SEARCH

SEARCH BY CITATION

Keywords:

  • Colposuspension;
  • laparoscopic;
  • open;
  • randomised

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

Objective  To compare the effectiveness and cost effectiveness of open and laparoscopic colposuspension in the treatment of stress urinary incontinence.

Design  A randomised controlled trial. Women were randomised between March 1999 and February 2002 and were seen for assessment at 6, 12 and 24 months, postoperatively.

Setting  Women were recruited from six gynaecology units in the UK.

Population  Women with proven stress urinary incontinence requiring surgery.

Methods  Open abdominal retropubic colposuspension or laparoscopic colposuspension.

Main outcome measures  Primary outcomes were subjective (satisfaction with outcome) and objective (negative 1-hour pad test). Secondary outcomes were operative and postoperative morbidity and quality of life. The study was powered to demonstrate noninferiority, i.e. that the absolute cure rate of laparoscopic colposuspension was no more than 15% below that of open colposuspension.

Results  A total of 291 women were randomised, with 24-month data on subjective and objective outcomes in 88 and 82.5%, respectively. The intention-to-treat analysis indicated no significant difference in cure rates between open and laparoscopic surgery. The objective cure rates for open and laparoscopic were 70.1 and 79.7%, respectively. Subjective cure rates by satisfaction were lower than objective cure; 54.6 and 54.9%, respectively, and there was considerable nonconcordance both ways.

Conclusions  Laparoscopic colposuspension is not inferior to open colposuspension in terms of curing stress urinary incontinence.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

Female urinary incontinence is a common clinical problem, which may impact negatively on quality of life. The prevalence of urinary stress incontinence in European women has been reported at around 20%, with 5% undergoing surgery.1 Around half of these cases are due to a faulty urethral sphincter mechanism, i.e. urodynamic stress incontinence.

The retropubic approach to surgery for stress urinary incontinence described by Burch et al.2 set the operative standard during the latter part of the 20th century. A systematic review of colposuspension published in 2005 indicated an objective cure rate of 78.4%,3 but duration of follow up was often short, ranging from 3 months to 15 years. The surgical morbidity associated with an abdominal incision stimulated the development of less-invasive approaches to colposuspension such as the needle suspension described by Stamey.4 Laparoscopic colposuspension was first described by Vancaillie and Schuessler5 in 1991 and has been followed by more than 50 articles in the literature. Many of these articles describe case series with small numbers and short follow up, making the value of laparoscopic colposuspension difficult to compare with the open procedure. A recent systematic review of laparoscopic colposuspension6 included five randomised controlled trials comparing laparoscopic with open colposuspension although three of these have only been published as abstracts. A Cochrane review in 20007 concluded that the long-term performance of laparoscopic colposuspension is uncertain, a view endorsed in a more recent Cochrane review.8 None of those trials had a sufficient number of cases or adequate length of follow up to determine with confidence whether or not laparoscopic colposuspension is inferior to the open procedure. The COLPO (COlposuspension; is Laparoscopic Preferable to Open?) Trial was set up to determine both the effectiveness and cost effectiveness of laparoscopic colposuspension compared with the open procedure at 24-month follow up. The results of the cost-effectiveness analysis are reported in the companion article.9

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

Patients and procedures

Women were recruited from six gynaecology units in the UK (Manchester, Glasgow, Basingstoke, Wolverhampton, London and Leeds) between March 1999 and February 2002. The units were selected on the basis of established experience with laparoscopic and open colposuspension. Eligible women were those in whom a colposuspension had been chosen to treat urodynamic stress incontinence. Women who had undergone previous anterior colporrhaphy were included. Exclusions comprised detrusor overactivity, previous retropubic surgery, grossly obese women considered unsuitable for any surgery and those unfit for general anaesthetic. Eligible women were offered entry into the trial and randomisation was undertaken following signed informed consent. Randomisation was performed by telephone at the University of York using random permuted block sizes 2 and 4 with a dispersed block throughout the string, stratified by centre, age > 50 years and previous bladder neck surgery.

Ethical approval and informed consent

The study was approved by the North West Multi-Centre Research Ethics Committee.

Trial protocol

A 1-hour pad test was carried out and body mass index was recorded at baseline. Urodynamic assessment included measures of flow, residual urine volume, dual channel subtracted cystometry and urethral closure pressure. A number of validated questionnaires were used to measure symptom severity, specific and generic health-related quality of life to allow cost–utility analysis. These included the UK version of Short Form-36 (SF-36),10 the Bristol Female Lower Urinary Tract Symptom questionnaire,11 Symptom Severity and Symptom Impact Index12 and the EQ-5D,13 which was used for the cost-effectiveness analysis.

There was no attempt to blind the surgical procedure as this was considered neither practical nor pragmatic. Standard operative procedure included antibiotic prophylaxis, skin preparation, suprapubic catheterisation and patient-controlled analgesic postoperatively. In all cases, a minimum of two sutures (Ethibond; Ethicon Inc., Somerville, NJ, USA) were placed on each side. No other surgical procedures were concomitantly performed. Following surgery, questionnaires were completed at 6 weeks (SF-36 and Urinary Symptom questionnaires only), 6, 12 and 24 months. A pad test was performed at each visit, a positive result being defined at >1.0 g/hour. Urodynamic assessment was routine at 6 months but thereafter was only performed if the pad test was positive.

Outcome measures

The primary outcome measures for cure were objective and subjective at 24 months postoperatively. Objective cure was a negative 1-hour pad test, and a subjective cure was a response of ‘perfectly happy/pleased’ to question 33 of the Bristol questionnaire (If you had to spend the rest of your life with your urinary symptoms as they are now, how would you feel?). The secondary outcomes were levels of operative morbidity, time to return to work and health economic costs to the NHS and patient. Quality of life was also measured using the physical and mental subscales of SF-36.

Statistical analysis

We wished to be able to demonstrate that laparoscopic colposuspension was noninferior to open surgery. We assumed an objective cure rate of 80% based on reported data for the open procedure and determined that whatever quality-of-life benefits and health economic benefits a laparoscopic colposuspension can offer, a cure rate of more than 65% would be acceptable.

A sample size of 145 in each arm and a two-group, large sample, normal approximation test of proportions with a one-sided 0.05 significance level would provide 80% power to reject the null hypothesis that the two procedures are not equivalent as defined (nQuery Advisor version 3; Statistical Solutions Limited, Saugus, MA, USA).

In the analysis of the trial outcome data, the main approach was intention to treat (ITT) using all available outcome data in Stata 8 (StataCorp, College Station, TX, USA). This involved fitting a random effects model to repeated measures using maximum likelihood (xtreg or xtlogit commands). Treatment differences were assumed to be constant over time (based on a preliminary look at the data), and the aim was to estimate the treatment effect common to all three follow-up times (i.e. 6, 12 and 24 months). As well as randomised treatment group and time, the covariates included treatment centre, previous bladder neck surgery (yes/no), age group (two levels: less or more than 50 years) and baseline Symptom Severity Index total score. Although we are primarily interested in noninferiority of the laparoscopic (closed) procedure, two-sided 95% confidence intervals are presented. The lower limit for this interval is equivalent to the lower limit of a one-sided 97.5% confidence interval.

When there are treatment switches (i.e. nonadherence to the randomisation), the ITT effect will be attenuated. That is, the treatment effect will be biased towards equivalence. We wish to demonstrate equivalence in terms of efficacy, not effectiveness, and this involves the estimation of the effects of treatment received. But because the effects of treatment received are subject to hidden selection effects (confounding), a naïve as treated analysis is likely to be flawed (this also applied to naïve per protocol analysis)14 Here, in order to evaluate the sensitivity of the findings to treatment switches, we used a two-stage instrumental variable (IV) model to estimate effects of treatment received, allowing for hidden confounding/selection effects. We refer to the resulting IV estimate as an adjusted treatment received (ATR) estimate. For the present analyses, the few women who received no surgery were dropped from the data file. Note that the ATR procedure correctly adjusts for missing outcome data, assuming that these are missing at random (MAR, or ignorable, using the terminology of Little and Rubin).15 Standard errors and confidence intervals for the treatment effects were estimated using the stratified (by Centre) bootstrap. Details of the use of this methodology in the analysis of data from clinical trials are reviewed in Dunn et al.16

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

Between April 1999 and February 2002, 291 women were recruited to the trial and 2-year follow up was completed in June 2004 (Figure 1). Baseline characteristics for each arm are shown in Table 1. Of the 144 women allocated to receive laparoscopic surgery, 11 received open surgery and 2 received no surgery at all. Of the 147 women randomised to receive open surgery, one underwent the laparoscopic procedure and three had no surgery at all. There were objective outcome data at 2 years for 123 (85.4%) in the laparoscopic arm and for 117 (79.6%) in the open arm. There were subjective outcome data at 2 years for 128 (88.9%) in the laparoscopic arm and 128 (87.5%) in the open arm. The objective and subjective cure rates are shown in Table 2 by ITT at 6, 12 and 24 months.

image

Figure 1. Progress of participants through the trial.

Download figure to PowerPoint

Table 1.  Baseline characteristics
 LaparoscopicOpen
Age (years)n= 144n= 147
Median50.550
>507267
Parityn= 105n= 112
023
11113
2–48289
≥5107
Previous bladder neck surgeryn= 144n= 147
Yes1010
Pad testn= 132n= 139
<5 g3435
5–10 g2319
10.1–20 g1224
>20.1 g6361
Table 2.  Subjective and objective cure rates for colposuspension at 24 months
 LaparoscopicOpen
CuredNo. tested%CuredNo. tested%
Objective
<1-g pad test
 6 months10512385.410912984.5
 12 months9011677.69011280.4
 24 months9812379.78211770.1
Subjective
Satisfaction (perfectly happy/pleased)
 6 months62/211226857/2013258.4
 12 months69/1713066.256/2013158.1
 24 months60/1313354.948/2313054.6
Symptoms (never leaks/leaks <1/month)
 6 months56/151235752/613243.9
 12 months52/1912855.453/2513060.0
 24 months39/3212855.448/2012853.1

Analysis of these results indicated no significant differences between laparoscopic and open colposuspension in terms of clinical outcomes. There are consistently higher rates of cure as judged objectively compared with subjective assessment. For the objective cure, the ITT effect was estimated by an odds ratio (OR) of 1.65 (95% CI: 0.74–3.67). The equivalent ATR (see Methods section) estimate was 1.87 (95% CI: 0.76–4.56). The point estimates (1.65 and 1.87) both indicate that the cure rate after laparoscopic (closed) surgery is better than after open surgery. This effect, however, is not statistically significant. The important results are the lower confidence limits (0.74 and 0.76). If we assume a cure rate of 80% after open surgery (as in the power calculations), then this implies that the cure rate after laparoscopic surgery is no less than 74%. This corresponds to a relative risk of 0.74/0.80 = 0.925. Although there were observed differences in treatment effects between centres, these were no more than could be expected by chance (test for treatment by centre interaction: χ2= 6.50 with 4 df). For the subjective cure, the ITT effect was estimated by an OR of 1.25 (95% CI: 0.60–2.64). The equivalent ATR estimate was 1.27 (95% CI: 0.57–2.84). There appears to be no difference in outcomes. Again the lower confidence limits (0.60 and 0.57) are the ones to concentrate on. If we assume that the cure rate in the open surgery group is 60%, then with an OR of 0.57 (the more pessimistic of the two lower limits) that in the laparoscopic group is no less than 46% (i.e. OR = 0.57 = (0.46 × 0.40)/(0.54 × 0.60)).

It is not surprising, in view of the very small amount of nonadherence to the randomised allocation, that the ITT and ATR estimates are so close. Treatment switches are not a major threat to the validity of the inference that the two procedures are equivalent. Both methods of analysis, however, assume that the missing outcome data are ignorable (in the sense defined by Little and Rubin15). In order to check the sensitivity of our findings to assumptions concerning the missing data mechanism, the 2-year outcomes were analysed using a worst-case scenario such that all of the missing 2-year outcomes for the laparoscopic procedures were assumed to be unsuccessful (n= 21) and all that missing for open procedures were successful (n= 30). Even with this highly unlikely scenario, there were no significant differences. For example, analysis of the observed 2-year cure rates (adjusting for the same baseline covariates) provided an OR estimate of 2.05 (95% CI: 1.06–3.95). Under the worst-case scenario, the estimate is 0.70 (95% CI: 0.40–1.21).

For the objective test (pad), there is a slow but statistically significant decrease over time in rates of cure common to both of the arms (ITT analysis, test for linear trend in log odds: P= 0.002) but no comparable drift in the subjective rates of cure (P= 0.982). The data presented in Table 2 suggest that the pattern of changes in recovery rates over time may be different in the two arms of the trial. We had no prior expectation of such differences, and formal tests for a time (coded as a qualitative factor) by treatment interaction produced nonsignificant results (P= 0.096 and P= 0.389 for objective and subjective outcomes, respectively).

Significantly fewer (23%) women who had laparoscopic surgery suffered 24-hour pain levels >6, compared with those after open surgery (40%). After allowance for the standard covariates (in an ITT analysis), the OR was 0.35 (95% CI: 0.20–0.62; P < 0.001).

Inability to pass urine meant that hospital stay was not significantly shorter for the laparoscopic group (5 versus 6 days for open surgery). Time to return to work was only applicable to about half of the sample as many women were retired or housewives.

Table 3 shows the comparative rates of significant morbidity. Bladder injury and bowel injury, although uncommon, were seen more frequently in the laparoscopic group. Wound infection was seen significantly more frequently following open colposuspension, but other infective morbidity was distributed equally.

Table 3.  Perioperative morbidity associated with colposuspension
ComplicationLaparoscopic, n (%)Open, n (%)
Return to theatre00
Bladder injury4 (2.8)1 (0.7)
Bowel injury1 (0.7)0
Haemorrhage >500 ml1 (0.7)2 (1.4)
Wound dehiscence02 (1.4)
Wound infection1 (0.7)11 (7.8)
Urine infection8 (5.7)7 (5.0)
Chest infection5 (3.5)7 (5.0)

Quality of life 6 weeks postoperatively was assessed using the Physical and Mental subscale scores of the SF-36. Descriptive statistics are shown in Table 4. There are no suggestions of any group differences.

Table 4.  Quality of life (SF-36 scores)
 LaparoscopicOpen
nMeanSDnMeanSD
Physical subscale
Baseline14270.8524.0614470.6024.22
6 weeks13275.4022.7013771.9023.06
6 months12779.9725.4913379.2723.61
12 months13081.1825.0313281.3923.61
24 months13279.3227.5912777.6027.74
Mental subscale
Baseline13565.6521.6514266.1119.54
6 weeks12766.5420.5613570.0916.26
6 months12771.7519.0613172.8518.62
12 months12771.9519.6213271.4021.05
24 months13269.5121.2112969.3822.65

Inability to pass urine meant that hospital stay was not significantly shorter for those in the laparoscopic surgery group. Time to return to work was only applicable to about half of the sample as many women were retired or housewives. The mean time to return to work was 10.54 weeks for the women in the open surgery group compared with 9.42 for those in the laparoscopic surgery group (median 9 and 8 weeks, respectively; range 3–26 weeks). This was not statistically significant (P= 0.72; ITT analysis using Cox proportional hazards model).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

This trial has confirmed that laparoscopic colposuspension is not clinically inferior to open colposuspension when both procedures were performed by skilled surgeons.

When calculating the required power for the trial, a 2-year cure rate of 80% was predicted for open colposuspension, and a noninferiority design was chosen because we did not believe there were any reasons why there should be any significant treatment differences in terms of achieving cure. In this trial, the cure rate at 2 years in terms of a negative pad test was 80% for laparoscopic and 70% for open colposuspension. When the outcome was measured subjectively, again there was no evidence that laparoscopic was inferior to open colposuspension. In the context of satisfaction, just more than 50% were either ‘perfectly happy or pleased’ with the outcome in both arms and an identical result was seen based on the symptom ‘never leaks or leaks less than once/month’. These results indicate that laparoscopic access enables exactly as precise a placing of sutures as does open surgery. The reporting of the primary outcomes at 2 years is a feature of this trial and allows assessment of maintenance of cure during the first 2 years of follow up. There is some evidence of ‘loss of cure’ when measured by pad test at 6, 12 and 24 months. When subjective outcomes are measured at the same time points, there is a trend for a loss of patient satisfaction with laparoscopic surgery and a gain in symptom control with open surgery but neither of these was significant.

One of the generally held beliefs of laparoscopic surgery is that it would be associated with longer operating times, less postoperative pain and shorter hospital stay, but for hospital stay, there was only a small advantage for laparoscopic surgery, with a median stay of 5 days compared with 6 for open surgery. It is quite possible that establishing normal voiding postoperatively has a bigger impact on length of stay in hospital than recovery from the surgery. The operating times were fairly similar in terms of ‘knife to skin’ to completion of surgery, there being a median time of 65 and 51 minutes for laparoscopic and open surgery, respectively, and postoperative pain was significantly less in the laparoscopic surgery group. The impact of these differences is considered more fully in the accompanying economic evaluation.

Complications of surgery were low generally, with a higher bladder injury rates in the laparoscopic procedure and a higher wound infection with open surgery as would be expected.

The difference between objective cure and subjective cure seen at each of the time points is consistent whether the latter is measured by overall satisfaction or by symptoms of leaking. This is an important finding not only for determining valid outcome measures for trials but also for making a robust diagnosis, which in incontinence trials is the basis for establishing eligibility. Not only is the subjective cure rate lower than that for objective cure but also there is nonconcordance both ways, in that a proportion of those who are objectively cured do not achieve subjective cure and vice versa. This suggests that there is a need to determine through qualitative research what outcomes matter most to women.

When the results of this trial are compared with other trials of colposuspension,17–21 several findings emerge. Eighty-two percent of all randomised women had 2-year follow up which ensures the reliability of the data. The centres involved in this trial included surgeons with considerable (more than 50 procedures) experience with both laparoscopic and open colposuspension. This means that the trial assesses the procedure and not the surgeon, and while it does not assess today’s generalisability of laparoscopic colposuspension, at least in skilled hands, it is as effective as open colposuspension in curing stress urinary incontinence. It is therefore appropriate for those who have mastered the technique to offer it as an alternative to an open procedure.

During the conduct of this trial, a randomised trial of tension-free vaginal tape (TVT procedure) has been reported from the UK, most recently with 2-year outcomes.22 This showed no significant difference in cure rate between open colposuspension and TVT procedure. When the outcomes are compared with the COLPO trial, objective cure in terms of a negative pad test was similar for those who were followed up at 2 years, although a higher proportion of COLPO women were evaluated at 2 years.

Since 1999 when the COLPO trial began, widespread adoption of the TVT suburethral procedure has occurred particularly in Europe. The benefits include reduced morbidity and hospital stay and faster recovery. The initial trial of TVT has been followed by other trials23,24 that confirmed its efficacy when compared with colposuspension. Furthermore, longer term data from follow-up studies of between 5 and 10 years have confirmed that cure appears to be maintained.

Initial fears about mesh erosion has not been confirmed, although longer term data on larger number of women will be needed to provide greater reassurance about this. Because of the benefits of the TVT procedure, it has largely replaced colposuspension as the treatment of choice in the UK over the past 2 or 3 years.

Despite this trend in surgical practice, this trial provides robust evidence that the surgeons trained in this technique can offer women this procedure with confidence. This result of this trial emphasises the importance of considering the cost implications of each procedure in terms of ‘good value for money’. COLPO participants are being followed up for 4 years to provide long-term effectiveness and cost-effectiveness data. The discrepant results between conventional objective outcomes and subjective outcomes are significant and suggest that further qualitative research is required to identify outcomes that matter most to women with urinary incontinence.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

Independent trial steering committee: Prof J Neilson (Chair), Dr Cathy McGrother, Dr Paula Williamson, Mr Jerry Jarvis and Ms E Adams. Data monitoring and ethics committee: Prof DR Jones (Chair), Mr P Abrams and Mr P Hilton.

Contribution to authorship

COLPO study group: H.C.K., principal investigator, conception and design of study, coordination, drafting of article; G.D., statistician, contributed to trial design, data analysis and writing the article; V.L., research fellow, acquired a substantial proportion of the clinical data, provided supervision and support to peripheral recruiting centres and contributed to critical revision of the article; F.R., research fellow, acquired a substantial proportion of the clinical data, provided supervision and support to peripheral recruiting centres and contributed to critical revision of the article; L.N., data manager, acquired a substantial proportion of the data and contributed to critical revision of the article; A.R.B.S., clinical investigator, contributed to the trial design and writing the article; David Torgerson, health economist, contributed to trial design; Andrea Manca, health economist, data analysis; Jo Dumville, health economist, data analysis; Nadia Ali, research fellow, provided care for trial subjects; Mr Alfred Cutner, clinical investigator and Mr John Osborne, clinical investigator, Elizabeth Garrett Anderson Hospital, London; Mr Tim Sayer, clinical investigator, North Hampshire Hospital NHS Trust, Basingstoke; Mr Charles Cox, clinical investigator, Royal Wolverhampton Hospitals NHS Trust, Wolverhampton; Dr Robert Hawthorn, clinical investigator and Dr Ian Ramsay, clinical investigator, Southern General Hospitals NHS Trust, Glasgow; Dr Christine Landon, clinical investigator, Leeds General Infirmary, Leeds; Margaret Hunter, research nurse (Glasgow), provided care for trial subjects; Julie Davies, research nurse (Wolverhampton), provided care for trial subjects; Mary Remy, research nurse (London), provided care for trial subjects; Jackie Moody, research nurse (Basingstoke), provided care for trial subjects and Sylvia Skerrow, research nurse (Leeds), provided care for trial subjects.

Funding

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References

This study was funded by the Medical Research Council (MRC). The MRC has not been involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data and preparation, review or approval of the manuscript. V.L., F.R. and L.N. salaries during the period of their research employment were funded by the MRC.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Funding
  9. References
  • 1
    Hunskaar S, Lose G, Sykes D, Voss S. The prevalence of urinary incontinence in women in four European countries. BJU Int 2004;93:32430.
  • 2
    Burch JC, Moehrer B, Carey M, Wilson D, Stamey TA, Vancaillie TG, et al. Cooper’s ligament urethrovesical suspension for stress incontinence. Am J Obstet Gynecol 1968;100:76474.
  • 3
    Smith A, Daneshgari F, Dmochowski R, Milani R, Miller M, Paraiso M, et al. Surgery for urinary incontinence in women. In: AbramsP, CardozoL, KouryS, WeinA, editors. Incontinence. Plymouth: Health Publications Ltd., 2002, p. 1297370.
  • 4
    Stamey TA. Endoscopic suspension of the vesical neck for urinary incontinence. Surg Gynecol Obstet 1973;136:54754.
  • 5
    Vancaillie TG, Schuessler W. Laparoscopic bladderneck suspension. J Laparoendosc Surg 1991;1:16973.
  • 6
    Moehrer B, Carey M, Wilson D. Laparoscopic colposuspension: a systematic review. BJOG 2003;110:2305.
  • 7
    Moehrer B, Ellis G, Carey M, Wilson PD. Laparoscopic colposuspension for urinary incontinence in women. Cochrane Database Syst Rev 2000;CD002239.
  • 8
    Lapitan MC, Cody DJ, Grant AM. Open retropubic colposuspension for urinary incontinence in women. Cochrane Database Syst Rev 2005;CD002912.
  • 9
    Dumville J, Manca A, Kitchener H, Smith A, Nelson L, Torgerson D. Cost-effectiveness analysis of open colposuspension versus laparoscopic colposuspension in the treatment of urodynamic stress incontinence. BJOG 2006;113:1014–22.
  • 10
    Ware JE, Kosinski M, Keller SD. SF-36 Physical and Mental Health Summary Scales: A User Manual. Boston, MA: New England Medical Centre, 1994.
  • 11
    Jackson S, Donovan J, Brookes S, Eckford S, Swithinbank L, Abrams P. The Bristol Female Lower Urinary Tract Symptoms questionnaire: development and psychometric testing. Br J Urol 1996;77:80512.
  • 12
    Black N, Griffiths J, Pope C. Development of a symptom severity index and a symptom impact index for stress incontinence in women. Neurourol Urodyn 1996;15:63040.
  • 13
    Kind P. The EuroQol instrument: an index of health-related quality of life. In: SpikerB, editor. Quality of Life and Pharmacoeconomics in Clinical Trials. Philadelphia, PA: Lippincott-Raven; 1996. p. 191201.
  • 14
    White IR. Uses and limitations of randomization-based efficacy estimators. Stat Methods Med Res 2005;14:32747.
  • 15
    Little RJA, Rubin DB. Statistical Analysis with Missing Data, 2nd edn. Hoboken, NJ: John Wiley & Sons, 2002.
  • 16
    Dunn G, Maracy M, Tomenson B. Estimating treatment effects from randomized clinical trials with noncompliance and loss to follow-up: the role of instrumental variable methods. Stat Methods Med Res 2005;14:36995.
  • 17
    Burton G. A three year prospective randomised urodynamic study comparing open and laparoscopic colposuspension. Neurourol Urodyn 1997;16:3534.
  • 18
    Carey M, Rosamilia A, Maher C, Cornish A, Murray C, Ugoni A, et al. Laparoscopic versus open colposuspension: a prospective multicentre randomised single blind trial. Neurourol Urodyn 2000;19:38991.
  • 19
    Fatthy H, El Hao M, Samaha I, Abdallah K. Modified Burch colposuspension: laparoscopy versus laparotomy. J Am Assoc Gynecol Laparosc 2001;8:99106.
  • 20
    Su TH, Wang KG, Hsu CY, Wei HJ, Hong BK. Prospective comparison of laparoscopic and traditional colposuspensions in the treatment of genuine stress incontinence. Acta Obstet Gynecol Scand 1997;76:57682.
  • 21
    Summitt RL Jr, Lucente V, Karram MM, Shull BL, Bent AE. Randomized comparison of laparoscopic and transabdominal Burch urethropexy for the treatment of genuine stress incontinence. Obstet Gynecol 2000;95 (Suppl 1):S2.
  • 22
    Ward KL, Hilton P. A prospective multicenter randomized trial of tension-free vaginal tape and colposuspension for primary urodynamic stress incontinence: two-year follow-up. Am J Obstet Gynecol 2004;190:32431.
  • 23
    Liapis A, Bakas P, Creatsas G. Burch colposuspension and tension-free vaginal tape in the management of stress urinary incontinence in women. Eur Urol 2002;41:46973.
  • 24
    Wang AC, Chen MC. Comparison of tension-free vaginal taping versus modified Burch colposuspension on urethral obstruction: a randomized controlled trial. Neurourol Urodyn 2003;22:18590.