Does adding acupuncture improve clinical pregnancy rates after recurrent implantation failure?
The review suggests a possible improvement, but does not establish a reliable treatment benefit. Across 14 trials involving 1,428 patients, the authors report more clinical pregnancies with acupuncture-related treatments added to embryo transfer than with transfer without these treatments or with sham treatment: risk ratio 1.73, with a 95% confidence interval of 1.51–1.99. This measures pregnancy establishment, not delivery of a baby. Only two trials used sham controls, and allocation concealment was unclear in 11. Consequently, differences in participant selection, additional care or treatment expectations cannot be adequately separated from the specific intervention’s effects. The pooled result should not be interpreted as a dependable individual forecast.
Does the reported live-birth benefit use comparable denominators across trials?
Not consistently in the published forest plot, which materially limits interpretation of this patient-important outcome. Only four trials contributed live-birth results. For You 2023, the plot uses denominators of 13 and 5—the clinical pregnancy counts—although Table 1 lists 35 participants per group. Other trials use their allocated group sizes. Births among established pregnancies and births among all treated patients answer different questions and should not be pooled as interchangeable outcomes. The miscarriage analysis shows a similar denominator inconsistency. Checking the original trial reports and reanalysing consistently defined outcomes is necessary before relying on the claimed live-birth increase or miscarriage reduction.
Does this review identify an effective treatment schedule and establish its safety?
Neither an optimal schedule nor adequate safety is established. The review combines manual and electrical acupuncture, skin-surface electrical stimulation, auricular pressure and moxibustion, delivered over markedly different schedules. Its suggestion that more than ten sessions or treatment around embryo transfer produces greater benefit comes from comparisons between studies, not random allocation to different schedules; other study differences could explain those findings. Treatment details were also insufficient to identify a reproducible optimal programme. Only two trials reported adverse events, including itching and fainting in one. Sparse reporting cannot establish safety, and the findings do not justify treating these different interventions as equally effective or interchangeable.
Assessed 18 Sept 2026.
