Adverse effects of spinal manipulation

Professor Ernst has published a so-called ‘systematic’ review¹ of adverse events following spinal manipulation based on low quality evidence including, for the most part, case reports (JRSM 2007;100:330-338). The whole point of systematic reviews is to get away from an author's often strong prior beliefs and opinions, and instead present an impartial and even-handed review of the evidence. Essential ingredients are that the evidence is critically appraised and the recommendations based firmly on the quality of the evidence presented. In this case, all of the evidence was included indiscriminately and the findings of a causal link between spinal manipulation and adverse events based on unsupportive low-level evidence. Moreover, Professor Ernst makes a judgement on this apparent causation as either ‘certain’, ‘likely’ or ‘possible’. There are no criteria presented in this paper for how such judgements were made.

Of course, the absence of high-quality evidence is not to say there is no risk associated with spinal manipulation: there are safety issues with all interventions. What we need to know is the nature and the size of those risks. In the UK, there are estimated to be well over two million cervical spine manipulations by chiropractors each year.² Given the number of cervical spine manipulations done worldwide it is likely, even if under-reporting is as high as Professor Ernst suggests, that the level of risk of a serious adverse event is extremely low. To turn Professor Ernst's point on under-reporting and publication bias around, could it be that journals of medicine are unlikely to publish findings which might be considered ‘positive’? Good evidence on safety is of paramount importance; what we have here is poor quality evidence and unsubstantiated claims masquerading as a systematic review.