Some people will probably think that I am obsessed with writing about the risk of chiropractic. True, I have published quite a bit on this subject, both in the peer-reviewed literature as well as on this blog – but not because I am obsessed; on this blog, I will re-visit the topic every time a relevant new piece of evidence becomes available because it is indisputably such an important subject. Writing about it might prevent harm.
So far, we know for sure that mild to moderate as well as serious complications, including deaths, do occur after chiropractic spinal manipulations, particularly those of the upper spine. What we cannot say with absolute certainty is whether they are caused by the treatment or whether they happened coincidentally. Our knowledge in this area relies mostly on case-reports and surveys which, by their very nature, do not allow causal inferences. Therefore chiropractors have, in the past, been able to argue that a causal link remains unproven.
A brand-new blinded parallel group RCT might fill this gap in our knowledge and might reject or establish the notion of causality once and for all. The authors’ objective was to establish the frequency and severity of adverse effects from short term usual chiropractic treatment of the spine when compared to a sham treatment group. They thus conducted the first ever RCT with the specific aim to examine the occurrence of adverse events resulting from chiropractic treatment. It was conducted across 12 chiropractic clinics in Perth, Western Australia. The participants comprised 183 adults, aged 20-85, with spinal pain. Ninety two participants received individualized care consistent with the chiropractors’ usual treatment approach; 91 participants received a sham intervention. Each participant received two treatment sessions.
Completed adverse questionnaires were returned by 94.5% of the participants after the first appointment and 91.3% after the second appointment. Thirty three per cent of the sham group and 42% of the usual care group reported at least one adverse event. Common adverse events were increased pain (sham 29%; usual care 36%), muscle stiffness (sham 29%; usual care 37%), headache (sham 17%; usual care 9%). The relative risk was not significant for either adverse event occurrence (RR = 1.24 95% CI 0.85 to 1.81); occurrence of severe adverse events (RR = 1.9; 95% CI 0.98 to 3.99); adverse event onset (RR = 0.16; 95% CI 0.02 to 1.34); or adverse event duration (RR = 1.13; 95% CI 0.59 to 2.18). No serious adverse events were reported.
The authors concluded that a substantial proportion of adverse events following chiropractic treatment may result from natural history variation and non-specific effects.
If we want to assess causality of effects, we have no better option than to conduct an RCT. It is the study design that can give us certainty, or at least near certainty – that is, if the RCT is rigorous and well-made. So, does this study reject or confirm causality? The disappointing truth is that it does neither.
Adverse events were clearly more frequent with real as compared to sham-treatment. Yet the difference failed to be statistically significant. Why? There are at least two possibilities: either there was no true difference and the numerically different percentages are a mere fluke; or there was a true difference but the sample size was too small to prove it.
My money is on the second option. The number of patients was, in my view, way too small for demonstrating differences in frequencies of adverse effects. This applies to the adverse effects noted, but also, and more importantly, to the ones not noted.
The authors state that no serious adverse effects were observed. With less that 200 patients participating, it would have been most amazing to see a case of arterial dissection or stroke. From all we currently know, such events are quite rare and occur perhaps in one of 10 000 patients or even less often. This means that one would require a trial of several hundred thousand patients to note just a few of such events, and an RCT with several million patients to see a difference between real and sham treatment. It seems likely that such an undertaking will never be affordable.
So, what does this new study tell us? In my view, it is strong evidence to suggest a causal kink between chiropractic treatment and mild to moderate adverse effects. I dose not prove it, but merely suggests it – yet I am fairly sure that chiropractors, once again, will not agree with me.