The safety of acupuncture during pregnancy: a systematic review

This systematic review represents the first comprehensive study on the safety of acupuncture during pregnancy, analyzing 105 studies that included 2,460 pregnant women who underwent approximately 22,283 treatment sessions. The primary objective was to identify and evaluate adverse events associated with acupuncture treatment in pregnant women, filling an important knowledge gap about the safety of this therapy in this special population. The investigators conducted extensive searches across multiple international and Korean databases, including Medline, Embase, Cochrane, and CINAHL, from inception through February 2013. Studies of any design that reported original data on acupuncture with needling and/or moxibustion for any condition in pregnant women were included, excluding studies on labor, abortion, assisted reproduction, or postpartum conditions.

The methodology included rigorous evaluation of adverse-event severity using standardized criteria and causality analysis to determine the relationship between events and acupuncture treatment. Results showed that only 25.7% of included studies reported details on adverse events, highlighting a significant problem in the literature regarding inadequate reporting of side effects. Of the 429 adverse events identified, the great majority were classified as mild (322 events), including primarily needle-site pain (48 cases) and local bleeding (40 cases). Other mild events included hematoma, fatigue, headache, drowsiness, dizziness, and discomfort at needling points.

Only 6 events were classified as moderate, mainly fainting episodes. Importantly, the 99 serious adverse events identified—which included maternal complications such as hypertension, preeclampsia, preterm birth, and spontaneous abortion, as well as fetal complications such as congenital defects—were all assessed as unlikely to have been caused by acupuncture. The overall incidence rate of adverse events was 1.9% (193 per 10,000 sessions), and when limited only to events considered certainly, probably, or possibly related to acupuncture, the incidence was 1.3% (131 per 10,000 sessions). These values are comparable to or lower than those reported in the general population receiving acupuncture.

The most frequently treated conditions were low back and/or pelvic pain (9 studies) and fetal malposition (7 studies). Treatments ranged from 2 to 40 sessions over 5 days to 8 weeks, with most studies using points along the 12 main meridians. The quality of adverse-event reporting was generally poor, with only one study providing a clear definition of adverse events and only 40% reporting methods for collecting safety data. The clinical implications are significant, as this study provides reassuring evidence on the safety of acupuncture during pregnancy when properly applied by qualified clinicians.

Considering that pregnant women have limited treatment options because of concerns about medications, acupuncture emerges as a safe alternative for several conditions. The absence of serious adverse events directly related to acupuncture is particularly relevant in the context of controversies regarding 'forbidden' points during pregnancy. Limitations include inadequate reporting of adverse events in the literature, possible publication bias, and heterogeneity of the included studies. The authors acknowledge that there may be underestimation of the true incidence of adverse events because of inadequate reporting in the primary studies.