Clinical effectiveness and cost-effectiveness of surgical options for the management of anterior and/or posterior vaginal wall prolapse: two randomised controlled trials within a comprehensive cohort study – results from the PROSPECT Study

Abstract

BackgroundThe use of mesh in prolapse surgery is controversial, leading to a number of enquiries into its safety and efficacy.ObjectiveTo compare synthetic non-absorbable mesh inlay, biological graft and mesh kit with a standard repair in terms of clinical effectiveness, adverse effects, quality of life (QoL), costs and cost-effectiveness.DesignTwo randomised controlled trials within a comprehensive cohort (CC) study. Allocation was by a remote web-based randomisation system in a 1 :1 : 1 ratio (Primary trial) or 1 : 1 : 2 ratio (Secondary trial), and was minimised on age, type of prolapse repair planned, need for a concomitant continence procedure, need for a concomitant upper vaginal prolapse procedure and surgeon. Participants and outcome assessors were blinded to randomisation; participants were unblinded if they requested the information. Surgeons were not blinded to allocated procedure.SettingThirty-five UK hospitals.ParticipantsPrimary study: 2474 women in the analysis (including 1348 randomised) having primary anterior or posterior prolapse surgery.Secondary study: 398 in the analysis (including 154 randomised) having repeat anterior or posterior prolapse surgery.CC3: 215 women having either uterine or vault prolapse repair.InterventionsAnterior or posterior repair alone, or with mesh inlay, biological graft or mesh kit.Main outcome measuresProlapse symptoms [Pelvic Organ Prolapse Symptom Score (POP-SS)]; prolapse-specific QoL; cost-effectiveness [incremental cost per quality-adjusted life-year (QALY)].ResultsPrimary trials: adjusting for baseline and minimisation covariates, mean POP-SS was similar for each comparison {standard 5.4 [standard deviation (SD) 5.5] vs. mesh 5.5 (SD 5.1), mean difference (MD) 0.00, 95% confidence interval (CI) –0.70 to 0.71; standard 5.5 (SD 5.6) vs. graft 5.6 (SD 5.6), MD –0.15, 95% CI –0.93 to 0.63}. Serious non-mesh adverse effects rates were similar between the groups in year 1 [standard 7.2% vs. mesh 7.8%, risk ratio (RR) 1.08, 95% CI 0.68 to 1.72; standard 6.3% vs. graft 9.8%, RR 1.57, 95% CI 0.95 to 2.59]. There were no statistically significant differences between groups in any other outcome measure. The cumulative mesh complication rates over 2 years were 2 of 430 (0.5%) for standard repair (trial 1), 46 of 435 (10.6%) for mesh inlay and 2 of 368 (0.5%) for biological graft. The CC findings were comparable. Incremental costs were £363 (95% CI –£32 to £758) and £565 (95% CI £180 to £950) for mesh and graft vs. standard, respectively. Incremental QALYs were 0.071 (95% CI –0.004 to 0.145) and 0.039 (95% CI –0.041 to 0.120) for mesh and graft vs. standard, respectively. A Markov decision model extrapolating trial results over 5 years showed standard repair had the highest probability of cost-effectiveness, but results were surrounded by considerable uncertainty.Secondary trials: there were no statistically significant differences between the randomised groups in any outcome measure, but the sample size was too small to be conclusive. The cumulative mesh complication rates over 2 years were 7 of 52 (13.5%) for mesh inlay and 4 of 46 (8.7%) for mesh kit, with no mesh exposures for standard repair.ConclusionsIn women who were having primary repairs, there was evidence of no benefit from the use of mesh inlay or biological graft compared with standard repair in terms of efficacy, QoL or adverse effects (other than mesh complications) in the short term. The Secondary trials were too small to provide conclusive results.LimitationsWomen in the Primary trials included some with a previous repair in another compartment. Follow-up is vital to identify any long-term potential benefits and serious adverse effects.Future workLong-term follow-up to at least 6 years after surgery is ongoing to identify recurrence rates, need for further prolapse surgery, adverse effects and cost-effectiveness.TriaI registrationCurrent Controlled Trials ISRCTN60695184.FundingThis project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full inHealth Technology Assessment; Vol. 20, No. 95. See the NIHR Journals Library website for further project information.

Primary surgical management of anterior pelvic organ prolapse: a systematic review, network meta‐analysis and cost‐effectiveness analysis

“…Our findings are consistent with a previous UK analysis, which also found mesh augmentation to be cost‐ineffective . The findings of a second economic evaluation were inconclusive; however, the results are not directly comparable because they included women with AC and/or posterior colporrhaphy …”

“…The risk of surgically managed recurrence following a secondary repair was based on an observational cohort study . This study did not report the anatomical recurrence rate so this was taken from a UK‐based RCT . The annual probabilities were estimated as described above for the primary repair.…”

“…The cost associated with conservative management was obtained from a UK‐based RCT that included treatment with pelvic floor exercises, estrogens and pessaries . It was assumed that only half of women experiencing recurrence would require treatment; symptoms in other women were not severe enough to require treatment for their POP …”

“…Utility values were required for active POP, resolved POP, recurrent POP and complications. Utility estimates were derived from the published UK RCT that reported the EuroQol (EQ‐5D‐3L) utility scores, estimated using the UK Time Trade‐Off Tariff …”

How to Appropriately Extrapolate Costs and Utilities in Cost-Effectiveness Analysis

“…There are also examples of studies which have used expert opinion [13,14] or analyst assumptions [15,16]. For many cost-effectiveness models there is insufficient information reported to reliably gauge the methods used to extrapolate healthcare costs and utilities for health states and clinical events, including for example Glazener et al [17], where it is not clear how costs observed at 2 years within the PROSPECT study were extrapolated over the 5 and 10 year time horizons.. In such instances the focus seems to be on describing extrapolation methods used for the effectiveness evidence while glossing over the assumptions and methods used to extrapolate costs and utilities, let alone assessing the implications of these assumptions in term of any decision uncertainty.…”