Adverse events (AE) are a concern for practitioners utilizing cervical manipulation or mobilization. While efficacious, these techniques are associated with rare but serious adverse events. Five bibliographic databases (PubMed, CINAHL, PEDro, AMED, EMBASE) and the gray literature were searched from 1998 to 2009 for any AE associated with cervical manipulation or mobilization for neck pain. Randomized controlled trials (RCTs), prospective or cross-sectional observational studies were included. Two independent reviewers conducted study selection, method quality assessment and data abstraction. Pooled relative risks (RR) were calculated. Study quality was assessed using the Cochrane system, a modified Critical Appraisal Skills Program form and the McHarm scale to assess the reporting of harms. Seventeen of 76 identified citations resulted in no major AE. Two pooled estimates for minor AE found transient neurological symptoms [RR 1.96 (95% CI: 1.09-3.54) p < 0.05]; and increased neck pain [RR 1.23 (95% CI: 0.85-1.77) p > .05]. Forty-four studies (58%) were excluded for not reporting AE. No definitive conclusions can be made due to a small number of studies, weak association, moderate study quality, and notable ascertainment bias. Improved reporting of AE in manual therapy trials as recommended by the CONSORT statement extension on harms reporting is warranted.
Copyright 2010 Elsevier Ltd. All rights reserved.