Safety of Acupuncture: Overview of Systematic Reviews

This comprehensive overview analyzed 17 systematic reviews published between 1996 and 2015 to examine all reported adverse events associated with acupuncture and related therapies. The study included reviews that covered a median of 43 primary studies and 279 patients, spanning various modalities such as traditional acupuncture, electroacupuncture, auriculotherapy, cupping, and moxibustion. The search was conducted in the MEDLINE and EMBASE databases from inception through December 2015. The methodological quality of the included reviews was assessed using the AMSTAR instrument, revealing overall median quality.

Only one review provided an a priori protocol, and only one adequately appraised the scientific quality of the included primary studies. The investigators identified four main categories of adverse events. The first category comprised injuries to organs or tissues, reported in 13 reviews with a median of 36 cases and 4 deaths. The most common events included pneumothorax (median 25.5 cases), central nervous system or spinal cord injuries (median 13 cases), subarachnoid or intracranial hemorrhage (median 35 cases), and cardiac tamponade (median 7 cases).

The second category involved infections, documented in 11 reviews with a median of 17 cases and 0.5 deaths. The most frequent infections were hepatitis (median 94 cases), abscesses (median 7 cases), tetanus (median 6 cases), and ear infections. The third category covered local adverse events or reactions, reported in 12 reviews with a median of 8.5 cases and no deaths. These included contact dermatitis or local allergic reactions, argyria, local bleeding, local pain or tenderness, local burns, and hematomas.

The fourth category included other complications reported in 11 reviews with a median of 21 cases and no deaths, with dizziness or syncope being the most common (median 10.5 cases). The study revealed that, contrary to the common impression that acupuncture is harmless, both minor and serious adverse events can occur. Although serious events are rare, they warrant significant attention due to the potential for associated mortality. The authors identified several important limitations, including the inability to determine causality in most reported cases, lack of adequate follow-up of many adverse events, and substantial concerns about underreporting.

Many publications came from physicians treating interesting complications, potentially leading to publication bias. Some reviews found that no adverse events were reported by acupuncturists, raising questions about practitioners' awareness of complications. Clinical implications include the need for practitioners to pay adequate attention to patient risk stratification, the implementation of more rigorous sterilization practices, improvement of patient education on common and serious risks, and improvement in practitioners' recognition of acute complications. The study emphasizes that referrals should consider the credibility of acupuncturists' training and that patient safety should be a central component of acupuncture education.

The authors recommend the establishment of standardized reporting platforms for adverse events and better communication between physicians treating complications and the practitioners who administered the acupuncture.