Intrauterine human chorionic gonadotropin administration before embryo transfer (IHABT): an individual participant data meta-analysis of randomized controlled trials

BACKGROUND: Intrauterine administration of hCG has been considered as a promising IVF add-on before embryo transfer to improve fertility outcomes. A Cochrane review and four more recent systematic reviews all showed improved clinical pregnancy rates and/or live birth rates following intrauterine administration of hCG, however, a high unexplained heterogeneity was also present. OBJECTIVE AND RATIONALE: To investigate the effectiveness and safety of intrauterine administration of hCG before embryo transfer in participants undergoing IVF. Individual participant data meta-analysis (IPD-MA) is recognized as the gold standard for evidence synthesis due to its ability to harmonize the data and to investigate treatment-covariate interactions. In addition, with recent experiences of guideline development and systematic review production raising increasing concerns about the trustworthiness of randomized controlled trials (RCTs) in women's health research, an IPD-MA provides a unique opportunity to summarize the best available and most trustworthy evidence on this topic. SEARCH METHODS: We searched MEDLINE, Embase, Cochrane Gynaecology and Fertility Group Specialised Register, Cochrane Central Register of Controlled Trials, PsycINFO, and clinical trial registries without language restrictions up to January 2026. Inclusion criteria included RCTs comparing intrauterine administration of hCG before embryo transfer versus placebo or no intervention in participants undergoing IVF. The IPD Integrity tool and the TRACT checklist were used to evaluate the trustworthiness of studies with and without IPD, respectively. Both one-stage and two-stage random-effect IPD meta-analyses were performed with one-stage being the primary analysis. OUTCOMES: We detected 28 RCTs, of which 7 RCTs with IPD involving 2244 participants were included. All seven RCTs with IPD met trustworthiness criteria and six RCTs had overall low risk of bias. All RCTs without IPD did not meet trustworthiness criteria. IPD-MA showed intrauterine administration of hCG before embryo transfer did not improve live birth rates (7 RCTs, 2244 participants, odds ratio [OR] 0.99, 95% CI 0.83-1.19) or clinical pregnancy rates (7 RCTs, 2244 participants, OR 1.04, 95% CI 0.83-1.31). Studies without IPD showed different results from those with IPD for live birth (1.99, 0.72-5.50, P for interaction <0.001) and clinical pregnancy (1.87 (1.48-2.35), 17 RCTs without IPD, 3152 participants, P for interaction 0.005). WIDER IMPLICATIONS: Our IPD-MA has shown that intrauterine administration of hCG before embryo transfer is unlikely to improve the chance of clinical pregnancy and live birth. In the comparison between studies with IPD and without IPD, we found that none of the RCTs without IPD met trustworthiness criteria but showed a significant improvement in clinical pregnancy. We therefore suggest that intrauterine administration of hCG should not be offered as an IVF add-on in practice. REGISTRATION NUMBER: PROSPERO (CRD42020177397).