Laparoscopic colposuspension for urinary incontinence in women
Abstract
Background
Stress urinary incontinence (SUI) imposes significant health and economic burden on society and the women affected. Laparoscopic colposuspension was one of the first minimal access operations for the treatment of women with SUI, with the presumed advantages of avoiding major incisions, shorter hospital stays and quicker return to normal activities.
Objectives
To determine the effects of laparoscopic colposuspension for urinary incontinence in women.
Search methods
We searched the Cochrane Incontinence Group Trials Register (searched 2 July 2009), and sought additional trials from other sources and by contacting study authors for unpublished data and trials.
Selection criteria
Randomised or quasi‐randomised controlled trials in women with symptomatic or urodynamic diagnosis of stress or mixed incontinence that included laparoscopic surgery as the intervention in at least one arm of the studies.
Data collection and analysis
The review authors evaluated trials for methodological quality and their appropriateness for inclusion in the review. Two review authors extracted data and another cross checked them. Where appropriate, we calculated a summary statistic.
Main results
We identified 22 eligible trials. Ten involved the comparison of laparoscopic with open colposuspension. Whilst the women's subjective impression of cure seemed similar for both procedures, in the short‐ and medium‐term follow‐up, there was some evidence of poorer results of laparoscopic colposuspension on objective outcomes. The results showed trends towards fewer perioperative complications, less postoperative pain and shorter hospital stay for laparoscopic compared with open colposuspension, however, laparoscopic colposuspension was more costly.
Eight studies compared laparoscopic colposuspension with newer 'self‐fixing' vaginal slings. There were no significant differences in the reported short‐ and long‐term subjective cure rates of the two procedures but objective cure rates at 18 months favoured slings. We observed no significant differences for postoperative voiding dysfunction and perioperative complications. Laparoscopic colposuspension had a significantly longer operation time and hospital stay.
We found significantly higher subjective and objective one‐year cure rates for women randomised to two paravaginal sutures compared with one suture in a single trial. Three studies compared sutures with mesh and staples for laparoscopic colposuspension and showed a trend towards favouring the use of sutures.
Authors' conclusions
Currently available evidence suggests that laparoscopic colposuspension may be as good as open colposuspension at two years post surgery. However, the newer vaginal sling procedures appear to offer even greater benefits, better objective outcomes in the short term and similar subjective outcomes in the longer term. If laparoscopic colposuspension is performed, the use of two paravaginal sutures appears to be the most effective method. The place of laparoscopic colposuspension in clinical practice should become clearer when there are more data available describing long‐term results. A brief economic commentary (BEC) identified three studies suggesting that tension‐free vaginal tape (TVT) may be more cost‐effective compared with laparoscopic colposuspension but laparoscopic colposuspension may be slightly more cost‐effective when compared with open colposuspension after 24 months follow‐up.
PICOs
Plain language summary
Keyhole (laparoscopic) surgery through the abdomen for treating urinary incontinence in women
Urinary incontinence is a common and often debilitating problem for many women. Around a third of women of child‐bearing age are incontinent during physical exertion or when they cough, laugh or sneeze. When such 'stress' incontinence persists despite non‐surgical treatment, surgery is often recommended. A significant amount of a woman's and their family's income can be spent on management of stress urinary incontinence.
Laparoscopic colposuspension is an operation carried out through a small incision in the abdomen to hold and support the tissues around the neck of the bladder. Women recover more quickly from laparoscopic colposuspension for urinary incontinence than from traditional, open surgery, with similar initial improvement. Longer‐term success rates may be lower but this is uncertain. However, when laparoscopic colposuspension is compared with newer 'self‐fixing' sling procedures, it appears that in the short term the sling procedures offer the greater benefits of minimal access techniques with similar, if not better cure rates.
The review of trials found that both traditional colposuspension and newer 'self‐fixing' slings had technically better results in the short term when compared with laparoscopic colposuspension. However, women's experience of improvement, both in the short and long term was similar for each type of operation. Using two stitches in laparoscopic colposuspensions was better than both one suture or the use of mesh. A trend was shown towards better outcomes for the laparoscopic operation when compared to open surgery, such as less postoperative pain, shorter hospital stay, quicker time to return to normal activities and shorter duration of catheterisation. When the laparoscopic technique was compared to the newer vaginal sling procedures, all the aforementioned trends were in favour of the sling procedure.
In terms of costs, a non‐systematic review of economic studies suggested that tension‐free vaginal tape and open colposuspension would be cheaper than laparoscopic colposuspension.
The value of the review is limited by the size and quality of the trials and the few data about long‐term results.
Authors' conclusions
Background
Urinary incontinence is a common and often debilitating problem for many adult women. Some degree of stress incontinence affects around a third of women of child‐bearing age (Wilson 1996). Continence is achieved by a combination of normal anatomical and physiological properties of the bladder, urethra and sphincter, the pelvic floor, and also the nervous system co‐ordinating these organs. Disruption to any of these components may lead to incontinence.
There are different types of urinary incontinence. 'Stress urinary incontinence' (SUI) is defined by the symptom of involuntary loss of urine associated with physical exertion and activities that increase intra‐abdominal pressure. The International Continence Society defines 'urodynamic stress incontinence' (USI) as the involuntary leakage of urine during increased abdominal pressure in the absence of a detrusor contraction, noted during filling cystometry (Abrams 2002). Therefore, diagnosis based on this definition requires urodynamic investigation. 'Urge urinary incontinence' is defined as involuntary loss of urine accompanied by or immediately preceded by urgency. Detrusor overactivity (DO) is a diagnosis of involuntary detrusor muscle contractions that are not due to neurological disorders; and the diagnosis must be made using urodynamic investigations (Abrams 2002). Mixed urinary incontinence is the complaint of involuntary leakage associated with urgency and also with exertion, effort, sneezing or coughing (Abrams 2002). This review includes women with SUI, proven urodynamic stress incontinence, women with symptoms of stress plus urge incontinence or other urinary symptoms diagnosed clinically or urodynamic stress incontinence plus detrusor overactivity diagnosed using urodynamics.
Stress urinary incontinence constitutes a huge financial economic burden to society. In the USA, the annual total direct costs of urinary incontinence in both men and women is over USD 16 billion (1995 USD) (Chong 2011) with societal costs of USD 26.2billion (1995 USD) (Wagner 1998). Approximately, USD 13.12 billion (1995 USD) of the total direct costs of urinary incontinence is spent on SUI (Chong 2011; Kunkle 2015). About 70% of this USD 13.12 billion is borne by people with SUI, mainly through routine care (purchasing pads and disposable underwear (diapers), laundry and dry cleaning). This constitutes a significant individual financial burden. Of the remaining 30% of costs, 14% is spent on nursing home admission, 9% on treatment, 6% on addressing complications and 1% on diagnosis (Chong 2011).
A study in the USA reported that about 1% of the median annual household income (USD 50,000 to USD 59,999 in 2006) was spent by women on incontinence management. This study estimated that women spent an annual mean amount of USD 751 to USD 1277 (2006 USD) on incontinence. This cost increases with the severity of the symptoms (Subak 2008). The indirect cost associated with SUI exerts a social and psychological burden that is unquantifiable (Chong 2011; Kilonzo 2004), nevertheless, Birnbaum 2004 estimated that the annual average direct medical costs of SUI for one year (1998 USD) was USD 5642 and USD 4208 for indirect workplace costs. The cost of management and treatment of SUI appears to have increased over time due to increasing prevalence and increased desire for improved quality of life. This, in turn, has resulted from improved recognition of the condition, as well as increased use of surgical and non‐surgical managements.
Non‐surgical treatments for SUI include conservative (Hay‐Smith 2006; Hay‐Smith 2008; Herbison 2002; Wallace 2004) and pharmacological therapies (Hay‐Smith 2005; Nabi 2006). These interventions are the subject of separate Cochrane Reviews. This review is one of a series of interrelated Cochrane Reviews of surgical approaches to the management of urinary incontinence. We refer the reader to another review in the series (Lapitan 2009) for further background information about the mechanisms of urinary incontinence in women, the principal categories of incontinence and the broad approaches to management.
Surgical procedures to remedy SUI generally aim to lift and support the urethro‐vesical junction between the urethra and the bladder and increase bladder outlet resistance. There is disagreement, however, regarding the precise mechanism by which continence is achieved after surgery. The choice of procedures is often influenced by co‐existent problems, a surgeon's specialty or preference and the physical features of the person affected. Numerous surgical methods have been described but essentially they fall into seven categories:
-
open abdominal retropubic urethropexy (e.g. colposuspension (Burch), Marshall‐Marchetti‐Krantz) (Lapitan 2009);
-
laparoscopic retropubic colposuspension (this review);
-
anterior vaginal repair (anterior colporrhaphy) (e.g. Kelly, Pacey) (Glazener 2001);
-
suburethral slings (traditional and newer 'self‐fixing') (Bezerra 2005; Ogah 2009);
-
needle suspensions (e.g. Pereyra, Stamey) (Glazener 2004);
-
periurethral injections (Keegan 2007); and
-
artificial sphincters.
This review concentrates on laparoscopic retropubic colposuspension; other Cochrane Reviews address six of the other categories.
Laparoscopic incontinence procedures were first introduced in the early 1990s (Vancaillie 1991) with the advantage to women that they avoided the major incisions of conventional open surgery. It was claimed that this would result in shorter lengths of hospital stay and shorten the time to return to normal activities.
The most popular laparoscopic procedure for urinary incontinence is laparoscopic colposuspension. As in open colposuspension, sutures are inserted into the paravaginal tissues on either side of the bladder neck and then attached to the ileopectineal ligaments on the same side. There are, however, technical variations in surgery in respect of the laparoscopic approach (transperitoneal into the abdominal cavity or extraperitoneal) and in the number and types of sutures, the site of anchor and the use of mesh and staples (Jarvis 1999).
Objectives
To determine the effects of laparoscopic colposuspension for urinary incontinence in women.
We tested the following hypotheses.
-
Laparoscopic colposuspension is better than no treatment or sham operation for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than conservative interventions (e.g. pelvic floor muscle training, electrical stimulation, cones, biofeedback) for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than open colposuspension (abdominal surgery) for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than needle suspension (abdominal and vaginal surgery) for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than traditional sling procedures (abdominal and vaginal surgery) for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than newer 'self‐fixing' sling procedures (abdominal and vaginal surgery) for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than anterior vaginal repair for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Laparoscopic colposuspension is better than periurethral injections for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
-
Some methods of laparoscopic colposuspension are better than others for the management of urodynamic stress incontinence (urodynamic diagnosis) or for symptoms of stress or mixed incontinence (clinical diagnosis).
Methods
Criteria for considering studies for this review
Types of studies
All randomised or quasi‐randomised controlled trials studying effects of treatment for stress or mixed urinary incontinence in women, where at least one management arm involved laparoscopic colposuspension.
Types of participants
Adult women with urinary incontinence who were diagnosed as having:
-
urodynamic stress incontinence (urodynamic diagnosis); or
-
stress incontinence (clinical diagnosis); or
-
mixed incontinence (stress incontinence plus other urinary symptoms such as urge incontinence).
The definitions for the above diagnoses were those used by the authors of the trial reports.
Types of interventions
At least one arm of a study must have involved laparoscopic colposuspension to treat urinary incontinence.
Comparison interventions were those described in the eight hypotheses above.
Types of outcome measures
We sought data for the following measures of outcome with subjective cure considered as the primary measure.
A. Women's observations
1. Subjective cure within 18 months (women's perception of cure)
2. Subjective cure after 18 months and within five years
3. Subjective cure after five years
B. Quantification of symptoms
4. Incontinent episodes (from self‐completed bladder chart)
5. Pad tests of quantified leakage (mean volume or weight of urine loss)
C. Clinicians' measures
6. Stress testing (alone or at cystometrogram)
7. Urge symptoms or urge incontinence (clinical diagnosis without urodynamics)
8. De novo detrusor instability (urodynamic diagnosis)
9. Voiding dysfunction or difficulty within 18 months, with or without urodynamic confirmation (residual urine estimation, flowmetry)
10. Urodynamically assessed cure (the absence of urinary stress incontinence on filling cystometry)
D. Quality of life
11. General health status measures e.g. Short Form 36 (Ware 1993) or specific instruments designed to assess incontinence
E. Surgical outcome measures
12. Perioperative surgical complications e.g. infection, haemorrhage
13. Pain or analgesia requirements
14. Length of inpatient stay
15. Time to return to normal activity level
16. Repeat incontinence surgery
F. Other outcomes
17. Non‐prespecified outcomes judged important when performing the review
Search methods for identification of studies
We did not impose any language or other limits on the searches.
Electronic searches
This review has drawn on the search strategy developed for Cochrane Incontinence as a whole. We have identified relevant trials primarily from the Cochrane Incontinence Trials Register. The methods used to derive this, including the search strategy, are described under the Group's details in the Cochrane Library. For more details please see the ‘Specialized Register’ section of the Group's module in the Cochrane Library). The register contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, CINAHL, and handsearching of journals and conference proceedings.
We searched the Cochrane Incontinence Trials Register using the Group's own keyword system. The search terms used were as follows.
{design.rct* or design.cct*}
AND
{TOPIC.URINE.INCON*}
AND
{INTVENT.SURG.LAP*}
All searches were of the keyword field of Reference Manager 2002. The date of the most recent search of the register for this review was 2 July 2009. The trials in the Incontinence Group Specialised Register are also contained in CENTRAL.
We performed additional searches for the brief economic commentary (BEC). These were conducted in MEDLINE (1 January 1946 to March 2017), Embase (1 January 1980 to 2017 Week 12) and NHS EED (1st Quarter 2016). All searches were conducted on 6 April 2017. Details of the searches run and the search terms used can be found in Appendix 1.
Searching other resources
We conducted the following, additional searches for this review: we checked all reference lists of identified trials and other relevant articles; we contacted authors and trialists in the field to identify any additional or unpublished data or studies.
Data collection and analysis
We identified randomised and quasi‐randomised trials using the above search strategy. We excluded studies from the review if they were not randomised or quasi‐randomised controlled trials for incontinent people or if they made comparisons other than those prespecified. Excluded studies are listed with reasons for their exclusion.
The review authors evaluated the reports of all possibly eligible studies for methodological quality and appropriateness for inclusion, without prior consideration of the results. Each review author assessed methodological quality using the Cochrane Incontinence assessment criteria, which include quality of random allocation and concealment, description of dropouts and withdrawals, analysis by intention to treat and blinding during treatment and at outcome assessment. We resolved any differences of opinion by discussion with a third party.
At least two review authors independently undertook data extraction of individual studies. Where data may have been collected but not reported, we sought clarification from the trialists. We planned to group trial data by type of incontinence, either urodynamic stress incontinence, based on a urodynamic diagnosis or stress or mixed incontinence, based upon a symptom classification. In the event, however, all included trials were limited to women with urodynamic stress incontinence.
Included trial data were processed as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2005). When appropriate, we carried out meta‐analysis. For categorical outcomes, we derived a risk ratio (RR) by relating the number reporting an outcome to the number at risk in each group. For continuous variables we used means and standard deviations to derive a mean difference and combined studies to get a weighted mean difference (WMD). Where appropriate, we used a fixed‐effect model to calculate the combined estimates and their 95% confidence intervals (CI). We examined heterogeneity between trial results using visual inspection of the forest plot and taking into account the results of Chi2 tests for heterogeneity and I2 tests (Higgins 2003). Due to the small number of studies, we could not carry out sensitivity analyses for methodological quality.
We subgrouped trial data by type of material used for the laparoscopic colposuspension and the technique used for the self‐fixing sling: either mesh or sutures and tension‐free vaginal tape (TVT) or suprapubic arc (SPARC).
Results
Description of studies
Included and excluded studies
We identified 22 randomised trials of surgery for urinary incontinence, with laparoscopic colposuspension in at least one arm, to be included in this review. There were 14 additional studies to the eight studies identified in 2000 for the original review (Burton 1997; Carey 2000; Fatthy 2001; Persson 2000; Ross 1995; Su 1997; Summitt 2000; Wallwiener 1995). We excluded three studies, two because they were not randomised (Lee 1997; Lernis 1997) and the third study because it compared two‐dimensional view laparoscopy with three‐dimensional view laparoscopy for different gynaecological operations and only three participants had colposuspension, all randomly allocated to the same group (Koster 1996).
Publication type
Two of the 22 included studies had published versions in both English and either German (Wallwiener 1995) or Italian (Adile 2001). Six studies were reported only as abstracts (Adile 2001; Burton 1997; Maher 2004; Mirosh 2005; Morris 2001; Summitt 2000). Where more detailed study information was required we contacted the study authors and used the information gained (some currently unpublished) in the review.
Sample characteristics
Sample sizes ranged from 20 to 300, with the majority of the trials having fewer than 50 women in each arm. Judged on the study reports, all participants had prior urodynamic investigation although this was not always explicitly stated. All studies, therefore, appeared to have been undertaken on women with urodynamic stress incontinence. The inclusion and exclusion criteria varied in other respects. All but two studies (Morris 2001; Ustun 2003) stated their exclusion criteria. Fifteen trials excluded women with previous anti‐incontinence surgery; six did not (Carey 2000; Fatthy 2001; Foote 2004; Kitchener 2006; Maher 2004; Wallwiener 1995) and for one of these studies recurrent urinary stress incontinence was an inclusion criteria (Maher 2004). Twelve trials excluded women who had had previous retropubic surgery (Adile 2001; Ankardal 2004; Ankardal 2005; Burton 1997; Carey 2000; Foote 2004; Kitchener 2006; Mirosh 2005; Persson 2000; Su 1997; Summitt 2000; Zullo 2001). Nine studies excluded women with detrusor overactivity (mixed urinary incontinence) (Fatthy 2001; Foote 2004; Kitchener 2006; Mirosh 2005; Paraiso 2004; Ross 1995; Su 1997; Summitt 2000; Valpas 2004) and one study excluded women with urge incontinence but included women with urgency symptoms (Persson 2000). Ten studies excluded women with varying degrees of pelvic organ prolapse: greater than grade one uterine prolapse or cystocele (Persson 2000; Su 1997; Wallwiener 1995), greater than or equal to stage two (Persson 2002; Zullo 2001), any significant prolapse (Foote 2004; Maher 2004; Paraiso 2004) and grade three prolapse or greater (Fatthy 2001; Mirosh 2005). Eight studies excluded women who required coincident additional gynaecological operations (Adile 2001; Ankardal 2004; Ankardal 2005; Foote 2004; Mirosh 2005; Persson 2002; Summitt 2000; Valpas 2004), making the trial intervention a sole procedure.
Comparison of interventions
Ten studies compared laparoscopic colposuspension with open colposuspension (Ankardal 2004; Ankardal 2005; Burton 1997; Carey 2000; Cheon 2003; Fatthy 2001; Kitchener 2006; Morris 2001; Su 1997; Summitt 2000) but they were not consistent in either the number or type of mesh or sutures used. Burton 1997 used absorbable Dexon sutures, whereas six trials used non‐absorbable sutures, mainly Ethibond (Ankardal 2004; Ankardal 2005; Cheon 2003; Kitchener 2006; Su 1997; Summitt 2000) or polypropylene sutures (Fatthy 2001). All studies with the exception of two (Ankardal 2004; Su 1997) used the same number and type of suspension material for both the laparoscopic and open procedures. In the study by Ankardal and colleagues (Ankardal 2004) a polypropylene mesh and Titan staples were used for the laparoscopic colposuspensions and sutures for the open colposuspensions. Unpublished information by Su and colleagues (Su 1997) revealed that the paravaginal fascia was held by one suture on each side in the laparoscopic procedure and by three sutures each side in the open procedure. The surgeons in three studies (Burton 1997; Cheon 2003; Fatthy 2001) had performed fewer than 20 laparoscopic colposuspensions before starting the trial; other studies either reported surgeons as being senior gynaecologists with extensive experience in both procedures (Ankardal 2004; Ankardal 2005; Carey 2000; Kitchener 2006; Su 1997) or did not report the experience of the surgeons (Morris 2001; Summitt 2000).
Eight studies (Adile 2001; Foote 2004; Maher 2004; Mirosh 2005; Paraiso 2004; Persson 2002; Ustun 2003; Valpas 2004) compared laparoscopic colposuspension with self‐fixing vaginal mesh slings. Seven of these studies used a tension‐free vaginal tape (TVT) for the vaginal sling technique and one study (Foote 2004) used SPARCTM. One study (Valpas 2004) used TackerTM mesh for the laparoscopic colposuspension whereas four studies (Foote 2004; Paraiso 2004; Persson 2002; Ustun 2003) reported using two single‐bite non‐absorbable sutures each side. The main differences between the interventions in this group was the type of anaesthesia used. In all trials the laparoscopic colposuspensions were performed under a general anaesthetic. The self‐fixing sling operations were performed under local anaesthetic with sedation in two studies (Persson 2002; Valpas 2004); under regional anaesthesia (spinal or epidural) in two studies (Adile 2001; Mirosh 2005) and a combination of general, regional and local with sedation in two other studies (Paraiso 2004; Ustun 2003).
Five studies (Ankardal 2005; Persson 2000; Ross 1995; Wallwiener 1995; Zullo 2001) compared different operative techniques or approaches for laparoscopic colposuspension. Of these, Ankardal 2005; Ross 1995 and Zullo 2001 compared polypropylene mesh fixed with staples or tacks with EthibondTM or Gore‐TexTM sutures; Persson 2000 compared two single‐bite Gore‐Tex sutures with one double‐bite suture on each side of the urethra; and Wallwiener 1995 compared extraperitoneal with transperitoneal access using Gore‐Tex sutures or mesh and staples.
Of the included 22 studies just over half stated the method of approach used for the laparoscopic colposuspension. Ten studies used the transperitoneal approach (Ankardal 2004; Burton 1997; Carey 2000; Cheon 2003; Foote 2004; Persson 2000; Ross 1995; Su 1997; Summitt 2000; Zullo 2001), three studies (Fatthy 2001; Paraiso 2004; Valpas 2004) used the extraperitoneal approach and Wallwiener 1995 compared a transperitoneal with an extraperitoneal approach.
Outcome measures
Fifteen studies reported subjective cure measures of operative success (Ankardal 2004; Ankardal 2005; Carey 2000; Cheon 2003; Fatthy 2001; Foote 2004; Maher 2004; Morris 2001; Persson 2000; Persson 2002; Kitchener 2006; Su 1997; Ustun 2003; Valpas 2004; Zullo 2001) although they used different instruments and scales to assess cure.
Thirteen studies performed objective measures in the form of a pad test or reported incontinent episodes (Ankardal 2004; Ankardal 2005; Burton 1997; Cheon 2003; Fatthy 2001; Foote 2004; Kitchener 2006; Morris 2001; Paraiso 2004; Persson 2000; Persson 2002; Su 1997; Valpas 2004). Eleven studies (Burton 1997; Carey 2000; Cheon 2003; Fatthy 2001; Maher 2004; Paraiso 2004; Ross 1995; Summitt 2000; Ustun 2003; Zullo 2001) used urodynamic measurements pre and postoperatively, but put their emphasis on different parameters, not all using the measurements as their definition of cure. All but three studies reported the number of perioperative complications and types of complications (Maher 2004; Morris 2001; Wallwiener 1995). These three studies reported either no major complications or that the number of complications were similar in each arm.
Ten studies (Ankardal 2004; Ankardal 2005; Carey 2000; Cheon 2003; Foote 2004; Kitchener 2006; Maher 2004; Mirosh 2005; Paraiso 2004; Valpas 2004) assessed quality of life as an outcome, however all the studies used various assessment questionnaires (for example (Short Form 36 (SF‐36), King's Health Questionnaire (KHQ), Incontinence Impact Questionnaire (IIQ), Urogenital Distress Inventory (UDI)) and in differing combinations. Two studies (Cheon 2003; Mirosh 2005) that assessed quality of life did not report which validated questionnaire they used and did not report the results.
All studies with the exception of one (Morris 2001) reported follow‐up outcomes within 18 months. Three studies (Burton 1997; Kitchener 2006; Zullo 2001) had a follow‐up time longer than 18 months and within five years and three studies (Burton 1997; Morris 2001; Paraiso 2004) reported on follow up after five years.
Further characteristics of the trials are reported in the tables 'Characteristics of included studies' and 'Characteristics of excluded studies'.
Risk of bias in included studies
Allocation concealment
Ten trials used an adequately concealed group allocation (Ankardal 2004; Ankardal 2005; Burton 1997; Cheon 2003; Fatthy 2001; Kitchener 2006; Summitt 2000; Paraiso 2004; Persson 2002; Persson 2000) while in seven trials no details were given about concealment of group allocation (Carey 2000; Foote 2004; Maher 2004; Ross 1995; Valpas 2004; Wallwiener 1995; Zullo 2001). In one study a random sequence of group allocation had initially been achieved by computer‐generated random numbers in sealed, opaque envelopes; then four participants, who were not willing to undergo laparoscopic colposuspension, were enrolled to the open colposuspension group, the respective next participant was assigned to the laparoscopic procedure and the following participants went back to the sequence of random numbers (Su 1997).
Blinding
Carey 2000 blinded participants and assessors, where the abdominal wounds were dressed in theatre in an identical fashion so that both participants and ward staff were masked to the procedure performed until removal of dressings prior to discharge. In one study participants could refuse one treatment and were allocated to the other group (Su 1997). In four studies assessors were blind to the type of operation performed but the participants were not (Burton 1997; Fatthy 2001; Zullo 2001; Persson 2002). Four study reports mentioned that assessors were not blinded (Persson 2000; Su 1997; Kitchener 2006; Ankardal 2004).
Data‐dependent stopping
Three trials were stopped earlier than intended. Persson 2000 set out to enrol 280 participants but the trialists became convinced during follow‐up that one trial arm had a higher cure rate. For ethical reasons they therefore performed an interim analysis on the objective one‐year cure rate on 108 participants, which confirmed a significantly higher cure rate in one group. Thereafter, further enrolment was stopped and only the 161 participants who had already been recruited were followed up and the analysis was adjusted for the early stopping. Su 1997 employed early stopping rules with evaluation points, which were determined before the beginning of the trial. The trial was stopped at the second evaluation point but no further explanation was given; 92 of 152 intended participants were recruited. In Valpas 2004 the number of women recruited fell short of the target owing to limitations of time and recruiting problems and for these reasons this study was also stopped prematurely. However, in this study the groups were still considered to be comparable and the data were analysed without adjustment after stopping the study.
Dropouts and losses to follow‐up
Eleven studies reported losses to follow‐up and dropouts (Ankardal 2004; Ankardal 2005; Burton 1997; Fatthy 2001; Maher 2004; Paraiso 2004; Persson 2000; Persson 2002; Kitchener 2006; Valpas 2004; Zullo 2001) that ranged from 1 to 22 participants across the studies.
Effects of interventions
Hypothesis 1: laparoscopic colposuspension versus no treatment or sham operation
No eligible studies were found.
Hypothesis 2: laparoscopic colposuspension versus conservative management
No eligible studies were found.
Hypothesis 3: laparoscopic colposuspension versus open colposuspension
Ten trials compared laparoscopic with open colposuspension (Ankardal 2004; Ankardal 2005; Burton 1997; Carey 2000; Cheon 2003; Fatthy 2001; Kitchener 2006; Morris 2001; Su 1997; Summitt 2000). All had different lengths of follow‐up: Carey et al (Carey 2000) for six months; Su et al (Su 1997), Ankardal et al studies (Ankardal 2004; Ankardal 2005), Cheon et al (Cheon 2003) and Summitt et al (Summitt 2000) for one year; Fatthy et al (Fatthy 2001) for six and 18 months; Smith et al (Kitchener 2006) for six, 12 and 24 months; and Burton et al (Burton 1997) for six months, one year, three years and five years. Outcome data for 'six months to 18 months' were therefore available for eight studies. Longer‐term data, over five years, were only available for two studies (Burton 1997; Morris 2001) and the study by Morris et al had only five to seven year follow‐up data with no earlier follow‐up results reported. The ability to synthesise data was also limited by the variable tests and definitions used to measure subjective and objective outcomes across the trials and failure to report standard deviations. For these reasons some data are reported in the 'Additional tables' (Tables 1‐9).
Women's observations
Subjective cure rates ranged from 58% to 96% in the open and 62% to 100% in the laparoscopic group within the 18 months follow‐up, with a non‐significant 5% lower relative subjective cure rate for laparoscopic colposuspension (Comparison 01.01.01; RR 0.95, 95% CI 0.90 to 1.00). Because of concerns about the effects of problems with randomisation and an inconsistent use of sutures for the different operations in the study Su 1997, we re‐ran the analysis after excluding this study and this did not change the results. The statistically significant heterogeneity mostly reflected the results of one Ankardal trial (Ankardal 2004), for which there were concerns that post randomisation withdrawals may have introduced bias. We could not include the results of the Burton 1997 trial in these analyses because of the way they had presented the data. In this trial (Table 1), based on visual analogue scales of 'cure', the open group had better results; increasingly so at three and five years. The Kitchener 2006 trial reported subjective cure rates at two years and there were no differences between the laparoscopic and open groups (Comparison 01.01.02; RR 1.00, 95% CI 0.81 to 1.25). The study Morris 2001 reported data from five to seven years of follow‐up, however, this differed from the results reported in Burton 1997 as they showed that the laparoscopic colposuspension group had a significantly higher subjective cure rate than the open colposuspension group (Comparison 01.01.03; RR 1.53, 95% CI 1.00 to 2.35).
| 6/12 | 1 year | 3 years | 5 years | |
| open | 10.0 | 9.8 | 9.6 | 9.4 |
| laparoscopic | 7.8 | 6.7 | 4.8 | 4.4 |
Quantification of symptoms
Two studies used urinary diaries and pad tests to quantify symptoms at follow‐up (Burton 1997; Fatthy 2001). The results for the Burton 1997 are shown in the Additional tables (Table 2 and Table 3). Although both groups improved significantly after surgery this was less in the laparoscopic group. Furthermore, there was marked deterioration over time in this group such that by five years the mean number of incontinent episodes in the laparoscopic group was approaching the preoperative level. In contrast, there was no apparent difference between the groups in incontinent episodes at 18 months in Fatthy's study (Fatthy 2001) (Comparison 01.02; MD ‐0.12 episodes, 95% CI ‐0.68 to 0.92).
| pre‐op | 6/12 | 1 year | 3 years | 5 years | |
| open | 12 | 0 | 2 | 2 | 3 |
| laparoscopic | 13 | 2 | 6 | 8 | 9 |
| pre‐op | 6/12 | 1 year | 3 years | 5 years | |
| open | 22 | 1 | 2 | 3 | 5 |
| laparoscopic | 24 | 5 | 12 | 14 | 15 |
Clinicians' measures
All 10 studies assessed leakage on a clinical stress test and they reported the data as objective cure rates. Six studies used negative urodynamic testing as their definition of cure and four studies (Ankardal 2004; Ankardal 2005; Kitchener 2006; Morris 2001) used a negative pad test as their definition of objective cure. Overall, objective cure rate within 18 months showed a statistically significant reduction of RR in laparoscopic colposuspension, thus favouring open colposuspension (Comparison 01.03.01; RR 0.91, 95% CI 0.86 to 0.96). The analysis for objective cure rates between 18 months and five years showed no significant differences between laparoscopic and open colposuspensions (Comparison 01.03.02; RR 1.01, 95% CI 0.88 to 1.16). There was, however, evidence of heterogeneity with the smaller Burton (Burton 1997) trial greatly favouring open colposuspension. Similar results were seen with the analysis of objective cure rates after five years' follow‐up as Burton 1997 reported objective cure rates favouring the open technique, which was opposite to the findings from Morris 2001 (Comparison 01.03.03; RR 0.94, 95% CI 0.69 to 1.27).
There were no statistically significant differences in de novo detrusor overactivity within 18 months (Comparison 01.04.01; RR 1.29, 95% CI 0.72 to 2.30). Longer‐term follow‐up data (between 18 months and five years and after five years) also showed no statistically significant difference (Comparisons 01.04.02 and 01.04.03; RR 1.00 and 1.22, 95% CI 0.07 to 15.12 and 0.40 to 3.75).
Similarly, there were no statistically significant differences in voiding dysfunction (Comparison 01.05; RR 0.82, 95% CI 0.50 to 1.35). Burton 1997 recorded higher residual urinary volumes at follow up in the laparoscopic group whereas in Fatthy 2001 the residual volumes were similar in the two groups (Table 4).
| Study | pre‐op | 6/12 | 1 year | 18/12 | 3 years | 5 years |
| Burton 1997 | ||||||
| open | 24 | 33 | 30 | 26 | 32 | |
| laparoscopic | 21 | 42 | 40 | 42 | 49 | |
| Fatthy 2001 | ||||||
| open | 23.9 | 30.95 | ||||
| laparoscopic | 26.65 | 32.85 |
Six studies used urodynamic investigations to assess cure objectively. Overall, there was a significantly higher success rate following open colposuspension (Comparison 01.06; RR 0.91, 95% CI 0.85 to 0.99). After exclusion of Su 1997 data the results still favoured open colposuspension (RR 0.92, 95% CI 0.84 to 1.00).
Quality of life
Five studies measured quality of life using a variety of questionnaires (Ankardal 2004; Ankardal 2005; Carey 2000; Cheon 2003; Kitchener 2006). Carey 2000 used the SF36 (Short Form 36), SUDI (Short Urogenital Distress Inventory) and IIQ questionnaires; both the SUDI and IIQ scores equally improved postoperatively in the two groups but no further details were available. Cheon 2003 assessed quality of life but details of the instrument used and further results were lacking in the study report. In both their studies, Ankardal and colleagues used a visual analogue score (VAS) to assess the bother symptoms caused and their impact on different quality‐of‐life domains (Ankardal 2004; Ankardal 2005). In Ankardal 2004, the scores had improved in all domains following surgery, in both groups; however, the improvement was greater in the open colposuspension group with regard to physical activity one year following surgery (Table 5). The study by Smith and colleagues (Kitchener 2006) used a number of validated questionnaires; these included SF‐36, the Bristol Female Lower Urinary Tract Symptom Questionnaire (BFLUTS), Symptom Severity and Symptom Impact Index and the EQ‐5D. Results for the SF‐36 and EQ‐5D were reported and showed that both treatment arms had improved scores after two years with no suggestion of any group differences (Table 6).
| Domain of QoL | laparoscopic colpo | Open colpo | p value |
| Physical activity | 63 (27‐81) | 74 (52‐84) | <0.05 |
| Working ability | 25 (4‐54) | 31 (10‐60) | NS |
| Social life | 37 (15‐66) | 38 (10‐65) | NS |
| Sexual life | 13 (1‐41) | 9 (0‐37) | NS |
| median (range) | median (range) |
| Score at 24 months | Laparoscopic colpoe | Open colpo |
| EQ‐5D values | 0.844 (0.259) | 0.825 (0.270) |
| SF‐36 Physical sub‐scale | 79.32 (27.59) | 77.60 (27.74) |
| SF‐36 Mental sub‐scale | 69.51 (21.21) | 69.38 (22.65) |
Surgical outcome measures
There were significantly fewer perioperative complications in the laparoscopic colposuspension group (Comparison 01.07; RR 0.74, 95% CI 0.58 to 0.96). However, 21 bladder injuries were reported amongst 521 laparoscopic procedures compared with 10 amongst 507 open operations (Comparison 01.08; RR 1.88, 95% CI 0.93 to 3.83). Two studies reported cases of laceration to the obturator vein during laparoscopic colposuspension (Carey 2000; Summitt 2000).
Women who underwent a laparoscopic colposuspension appeared to have significantly less pain and needed less postoperative analgesia but data from the studies were not presented in a form suitable for quantitative synthesis (Burton 1997; Carey 2000; Cheon 2003; Fatthy 2001; Kitchener 2006; Su 1997). All studies with the exception of Morris 2001 reported the length of hospital stay to be longer for open colposuspension. Six studies reported data in such a way that could be analysed (Comparison 01.09; WMD 0.86 days, 95% CI 0.65 to 1.06) so favouring laparoscopic colposuspension due to a shorter hospital stay. There was, however, significant heterogeneity between the trials in this respect. The time to return to normal activities, where reported, was longer for open colposuspension (Table 7). Three women (10%) in the laparoscopic group in Burton 1997 had repeat incontinence surgery compared with none in the open group; none of the other studies reported this outcome.
| Carey (mean) | Fatthy (mean) | Summitt (mean) | Cheon (mean) | |
| open | 21.8 | 31.5 | 37.3 | 29.3 |
| laparoscopic | 17.3 | 8.5 | 35.5 | 22.2 |
Other outcomes
Laparoscopic surgery took significantly longer than open colposuspension in the majority of the studies (Comparison 01.11; MD 14.06 minutes, 95% CI 11.39 to 16.73). Although there was significant statistical heterogeneity the results of the three largest trials (contributing 85% of the weight) were consistent in suggesting a 14 to 17 minute difference. The estimated blood loss was higher in the open groups in all studies that reported these data (Table 8). The duration of catheterisation was longer after open than after laparoscopic colposuspension in Ankardal 2004; Ankardal 2005; Burton 1997; Carey 2000; Cheon 2003; and Su 1997 but was similar for both procedures in Summitt 2000 (Table 9).
| Burton(median+range) | Carey (mean) | Fatthy (mean+SD) | Su (mean +SD) | Summitt (mean) | Ankardal (mean) | Cheon (mean) | |
| open | 261 (120 ‐ 410) | 171 | 240.5 (35.5) | 134.3 (102.0) | 131.9 | 105 | 327 |
| laparoscopic | 190 (180 ‐ 290) | 125 | 42.75 (7.2) | 59.3 (42.1) | 112 | 35 | 125 |
| Burton (mean) | Carey (mean) | Su (mean + SD) | Summitt (mean) | Ankardal (mean + SD) | Cheon (mean +SD) | |
| open | 1.7 | 4.9 | 6.8 (2.3) | 2.7 | 4.9 (8.3) | 3.7 (2.8) |
| laparoscopic | 1.33 | 4.4 | 3.9 (1.9) | 2.7 | 1.9 (2.4) | 3.0 (1.6) |
Hypothesis 4: laparoscopic colposuspension versus needle suspension
No eligible studies were found.
Hypothesis 5: laparoscopic colposuspension versus traditional sling procedures
No eligible studies were found.
Hypothesis 6: laparoscopic colposuspension versus newer 'self‐fixing' sling procedures
Eight studies compared laparoscopic colposuspension with newer 'self‐fixing' sling procedures (Adile 2001; Foote 2004; Mirosh 2005; Maher 2004; Paraiso 2004; Persson 2002; Ustun 2003; Valpas 2004). Within these eight studies there were variations in the techniques (TVT or SPARC slings) and type of material (sutures or mesh) used for both of the procedures; for this reason we have displayed subgroups within the comparison graphs. All studies reported data within 18 months and the majority of the studies had follow‐up at one year postoperatively (Foote 2004; Paraiso 2004; Persson 2002; Mirosh 2005; Valpas 2004). The ability to synthesise data was limited in some areas by the variable tests and definitions used for subjective and objective outcomes and for this reason some data were reported in the Additional tables. A published supplementary version of the study by Paraiso reported long‐term follow‐up (four to eight years) (Paraiso 2004).
Women's observations
Subjective cure rates were near equal in all studies that compared laparoscopy using sutures with vaginal slings, but Valpas 2004 used a mesh for the laparoscopic colposuspension and the subjective cure rate favoured the sling procedure. Overall, there was no statistically significant difference in the reported subjective cure rates between laparoscopic colposuspension and vaginal sling procedures within 18 months (Comparison 02.01; RR 0.91, 95% CI 0.80 to 1.02). This appeared to remain the case in the longer‐term follow‐up (four to eight years), as TVT was reported to have similar subjective cure rates as laparoscopic colposuspension (Paraiso 2004) (Comparison 02.10; RR 1.18, 95% CI 0.36 to 3.81).
Quantification of symptoms
Two studies (Foote 2004; Paraiso 2004) reported the mean number of incontinent episodes per week in each group. Although the reporting was incomplete and not statistically significant, both the means were lower in the laparoscopic groups (Comparison 02.02; WMD ‐1.40, 95% CI ‐3.30 to 0.50). Valpas 2004 reported pad test data but not in a way suitable for analysis.
Clinicians' measures
All but one of the studies (Mirosh 2005) assessed objective cure rates within 18 months . The definitions of objective cure varied across the studies as they used negative pad tests (Persson 2002; Valpas 2004), negative clinical stress tests (Valpas 2004), negative urodynamic tests (Maher 2004; Paraiso 2004; Ustun 2003) or no leaks recorded on urinary diary (Foote 2004). The objective cure rate was higher for TVT when compared with laparoscopy using sutures (Comparison 02.03.01; RR 0.92, 95% CI 0.85 to 0.99) and also in Valpas 2004, when TVT was compared with laparoscopy using mesh (Comparison 02.03.02; RR 0.66, 95% CI 0.51 to 0.86). The report from Adile 2001 was unclear about the exact timing of the follow‐up, which could have been any time from six to 36 months. When we re‐ran the analysis excluding this trial the effect estimate was unchanged but was no longer statistically significant (RR 0.92, 95% CI 0.83 to 1.02). In Foote 2004 there was no statistically significant difference between the groups for the objective cure rates within 18 months when laparoscopy (67% objective cure) was compared with SPARC sling (65% objective cure) (Comparison 02.03.03; RR 1.05, 95% CI 0.72 to 1.52). Overall, laparoscopic colposuspension procedures had statistically significantly lower objective cure rates (RR 0.88, 95% CI 0.81 to 0.95).
Only two studies reported urge symptoms (Foote 2004; Valpas 2004) and showed that urgency was more common in the sling groups (Table 10). There was no difference in de novo detrusor overactivity within 18 months (Comparison 02.04; RR 0.80, 95% CI 0.34 to 1.88); three of the studies reporting this outcome (Adile 2001; Paraiso 2004; Ustun 2003) had higher rates of overactivity in the vaginal sling group but tMaher 2004 reported the opposite, with overactivity being more common in the laparoscopic colposuspension group. Similarly, there was no apparent overall difference in voiding dysfunction (Comparison 02.05; RR 1.06, 95% CI 0.47 to 2.41). However, the numbers in each of the groups with these outcomes were small.
| Study | Laparoscopic colpo | Vaginal sling |
| Foote 2004 | 3/40 | 6/40 (SPARC) |
| Valpas 2004 | 0/51 | 2/70 (TVT) |
Urodynamic investigations were used to assess cure objectively in three studies (Maher 2004; Paraiso 2004; Ustun 2003) and showed no difference between the two procedures (Comparison 02.06; RR 0.91, 95% CI 0.80 to 1.03).
Quality of life
Five studies assessed quality of life (QoL) as an outcome (Foote 2004; Maher 2004; Mirosh 2005; Paraiso 2004; Valpas 2004). These studies each used a varying number of validated questionnaires (UDI, IIQ, SF‐36, KHQ and visual analogue scores) and reported the data in different ways so that the results could not be combined. Mirosh 2005 gave no details of the quality‐of‐life instrument used and just reported the QoL scores to be similar in each group at one‐year postsurgery. Foote 2004; Maher 2004; and Paraiso 2004 all reported a significant improvement in QoL measures in both laparoscopic and vaginal sling groups but no differences between the two groups. In one study this improvement was reported to be maintained for the longer‐term follow‐up of four to eight years (Paraiso 2004). Only one study (Valpas 2004) reported a significant difference between the two procedures, in VAS scores and in the majority of the domains in King's Health Questionnaire scores at one year postsurgery; this difference was in favour of the TVT procedure.
Surgical outcome measures
Seven of the studies reported perioperative complications. Only Maher 2004 was lacking this information. However, this study did report that the postoperative complications were similar in each group. There was no difference in the perioperative complication rates between laparoscopic colposuspension and vaginal sling procedures (Comparison 02.07; RR 0.99, 95% CI 0.60 to 1.64).
Only two studies assessed the use of postoperative analgesia and pain relief (Paraiso 2004; Valpas 2004). Valpas 2004 reported that the use and doses of analgesics were lower in the TVT group, whereas Paraiso 2004 reported that the length of time (in hours) that patient‐controlled analgesia was used was equal in both groups.
The length of inpatient stay significantly differed between the two groups although by only one day (Comparison 02.08; MD 1.10, 95% CI 0.79 to 1.41) and favoured the vaginal sling for shortest hospital stay. The time to return to normal activities, where reported, was longer for laparoscopic colposuspension (Table 11).
| Study | Laparoscopic colpo | Vaginal sling |
| Foote 2004 | 3.8 weeks | 2.8 weeks |
| Maher 2004 | 25 days | 17.9 days |
| Valpas 2004 | 24 days | 15 days |
Persson 2002 reported re‐operation rates at one year, with three out of 38 women in the TVT group and one out of 32 woman in the laparoscopic colposuspension group requiring repeat surgery for non‐cure.
Other outcomes
Laparoscopic surgery took significantly longer than the vaginal sling surgery, by an average of 20 minutes (Comparison 02.09; MD 20.31 minutes, 95% CI 16.75 to 23.86). There was significant statistical heterogeneity reflecting the two trials (Mirosh 2005; Ustun 2003) that contributed least weight to the comparison. The duration of indwelling catheterisation was reported to be longer in the laparoscopic colposuspension group in three trials (Maher 2004; Ustun 2003; Valpas 2004) and longer in the TVT group in one trial Paraiso 2004; we have displayed these results in the Additional tables (Table 12).
| Study | Lap colpo | Vaginal sling (TVT) |
| Maher 2004 | 2.7 | 1.4 |
| Unstan 2003 | 3 (range 1‐5) | 1 (range 0‐7) |
| Valpas 2004 | 24.4 hours | 9.2 hours |
| Paraiso 2004 | 4.9 | 5.2 |
Hypothesis 7: laparoscopic colposuspension versus anterior repair
No eligible studies were found.
Hypothesis 8: laparoscopic colposuspension versus periurethral injections
No eligible studies were found.
Hypothesis 9: one method of laparoscopic colposuspension is better than another method
Five studies looked at different methods of laparoscopic colposuspension. One compared one double‐bite suture with two single‐bite sutures each side of the urethra (Persson 2000), three compared mesh and staples with sutures (Ankardal 2005; Ross 1995; Zullo 2001) and the final study compared two different methods of laparoscopic approach (Wallwiener 1995).
i) One versus two sutures
Women's observations
Persson 2000 reported subjective cure rates using the women's description of: cure, improvement and non‐improvement. A significantly greater number of women felt cured (89%) in the two‐suture compared to the one‐suture group (65%) (Comparison 03.01.01; RR 1.37, 95% CI 1.14 to 1.64).
Quantification of symptoms
Cure was quantified by an 'ultrashort' pad test assessment. The number of cured women was again significantly higher in the two‐suture group compared to one‐suture (Comparison 03.03.01; RR 1.42, 95% CI 1.14 to 1.77) with an 83% cure rate for two sutures compared with a 58% cure rate for one suture.
Clinicians' measures
Four of 83 women in the two‐suture group compared with six of 78 in the one‐suture group had postoperative urge symptoms. The equivalent numbers for voiding dysfunction were 3 and 1 women, respectively (Comparison 03.06.01; RR 2.82, 95% CI 0.30 to 26.54). The mean residual urinary volume at follow up was 25 mL in the one‐suture group and 30 mL in the two‐suture group.
Quality of life
No data reported.
Surgical outcome measures
Fourteen women in the one‐suture group and 15 in the two‐suture group had a perioperative surgical complication (Comparison 03.07.01; RR 0.88, 95% CI 0.45 to 1.70). One woman in the one‐suture group had pubic bone osteitis after the operation. The median length of hospital stay was 1.7 days in both groups. One women in the two‐suture group had repeat incontinence surgery.
Other outcomes
The median operating time was 17 minutes longer among those allocated to two sutures. Among women without additional surgery, mean blood loss was similar in the two groups. The mean time of catheterisation was 2.5 days in both groups. One year after surgery one woman in the one‐suture group and four in the two‐suture group were lost to follow‐ up. Of the remainder, four in the one‐suture group and seven in the two‐suture group did not provide a follow‐up pad test. The cure rates given were for the number of women at follow‐ up.
ii) Sutures versus mesh and staples
Women's observations
Ross 1995 did not report subjective methods of cure or improvement. Ankardal 2005 and Zullo 2001 reported significantly better subjective cure rates with sutures than mesh within 18 months (Comparison 03.01.02 and 02.02; RR 1.28, 95% CI 1.11 to 1.47).
Quantification of symptoms
No data quantifying leakage were reported.
Clinicians' measures
All studies reported failure of objective cure as leakage on the clinical stress test. In Zullo 2001 there were more failures in the mesh and staples group; analysis of the trials showed a significantly better objective cure rate in the suture group within 18 months (Comparison 03.03.02; RR 1.20, 95% CI 1.07 to 1.35 ). Objective cure data were also reported at three years' follow‐up in Zullo 2001 and again, tended to favour the method using sutures, but this was not significant (Comparison 03.04; RR 1.50, 95% CI 0.83 to 2.70). Two studies measured de novo detrusor overactivity and found no difference between the groups (Ross 1995; Zullo 2001) (Comparison 03.05; RR 0.72, 95% CI 0.17 to 3.06).
Quality of life
No data reported.
Surgical outcome measures
There were no statistically significant differences between the number of postoperative complications in two studies (Ross 1995; Zullo 2001) however, the meta‐analysis included the study by Ankardal et al (Ankardal 2005) and resulted in significantly fewer complications in the mesh group (Comparison 03.07.02; RR 1.94, 95% CI 1.09 to 3.48). In both Ankardal 2005 and Ross 1995 there were more (four and two, respectively) bladder perforations in the suture group than in the mesh group (one perforation) and in Zullo 2001 there was one in each of the groups. In two studies (Ross 1995; Zullo 2001), length of hospital stay was the same regardless of whether sutures or mesh were used. In Ankardal 2005, the length of hospital stay was less in the mesh group (Table 13). In Ross 1995 one woman in the mesh group and two women in the suture group had repeat incontinence surgery.
| Mean +/‐ SD | Sutures | Mesh |
| Time of Surgery (min) | 84+/‐30 | 74+/‐25 |
| Time with bladder drainage (days) | 6.2+/‐9.2 | 1.9+/‐2.5 |
| Length of stay in hospital (days) | 3.3+/‐2.5 | 2.1+/‐1.3 |
Other outcomes
Not applicable.
iii) Transperitoneal versus extraperitoneal laparoscopic colposuspension
One study compared a transperitoneal approach with an extraperitoneal approach for laparoscopic colposuspension (Wallwiener 1995), using a mixture of sutures or mesh stapler fixation, but it was not clear which method of fixation was used with which operation. The sample size was small (n = 22).
Women's observations
Subjective assessment by the participants was a criterion for evaluating outcome but no further information was given.
Quantification of symptoms
No data reported.
Clinicians' measures
Urodynamic assessment, clinical findings and perineal sonography to control the anatomic correction of bladder neck descent were performed but no data were given. Overall, 18 out of 22 women were subjectively and objectively cured, one was unchanged and one was worse. One woman had postoperative detrusor overactivity.
Quality of life
No data reported.
Surgical outcome measures
Two perioperative complications were reported of which one was a bladder injury.
Other outcomes
Surgery lasted between 45 to 105 minutes. The length of follow‐up varied between two and 12 months.
Discussion
The 22 eligible trials that were identified addressed only three of the pre‐stated hypotheses. In particular, there were no comparisons with non‐surgical management. Nevertheless, the three hypotheses that were addressed were the most obvious areas of clinical uncertainty. These were whether colposuspension should be performed laparoscopically rather than by open surgery; how laparoscopic colposuspension compares with other, newer minimal access procedures (self‐fixing slings); and what laparoscopic techniques should be used.
To supplement the main systematic review of effects, we sought to identify economic evaluations that have compared laparoscopic colposuspension with any of the other main categories of surgical procedures listed in the background section. A supplementary search in NHS EED, MEDLINE and Embase identified two such economic evaluations.
Hypothesis 3: laparoscopic colposuspension versus open colposuspension
Eight of the 10 trials comparing laparoscopic with open colposuspension were of good quality (Ankardal 2004; Ankardal 2005; Burton 1997; Carey 2000; Cheon 2003; Fatthy 2001; Kitchener 2006; Summitt 2000). Burton 1997 had the potentially confounding factors of using absorbable sutures and the surgeon only having carried out a relatively small number of laparoscopic colposuspensions (fewer than 20) before commencing the trial. Cheon 2003 also involved surgeons with relatively little experience, only 15 previous laparoscopic colposuspensions. These factors may have influenced the results, in particular since there is believed to be a definite, albeit relatively steep, learning curve associated with laparoscopic colposuspension. The first of Ankardal's studies in 2004 (Ankardal 2004) differed from the other studies, as they used a polypropylene mesh and staples for the laparoscopic procedure rather than sutures; this may explain why this study has cure rates in favour of the open technique. However, if this study is removed from the analysis, the overall results do not change. Su 1997 had methodological problems with corrupted randomisation, confounding factors of performance of additional surgery in some participants, and the use of a different number of sutures for laparoscopic colposuspension (one suture) and open colposuspension (three sutures). As shown elsewhere in the review, the number of sutures used appears to have a significant influence on the cure rate, with more sutures resulting in a significantly higher success rate (Persson 2000) (see below). For these reasons, we have repeated some meta‐analyses in this review after excluding data from Su 1997. This, again, did not have an effect on the meta‐analysis results.
Three trials currently have data beyond 18 months' follow‐up (Burton 1997; Morris 2001; Kitchener 2006). Burton 1997 suggested poorer long‐term results after laparoscopic surgery. This finding should be interpreted cautiously, however, as there are concerns that the surgeon's laparoscopic performance may have been suboptimal, as he had performed few laparoscopic colposuspensions when the trial started. The Morris 2001 results disagree and suggest poorer long‐term cure rates for open colposuspension. However this trial has a very limited study report, in abstract form, and only reports five to seven years' follow up, with no earlier data. The data from Kitchener 2006 is more reliable, as it was a larger trial with multiple operators and the follow‐up at two years showed no significant differences between the cure rates of the two procedures.
The data that are available up to a maximum of 18 months showed some inconsistencies. Outcomes assessed by the women participating (arguably the most important outcome) appear equally good in the two groups, whereas objectively and urodynamically‐assessed cure rates were lower in the laparoscopic group.
No significant differences between the two groups were observed for postoperative urgency, voiding dysfunction or de novo detrusor overactivity. Other short‐term outcomes have shown lower perioperative complication rates, longer operation times, less postoperative pain, shorter hospital stay and a quicker overall recovery with laparoscopic colposuspension. All of these are consistent with reviews of other laparoscopic operations (McCormack 2001).
It is noteworthy that this rigorous systematic review has shown significant differences, in comparison with a published conventional review in the British Medical Journal. Thakar and Stanton (Thakar 2000) quoted laparoscopic colposuspension to be 20% less successful than open colposuspension, which is not consistent with our findings. They quoted two references of which one was to a trial of poor methodological quality.
Economic evidence
A cost‐effectiveness analysis by Dumville 2006, alongside a randomised controlled trial, compared open colposuspension with laparoscopic colposuspension in women with SUI in the UK. The cost analysis was conducted from the perspective of the UK NHS and healthcare resource use (2002 to 2003, GBP) relating to the surgery, associated hospital stay and first six months after hospital discharge were collected prospectively for each participant. The data required for the calculation of quality adjusted life years (QALYs) were collected prospectively using the EQ‐5D‐3L questionnaire at baseline, 6, 12 and 24 months.
The economic evaluation was for both a six‐month and up‐to‐24‐month time horizon. Healthcare resource use over six‐month follow‐up resulted in costs of GBP 1805 for the laparoscopic arm and GBP 1433 for the open arm (differential mean cost GBP 372, 95% credibility interval (CrI) 274 to 471). At six months, QALYs were slightly higher on average in the laparoscopic arm relative to the open arm (0.005, 95% CrI –0.012 to 0.023). On average the incremental cost per extra QALY provided by the laparoscopic approach was GBP 74,400 at six months. At 24 months, the laparoscopic arm again had higher mean QALYs compared to the open‐surgery group (0.04, 95% CrI ‐0.009 to 0.086). If the laparoscopic colposuspension did not incur any significant additional costs after six months compared with open colposuspension, then the incremental cost per extra QALY reduced to GBP 9300 at 24 months. The probability that laparoscopic colposuspension is cost effective was 86% when the decision maker was willing to pay up to GBP 30,000 for an additional QALY. Dumville 2006 concluded that laparoscopic colposuspension is not cost effective when compared with open colposuspension during the first six months but the additional QALYs might be judged to be worth the additional cost after 24‐months' follow‐up.
Hypothesis 6: laparoscopic colposuspension versus newer 'self‐fixing' sling procedures
The eight trials that we included in this review comparing laparoscopic colposuspension with sling procedures involved only newer, 'self‐fixing' slings. They were generally of good quality, with the exception of two (Adile 2001; Mirosh 2005). Adile 2001 gave no details of the randomisation process and the exact timing of the follow‐up was unclear, being anywhere between 6 to 36 months. Mirosh 2005 had no description of the randomisation or allocation process and had only a very small number of participants in each group (14 women in the laparoscopic group and 16 women in the TVT group). For these reasons, the data from these two trials must be interpreted with caution. Maher 2004 only recruited women with recurrent urinary stress incontinence and this different inclusion criterion may have influenced the results from this trial.
As the 'self‐fixing' vaginal slings are relatively new procedures, long‐term follow‐up studies are lacking. Only one study reported long‐term follow‐up of four to eight years (Paraiso 2004). Overall, this review has shown that the subjective cure rates are equally good for both of these minimal access techniques, both in the short and long term. However, the objective cure rate within 18 months appears to favour the vaginal sling technique, especially TVT, when compared to laparoscopy using a mesh or sutures. However, three studies assessed the urodynamic cure rate (Maher 2004; Paraiso 2004; Ustun 2003), all comparing laparoscopic sutures with TVT, and there was no significant difference in the outcome rates. There were no significant differences between the two procedures for the rates of postoperative voiding dysfunction, de novo detrusor overactivity and perioperative complications.
Laparoscopy has a longer operation time, longer hospital stay and slower return to normal activities. The re‐operation rates at one year were reported in one trial (Persson 2002) and TVT had a higher re‐operation rate. Longer‐term data are now emerging with regard to laparoscopic colposuspension and there is a need for more long‐term follow‐up of 'self‐fixing' sling procedures. One study in this review (Paraiso 2004) reported long‐term follow‐up and concluded that TVT has similar long‐term satisfaction rates to laparoscopic colposuspension. However, the study also reported that a substantial proportion of subjects had some degree of urinary incontinence at four to eight years following surgery, but the majority was not bothersome. Like laparoscopic surgery, the newer vaginal slings have all the advantages of minimal access techniques, and this review has shown that they are as, if not more, clinically effective than laparoscopic colposuspension. As the skill needed to perform laparoscopic colposuspension is greater than that needed for the 'self‐fixing' slings, many surgeons training in the techniques of surgery for urinary stress incontinence are more likely to become efficient in the sling procedures. While there is still a trend in favour of the sling procedures, more studies reporting longer‐term data are needed as unanticipated and even anticipated complications may arise from these newer procedures.
Economic evidence
Valpas 2006 reported a cost‐effectiveness analysis alongside a randomised controlled trial that compared laparoscopic mesh colposuspension with TVT as a primary surgical treatment in women with SUI over a follow‐up period of one year in Finland. The primary outcome was negative stress test and 48‐hour pad test (<8g/48h) and secondary outcomes were health‐related quality of life. These were collected prospectively during the main period of hospitalisation, six weeks after surgery, and after one year of follow‐up. Cost items included costs of the treatments in each treatment arm, other hospital costs, and productivity costs.
Primary and secondary outcomes in the TVT group were significantly better than laparoscopic mesh colposuspension as measured by the negative stress test (60 versus 29; 95% CI for change between the groups 12.7 to 43.9), visual analogue scale (VAS) (0 = no inconvenience at all; 10 = maximal inconvenience) (0.8 versus 2.4; 95% CI for change between the groups 0.65 to 2.07), and Urinary Incontinence Severity Score (UISS) (1.1 versus 2.8; 95% CI for change between the groups 0.27 to 2.94). The between‐group difference in the 48‐hour pad test did not reach statistical significance (3 versus 12.4; 95% CI for change between the groups ‐2.8 to 30.4; P = 0.105). The total costs per participant at one year of follow‐up for laparoscopic mesh colposuspension was EUR 3262 while TVT cost EUR 2081 (2000 EUR). Valpas 2006 concluded that when the VAS or UISS are used as the outcome measure, TVT is more cost‐effective than laparoscopic mesh colposuspension over a follow‐up period of one year (P < 0.0001).
Another cost‐effectiveness analysis (Cody 2003) compared TVT with Burch colposuspension, laparoscopic colposuspension, traditional suburethral retropubic sling procedure and injectables. The study utilised clinical data from a systematic review of RCTs conducted up to mid‐2002 and the results modelled for a time horizon of up to 10 years. The analysis of costs and resources used was based on the UK payer’s perspective (UK National Health Service). Based on clinical evidence, this study assumed that traditional sling effectiveness was equivalent to open colposuspension, that the effectiveness of laparoscopic colposuspension was equivalent to or possibly worse than open colposuspension, and that use of injectables was the least effective procedure. This study reported the cost (2001 GBP) to be GBP 1058 per woman with an average of 2.9 days hospital stay for TVT, GBP 1317 per woman and average hospital stay of 4.6 days for laparoscopic colposuspension, GBP 1301 per woman and average hospital stay of 7.1 days for open colposuspension, GBP 1340 per woman and average hospital stay of 7.2 days for traditional sling and GBP 1305 per woman with an average hospital stay of 2 days for injectables. The study concluded that TVT was more likely to be considered cost‐effective compared with the other surgical procedures based on the assumptions that traditional slings have the same effectiveness as open colposuspension and are also more costly; that laparoscopic colposuspension has the same or lower effectiveness as open colposuspension and similar costs; and that injectable agents are less effective than TVT but of greater cost.
We did not subject these two identified economic evaluations to critical appraisal and we do not attempt to draw any firm or general conclusions regarding the relative costs or efficiency of laparoscopic colposuspension in treatment of SUI. However, the economic evidence available suggests that laparoscopic colposuspension is not cost‐effective when compared with TVT in the treatment of SUI in women.
Hypothesis 9: different methods of laparoscopic colposuspension
i) One suture versus two sutures
A single study (Persson 2000) comparing different numbers of paravaginal sutures found a significantly higher objective one year cure rate (dry on 'ultrashort' pad test) for women randomised to two sutures compared with one suture, with a 83% cure rate for two sutures and a 58% cure rate for one suture. These findings were supported by a significantly higher subjective impression of cure in the two‐sutures group compared to the one‐suture group, with subjective cure rates of 89% and 65% respectively.
This study was of good quality, with adequate group allocation. However, the trial was stopped early as it became apparent at an interim analysis that two sutures were superior to one suture. Such data‐dependent stopping may lead to an exaggeration of the true effect size. Another concern is the potential confounding from one third of the women having additional surgery.
ii) Sutures versus mesh and staples
Sutures appeared to be more effective than mesh and staples. Two of the studies (Ross 1995; Zullo 2001) were small and the confidence intervals were wide. Cure was not clearly defined and it was unclear how the women were randomised in Ross 1995; we assumed that they used a negative cough stress test as a proof of cure, which was found to be similar for both groups. They did not report subjective measures. The trialists set out to prove equivalence between the two techniques as they found the technique of mesh and staples much easier to apply than sutures. However, the methodological problems mean that this question was not addressed reliably by this study. Zullo 2001 was of a similar size with clearly stated definitions of both objective and subjective cure, and adequate randomisation and concealment of allocation. This study reported outcomes in favour of sutures. The analysis with the addition of Ankardal 2005, which is a larger study and of good quality, showed significant subjective and objective outcomes favouring the use of sutures over mesh. However, Ankardal 2005 had fewer perioperative complications (mainly bladder perforations) with the mesh and staples technique, making the overall analysis results favour the use of mesh for perioperative complications.
iii) Transperitoneal versus extraperitoneal laparoscopic colposuspension
Wallwiener 1995, comparing transperitoneal with extraperitoneal laparoscopic colposuspension, was of poor quality and had very small numbers of participants. It is unclear how randomisation was achieved and it was difficult to interpret which suture material was used for which operation. Their cure rate was expressed as an overall 'subjective and objective' cure. Women's evaluation of cure was part of the outcome measures but it is unclear how it contributed. The cure rate was not given separately for the two groups. All the above make valid conclusions from this study impossible.
Economic evidence
No relevant economic evidence was identified for the brief economic commentary.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 1 Subjective cure.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 2 Incontinent episodes over 24 hours within 18 months.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 3 Objective cure on clinical testing.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 4 De novo detrusor overactivity (urodynamic diagnosis).
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 5 Voiding dysfunction within 18 months.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 6 Urodynamic cure within 18 months.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 7 Perioperative complications.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 8 Bladder perforations.
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 9 Length of inpatient stay (days).
Comparison 1 Laparoscopic urethropexy vs open colposuspension, Outcome 10 Operation time (minutes).
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 1 Subjective cure within 18 months.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 2 Incontinent episodes per week within 18 months.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 3 Objective cure on clinical testing within 18 months.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 4 De novo detrusor overactivity within 18 months.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 5 Voiding dysfunction within 18 months.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 6 Urodynamic cure within 18 months.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 7 Perioperative complications.
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 8 Length of inpatient stay (days).
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 9 Operation time (minutes).
Comparison 2 Laparoscopc urethropexy vs 'self‐fixing' sling procedures, Outcome 10 Subjective cure 4‐8 year follow‐up.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 1 Subjective cure within 18 months.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 2 Subjective cure between 18 months and 5 years.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 3 Objective cure on clinical testing within 18 months.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 4 Objective cure on clinical testing between 18 months and 5 years.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 5 De novo detrusor overactivity within 18 months.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 6 Voiding dysfunction within 18 months.
Comparison 3 Different methods of laparoscopic colposuspension, Outcome 7 Perioperative complications.
| 6/12 | 1 year | 3 years | 5 years | |
| open | 10.0 | 9.8 | 9.6 | 9.4 |
| laparoscopic | 7.8 | 6.7 | 4.8 | 4.4 |
| pre‐op | 6/12 | 1 year | 3 years | 5 years | |
| open | 12 | 0 | 2 | 2 | 3 |
| laparoscopic | 13 | 2 | 6 | 8 | 9 |
| pre‐op | 6/12 | 1 year | 3 years | 5 years | |
| open | 22 | 1 | 2 | 3 | 5 |
| laparoscopic | 24 | 5 | 12 | 14 | 15 |
| Study | pre‐op | 6/12 | 1 year | 18/12 | 3 years | 5 years |
| Burton 1997 | ||||||
| open | 24 | 33 | 30 | 26 | 32 | |
| laparoscopic | 21 | 42 | 40 | 42 | 49 | |
| Fatthy 2001 | ||||||
| open | 23.9 | 30.95 | ||||
| laparoscopic | 26.65 | 32.85 |
| Domain of QoL | laparoscopic colpo | Open colpo | p value |
| Physical activity | 63 (27‐81) | 74 (52‐84) | <0.05 |
| Working ability | 25 (4‐54) | 31 (10‐60) | NS |
| Social life | 37 (15‐66) | 38 (10‐65) | NS |
| Sexual life | 13 (1‐41) | 9 (0‐37) | NS |
| median (range) | median (range) |
| Score at 24 months | Laparoscopic colpoe | Open colpo |
| EQ‐5D values | 0.844 (0.259) | 0.825 (0.270) |
| SF‐36 Physical sub‐scale | 79.32 (27.59) | 77.60 (27.74) |
| SF‐36 Mental sub‐scale | 69.51 (21.21) | 69.38 (22.65) |
| Carey (mean) | Fatthy (mean) | Summitt (mean) | Cheon (mean) | |
| open | 21.8 | 31.5 | 37.3 | 29.3 |
| laparoscopic | 17.3 | 8.5 | 35.5 | 22.2 |
| Burton(median+range) | Carey (mean) | Fatthy (mean+SD) | Su (mean +SD) | Summitt (mean) | Ankardal (mean) | Cheon (mean) | |
| open | 261 (120 ‐ 410) | 171 | 240.5 (35.5) | 134.3 (102.0) | 131.9 | 105 | 327 |
| laparoscopic | 190 (180 ‐ 290) | 125 | 42.75 (7.2) | 59.3 (42.1) | 112 | 35 | 125 |
| Burton (mean) | Carey (mean) | Su (mean + SD) | Summitt (mean) | Ankardal (mean + SD) | Cheon (mean +SD) | |
| open | 1.7 | 4.9 | 6.8 (2.3) | 2.7 | 4.9 (8.3) | 3.7 (2.8) |
| laparoscopic | 1.33 | 4.4 | 3.9 (1.9) | 2.7 | 1.9 (2.4) | 3.0 (1.6) |
| Study | Laparoscopic colpo | Vaginal sling |
| Foote 2004 | 3/40 | 6/40 (SPARC) |
| Valpas 2004 | 0/51 | 2/70 (TVT) |
| Study | Laparoscopic colpo | Vaginal sling |
| Foote 2004 | 3.8 weeks | 2.8 weeks |
| Maher 2004 | 25 days | 17.9 days |
| Valpas 2004 | 24 days | 15 days |
| Study | Lap colpo | Vaginal sling (TVT) |
| Maher 2004 | 2.7 | 1.4 |
| Unstan 2003 | 3 (range 1‐5) | 1 (range 0‐7) |
| Valpas 2004 | 24.4 hours | 9.2 hours |
| Paraiso 2004 | 4.9 | 5.2 |
| Mean +/‐ SD | Sutures | Mesh |
| Time of Surgery (min) | 84+/‐30 | 74+/‐25 |
| Time with bladder drainage (days) | 6.2+/‐9.2 | 1.9+/‐2.5 |
| Length of stay in hospital (days) | 3.3+/‐2.5 | 2.1+/‐1.3 |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Subjective cure Show forest plot | 8 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Within 18 months | 7 | 1025 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.95 [0.90, 1.00] |
| 1.2 Between 18 months and 5 years | 1 | 263 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.00 [0.81, 1.25] |
| 1.3 After 5 years | 1 | 64 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.53 [1.00, 2.35] |
| 2 Incontinent episodes over 24 hours within 18 months Show forest plot | 2 | 133 | Mean Difference (IV, Fixed, 95% CI) | ‐0.12 [‐0.92, 0.68] |
| 3 Objective cure on clinical testing Show forest plot | 10 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 3.1 Within 18 months | 9 | 1087 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.86, 0.96] |
| 3.2 Between 18 months and 5 years | 2 | 290 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.88, 1.16] |
| 3.3 After 5 years | 2 | 107 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.94 [0.69, 1.27] |
| 4 De novo detrusor overactivity (urodynamic diagnosis) Show forest plot | 6 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 4.1 Within 18 months | 5 | 512 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.29 [0.72, 2.30] |
| 4.2 Between 18 months and 5 years | 1 | 50 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.0 [0.07, 15.12] |
| 4.3 After 5 years | 2 | 107 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.22 [0.40, 3.75] |
| 5 Voiding dysfunction within 18 months Show forest plot | 4 | 439 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.82 [0.50, 1.35] |
| 6 Urodynamic cure within 18 months Show forest plot | 6 | 564 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.85, 0.99] |
| 7 Perioperative complications Show forest plot | 9 | 1182 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.74 [0.58, 0.96] |
| 8 Bladder perforations Show forest plot | 7 | 1028 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.88 [0.93, 3.83] |
| 9 Length of inpatient stay (days) Show forest plot | 6 | mean difference (Fixed, 95% CI) | 0.86 [0.65, 1.06] | |
| 10 Operation time (minutes) Show forest plot | 7 | Mean difference (Fixed, 95% CI) | 14.06 [11.39, 16.73] | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Subjective cure within 18 months Show forest plot | 5 | 377 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.80, 1.02] |
| 1.1 Lap (sutures) vs TVT | 3 | 196 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.86, 1.19] |
| 1.2 Lap (mesh) vs TVT | 1 | 121 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.71 [0.55, 0.91] |
| 1.3 Lap vs SPARC sling | 1 | 60 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.99 [0.75, 1.29] |
| 2 Incontinent episodes per week within 18 months Show forest plot | 2 | 143 | Mean Difference (IV, Fixed, 95% CI) | ‐1.4 [‐3.30, 0.50] |
| 2.1 Lap (sutures) vs TVT | 1 | 63 | Mean Difference (IV, Fixed, 95% CI) | ‐1.4 [‐3.30, 0.50] |
| 2.2 Lap vs SPARC sling | 1 | 80 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] |
| 3 Objective cure on clinical testing within 18 months Show forest plot | 7 | 575 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.88 [0.81, 0.95] |
| 3.1 Lap (sutures) vs TVT | 5 | 394 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.92 [0.85, 0.99] |
| 3.2 Lap (mesh) vs TVT | 1 | 121 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.66 [0.51, 0.86] |
| 3.3 Lap vs SPARC sling | 1 | 60 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.05 [0.72, 1.52] |
| 4 De novo detrusor overactivity within 18 months Show forest plot | 4 | 326 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.80 [0.34, 1.88] |
| 4.1 Lap (sutures) vs TVT | 4 | 326 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.80 [0.34, 1.88] |
| 5 Voiding dysfunction within 18 months Show forest plot | 5 | 412 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.06 [0.47, 2.41] |
| 5.1 Lap (sutures) vs TVT | 3 | 211 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.15 [0.44, 2.99] |
| 5.2 Lap (mesh) vs TVT | 1 | 121 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.37 [0.20, 9.42] |
| 5.3 Lap vs SPARC sling | 1 | 80 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.33 [0.01, 7.95] |
| 6 Urodynamic cure within 18 months Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 6.1 Lap (sutures) vs TVT | 3 | 189 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.80, 1.03] |
| 7 Perioperative complications Show forest plot | 7 | 514 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.99 [0.60, 1.64] |
| 7.1 Lap (sutures) vs TVT | 5 | 350 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.88 [0.48, 1.60] |
| 7.2 Lap (mesh) vs TVT | 1 | 121 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.92 [0.65, 5.71] |
| 7.3 Lap vs SPARC sling | 1 | 43 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.38 [0.04, 3.40] |
| 8 Length of inpatient stay (days) Show forest plot | 5 | Mean difference (Fixed, 95% CI) | 1.10 [0.79, 1.41] | |
| 8.1 Lap (sutures) vs TVT | 3 | Mean difference (Fixed, 95% CI) | 0.95 [0.57, 1.32] | |
| 8.2 Lap (mesh) vs TVT | 1 | Mean difference (Fixed, 95% CI) | 1.1 [0.46, 1.74] | |
| 8.3 Lap vs SPARC sling | 1 | Mean difference (Fixed, 95% CI) | 2.5 [1.36, 3.64] | |
| 9 Operation time (minutes) Show forest plot | 6 | Mean difference (Fixed, 95% CI) | 20.31 [16.75, 23.86] | |
| 9.1 Lap (sutures) vs TVT | 4 | Mean difference (Fixed, 95% CI) | 21.60 [17.29, 25.92] | |
| 9.2 Lap (mesh) vs TVT | 1 | Mean difference (Fixed, 95% CI) | 18.0 [7.54, 28.46] | |
| 9.3 Lap vs SPARC sling | 1 | Mean difference (Fixed, 95% CI) | 17.3 [9.43, 25.17] | |
| 10 Subjective cure 4‐8 year follow‐up Show forest plot | 1 | 53 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.18 [0.36, 3.81] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Subjective cure within 18 months Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Two sutures vs one suture | 1 | 158 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.37 [1.14, 1.64] |
| 1.2 Sutures vs mesh | 2 | 180 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.28 [1.11, 1.47] |
| 2 Subjective cure between 18 months and 5 years Show forest plot | 1 | 56 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.16 [0.76, 1.75] |
| 3 Objective cure on clinical testing within 18 months Show forest plot | 4 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 3.1 Two sutures vs one suture | 1 | 149 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.42 [1.14, 1.77] |
| 3.2 Sutures vs mesh | 3 | 243 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.20 [1.07, 1.35] |
| 4 Objective cure on clinical testing between 18 months and 5 years Show forest plot | 1 | 52 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.5 [0.83, 2.70] |
| 5 De novo detrusor overactivity within 18 months Show forest plot | 2 | 122 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.72 [0.17, 3.06] |
| 6 Voiding dysfunction within 18 months Show forest plot | 1 | 161 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.82 [0.30, 26.53] |
| 6.1 Two sutures vs one suture | 1 | 161 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.82 [0.30, 26.53] |
| 7 Perioperative complications Show forest plot | 4 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 7.1 Two sutures vs one suture | 1 | 161 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.88 [0.45, 1.70] |
| 7.2 Sutures vs mesh | 3 | 260 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.94 [1.09, 3.48] |
