Marlex, Amid suspension (2.4%) no definition B, 2y type I) A+B, 6m-5y Ng 200461 Abdominal (open) 3/60 (5.0%) Sacrospinous 13/48 Vault prolapse A, 18m (1-48) sacrocolpopexy, non-absorbable fixation (27.1%) >= grade II B, 13m (1-29) synthetic mesh (Gore-Tex, Amid type II) Case series Brizzolara Sacrocolpopexy, non-absorbable 3/124 (2.4%) - - Vault prolapse 36m (0-74) 200330 synthetic mesh (Prolene) or (no definition) biological graft (allograft) Culligan 200234 Sacrocolpopexy, synthetic mesh 0/44 (0%) - - Apical failures >=4y (NR details) Higgs 2005b49 Abdominal (open) 2/64 (3.1%) - - Vault proalpse 45m (15) sacrocolpopexy, a mixture of >=stage 2 absorbable and non-absorbable mesh/graft Lindeque Abdominal (open) 3/262 (1.1%) - - Vault prolapse, All >=16m 200255 sacrocolpopexy, a mixture of no definition absorbable and non-absorbable mesh/graft De novo prolapse at other sites Case series Bradley 200728 Abdominal (open) 3/193 (1.6%) - - Posterior 1y sacrocolpopexy, mixture of biological and non-absorbable mesh Further operation deeded for prolapse (recurrent or de novo) Case series Higgs 2005b49 Abdominal (open) 13/90 (14.4%) - - NR 45m (15) sacrocolpopexy, a mixture of absorbable and non-absorbable mesh/graft Lindeque Abdominal (open) 4/262 (1.5%) - - NR All >=16m 200255 sacrocolpopexy, a mixture of absorbable and non-absorbable mesh/graft

175 Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) persistent urinary incontinence in women having this symptom at baseline Case series Bradley 200728 Abdominal (open) 43/169 (25.4%) - - Persistant 1y sacrocolpopexy, mixture of bothersome biological and non-absorbable irritative mesh symptoms persistent bowel symptoms in women having these symptoms at baseline Case series Bradley 200728 Abdominal (open) 7/54 (13.0%) - - Need of digital 1y sacrocolpopexy, mixture of assistance biological and non-absorbable mesh persistent sexual problems in women having these problems at baseline Case series Bradley 200728 Abdominal (open) 25/60 (41.7%) - - Dyspareunia 1y sacrocolpopexy, mixture of biological and non-absorbable mesh

176 APPENDIX 11 Summary of outcomes: sacrocolpopexy, safety

Table Uterine prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Blood loss, ml, mean (SD)/median (range) Non-rand. comp. Costantini A, sacrohysteropexy A, 200 ml, n=36 - - - - 200533 B, hysterectomy+sacrocolpopexy B, 325 ml, n=39 Case series Wu 200679 Hysterectomy+sacrocolpopexy 328 (163) ml, - - - - n=101 Number of women with blood loss requiring blood transfusion Non-rand. comp. Costantini A, sacrohysteropexy A, 2/36 (5.6%) - - Transfusion - 200533 B, hysterectomy+sacrocolpopexy B, 2/39 (5.1%) required Damage to surrounding organs during operation (no studies reported this outcome) Mesh erosion RCTs (abstracts) Braun 200729 Hysterectomy+sacrocolpopexy, 1/23 (4.3%) Vaginal Not Mesh erosion NR, 33m (20-41) combination of absorbable and hysterectomy applicable non-absorbable material (Vypro) Non-rand. comp. Costantini A, sacrohysteropexy A, 0/36 (0%) - - Mesh erosion NR, 51m (12-115) 200533 B, hysterectomy+sacrocolpopexy B, 3/39 (7.7%) Both A and B used non-absorbable synthetic mesh (Marlex, Amid type I) Griffis 200645 A, hysterectomy+ total A, 8/76 (10.5%) - - Mesh erosion Mean time to hysterectomy B, 1/28 (3.6%) occur, 10m, B, hysterectomy+ supracervical median 4m, range hysterectomy 0.4-53.3m Both A and B used non-absorbable synthetic mesh (Amid type I or III) Case series Wu 200679 Hysterectomy+sacrocolpopexy, 7/101 (6.9%) - - Mesh erosion Median time to non-absorbable synthetic mesh occur, 8.4m (Amid II or III) (range 1.4-13) Operation for mesh erosion Non-rand. comp. Costantini A, sacrohysteropexy A, 0/36 (0%) - - Vaginal revision - 200533 B, hysterectomy+sacrocolpopexy B, 3/39 (7.7%) Both A and B used non-absorbable synthetic mesh (Marlex, Amid type I) Griffis 200645 A, hysterectomy+ total A, 4/76 (5.3%) - - Excision in the - hysterectomy B, 0/28 (0%) operation room. B, hysterectomy+ supracervical The others (A hysterectomy 4/76, B 1/28) Both A and B used non-absorbable were managed synthetic mesh (Amid type I or III) by estrogen cream, antibiotic cream, silver nitrate, or local excision in the office) De novo urinary symptoms in women who did not have these symptoms at baseline De novo bowel symptoms in women who did not have these symptoms at baseline De novo sexual problems in women who did not have these symptoms at baseline Infection Non-rand. comp. Costantini A, sacrohysteropexy A, 0/36 (0%) - - Wound infection NR, 51m (12-115) 200533 B, hysterectomy+sacrocolpopexy 1/36 (2.8%) Fever B, 2/39 (5.1%) 1/39 (2.6%) Other serious adverse effects RCTs (abstracts) Braun 200729 Hysterectomy+sacrocolpopexy 1/23 (4.3%) Vaginal 0/24 (0%) Incisional hernia NR, follow up 33m hysterectomy (20-41) Non-rand. comp. Costantini A, sacrohysteropexy A, 2/36 (5.6%) - - Incisional hernia NR, follow up 51m 200533 B, hysterectomy+sacrocolpopexy B, 1/39 (2.6%) (12-115)

177 Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Other adverse effects RCTs (abstracts) Braun 200729 Hysterectomy+sacrocolpopexy 1/23 (4.3%) Vaginal 0/24 (0%) Subaponeurotic NR, 33m (20-41) hysterectomy hematoma Non-rand. comp. Costantini A, sacrohysteropexy A, 2/36 - - Perivesical NR, 51m (12-115) 200533 B, hysterectomy+sacrocolpopexy 4/36 haematoma; B, 4/39 Voiding 1/39 dysfunction.

178 Table Vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Blood loss, ml, mean (SD)/median (range) RCTs Culligan Abdominal (open) A, 247 (148) ml, - - - - 200535 sacrocolpopexy: n=54 A, non-absorbable B, 265 (261), n=46 synthetic mesh B, absorbable biological graft Maher 200457 Sacrocolpopexy, non- 362 (239) ml, 100- Sacrospinous 306 (201) ml, - - absorbable synthetic 1100, n=47 colpopexy 100-1000, mesh n=48 Non-rand. comp. - - Marcickiewicz Sacrocolpopexy, non- 64 (0-1000) ml, n=60 Sacrospinous 61 (0-400) ml, - - 200758 absorbable synthetic (laparoscopic) colpopexy n=51 mesh Paraiso 200564 Sacrocolpopexy, a A, 172 (166), 0-1000 - - - - mixture of absorbable ml, n=56 and non-absorbable (laparoscopic) mesh: B, 234 (149), 50-700 A, laparoscopic ml , n=61 B, abdominal open (abdominal, open) Case series Wu 200679 Abdominal (open) 298 (193) ml, n=212 - - - - sacrocolpopexy, non- absorbable synthetic mesh Number of women with blood loss requiring blood transfusion RCTs Culligan Abdominal (open) A, 1/54 - - Transfusion needed. - 200535 sacrocolpopexy: B, 0/46 A, non-absorbable synthetic mesh B, absorbable biological graft Maher 200457 Sacrocolpopexy, non- 1/47 (2.1%) Sacrospinous 1/48 (2.1%) Transfusion needed. - absorbable synthetic colpopexy mesh Non-rand. comp. Marcickiewicz Laparoscopic 1/60 (1.7%) Sacrospinous 2/51 (3.9%) Transfusion needed. - 200758 sacrocolpopexy, non- colpopexy absorbable synthetic mesh Paraiso 200564 Sacrocolpopexy, a A, 1/56 - - Transfusion needed. - mixture of absorbable B, 1/61 and non-absorbable mesh: A, laparoscopic B, abdominal open Case series Fedorkow Sacrocolpopexy, non- 0/149 - - Intraoperative - 199340 absorbable synthetic hemorrhage mesh Damage to surrounding organs during operation RCTs Culligan Abdominal (open) A, 1/54 (1.9%) - - Bladder injury - 200535 sacrocolpopexy: B, 0/46 A, non-absorbable synthetic mesh B, absorbable biological graft Maher 200457 Sacrocolpopexy, non- 1/47 (2.1%) Sacrospinous 1/48 (2.1%) Cystotomy - absorbable synthetic colpopexy mesh Non-rand. comp. Marcickiewicz Laparoscopic 2/60 Sacrospinous 2/51 Bowel injury; - 200758 sacrocolpopexy, non- 1/60 colpopexy 0/51 Puncture of the inferior absorbable synthetic 1/60 0/51 epigastric artery mesh (In total, 4/60, 6.7%) (In total, 2/51, required laparotomy; 3.9%) Colon perforation required laparotomy

179 Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Paraiso 200564 Sacrocolpopexy, a A, 1/56 - - Cystotomy/bladder - mixture of absorbable 1/56 sutures; and non-absorbable (In total, 2/56, 3.6%) Enterotomy mesh: B, 2/61 A, laparoscopic 2/61 B, abdominal open (In total, 4/61, 6.6%) Case series Higgs 2005a50 Laparoscopic 2/103 (1.9%) - - Bowel perforation - sacrocolpopexy, non- absorbable synthetic mesh Mesh erosion RCTs Culligan Abdominal (open) A, 2/54 - - Mesh erosion NR, 1y 200535 sacrocolpopexy: B, 0/46 A, non-absorbable synthetic mesh B, absorbable biological graft Maher 200457 Sacrocolpopexy, non- 1/47 (2.1%) Sacrospinous Not applicable Mesh erosion NR, 24m (6-60) absorbable synthetic colpopexy mesh Non-rand. comp. Govier 200543 Sacrocolpopexy, non- A, 0/24 - - Mesh erosion Detected at 4, 7, absorbable synthetic B, 4/21 12, and 20m mesh: A, polypropylene; respectively. B, silicone-covered polyethylene Paraiso 200564 Sacrocolpopexy, a A, 2/56 - - Mesh erosion NR, A, 14m (1-46) mixture of absorbable B, 1/61 B, 16m (1-73) and non-absorbable mesh: A, laparoscopic B, abdominal open Case series Fedorkow Sacrocolpopexy, non- 0/149 (0%) - - Mesh erosion NR follow-up time 199340 absorbable synthetic mesh Griffis 200645 Sacrocolpopexy, non- 16/196 (8.2%) - - Mesh erosion Occurred in a absorbable synthetic mean of 10.4m, mesh median 4m, range 0.4-54m Higgs 2005a50 Sacrocolpopexy, non- 9/103 (8.7%) - - Mesh erosion All presented absorbable synthetic >=6m after mesh surgery, with some occurring over 3y later Wu 200679 Sacrocolpopexy, non- 10/212 (4.7%) - - Mesh erosion Median (range) absorbable synthetic time to occure, mesh 8.4m (1.4-13.1) Operation need for mesh erosion Non-rand. comp. Govier 200543 Sacrocolpopexy, non- A, 0/24 - - In B, 4 mesh erosion, 1 One was removed absorbable synthetic B, 5/21 mesh infection transvaginally, mesh: A, polypropylene; four was B, silicone-covered abdominally polyethylene Paraiso 200564 Sacrocolpopexy, a A, 2/56 - - In A, one was vaginally - mixture of absorbable B, 1/61 removed, one was and non-absorbable laparoscopically mesh: removed; in B, it was A, laparoscopic vaginally removed. B, abdominal open Case series Fedorkow Sacrocolpopexy, non- 0/149 (0%) - - - - 199340 absorbable synthetic mesh Griffis 200645 Sacrocolpopexy, non- 5/196 (2.6%) - - Excision in the - absorbable synthetic operation room, the mesh others (11/196) were managed with estrogen cream, antibiotic cream, silver nitrate, or local excision in the office.

180 Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Higgs 2005a50 Sacrocolpopexy, non- 5/103 (4.9%) - - Excision, 1 - absorbable synthetic laparoscopically, 4 mesh vaginal excisions, the others (4/103) declined to trim. De novo urinary symptoms in women who did not have these symptoms at baseline RCTs Maher 200457 Sacrocolpopexy, non- 2/22 (19/47 Sacrospinous 8/24 (no SUI 24m (6-60) absorbable synthetic paravaginal repair) colpopexy paravaginal mesh repair, authors mentioned the difference may be due to the differences in paravaginal repairs) De novo bowel symptoms in women who did not have these symptoms at baseline De novo sexual problems in women who did not have these symptoms at baseline Infection RCTs Culligan Abdominal (open) A, 2/54 - - Postoperative fever 1y 200535 sacrocolpopexy: B, 2/46 A, non-absorbable synthetic mesh B, absorbable biological graft Maher 200457 Sacrocolpopexy, non- 1/47 (2.1%) Sacrospinous 1/48 (2.1%) Infection 24m (6-60) absorbable synthetic colpopexy mesh Non-rand. comp. Govier 200543 Sacrocolpopexy, non- A, 0/24 - - Mesh infection 23m (16-41) absorbable synthetic B, 1/21 mesh: A, polypropylene; B, silicone-covered polyethylene Marcickiewicz Laparoscopic 3/60 (5.0%) Sacrospinous 4/51 (7.8%) Urinary tract infection A, 34m (13-60) 200758 Sacrocolpopexy, non- colpopexy B, 38m (7-108) absorbable synthetic mesh Paraiso 200564 Sacrocolpopexy, a A, 6/56 - - Wound infection A, 14m (1-46) mixture of absorbable B, 2/61 B, 16m (1-73) and non-absorbable mesh: A, laparoscopic B, abdominal open

Case series Fedorkow Sacrocolpopexy, non- 22/149 - - Urinary tract infection; NR 199340 absorbable synthetic 2/149 Wound infection; mesh 8/149 Febrile morbidity, 0/149 occurred in 5d; 4/149 Pelvic abscess; 0/149 Pneumonia; (In total, 36/149, Delayed sepsis. 24.2%) Higgs 2005a50 Sacrocolpopexy, non- 2/103 (1.9%) - - Urinary tract infection 66m (37-124) absorbable synthetic mesh Other serious adverse effects RCTs Culligan Abdominal (open) A, 1/54 - - Pulmonary embolism; NR, 1y 200535 sacrocolpopexy: 2/54 Ileus; A, non-absorbable 8/54 Wound breakdown. synthetic mesh (In total, 11/54, B, absorbable biological 20.4%) graft B, 0/46 0/46 5/46 (In total, 5/46, 10.9%) Maher 200457 Sacrocolpopexy, non- 2/47 (4.3%) Sacrospinous 0/43 Incisional hernia NR, 24m (6-60) absorbable synthetic colpopexy mesh

181 Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Non-rand. comp. Govier 200543 Sacrocolpopexy, non- A, 0/24 - - Osteomyelitis NR absorbable synthetic B, 0/21 A, 12m (1-38) mesh: A, polypropylene; B, 23m (16-41) B, silicone-covered polyethylene Paraiso 200564 Sacrocolpopexy, a A, 1/56 - - DVT; NR mixture of absorbable 0/56; Cardiac complications; A, 14 (12), 1-46m and non-absorbable 1/56; Ventral hernia; Ileus; B, 16 (18), 1-73m mesh: 0/56; Small bowel A, laparoscopic 1/56 obstruction. B, abdominal open (In total, 3/56, 5.4%) B, 1/61 2/61 (one atrial fibrillation and mitral valve disease, one atrial fibrillation with heart failure); 2/61; 2/61; 2/61 (In total, 9/61, 15%) Other adverse effects RCTs Maher 200457 Sacrocolpopexy, non- 0/47 Sacrospinous 1/48 Rectovaginal NR, 24m (6-60) absorbable synthetic 0/47 colpopexy 1/48 hematoma; mesh (In total, 2/48, Vaginal pain of 4.2%) undetermined origin. Case series Fedorkow Sacrocolpopexy, non- 0/149 - - Pelvic hematoma NR 199340 absorbable synthetic mesh Higgs 2005a50 Sacrocolpopexy, non- 5/103 - - Wound haematoma; Immediate post- absorbable synthetic 1/103 Hypercarboxaemia operation mesh 1/103 (overnight ICU); 3/103 Hyponatremia (In total, 10/103, (overnight ICU); 10%) Sciatica requiring physiotherapy.

182 Table Uterine and/or vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Blood loss, ml, mean (SD)/median (range) RCTs Lo 199856 Abdominal (open) sacrocolpopexy, 150 (137), n=52 Sacrospinous 448 (258), - - non-absorbable synthetic mesh ligament n=66 (Mersilene) suspension Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 162ml, n=25 - - - - A, non-absorbable synthetic mesh B, 150ml, n=27 (polypropylene or polytetrafluoroethylene) B, biological graft (Pelvicol) Gregory Abdominal (open) sacrocolpopexy: A, 276 (153) ml, - - - - 200544 A, non-absorbable synthetic mesh n=49 (Marlex, Amid type I) B, 298 (162), B, biological graft (fascia lata) n=33 Hardiman Sacrocolpopexy, non-absorbable 745 (100-5000) Sacrospinous 567 (100- - - 199647 synthetic mesh (Marlex, Amid type I) ml, n=80 suspension 2700), n=125

Hsiao 200751 A, laparoscopic sacrocolpopexy, A, 83 (20-200) - - - - non-absorbable synthetic mesh ml, n=25 (Prolene or silicon) B, 195 (100-500) B, abdominal (open) ml, n=22 sacrocolpopexy, non-absorbable synthetic mesh (Prolene [Amid I], silicon [Amid IV], or Gore-Tex [Amid type II]) or biological graft (fascia lata) Ng 200461 Abdominal (open) sacrocolpopexy, 557 (50-2700) ml, Sacrospinous 239 (5-700) - - non-absorbable synthetic mesh n=113 fixation (the ml, n=64 (Gore-Tex, Amid type II) authors mentioned that the difference in blood loss was due to contaminant procedure) Young 200480 Laparoscopic sacrocolpopexy, 40 (10-300) ml, A, sacrospinous A, 182 (10- - - mixture of biological and non- n=92 fixation 300), absorbable mesh B, high n=187 uterosacrosuspen B, 280 (40- sion 1000), n=70 Case series Culligan Sacrocolpopexy, synthetic mesh 329 (203), 50- - - - - 200234 (NR details) 1300 ml, n=245

Synder 199171 Abdominal (open) sacrocolpopexy, a 302 (30-2600) ml, - - - - mixture of absorbable and non- n=132 absorbable mesh/graft Number of women with blood loss requiring blood transfusion RCTs Benson Sacrocolpopexy 2/38 (5.3%) Bilateral 0/42 (0%) Transfusion needed - 199627 sacrospinous suspension Non-rand. comp. Hardiman Sacrocolpopexy, non-absorbable 1/80 (1.3%) Sacrospinous 0/125 (0%) Transfusion needed - 199647 synthetic mesh (Marlex, Amid type I) suspension Hsiao 200751 A, laparoscopic sacrocolpopexy, A, 0/25 Transfusion needed. non-absorbable synthetic mesh B, 0/22 (Prolene or silicon) B, abdominal (open) sacrocolpopexy, non-absorbable synthetic mesh (Prolene [Amid I], silicon [Amid IV], or Gore-Tex [Amid type II]) or biological graft (fascia lata) Ng 200461 Abdominal (open) sacrocolpopexy, 13/113 (11.5%) Sacrospinous 2/64 (3.1%) Transfusion needed - non-absorbable synthetic mesh fixation (Gore-Tex, Amid type II) Sze 199973 Abdominal (open) sacrocolpopexy 0/56 (0%) Sacrospinous 2/61(3.3%) Transfusion needed - fixation

183 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Young 200480 Laparoscopic sacrocolpopexy, 3/92 (3.3%) A, sacrospinous A, 2/187 Excessive - mixture of biological and non- fixation (1.1%) hemorrhage absorbable mesh) B, high B, NA uterosacrosuspen sion Case series Bensinger Sacrocolpopexy, non-absorbable 2/115 (1.7%) - - Transfusion needed - 200526 synthetic mesh (polypropylene, NR trade name) Brizzolara Sacrocolpopexy, non-absorbable 1/124 (0.8%) - - Transfusion needed - 200330 synthetic mesh (Prolene) or biological graft (allograft) Culligan Sacrocolpopexy, synthetic mesh 2/245 (0.8%) - - Transfusion needed - 200234 (NR details) De Vries Abdominal (open) sacrocolpopexy, 5/101 (5.0%) - - Hemorrhage - 199536 non-absorbable synthetic mesh >=750ml (Mersilene, Amid type III) Lindeque Abdominal (open) sacrocolpopexy, a 3/262 (1.2%) - - Transfusion needed - 200255 mixture of absorbable and non- absorbable mesh/graft Synder 199171 Abdominal (open) sacrocolpopexy, a 23/132 (17.4%) - - >500ml blood loss - mixture of absorbable and non- absorbable mesh/graft Timmons Abdominal (open) sacrocolpopexy, a 2/163 (1.2%) - - Severe hemorrhage - 199275 mixture of absorbable and non- absorbable mesh/graft Damage to surrounding organs during operation RCTs Benson Sacrocolpopexy 1/38 Bilateral 0/42 Enterotomy; - 199627 1/38 sacrospinous 1/42 Cystotomy; 1/38 suspension 0/42 Obturator nerve (In total, 3/38, (In total, injury 7.9%) 1/42, 2.4%) Lo 199856 Abdominal (open) sacrocolpopexy, 1/52 (1.9%) Sacrospinous 0/66 Urethral injury - non-absorbable synthetic mesh ligament (Mersilene) suspension Non-rand. comp. Ng 200461 Abdominal (open) sacrocolpopexy, 1/113 Sacrospinous 0/64 Bladder injury - non-absorbable synthetic mesh 0/113 fixation 1/64 Rectal mucosa injury (Gore-Tex, Amid type II) (In total, 1/113, (In total, 0.9%) 1/64, 1.6%)

Young 200480 Laparoscopic sacrocolpopexy, 1/92 A, sacrospinous A, 0/187; Enterotomy - mixture of biological and non- 0/92 fixation 4/187; Proctotomies absorbable mesh 0/92 B, high 1/187; Cystotomy 1/92 uterosacrosuspen 0/187; Small bowel serosal 0/92 sion 0/187 tear (In total, 2/92, (In total, Rectal laceration 2.2%) 5/187, 2.7%) B, 1/70; NR; NR; NR; 1/70 (In total, 2/70, 2.9%)

(A and B in average, 2.8%) Case series Bensinger Sacrocolpopexy, non-absorbable 2/121 - - Cystotomy - 200526 synthetic mesh (polypropylene, NR 1/121 Small bowel trade name) (In total, 3/121, laceration 2.5%) Brizzolara Sacrocolpopexy, non-absorbable 1/124 (0.8%) - - Ureteral - 200330 synthetic mesh (Prolene) or biological graft (allograft) Culligan Sacrocolpopexy, synthetic mesh 1/245 (0.4%) - - Cystotomy - 200234 (NR details)

184 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) De Vries Abdominal (open) sacrocolpopexy, 6/101 - - Bladder lesion - 199536 non-absorbable synthetic mesh 2/101 Intestine lesion (Mersilene, Amid type III) (In total, 8/101, 7.9%) Higgs 2005b49 Abdominal (open) sacrocolpopexy, a 5/148 - - Bladder injury - mixture of absorbable and non- 1/148 Vascular injury absorbable mesh/graft 2/148 Bowel injury (In total 8/148, 5.4%) Lindeque Abdominal (open) sacrocolpopexy, a 3/262 - - Minor bladder injury - 200255 mixture of absorbable and non- 1/262 Minor small-bowel absorbable mesh/graft 0/262 injury (In total, 4/262, Vessel injury 1.5%) Mesh erosion Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 0/25 - - - A, non-absorbable synthetic mesh B, 0/27 (polypropylene or polytetrafluoroethylene) B, biological graft (Pelvicol) Bai 200622 Sacrocolpopexy, non-absorbable 4/34 (11.8%) Uterosacral Not Mesh erosion NR, 1y synthetic mesh (polyethylene colpopexy applicable tetraphalate, Mersilene [Amid type III]) Begley 200525 A, open or laparoscopic A, 0/24 - - Mesh erosion Mean time to sacrocolpopexy, non-absorbable B, 3/33 occure: synthetic mesh (polypropylene, C, 4/41 B, 5m (3-8.); Amid type I); D, 0/14 C, 11m (4-22) B, abdominal (open) sacrocolpopexy, non-absorbable synthetic mesh (polypropylene, Amid type II); C, open or laparoscopic sacrocolpopexy, non-absorbable synthetic mesh (silicon-covered polyester); D, abdominal (open) sacrocolpopexy, biological graft (fascia lata) Gregory Abdominal (open) sacrocolpopexy: A, 0/19 - Mesh erosion NR 200544 A, non-absorbable synthetic mesh B, 0/18 A, 26 (10-48) (Marlex, Amid type I) B, 21 (11-34) B, biological graft (fascia lata) Hardiman Sacrocolpopexy, non-absorbable 0/80 (0%) Sacrospinous - Mesh erosion A, 4y 199647 synthetic mesh (Marlex, Amid type I) suspension B, 2y A+B, 6m-5y Hsiao 200751 A, laparoscopic sacrocolpopexy, A, 3/25 (all - - Mesh extrusion NR non-absorbable synthetic mesh silicone mesh) A, 6m (Prolene or silicon) B, 2/22, one B, 10m B, abdominal (open) silicone mesh, sacrocolpopexy, non-absorbable one Gore-Tex synthetic mesh (Prolene [Amid I], silicon [Amid IV], or Gore-Tex [Amid type II]) or biological graft (fascia lata) Ng 200461 Abdominal (open) sacrocolpopexy, 1/60 (1.7%) Sacrospinous - Mesh/suture erosion NR non-absorbable synthetic mesh fixation 18m (1-48) (Gore-Tex, Amid type II) Visco 200177 A, sacrocolpopexy, non-absorbable A, 5/155 - - Mesh erosion Median time synthetic mesh (Mersilene [Amid B, 4/88 occurred: type III] or Gore-Tex [Amid type II] C, 4/25 Sacrocolpope B, C, D, Sacrocolpoperineopexy, D, 2/5 xy: 15m (2- non-absorbable synthetic mesh 33); (Mersilene [Amid type III] or Gore- A, 12m (3-40) Tex [Amid type II): B, 9m (5-15) B, abdominal C, 4m (2-7) C, vaginal+abdominal D, vaginal+abdominal with mesh placed down to the vaginal field Case series Bensinger Sacrocolpopexy, non-absorbable 4/62 (6.5%) - - Mesh erosion 3 occurred 5- 200526 synthetic mesh (polypropylene, NR 10m, one trade name) occurred at 2y

185 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Brizzolara Sacrocolpopexy, non-absorbable 1/124 - - Mesh erosion Occurred at 200330 synthetic mesh (Prolene) or 1/124 Cuff granulation 3y biological graft (allograft) (In total, 2/124, Occurred at 1.6%) 3m Culligan Sacrocolpopexy, synthetic mesh 6/245 (2.4%) - - Mesh erosion >=4y 200234 (NR details) De Vries Abdominal (open) sacrocolpopexy, 1/101 (1.0%) - - Coming off of the NR, 4y (1-13) 199536 non-absorbable synthetic mesh mesh (Mersilene, Amid type III) Higgs 2005b49 Abdominal (open) sacrocolpopexy, a 4/183 (2.2%) - - Mesh erosion 2 occurred in mixture of absorbable and non- a y, 1 between absorbable mesh/graft 1y-18m. Lindeque Abdominal (open) sacrocolpopexy, a 10/262 (3.8) - - Mesh erosion All occurred in 200255 mixture of absorbable and non- a year absorbable mesh/graft Synder 199171 Abdominal (open) sacrocolpopexy, a 4/132 (3.0%) - - Mesh erosion NR, 43m (1m- mixture of absorbable and non- 17y) absorbable mesh/graft Operation for mesh erosion Non-rand. comp. Bai 200622 Sacrocolpopexy, non-absorbable 2/34 (5.9%) - - Mesh was removed - synthetic mesh (polyethylene transvaginally (the tetraphalate, Mersilene [Amid type others, 2/34 III]) erosions were treated with estrogen) Begley 200525 A, open or laparoscopic A, 0/24 - - B, all partial graft - sacrocolpopexy, non-absorbable B, 3/33 (9.1%) resection, all synthetic mesh (polypropylene, C, 4/41 (9.8%) transvaginally; Amid type I); D, 0/14 C, partial or B, abdominal (open) complete removal, sacrocolpopexy, non-absorbable more complicated, 3 synthetic mesh (polypropylene, via laparotomy and 1 Amid type II); transvaginally. C, open or laparoscopic sacrocolpopexy, non-absorbable synthetic mesh (silicon-covered polyester); D, abdominal (open) sacrocolpopexy, biological graft (fascia lata) Hsiao 200751 A, laparoscopic sacrocolpopexy, A, 3/25 (12.0%) - - Mesh removal - non-absorbable synthetic mesh B, 2/22 (9.1%) (Prolene or silicon) B, abdominal (open) sacrocolpopexy, non-absorbable synthetic mesh (Prolene [Amid I], silicon [Amid IV], or Gore-Tex [Amid type II]) or biological graft (fascia lata) Visco 200177 A, sacrocolpopexy, non-absorbable A, 5/155 (3.2%) - - 12/15 had single - synthetic mesh (Mersilene [Amid B, 4/88 (4.6%) outpatient excision, type III] or Gore-Tex [Amid type II] C, 4/25 (16.0%) 2/25 had two vaginal B, C, D, Sacrocolpoperineopexy, D, 2/5 (40.0%) procedures, 1/25 non-absorbable synthetic mesh had four excisions (Mersilene [Amid type III] or Gore- including a Tex [Amid type II): transdominal B, abdominal resection. C, vaginal+abdominal D, vaginal+abdominal with mesh placed down to the vaginal field Case series Bensinger Sacrocolpopexy, non-absorbable 4/62 (6.5%) - - 3 simple excision, 1 - 200526 synthetic mesh (polypropylene, NR abdominal resection trade name) Brizzolara Sacrocolpopexy, non-absorbable 1/124 (0.8%) - - Office resection 0 - 200330 synthetic mesh (Prolene) or biological graft (allograft) Culligan Sacrocolpopexy, synthetic mesh 5/245 (2.0%) - - Mesh removal - 200234 (NR details) Higgs 2005b49 Abdominal (open) sacrocolpopexy, a 4/148 (2.7%) - - 3 excised vaginally, - mixture of absorbable and non- 1 removed absorbable mesh/graft abdominally

186 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Lindeque Abdominal (open) sacrocolpopexy, a 10/262 (3.8%) - - 7 mesh removal - 200255 mixture of absorbable and non- through laparotomy, absorbable mesh/graft 3 removal transvaginally. Synder 199171 Abdominal (open) sacrocolpopexy, a 4/132 (3.0%) - - Surgical removal - mixture of absorbable and non- absorbable mesh/graft De novo urinary symptoms in women who did not have these symptoms at baseline RCTs Lo 199856 Abdominal (open) sacrocolpopexy, 2/52 (3.8%) Sacrospinous 1/66 (1.5%) Urinary incontinence 2.1y (1-5.2y) non-absorbable synthetic mesh suspension (Mersilene) Non-rand. comp. Ng 200461 Abdominal (open) sacrocolpopexy, 4/89 (4.5%) Sacrospinous 0/52 Genuine stress A, 18m (1-48) non-absorbable synthetic mesh fixation incontinence B, 13m (1-29) (Gore-Tex, Amid type II) Case series Bradley 200728 Abdominal (open) sacrocolpopexy, 12/131 (9.2%) - - Bothersome irritative 1y mixture of biological and non- symptoms absorbable mesh De novo bowel symptoms in women who did not have these symptoms at baseline Case series Bradley 200728 Abdominal (open) sacrocolpopexy, 2/178 (1.1%) - - Need of digital 1y mixture of biological and non- assistance absorbable mesh De novo sexual problems in women who did not have these symptoms at baseline RCTs Lo 199856 Abdominal (open) sacrocolpopexy, 1/11 (9.1%) Sacrospinous 7/18 (39%) Dyspareunia 2.1y (1-5.2y) non-absorbable synthetic mesh suspension (Mersilene) Case series Bradley 200728 Abdominal (open) sacrocolpopexy, 11/76 (14.5%) - - dyspareunia 1y mixture of biological and non- absorbable mesh Infection RCTs Benson Sacrocolpopexy 0/38 Bilateral 9/42 Urinary tract - 199627 3/38 sacrospinous 0/42 infection; 8/38 suspension 4/42 Wound infection; (In total, 11/38, (In total, Febrile morbidity. 28.9%) 13/42, 31.0%) Lo 199856 Abdominal (open) sacrocolpopexy, 2/52 Sacrospinous 4/66 Urinary tract - non-absorbable synthetic mesh 1/52 suspension) 0/66 infection; (Mersilene) 0/52 3/66 Retroperitoneal (In total, 3/52, (In total, infection; 5.8%) 7/66, Perineal infection. 10.6%)

Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 0/25; - - Wound infection; - A, non-absorbable synthetic mesh 2/25; Cystitis; (polypropylene or 1/25 Fever >3d after polytetrafluoroethylene) (In total, 3/25, surgery; B, biological graft (Pelvicol) 12%) B, 3/27; 5/27; 12/27 (In total, 20/27, 74.1%) Hardiman Sacrocolpopexy, non-absorbable 2/80 Sacrospinous 10/125 Urinary tract - 199647 synthetic mesh (Marlex, Amid type I) 5/80 suspension 13/125 infection (In total, 7/80, (In total Febrile morbidity 8.8%) 23/125, 18.4%) Ng 200461 Abdominal (open) sacrocolpopexy, 12/113 Sacrospinous 7/64 Urinary tract - non-absorbable synthetic mesh 6/113 fixation 0/64 infection (Gore-Tex, Amid type II) 59/113 18/64 Wound infection (In total, 68.1%) (In total, Pyrexia (>37.5C) 25/64, 39.1%) Sze 199973 Abdominal (open) sacrocolpopexy 7/56 (12.5%) Sacrospinous 5/61 (7.8%) Febrile morbidity - fixation

187 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Case series Bensinger Sacrocolpopexy, non-absorbable 1/115 - Infection - 200526 synthetic mesh (polypropylene, NR 11/115 Febrile morbidity trade name) Brizzolara Sacrocolpopexy, non-absorbable 1/124 - - Wound infection - 200330 synthetic mesh (Prolene) or biological graft (allograft) De Vries Abdominal (open) sacrocolpopexy, 17/101 - - Wound infection - 199536 non-absorbable synthetic mesh (Mersilene, Amid type III) Other serious adverse effects RCTs Benson Sacrocolpopexy 1/38 Bilateral 0/42 Ileus; NR 199627 0/38 sacrospinous 1/42 Development of 2.5y (1 – 5.5) (In total, 1/38, suspension (In total, vaginal band 2.6%) 1/42, 2.4%) requiring excision (NR detail); Lo 199856 Abdominal (open) sacrocolpopexy, 3/52 Sacrospinous 0/66 Intestinal ileus; NR non-absorbable synthetic mesh 0/52 suspension 1/66 Rectovaginal fistula 2.1y (1-5.2) (Mersilene) (In total, 3/52, (In total, 5.8%) 1/66, 1.5%) Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 0/25 Re-admission, NR NR, A, non-absorbable synthetic mesh B, 4/27 (14.8%) reasons A, 4.3y (polypropylene or B, 2.5y polytetrafluoroethylene) B, biological graft (Pelvicol) Hardiman Sacrocolpopexy, non-absorbable 0/80 (0%) Sacrospinous 0/125 (0%) Thrombo-embolism NR 199647 synthetic mesh (Marlex, Amid type I) suspension A, 4y B, 2y A+B, 6m-5y Hsiao 200751 A, laparoscopic sacrocolpopexy, A, 0/25; Prolonged NR non-absorbable synthetic mesh 1/25; postoperative ileus; A, 6m (Prolene or silicon) (In total, 1/25, Pneumonia B, 10m B, abdominal (open) 4.0%) sacrocolpopexy, non-absorbable B, 3/22; synthetic mesh (Prolene [Amid I], 0/22 silicon [Amid IV], or Gore-Tex [Amid (In total, 3/22, type II]) or biological graft (fascia 13.6%) lata) Ng 200461 Abdominal (open) sacrocolpopexy, 1/113 Sacrospinous 0/64 DVT; NR non-absorbable synthetic mesh 1/113 fixation 0/64 Myocardial A, 18m (1-48) (Gore-Tex, Amid type II) 1/113 0/64 infarction; B, 13m (1-29) 4/113 0/64 Congestive heart 2/113 1/64 failure; (In total, 9/113, (In total, Incisional hernia; 8.0%) 1/64, 1.6%) Wound breakdown Sze 199973 Abdominal (open) sacrocolpopexy 2/56 (3.6%) Sacrospinous 0/61 (0%) Ileus NR fixation A, 23m (4-51) B, 24m (7-72) Young 200480 Laparoscopic sacrocolpopexy, 1/92 A, sacrospinous A, 0/187 Incisonal hernia; NR, safety mixture of biological and non- 1/92 fixation 0/187 Small bowel registry in a absorbable mesh 0/92 B, high 0/187 obstruction; year time. 2/92 uterosacrosuspen 0/187 Ureteral obstruction (In total, 4/92, sion B, 0/70 required stent 4.3%) 0/70 placement; 1/70 Small bowel 0/70 obstruction required surgery (In total, 1/257, 0.4%) Case series Bensinger Sacrocolpopexy, non-absorbable 4/115 (2.7%) - - Partial small bowel NR 200526 synthetic mesh (polypropylene, NR obstruction/ileus; 13m (0.3-63) trade name) Brizzolara Sacrocolpopexy, non-absorbable 3/124 - - Small bowel Late clinic 200330 synthetic mesh (Prolene) or 2/124 obstruction Occurred at biological graft (allograft) 1/124 Incision hernia 18m (In total, 6/124, Abdominal hernia Late clinic 4.8%) mesh infection

188 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Culligan Sacrocolpopexy, synthetic mesh 2/245 - - DVT; NR 200234 (NR details) 1/245 PE; >=4y 2/245 Small bowel 2/245 obstruction requires 1/245 re-operation; (In total, 8/245, Ureteral obstruction ; 3.3%) Fascia dehenscence De Vries Abdominal (open) sacrocolpopexy, 4/101 - - Ileus NR 199536 non-absorbable synthetic mesh 2/101 Cicatricial hernia 4y (1-13) (Mersilene, Amid type III) (In total, 6/101, 5.9%) Higgs 2005b49 Abdominal (open) sacrocolpopexy, a 1/148 - - Vesico-cutaneous NR mixture of absorbable and non- 8/148 fistula 45m (15) absorbable mesh/graft (In total, 9/148, Incisional hernia 6.1%) Lindeque Abdominal (open) sacrocolpopexy, a 1/262 - - Pneumonia NR 200255 mixture of absorbable and non- 3/262 Prolonged ileus All >=16m absorbable mesh/graft 2/262 Arrhythmia 1/262 Left ventricular 3/262 failure (In total, 10/262, Intestinal obstruction 3.8%) Timmons Abdominal (open) sacrocolpopexy, a 1/163 - - Wound dehiscence NR, 33m (9m- 199275 mixture of absorbable and non- 6/163 Would separation 19y) absorbable mesh/graft (In total, 7/163, 4.3%) Other adverse effects RCTs Benson Sacrocolpopexy 2/38 Bilateral 0/42 Sciatica; NR, 2.5y (1 – 199627 1/38 sacrospinous 0/42 Phlebitis. 5.5) (In total, 3/38, suspension (In total, 7.9%) 0%)

Lo 199856 Abdominal (open) sacrocolpopexy, 0/52 Sacrospinous 2/66 Vaginal vault NR, 2.1y (1- non-absorbable synthetic mesh 0/52 suspension 17/66 strictures; 5.2) (Mersilene) (In total, 0%) (In total, Prolonged 19/66, catheterization 28.8%) (>=4d); Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 1/25; Vaginal vault NR, A, non-absorbable synthetic mesh 0/25 (In total, hematoma; A, 4.3y (polypropylene or 1/25, 4.0%) Wound hematoma. B, 2.5y polytetrafluoroethylene) B, 1/27; B, biological graft (Pelvicol) 1/27 (In total, 2/27, 7.4%) Hardiman Sacrocolpopexy, non-absorbable 4/80 (5.0%) Sacrospinous 4/125 Wound NR 199647 synthetic mesh (Marlex, Amid type I) suspension (3.2%) complications A, 4y B, 2y A+B, 6m-5y Ng 200461 Abdominal (open) sacrocolpopexy, 1/113 Sacrospinous 1/64 Pelvic haematoma NR non-absorbable synthetic mesh 1/113 fixation 0/64 Subacute intestinal A, 18m (1-48) (Gore-Tex, Amid type II) (In total, 2/113, (In total, obstruction B, 13m (1-29) 1.8%) 1/64, 1.6%) Sze 199973 Abdominal (open) sacrocolpopexy 0/56 Sacrospinous 5/61 Transient right NR 1/56 fixation 0/61 gluteal pain A, 23m (4-51) (In total, 1/56, (In total, Phlebitis B, 24m (7-72) 1.8%) 5/61, 8.2%) Young 200480 Laparoscopic sacrocolpopexy, 1/92 A, sacrospinous A, 0/187; Hematoma NR, safety mixture of biological and non- 0/92 fixation 2/187; Buttock pain registry in a absorbable mesh 0/92 B, high 1/187; Sciatic pain year time. 0/92 uterosacrosuspen 1/187 Foot drop (In total, 1/92, sion B, 0/70; 1.1%) 0/70; 1/70; 0/70; 0/70; 0/70

(In total, 4/257, 1.6%) Case series Bensinger Sacrocolpopexy, non-absorbable 3/115 - - Persistent vaginal NR 200526 synthetic mesh (polypropylene, NR 1/115 (In total, discharge; 13m (0.3-63)

189 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) trade name) 4/115, 3.5%) Vaginal bleeding. Brizzolara Sacrocolpopexy, non-absorbable 1/124 - - Delayed vaginal Resolved in 200330 synthetic mesh (Prolene) or (0.8%) healing 3m biological graft (allograft) Bradley 200728 Abdominal (open) sacrocolpopexy, 32/322 (9.9%) - - Gastrointestinal At 6w. mixture of biological and non- complications, e.g. absorbable mesh nausea, emesis, bloating or ileus Culligan Sacrocolpopexy, synthetic mesh 1/245 (0.4%) - - Femoral neuropathy NR, >=4y 200234 (NR details) Resolved over 3m NR, >=4y De Vries Abdominal (open) sacrocolpopexy, 1/101 - - Extra-peritoneal During 199536 non-absorbable synthetic mesh 3/101 haematoma surgery (Mersilene, Amid type III) 3/101 Wound haematoma Postoperative 1/101 Urinary retention NR, 4y (1-13) 8/101 (In total, Platzbanch NR 16/101, 15.8%) Problems of motion NR

190 APPENDIX 12 Summary of outcomes: sacrocolpopexy, operation time and hospital stay

Table Uterine prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time RCTs (abstract) Braun 200729 Hysterectomy+sacrocolpopexy 140min (100-240), n=23 Hysterectomy 90min (50-130), n=24 Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 89 (60-110) min, n=36 - - B, hysterectomy+sacrocolpopexy B, 115 (80-135) min, n=39 Hospital stay RCTs (abstract) Braun 200729 Hysterectomy+sacrocolpopexy 3.8d, n=23 Hysterectomy 2d, n=24 Case series Wu 200679 Hysterectomy+sacrocolpopexy 3.5 (1.9) d, n=101 - -

Table Vault prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time RCTs Culligan 200535 A, abdominal (open) sacrocolpopexy, A, 227 (63) min, n=54 - - non-absorbable synthetic mesh B, 233 (67) min, n=46 B, abdominal (open) sacrocolpopexy, absorbable biological graft Maher 200457 Sacrocolpopexy 106 (37), 45-100 min, n=47 Sacrospinous 76 (42), 26-300 min, colpopexy n=48 Non-rand. comp. Marcickiewicz 200758 Laparoscopic sacrocolpopexy 129 min, n=60 Sacrospinous 62 min, n=51 colpopexy Paraiso 200564 A, Laparoscopic sacrocolpopexy A, 269 (65), 150-467 min, n=56 - - B, abdominal (open) sacrocolpopexy, B, 218 (60), 45-372 min, n=61 Case series - - Higgs 2005a50 Laparoscopic sacrocolpopexy 107 (62-170) min, n=103 Hospital stay RCTs Maher 200457 Sacrocolpopexy 5 (2), 3-16 d, n=47 Sacrospinous 5 (1), 3-10 d, n=48 colpopexy Non-rand. comp. Marcickiewicz 200758 Laparoscopic sacrocolpopexy 4 (2-21), n=60 Sacrospinous 4 (1-18), n=51 colpopexy Paraiso 200564 A, Laparoscopic sacrocolpopexy A, 2 (1), 1-6 d, n=56 B, abdominal (open) sacrocolpopexy, B, 4 (2), 2-12 d, n=61 Case series Higgs 2005a50 Laparoscopic sacrocolpopexy Wu 200679 Abdominal (open) sacrocolpopexy 4d (2), n=212 - -

191 Table Uterine and/or vault prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time RCTs Benson 199627 Sacrocolpopexy 215 (47) min, n=38 Bilateral sacrospinous 196 (38) min, n=42 suspension Lo 199856 Abdominal (open) sacrocolpopexy, 2.63 (0.59)h, n=52 Sacrospinous suspension 2.36 (0.61)h, n=66 non-absorbable synthetic mesh (Mersilene) Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 1.48h, n=25 - - A, non-absorbable synthetic mesh B, 1.74h, n=27 (polypropylene or polytetrafluoroethylene) B, biological graft (Pelvicol) Gregory 200544 Abdominal (open) sacrocolpopexy: A, 3.5 (1.2) h, n=49 - - A, non-absorbable synthetic mesh B, 4.5 (1.5) h, n=33 (Marlex, Amid type I) (mentioned the longer time B, biological graft (fascia lata) in group B was due to concomitant procedures.) Hsiao 200751 A, laparoscopic sacrocolpopexy, non- A, 220 (142-331) min, - - absorbable synthetic mesh (Prolene n=25 or silicon) B, 185 (195-278) min, B, abdominal (open) sacrocolpopexy, n=22 non-absorbable synthetic mesh (Prolene [Amid I], silicon [Amid IV], or Gore-Tex [Amid type II]) or biological graft (fascia lata) Ng 200461 Abdominal (open) sacrocolpopexy, 133 (32-227) min, n=113 Sacrospinous fixation 78 (28-56) min, n=64 non-absorbable synthetic mesh (abdominal, open, Gore- (Gore-Tex, Amid type II) Tex, Amid type II) Young 200480 Laparoscopic sacrocolpopexy, 218 (120-460) min, n=92 A, sacrospinous fixation A, 182 (50-502) min, mixture of biological and non- B, high n=187 absorbable mesh uterosacrosuspension B, 118 (45-240) min, n=70 Case series Synder 199171 Abdominal (open) sacrocolpopexy, a 97 (50-210) min, n=132 - - mixture of absorbable and non- absorbable mesh/graft Hospital stay RCTs Benson 199627 Sacrocolpopexy 5.4 (1.1) d, n=38 Bilateral sacrospinous 5.1 (1.2) d, n=42 suspension Lo 199856 Abdominal (open) sacrocolpopexy, 7.24 (2.07) d, n=52 Sacrospinous suspension 8.77 (3.78) d, n=66 non-absorbable synthetic mesh (Mersilene) Non-rand. comp. Altman 200521 Sacrocolpopexy: A, 4.1d, n=25 - - A, non-absorbable synthetic mesh B, 3.9d, n=27 (polypropylene or polytetrafluoroethylene) B, biological graft (Pelvicol) Hsiao 200751 A, laparoscopic sacrocolpopexy, non- A, 1.2 (1-5) d, n=25 - - absorbable synthetic mesh (Prolene B, 3.3 (2-8) d, n=22 or silicon) B, abdominal (open) sacrocolpopexy, non-absorbable synthetic mesh (Prolene [Amid I], silicon [Amid IV], or Gore-Tex [Amid type II]) or biological graft (fascia lata) Ng 200461 Abdominal (open) sacrocolpopexy, 6 (2-28) d, n=113 Sacrospinous fixation 4 (2-14) d, n=64 non-absorbable synthetic mesh (Gore-Tex, Amid type II) Sze 199973 Abdominal (open) sacrocolpopexy 3.4 (0.9), n=56 Sacrospinous fixation 4.7 (1.1), n=61 Case series Synder 199171 Abdominal (open) sacrocolpopexy, a 5 (2-13) d, n=132 - - mixture of absorbable and non- absorbable mesh/graft

192 APPENDIX 13 Summary of outcomes: infracoccygeal sacropexy, efficacy

Table Uterine prolapse repair Mesh n/N (%) No mesh n/N Outcome reported Time that occurred (%) as (follow up) Subjective failure (no studies reported this outcome) Objective failure Non-rand. comp. Neuman 200760 A, infracoccygeal sacropexy A, 1/35 (2.9%) - - A, uterine prolapse, 30m (12-44) B, hysterectomy + infracoccygeal B, 0/44 (0%) no definition sacropexy B, vault prolapse, no definition Case series Sivaslioglu Infracoccygeal sacropexy 1/10 (10.0%) - - Uterine prolapse >= 16m (6) 200570 grade 2 De novo prolapse at other sites (no studies reported this outcome) Further operation deeded for prolapse (recurrent or de novo) Case series Sivaslioglu Infracoccygeal sacropexy 0/10 (0%) - - - 16m (6) 200570 persistent urinary symptoms in women having this symptom at baseline persistent bowel symptoms in women having these symptoms at baseline persistent sexual problems in women having these problems at baseline

Table Vault prolapse repair Mesh n/N (%) No mesh n/N Outcome reported as Time that occurred (%) (follow up) Subjective failure Case series Farnsworth Infracoccygeal sacropexy 8/91 (8.8%) - - <50% reduction in the 12m (2-24) 200239 severity of symptoms Objective failure Case series Petros 200163 Infracoccygeal sacropexy 4/40 (10.0%) - - Vault prolapse >= II All occurred in 24m, degree mean follow up 55m (48-60) Sivaslioglu Infracoccygeal sacropexy 0/20 (0%) - - Vault prolapse >=grade 16m (6) 200570 2 De novo prolapse at other sites (no studies reported this outcome) Further operation deeded for prolapse (recurrent or de novo) Case series Farnsworth Infracoccygeal sacropexy 2/91 (1.1%) - - Recurrent or new 12m (2-24) 200239 Petros 200163 Infracoccygeal sacropexy 12/40 (30.0%) - - 8 for de novo anterior 1-4.5y prolapse, 3 for vault prolapse, and 1 for rectocele Sivaslioglu Infracoccygeal sacropexy 0/20 (0%) - - NA 16m (6) 200570 persistent urinary symptoms in women having this symptom at baseline persistent bowel symptoms in women having these symptoms at baseline persistent sexual problems in women having these problems at baseline

193 Table Uterine and/or vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported Time that occurred as (follow up) Subjective failure Case series Hefni 200748 Infracoccygeal sacropexy 27/127 (21.3%) - - Prolapse symptoms 14m (2-26) Sentihes 200769 Infracoccygeal sacropexy 1/44 (2.3%) - - Feeling of vaginal 29m (9-47) lump Objective failure RCTs (abs.) De Tayrac Infracoccygeal sacropexy 1/21 (4.8%) Unilateral 0/20 (0%) Uterine prolapse, no A, 11m (2-34) 200674 sacrospinous definition B, 16m (2-32) suspension Case series Hefni 200748 Infracoccygeal sacropexy 14/127 (11.0%) - - Upper prolapse II-IV 14m (2-26) Jordaan 200653 Infracoccygeal sacropexy 2/8 (25.0%) - - Grade 3 or 4 13m vault/uterine prolapse Sentihes 200769 Infracoccygeal sacropexy Vardy 200776 Infracoccygeal sacropexy 1/53 (1.9%) - - POP-Q C or D point 1y score >-5 De novo prolapse at other sites Case series Hefni 200748 Infracoccygeal sacropexy 3/63 (4.8%) - - Cystocele 14m (2-26) Further operation deeded for prolapse (recurrent or de novo) Case series Hefni 200748 Infracoccygeal sacropexy 10/127 (7.9%) - - NR 14m (2-26) persistent urinary symptoms in women having this symptom at baseline Case series Sentihes 200769 Infracoccygeal sacropexy 5/17 (29.4%) - - SUI 29m (9-47) persistent bowel symptoms in women having these symptoms at baseline persistent sexual problems in women having these problems at baseline Case series Sentihes 200769 Infracoccygeal sacropexy 3/7 (42.3%) - - Persistent sexual 29m (9-47) symptoms

194 APPENDIX 14 Summary of outcomes: infracoccygeal sacropexy, safety

Table Uterine prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Blood loss, ml, mean (SD)/median (range) Case series Sivaslioglu Infracoccygeal sacropexy 200 (20), n=10 - - - - 200570 Number of women with blood loss requiring blood transfusion Case series Sivaslioglu Infracoccygeal sacropexy 0/10 - - Transfusion - 200570 required Damage to surrounding organs during operation (no studies reported this outcome) Case series Sivaslioglu Infracoccygeal sacropexy 0/10 - - Rectal - 200570 perforation Mesh erosion Non-rand. comp. Neuman 200760 A, infracoccygeal sacropexy A, 4/35 (11.4%) - - Tape protrusion NR, 30m (12-44) B, hysterectomy + infracoccygeal sacropexy B, 6/44 (13.6%) Operation need for mesh erosion Non-rand. comp. Neuman 200760 A, infracoccygeal sacropexy A, 4/35 (11.4%) - - Segmental tape - B, hysterectomy + infracoccygeal sacropexy B, 6/44 (13.6%) resection in the out-patient clinic De novo urinary symptoms in women who did not have these symptoms at baseline Non-rand. comp. Neuman 200760 A, infracoccygeal sacropexy A, 0/26 (0%) - - Incontinence 30m (12-44) B, hysterectomy + infracoccygeal sacropexy B, 0/29 (0%) De novo bowel symptoms in women who did not have these symptoms at baseline De novo sexual problems in women who did not have these symptoms at baseline Infection Other serious adverse effects Other adverse effects

195 Table Vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Blood loss, ml, mean (SD)/median (range) Case series Farnsworth 200239 infracoccygeal sacropexy 110 (20-400) ml, n=93 - - - - Foote 200741 infracoccygeal sacropexy 190 ml, n=52 - - - - Ghanbari 200642 infracoccygeal sacropexy All <130ml, n=15 - - - - Petros 200163 infracoccygeal sacropexy 120 (20-800) ml, n=75 - - - - Sivaslioglu 200570 infracoccygeal sacropexy 200 (20) ml, n=20 - - - - Number of women with blood loss requiring blood transfusion Case series Farnsworth 200239 infracoccygeal sacropexy 0/93 - - Transfusion needed. - Petros 200163 infracoccygeal sacropexy 0/75 - - Transfusion needed. - Sivaslioglu 200570 infracoccygeal sacropexy 0/20 - - Transfusion needed. - Damage to surrounding organs during operation Case series Farnsworth 200239 infracoccygeal sacropexy 1/93 (1.1% - - Rectal perforation - Ghanbari 200642 infracoccygeal sacropexy 0/15 (0%) - - Rectal perforation - Petros 200163 infracoccygeal sacropexy 2/75 (2.7%) - - Rectum perforation - Sivaslioglu 200570 infracoccygeal sacropexy 0/20 (0%) - - Rectal perforation - Mesh erosion RCTs (abstract) Meschia 200559 infracoccygeal sacropexy 2/30 (6.7%) Sacrospinous Not Mesh erosion NR, mean 24m fixation applicable Case series Farnsworth 200239 infracoccygeal sacropexy 5/49 nylon tape and - - Tape rejection; NR, 12m (2-24) 0/44 polypropylene Rectal tape erosion tape; 1/93 (In total, 6/93, 6.5%) Foote 200741 infracoccygeal sacropexy 11/52 (21.1%) - - Mesh erosion NR, mean 20wk Ghanbari 200642 infracoccygeal sacropexy 1/15 (6.7%) - - Tape rejection 6m postoperative Petros 200163 infracoccygeal sacropexy 4/75 (5.3%) - - Tape erosion Between 6-12m Operation need for mesh erosion Case series Petros 200163 infracoccygeal sacropexy 4/75 (5.3%) - - Removed in an office - De novo urinary symptoms in women who did not have these symptoms at baseline De novo bowel symptoms in women who did not have these symptoms at baseline De novo sexual problems in women who did not have these symptoms at baseline Infection Case series Farnsworth 200239 infracoccygeal sacropexy 5/93 - - Urinary tract infection, 12m (2-24) 0/93 all in <1w; (In total, 5/93, 5.4%) Pyrexin. Ghanbari 200642 infracoccygeal sacropexy 0/15 - - Urinary tract infection; NR 0/15 Pyrexia Petros 200163 infracoccygeal sacropexy 1/75 - - Unspecified; At d5. 0/75 Pyrexia (In total, 1/75, 1.3%) Other serious adverse effects Other adverse effects Case series Farnsworth 200239 infracoccygeal sacropexy 0/93 - - Significant brusing or 12m (2-24) hematoma Ghanbari 200642 infracoccygeal sacropexy 0/15 - - Haematoma or NR 0/15 significant bruising Petros 200163 infracoccygeal sacropexy 0/75 - - Hematomas 1-4.5y

196 Table Uterine and/or vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Blood loss, ml, mean (SD)/median (range) Case series Oliver 200662 infracoccygeal sacropexy 182 (12) ml, n=14 - - - - (uterus preserved) Number of women with blood loss requiring blood transfusion Case series Hefni 200748 infracoccygeal sacropexy 0/127 (0%) - - Transfusion needed - Oliver 200662 infracoccygeal sacropexy 0/14 (0%) - - Transfusion needed - (uterus preserved) Sentihes 200769 infracoccygeal sacropexy 1/44 (2.3%) - - Transfusion needed - Damage to surrounding organs during operation (no studies reported this outcome) RCTs (abstract) De Tayrac 200674 infracoccygeal sacropexy 0/21 Unilateral 0/20 Rectal injury - sacrospinous suspension Case series Hefni 200748 infracoccygeal sacropexy 0/127 - - Rectal, vesical or ureteric - injuries Oliver 200662 infracoccygeal sacropexy 0/14 - - Rectal perforation - (uterus preserved) Sentihes 200769 infracoccygeal sacropexy 1/44 (2.3%) - - Rectal injury - Vardy 200776 infracoccygeal sacropexy 0/286 - - Rectal injury - Mesh erosion RCTs (abstract) De Tayrac 200674 infracoccygeal sacropexy 0/21 Unilateral Not Mesh erosion 11m (2-34) sacrospinous applicable suspension Case series Hefni 200748 infracoccygeal sacropexy 21/127 - - Tape exposure Mean time to 1/127 Granulation tissue over the occur, 8m (3.5), (In total, 22/127, tape 1.5-20m 17.3%) NR, 14m (2-26) Petros 200565 Infracoccygeal sacropexy 1/67 (1.5%) - - Large granuloma NR, 9m (3-15) (used Tissue Fixation device) Sentihes 200769 infracoccygeal sacropexy 4/44 (9.0%) - - Vaginal erosions 3 at 6w, 1 at 6m, 1 at 16m Vardy 200776 infracoccygeal sacropexy 6/286 (2.1%) - - Mesh extrusion 3 within 6m, 3 6-12m Operation for mesh erosion Case series Hefni 200748 infracoccygeal sacropexy 21/127 (16.5%) - - Tape excision with vaginal - edges approximation Petros 200565 Infracoccygeal sacropexy 1/67 (1.5%) - - Excised sling - (used Tissue Fixation device) Sentihes 200769 infracoccygeal sacropexy 4/44 (9.1%) - - 2 mucosal closures, 2 - partial removal Vardy 200776 infracoccygeal sacropexy 6/286 (2.1%) - - 5 trimmed in the office, 1 - trimmed in the operating room De novo urinary symptoms in women who did not have these symptoms at baseline Case series Sentihes 200769 infracoccygeal sacropexy 0/27 (0%) - - SUI 29m (9-47) De novo bowel symptoms in women who did not have these symptoms at baseline De novo sexual problems in women who did not have these symptoms at baseline Infection RCTs (abstract) De Tayrac 200674 infracoccygeal sacropexy 0/21 Unilateral 0/20 Postoperative vaginal 11m (2-34) sacrospinous infection suspension Case series Hefni 200748 infracoccygeal sacropexy 2/127 - - Urinary tract infection - 1/127 Infection requires tape removal Oliver 200662 infracoccygeal sacropexy 1/14 - - Urinary tract infection - (uterus preserved) Petros 200565 Infracoccygeal sacropexy 6/67 - - Urinary tract infection - (used Tissue Fixation device) Vardy 200776 infracoccygeal sacropexy 4/286 - - Urinary tract infection - 1/286 Febrile morbidity 1/286 vaginitis Other serious adverse effects

197 Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Other adverse effects RCTs (abstract) De Tayrac 200674 infracoccygeal sacropexy 2/21 Unilateral 2/20 Haematomas Intra-operative sacrospinous or post- suspension operative Case series Hefni 200748 infracoccygeal sacropexy 1/127 - - Para-rectal haematoma NR 3/127 Vaginal wall adhesion 14m (2-26) (In total, 4/127, 3.1%) Oliver 200662 infracoccygeal sacropexy 0/14 - - Haematomas (uterus preserved) 1/14 Urine retention (In total, 1/14, 7.1%) Sentihes 200769 infracoccygeal sacropexy 1/44 (2.3%) - - Haematoma NR, 29m (9-47) Vardy 200776 infracoccygeal sacropexy 1/286 - - Hematoma NR, 1y 1/286 Pain at skin incision 1/286 Vaginal pressure (In total, 3/286, 1.0%)

198 Appendix 15 Summary of outcomes: infracoccygeal sacropexy, operation time and hospital stay

Table Uterine prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time Case series Sivaslioglu 200570 Infracoccygeal sacropexy 45 (8) min, n=10 - - Hospital stay Non-rand. comp. Neuman 200760 A, infracoccygeal sacropexy A, 1.5d, n=35 - - B, hysterectomy + infracoccygeal sacropexy B, 4.2d, n=79 Case series Sivaslioglu 200570 Infracoccygeal sacropexy 2 (1)d, n=10 - -

Table Vault prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time Case series Farnsworth 200239 Infracoccygeal sacropexy 60 (45-90) min, n=93 - - Foote 200741 Infracoccygeal sacropexy 66 min, n =52 - - Ghanbari 200642 Infracoccygeal sacropexy 30 (20-40) min, n=15 - - Petros 200163 Infracoccygeal sacropexy 30-60min, n=75 - - Sivaslioglu 200570 Infracoccygeal sacropexy 45 (8) min, n=20 - - Hospital stay Case series Farnsworth 200239 Infracoccygeal sacropexy All <24h, n=93 - - Foote 200741 Infracoccygeal sacropexy Mean 4 d, n=52 - - Ghanbari 200642 Infracoccygeal sacropexy All <24h, n=15 - - Petros 200163 Infracoccygeal sacropexy All <24h, n=15 - - Sivaslioglu 200570 Infracoccygeal sacropexy 2 (1) d, n=20 - -

Table Uterine and/or vault prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time RCTs (abstracts) De Tayrac 200674 Infracoccygeal sacropexy 88 (25) min, n=21 Unilateral sacrospinous 87 (18) min, n=20 suspension Case series Hefni 200748 Infracoccygeal sacropexy 46 (18.5) min, n=127 - - Oliver 200662 Infracoccygeal sacropexy 51 (8) min, n=14 - - Petros 200565 Infracoccygeal sacropexy (used Tissue 5-10 min, n=67 - - Fixation device) Sentihes 200769 Infracoccygeal sacropexy (inc. anterior 149 (90-240) min, n=44 - - transoburator hammock placement) Hospital stay Case series Hefni 200748 Infracoccygeal sacropexy 3 (1) d, n= 127 - - Oliver 200662 Infracoccygeal sacropexy 2d, n=14 - - Petros 200565 Infracoccygeal sacropexy (used Tissue 1.5 (1-3) d, n=67 - - Fixation device) Sentihes 200769 Infracoccygeal sacropexy (inc. anterior 5 (3-22) d, n=44 - - transoburator hammock placement)

199 APPENDIX 16 Summary of outcomes: uterine suspension sling, efficacy

Table Uterine prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported as Time that occurred (follow up) Subjective failure RCTs Roovers 200467 Sacrohysteropexy 16/41 (39.0%) Hysterectomy 5/41 Visited a doctor because of 1y (12.2%) prolapse symptoms Case series Barranger Sacrohysteropexy 1/30 (3.3) - - Symptomatic prolapse Occurred at 3y after 200324 treatment Joshi 199352 Suspend the uterus to 0/20 (0%) - - Prolapse symtpoms 6-30m the pectineal ligament Objective failure RCTs Roovers 200467 Sacrohysteropexy 2/38 (5.0%) Hysterectomy 2/40 Uterine/vault prolapse >= stage 1y (5.0%) II Non-rand. comp. Costantini A, sacrohysteropexy A, 0/36 (0%) - - Recurrent cervical/vault 51m (12-115) 200533 B, hysterectomy + B, 0/39 (0%) prolapse (<6cm above the sacrocolpopexy hymen) Case series Banu 199723 Sacrohysteropexy 0/19 (0%) - - NR 3-5y Barranger Sacrohysteropexy 1/30 (3.3%) - - Baden and Walker classification Occurred at 3y after 200324 used, but NR grade treatment Leron 200154 Sacrohysteropexy 1/13 (7.7%) - - 1st degree uterine prolapse, i.e. 16m (4-49) prolapse reached about 1cm above the introitus De novo prolapse at other sites (no studies reported this outcome) Further operation deeded for prolapse (recurrent or de novo) RCTs Roovers 200467 Sacrohysteropexy 9/41 (22.0%; Hysterectomy 1/41 - 1y 5 cystocele, 4 (2.4%) uterine prolapse) Non-rand. comp. Costantini A, sacrohysteropexy A, 0/36 (0%) - - - 51m (12-115) 200533 B, hysterectomy + B, 0/39 (0%) sacrocolpopexy Case series Barranger Sacrohysteropexy 1/30 (3.3%) - - - 3y after treatment 200324 persistent urinary incontinence in women having this symptom at baseline Case series Barranger Sacrohysteropexy 3/19 (15.8%) - - - 94.6m (8-160) 200324 persistent bowel symptoms in women having these symptoms at baseline Case series Barranger Sacrohysteropexy 1/5 (20.0%) - - - 94.6m (8-160) 200324 persistent sexual problems in women having these problems at baseline Case series Barranger Sacrohysteropexy 3/27 (11.1%) - - - 94.6m (8-160) 200324

Uterien suspension sling is not for vault prolapse repair.

200 APPENDIX 17 Summary of outcomes: uterine suspension sling, safety

Table Uterine prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Blood loss, ml, mean (SD)/median (range) RCTs Roovers 200467 Sacrohysteropexy 244 (52) ml, Hysterectomy 248 (34) ml, - - n=41 n=41 Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 200 ml, n=36 - - - - B, hysterectomy + sacrocolpopexy B, 325 ml, n=39 Case series Leron 200154 Sacrohysteropexy 271ml (50-800), - - - - n=13 Number of women with blood loss requiring blood transfusion RCTs Roovers 200467 Sacrohysteropexy 1/41 (2.4%) Hysterectomy 2/41 (4.9%) Transfusion required. - Non-rand. comp. - Costantini 200533 A, sacrohysteropexy 2/36 (5.6%) - 2/39 (5.1%) Transfusion required B, hysterectomy + sacrocolpopexy Case series Barranger 200324 Sacrohysteropexy 0/30 - - Transfusion required - Joshi 199352 Suspend the uterus to the pectineal 0/20 - - Transfusion required - ligament Damage to surrounding organs during operation (no studies reported this outcome) RCTs Roovers 200467 Sacrohysteropexy 0/41 (0%) Hysterectomy 1/41 (2.4%) Bowel lesion - Mesh erosion Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 0/36 (0%) - - Mesh erosion (non- NR, 51m B, hysterectomy + sacrocolpopexy B, 3/39 (7.7%) absorbable synthetic, (12-115) Marlex, Amid type I) Case series Barranger 200324 Sacrohysteropexy 1/30 (3.3%) - - Vaginal mesh erosion Occurred at 2y after treatment Operation for mesh erosion Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 0/36 (0%) - - Vaginal revision - B, hysterectomy + sacrocolpopexy B, 3/39 (7.7%) Case series Barranger 200324 Sacrohysteropexy 1/30 (3.3%) - - Vaginal excision - De novo urinary symptoms in women who did not have these symptoms at baseline De novo bowel symptoms in women who did not have these symptoms at baseline Case series Barranger 200324 Sacrohysteropexy 0/25 (3.3%) - - Bowel symmptoms - De novo sexual problems in women who did not have these symptoms at baseline Case series Barranger 200324 Sacrohysteropexy 1/2 (50.0%) - - Dyspareunia - Infection RCTs Roovers 200467 Sacrohysteropexy 0/41 Hysterectomy 1/41 Wound infection; NR, follow 2/41 0/41 Vault abscess during up, 1y 2/41 NA admission; 3/41 1/41 Infected implant (In total, 7/41, (In total, 2/41, requiring surgery, NR 17.1%) 4.9%) timing; Fever of unknown origin during admission. Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 0/36 (0%) - - Wound infection NR, 51m B, hysterectomy + sacrocolpopexy 1/36 (2.8%) Fever (12-115) B, 2/39 (5.1%) 1/39 (2.6%) Case series Barranger 200324 Sacrohysteropexy 1/30 (3.3%) - - Urinary tract infection NR 1/30 (3.3%) Wound infection

Other serious adverse effects RCTs

201 Mesh n/N (%) No mesh n/N (%) Outcome reported Time that as occurred (follow up) Roovers 200467 Sacrohysteropexy 1/41 (2.4%) Hysterectomy 0/41 (0%) Incisional peritoneal NR, follow hernia requiring up 1y surgery Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 2/36 (5.6%) - - Incisional hernia NR, follow B, hysterectomy + sacrocolpopexy B, 1/39 (2.6%) up 51m (12- 115) Case series Barranger 200324 Sacrohysteropexy 1/30 (3.3%) - - Post incisional The 1/30 (3.3%) abdominal hernia; intestinal Intestinal occlusion occlusion by the mesh occurred 4y after treatment Other adverse effects RCTs Roovers 200467 Sacrohysteropexy 8/41 Hysterectomy 8/41 Lower urinary tract NR, 1y 1/41 0/41 symptoms during 0/41 1/41 admission; (In total, 9/41, (In total, 9/41, Dullness upper leg 22.0%) 22.0%) during admission; Vaginal stricture requiring readmission for excision Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 2/36 Perivesical NR, 51m B, hysterectomy + sacrocolpopexy 4/36 haematoma; (12-115) (In total, 6/36, Voiding dysfunction. 16.7%) B, 4/39 1/39 (In total, 5/39, 12.8%) Case series Banu 199723 Sacrohysteropexy 2/19 (10.5%) - - Dysmenorrhae - during the 1st menstrual cycle Barranger Sacrohysteropexy 1/30 (3.3%) - - Retroperitoneal - 200324 1/30 (3.3%) haematoma; 1/30 (3.3%) Haemorrage of presacral vein, but no transufion needed; Sciatic pain

Uterien suspension sling is not for vault prolapse repair.

202 APPENDIX 18 Summary of outcomes: uterine suspension sling, operation time and hospital stay

Table Uterine prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time RCTs Roovers 200467 Sacrohysteropexy 97 (3.6) min, n=41 Hysterectomy 107 (4.7), n=41 Non-rand. comp. Costantini 200533 A, sacrohysteropexy A, 89 (60-110) min, n=36 - - B, hysterectomy + sacrocolpopexy B, 115 (80-135) min, n=39 Hospital stay RCTs Roovers 200467 Sacrohysteropexy 7.7 (0.2) d, n=41 Hysterectomy 7.6 (0.3) d, n=41

Uterien suspension sling is not for vault prolapse repair.

203 APPENDIX 19 Summary of outcomes: other mesh techniques or a mixture of mesh techniques, efficacy

No studies reported uterine prolapse repair or vault prolapse repair alone.

Table Uterine and/or vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Subjective failure Case series Su 200772 Sacrocolpoperinopexy 2/169 (1.2%) - - Symptoms 14m (1.5-24) Objective failure Case series Su 200772 Sacrocolpoperinopexy 1/169 (0.6%) - - Apical prolapse, 14m (1.5-24) stage >=II De novo prolapse at other sites (no studies reported this outcome) Further operation deeded for prolapse (recurrent or de novo) Case series Elneil 200538 Sacrocolpopexy, sacrohysteropexy, or 6/128 (4.7%) - - NR 19m (2-65) sacrocervicopexy Su 200772 Sacrocolpoperinopexy 0/169 - - NR 14m (1.5-24) persistent urinary symptoms in women having these symptom at baseline persistent bowel symptoms in women having these symptoms at baseline persistent sexual problems in women having these problems at baseline

204 APPENDIX 20 Summary of outcomes: other mesh techniques or a mixture of mesh techniques, safety

No studies reported uterine prolapse repair or vault prolapse repair alone.

Table Uterine and/or vault prolapse repair Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Blood loss, ml, mean (SD)/median (range) Number of women with blood loss requiring blood transfusion Non-rand. comp. Demirci 20 sacrohysterexy, 25 2/45 (4.4%) Sacrospinous 0/60 (0%) Transfusion - 200737 sacrocolpopexy, non- fixation needed absorbable synthetic mesh (polypropylene, Prolene, Amid type I) Damage to surrounding organs during operation (no studies reported this outcome) Non-rand. comp. Demirci 20 sacrohysterexy, 25 1/45 Sacrospinous 0/60 Bladder injury - 200737 sacrocolpopexy, non- 0/45 fixation 1/60 Rectal injury absorbable synthetic mesh (In total, 1/45, 2.2%) (In total, 1/60, (polypropylene, Prolene, 1.7%) Amid type I) Case series Elneil Sacrocolpopexy, 2/128 (1.6%) - - Bowel perforation - 200538 sacrohysteropexy, or sacrocervicopexy Su 200772 Sacrocolpoperinopexy 1/169 - - Enterotomy - 7/169 Cystotomies (In total, 8/169, 4.7%) Mesh erosion Non-rand. comp. Visco A, sacrocolpopexy, non- A, 5/155 - - Mesh erosion Median time 200177 absorbable synthetic mesh B, 4/88 occurred: (Mersilene [Amid type III] or C, 4/25 Sacrocolpopexy: Gore-Tex [Amid type II] D, 2/5 15m (2-33); B, C, D, A, 12m (3-40) Sacrocolpoperineopexy, non- B, 9M (5-15) absorbable synthetic mesh C, 4M (2-7) (Mersilene [Amid type III] or Gore-Tex [Amid type II): B, abdominal C, vaginal+abdominal D, vaginal+abdominal with mesh placed down to the vaginal field Case series Elneil Sacrocolpopexy, 3/128 (2.3%) - - Asymptomatic At 6w 200538 sacrohysteropexy, or vaginal mesh sacrocervicopexy erosion Su 200772 Sacrocolpoperinopexy (a 14/169 (8.3%) - - Mesh erosion Median time to mixture of absorbable and occurred was 3m non-absorbable mesh/graft Operation for mesh erosion Non-rand. comp. Visco A, sacrocolpopexy, non- A, 5/155 - - 12/15 had single 200177 absorbable synthetic mesh B, 4/88 (4.6%) outpatient (Mersilene [Amid type III] or C, 4/25 (16.0%) excision, 2/25 had Gore-Tex [Amid type II] D, 2/5 (40.0%) two vaginal B, C, D, procedures, 1/25 Sacrocolpoperineopexy, non- had four excisions absorbable synthetic mesh including a (Mersilene [Amid type III] or transdominal Gore-Tex [Amid type II): resection. B, abdominal C, vaginal+abdominal D, vaginal+abdominal with mesh placed down to the vaginal field

205 Mesh n/N (%) No mesh n/N (%) Outcome Time that reported as occurred (follow up) Case series Elneil Sacrocolpopexy, 3/128 (2.3%) - - Mesh trimming in - 200538 sacrohysteropexy, or the clinic sacrocervicopexy Su 200772 Sacrocolpoperinopexy (a 1/169 (0.6%) - - Further operation - mixture of absorbable and non-absorbable mesh/graft De novo urinary symptoms in women who did not have these symptoms at baseline Case series Su 200772 Sacrocolpoperinopexy (a 4/9 (44.4%) - - Urodynamic At 1y mixture of absorbable and incontinence non-absorbable mesh/graft De novo bowel symptoms in women who did not have these symptoms at baseline De novo sexual problems in women who did not have these symptoms at baseline Infection Non-rand. comp. Demirci 20 sacrohysterexy, 25 7/45 Sacrospinous 6/60 Urinary tract - 200737 sacrocolpopexy, non- 5/45 fixation 1/60 infection; absorbable synthetic mesh 4/45 2/60 Wound infection; (polypropylene, Prolene, 0/45 1/60 Febrile morbidity; Amid type I) (In total, 16/45, 35.6%) (In total, 10/60, Vaginal vault 16.7%) infection. Case series Elneil Sacrocolpopexy, 2/128 - - Wound infection - 200538 sacrohysteropexy, or sacrocervicopexy Other serious adverse effects Non-rand. comp. Demirci 20 sacrohysteropexy, 25 3/45 (6.7%) Sacrospinous 0/60 Wound NR, 6w 200737 sacrocolpopexy, non- fixation dehiscence postoperative absorbable synthetic mesh (polypropylene, Prolene, Amid type I) Case series Elneil Sacrocolpopexy, 1/128 (0.8%) - Wound Early 200538 sacrohysteropexy, or dehiscence postoperative sacrocervicopexy NR, 19m (2-65) Su 200772 Sacrocolpoperinopexy (a 1/169 - Pulmonary NR, 14m (1.5-24) mixture of absorbable and 4/169 oedema non-absorbable mesh/graft (In total, 5/169, 3.0%) Ileus Other adverse effects Non-rand. comp. Demirci 20 sacrohysterexy, 25 7/45 (15.6%) Sacrospinous 5/60 (8.3%) Urinary retension NR, 6w 200737 sacrocolpopexy, non- fixation >5d postoperative absorbable synthetic mesh (polypropylene, Prolene, Amid type I) Case series Elneil Sacrocolpopexy, 1/128 (0.8%) - - Wound Early 200538 sacrohysteropexy, or haematoma postoperative sacrocervicopexy NR, 19m (2-65) Su 200772 Sacrocolpoperineopexy (a 4/169 (2.4%) - - Wound cellulitis NR, 14m (1.5-24) mixture of absorbable and non-absorbable mesh/graft

206 APPENDIX 21 Summary of outcomes: other mesh techniques or a mixture of mesh techniques, operation time and hospital stay

No studies reported uterine prolapse repair or vault prolapse repair alone.

Table Uterine and/or vault prolapse repair Mesh Mean(SD)/ No mesh Mean(SD)/ median (range) median (range) Operation time Non-rand. comp. Demirci 200737 20 sacrohysterexy, 25 sacrocolpopexy, non- 192 (38), n=45 Sacrospinous 141 (28), n=60 absorbable synthetic mesh (polypropylene, fixation Prolene, Amid type I) Hospital stay Non-rand. comp. Demirci 200737 20 sacrohysterexy, 25 sacrocolpopexy, non- 7 (2), n=45 Sacrospinous 6 (2), n=60 absorbable synthetic mesh (polypropylene, fixation Prolene, Amid type I) Case series Su 200772 Sacrocolpoperinopexy (a mixture of absorbable 3.3d (0.8), 3-8, n=169 and non-absorbable mesh/graft

207