This individual patient data review found that laparoscopic uterosacral nerve ablation did not improve worst pain in comparison with diagnostic laparoscopy in women with chronic pelvic pain. There were no important variations in treatment effectiveness in any of the subgroups explored. The evidence base and review methods minimised the potential for bias and these findings are likely to be reliable.

To assess the effectiveness of laparoscopic uterosacral nerve ablation in chronic pelvic pain.

MEDLINE, EMBASE, PsycINFO and The Cochrane Library databases were searched from inception to August 2009. Search terms were reported in a prior review. Reference lists were also scrutinised for further eligible studies.

Randomised controlled trials that compared laparoscopic uterosacral nerve ablation against diagnostic laparoscopy in women of reproductive age with chronic pelvic pain or dysmenorrhoea were eligible for inclusion (and analysed provided that individual patient data was available). Trials with concomitant ablation of endometrial lesions were eligible if it was an option in both study groups. The primary outcome was worst pain experienced.

Trial populations were similar in most respects although one trial included women for primary dysmenorrhoea only. Treatment sometimes included concomitant laparoscopic excision or ablation of all visible endometriosis along with laparoscopic uterosacral nerve ablation whilst controls sometimes consisted of diagnostic laparoscopic surgery only or laparoscopic excision or ablation of all visible endometriosis.

Two reviewers independently assessed trial eligibility with discrepancies resolved by a third reviewer.

Trial quality was assessed in terms of: adequacy of sequence generation, allocation concealment, blinding of participants and assessors to treatment allocation, intention-to-treat and potential impact of losses to follow-up. Further assessment of data quality was undertaken upon receipt of individual patient data with discrepancies between individual patient data and reported aggregate data resolved with the original trialists.

A measure of worst pain level experienced was derived from individual patient data across different pain symptoms (such as dysmenorrhoea, non-cyclical pain, dyspareunia). Pain scores using the Visual Analogue Scale were standardised where necessary to anchor the scale between 0 (no pain) and 10 (worst pain imaginable). Associated covariates pre-specified in a protocol were also extracted in individual patient data form. Data was extracted based on intention-to-treat where this was specified.

Worst pain was treated as a continuous response variable and overall treatment effects were estimated using fixed-effect and random-effects multilevel models including trial, treatment and trial by treatment interaction. Multiple time points were incorporated as repeated measures but the longitudinal structure of the model was not reported. It was assumed that baseline event rate (expressed as the intercept) was included as a fixed-effect. Individual treatment interaction terms were added to this model to explore variation in treatment effect by presence or absence of visual pathology, site of pain (central versus peripheral), age group (below 30 years, 30 years or above) and parity (nulliparous versus parous). Centring, transformation and standardisation of covariates were not reported.

Five trials (862 women) met the inclusion criteria and all had individual patient data available. Methodological quality of the included studies was generally good although the smallest of these studies was included only in sensitivity analysis as randomisation was inadequately concealed. Four of the five studies had small sample sizes (21 to 180 patients) whilst the largest study consisted of 487 patients.

Worst pain was lower with no laparoscopic uterosacral nerve ablation than laparoscopic uterosacral nerve ablation at three months (0.49 points, 95% CI 0.12 to 0.86) but there was no difference in treatment effects at subsequent time points. Heterogeneity was not statistically significant in the fixed-effect model but was when random-effects were introduced. There was some evidence of interaction with pathology but the mean differences were not clinically significant and the confidence intervals of the point estimates overlap substantially. There was no evidence of interaction amongst the remaining subgroups.

Laparoscopic uterosacral nerve ablation does not result in improved chronic pelvic pain

Methods of searching, study selection, validity assessment and data extraction were all appropriate and minimised the potential for bias. The multilevel models were appropriate but it was difficult to assess their optimality in the absence of full specification. There may have been some potential for confounding individual-based within trial relationships with across trial group level averages. In addition, arbitrary dichotomisation of age leads to a loss of power, but these limitations were likely to have had a minimal impact on the results. The differences between fixed and random models suggest that the analysis was driven by the largest trial, but overall interpretation was consistent across all trials and syntheses. The generalisability of results from fewer than 500 patients was questionable, but this was not atypical of sample sizes in similar pain studies.

The conclusions of the review pertain to an outcome of worst pain and a comparator consisting of diagnostic laparoscopy. The authors' implications for research and practice rely on the assumptions that other comparators and outcomes were not pertinent. The authors' conclusions reflected the evidence presented and were likely to have been reliable.

National guidelines should have been updated to reflect the lack of clinical benefit of laparoscopic uterosacral nerve ablation. There was no requirement for further trials with laparoscopic uterosacral nerve ablation but research effort should be directed to alternative novel therapies.

Wellbeing of Women (UK).

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.