Critique of review of deaths after chiropractic, 5

To the Editor: In the July edition of the journal, Ernst reviewed a collection of 26 fatalities reported to be associated with chiropractic care (1). Although we applaud the author’s efforts to evaluate adverse events associated with chiropractic spinal manipulation, we fear that the study fell short of offering valuable information and, if anything, may distort our understanding of the benefits and harms of chiropractic spinal manipulation. First, resulting from a lack of clarity in the reporting of the study, we found it difficult to grasp what was done and how the author reached his conclusion. Although the author described the study as a systematic review, it clearly did not meet standards recommended by The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement (2). According to our review, the study met only two of the 27 PRISMA checklist items. The article is best described as a case series and although a case series is of value for generating hypotheses, it is not usually used for drawing conclusions about the efficacy or risks associated with an intervention. Most (90–95%) of high-velocity low-amplitude spinal manipulation procedures are performed by chiropractors (at least in the US) (3). Yet, as the author noted, osteopaths, physical therapists, medical physicians and others also perform these procedures. Ernst stated that spinal manipulation performed by chiropractors is less safe than that performed by other professionals and in support of this argument, he cited a study conducted by Terrett (4) that reported 17 deaths associated with chiropractic spinal manipulation and 13 deaths associated with spinal manipulation performed by other types of providers. However, because 90–95% of spinal manipulation is performed by chiropractors and 43% of the deaths reported in this study were caused by procedures performed by non-chiropractors, it appears as though chiropractors actually perform this procedure more safely than other professionals. Examination of the risks associated with spinal manipulation performed by different professionals might make an interesting subject for future enquiry. The author’s final conclusion is that the risks of spinal manipulation outweigh its benefits. There are two critical questions that must be answered to reach such a conclusion: (i) What are the benefits? and (ii) What are the risks? The first question was not addressed in this study. The study provided a summary of reported deaths after chiropractic, but without a relative rate (or at least a denominator) any meaningful interpretation of the results is nearly impossible. Herein is a denominator: in the US alone, over 100 million visits are made annually to chiropractors (5). Three deaths were reported during the last 10 years of the study, so for that most recent time period, the absolute risk could be estimated to be 3/10 per 100 million, or three deaths for every billion chiropractic encounters. Alternatively, over the full 75-year time period of the study, we could calculate a slightly higher estimate of mortality, equal to 26 per 7.5 billion encounters. This rate is so low that it cannot possibly be considered significant. An interesting flip side to the research question might be: by undergoing a course of chiropractic spinal manipulation, how many patients were able to avoid death by avoiding complications of surgical intervention? None.