Abstract
BACKGROUND: The effects of acupuncture on in vitro fertilization (IVF) outcomes remain controversial. And the variation in participant, interventions, outcomes studied, and trial design may relate to the efficacy of adjuvant acupuncture. METHODS: We searched digital databases for relevant studies, including Embase, PubMed, Cochrane Library and some Chinese databases up to December 2018, for randomized controlled trials (RCTs) evaluating the effects of acupuncture on women undergoing IVF. We included studies with intervention groups using needling, and control groups consisting of no acupuncture or sham (placebo) acupuncture. Primary outcomes were clinical pregnancy rate (CPR) and live birth rate (LBR). Meta-regression and subgroup analysis were conducted on the basis of eight pre-specified covariates to investigate the variances of the effects of adjuvant acupuncture on pregnancy rates and the sources of heterogeneity. RESULTS: Twenty-seven studies with 6116 participants were included. The pooled clinical pregnancy rate (CPR) from all of acupuncture groups was significantly greater than that of control groups (RR 1.21, 95% CI: 1.07-1.38), whereas the pooled live birth rate (LBR) was not. Meta-regression subgroup analysis showed a more significant benefit of acupuncture for repeated IVF cycle proportion (number of women with a history of prior unsuccessful IVF attempt divided by number of women included in each trial) ≥ 50% group (CPR: RR 1.60, 95% CI: 1.28-2.00; LBR: RR 1.42, 95% CI: 1.05-1.92), and this covariate explained most of the heterogeneity (CPR and LBR: adjusted R(2) = 100 and 87.90%). Similar results were found between CPR and number of acupuncture treatments (CPR: p = 0.002, adjusted R(2) = 51.90%), but not LBR. CONCLUSIONS: Our analysis finds a benefit of acupuncture for IVF outcomes in women with a history of unsuccessful IVF attempt, and number of acupuncture treatments is a potential influential factor. Given the poor reporting and methodological flaws of existing studies, studies with larger scales and better methodologies are needed to verify these findings.