Efficacy of ten interventions for endometriosis: A network meta-analysis.

Abstract

Clinical trials comparing the efficacy of different interventions for endometriosis are limited and controversial. The aim of the present study is to perform a network meta-analysis on the efficacy of various interventions for endometriosis. We searched PubMed and Ovid EMBASE through 1 June, 2018, for trials reporting the pain score and 1-year pregnancy rate of patients including at least one pair of direct control group. The mean difference of pain score, odds ratio of 1-year pregnancy rate, and their associated 95% credible intervals (CrI) were used to describe efficacy. The surface under the cumulative ranking curve (SUCRA) was calculated to illustrate the rank probability of various treatments for different outcomes, on the basis of network meta-analysis. Our meta-analysis enrolled six studies for the evaluation of reducing pain and 10 studies for the 1-year pregnancy rate. All involved trials were sufficiently powered with a low risk of bias. Expectant management (EM), progesterone (PR), and gonadotropin-releasing hormone (GnRH)-agonist (GN) were significantly effective to reduce pain when compared with the placebo; EM ranked the highest on the SUCRA curve. For the 1-year pregnancy rate, no significant difference between the interventions was evident. Ablation ranked the highest with a SUCRA value of 0.6328. The rank of EM was acceptable with a SUCRA value of 0.4452. Our experimental results need to be verified by more high-qualityrandomized controlled trial articles. Limited available evidence showed that EM, PR, and GN were significantly effective to reduce pain when compared with the placebo. Due to a lack of convincing evidence favoring surgery or medication, we recommend considering EM.