Re: Should we abandon cervical spine manipulation for mechanical neck pain? Yes
As the authors of the “Yes” position we are pleased that our paper has attracted wide attention and encouraged discussion. We would like to take the opportunity to address some of the key themes raised in these responses.
It is with regret that we read comments suggesting we have misrepresented the data and tried to stoke inter and intra-disciplinary rivalry. We have attempted from the start to provide a balanced and fair assessment of the available evidence, albeit one which in the end carries our own particular interpretation, just as all such reviews of the literature do. The article in question was never designed to be a systematic review of all the evidence. However, we do feel that our piece, unsystematic as it may be, is a fair reflection of the evidence and cannot be justly accused of extrapolation or misrepresentation.
While we acknowledge that there is uncertainty surrounding the causality of the association between manipulation and these events, we do not accept that the same uncertainty exists regarding the benefits. Indeed given the risk of bias of many of the trials included in the Cochrane review that we have cited, one might consider that the effect estimates for manipulation are likely to be overstated. We are not aware of any compelling evidence to the effect that manipulation confers unique benefit as part of a care package. Similarly there is no convincing evidence that unique responders to cervical manipulation techniques either exist or, if they do, can be reliably identified. The British Chiropractic Association has suggested that we have cherry picked low quality evidence. We believe that we have considered the key studies, but if the BCA are aware of important or more robust contradictory evidence then we would be very happy to consider it.
Dr Mann suggests that the UK BEAM trial provides evidence for the effectiveness of cervical manipulation, but since it was a trial of treatment for low back pain, this assertion is tenuous. However it is worth discussing this study as it highlights a number of important issues for manual therapists wishing to use high quality evidence rather than anecdote to inform their management of patients with neck pain. Firstly, the differences reported for the primary outcome measures were lower than the authors pre-set minimal clinically important difference. In addition, as is common to many trials of manual therapy for neck pain, the therapist, patients and the assessors were not blinded. It is likely that the already small effect sizes seen are exaggerated by the resulting biases. Most significantly in relation to the current discussion on high velocity manipulation techniques, it is important to recognise that clinicians were able to choose from a range of manual therapy techniques as well as mobilising and strengthening exercises. As is the case in many neck pain trials, effectiveness of a package of care provided by a manual therapist cannot be used to specifically endorse high velocity thrust manipulations, as they are often only part of a treatment package. It should also be noted that this package of care was designed and endorsed by the professional bodies representing osteopathy, chiropractic and physiotherapy in the UK and explicitly excluded the use of high velocity thrust techniques to the neck because of the chance of serious side effects.
We also do not accept that the existing data on risks, imperfect though it is, is uninformative. The fact that not all studies have shown an effect in the over 45's most likely reflects the other causes of these events in that population clouding the picture. Indeed finding the association in a group who would not normally be expected to experience such an event (i.e. <45 yrs) arguably makes the association more convincing. While attempts to develop screening procedures is commendable, given that the association between manipulation and VAD/stroke is found in this group who are less likely to present with key vascular risk factors, it seems unlikely that screening will achieve its goal satisfactorily. Reid and colleagues point out that in one survey  , 45% of these events may have been preventable, which still leaves a rather uncomfortable 55% that may not have been. Indeed acceptance of Cassidy and colleagues argument, that there are patients receiving cervical manipulation who present with a pre-existing dissection, only further illustrates that screening is far short of being effective.
We would suggest that comparisons with NSAIDs or surgical procedures are less useful. Beyond the issue of inadequate data for drawing fair comparisons and the likely underestimation of the true risks for manipulation, manual therapists might more simply ask how their own practice may be made safer. With a range of manual therapy approaches available to the practitioner, not performing one particular type of technique would impose little restriction on the therapist and would not be expected to seriously impact patient care. Indeed, the most meaningful comparison between manipulation and low velocity mobilisation for neck pain suggests that mobilisation may be associated with a lower recurrence rate .
Patients may wish to have their neck “clicked” but such demand is driven in part by the claims of the industry that promotes these therapies. After all, in possession of truly informed consent, why would one choose a technique that confers no unique benefit but poses a possible unique risk? We propose that rather than simply reacting to patient demand, manual therapists might use their skills to better persuade patients that lower risk alternatives are equally effective. We would also like to emphasise that the argument is not simply one of potential harm; as some responses have pointed out, many treatments carry risk. The issue is that although the risk is low, the potential consequences are severe, it is not possible to meaningfully mitigate this risk by screening and, as a standalone treatment, manipulation is not very effective and certainly no more effective than other treatments which are less likely to have such catastrophic consequences.
We also refute the allegation of bias towards any particular profession. Even a cursory read of the article would reveal that we have taken great pains to state that our comments are about a specific treatment technique only, one that is used by many different professions. At no point did we attack any particular professional group – on the contrary, we have made efforts to make this point clear and to correct some of the misquotes that have been attributed to us by various media outlets. It was not our decision to press release the paper, nor could we control the subsequent media coverage. We feel that we should be judged on what we published in the BMJ, not on how the media have chosen to interpret it. In the end, it remains our opinion that cervical spinal manipulation is unnecessary and inadvisable given the other options available.
1. Leaver AM, Maher CG, Herbert RD, Latimer J, McAuley JH, Jull G, Refshauge KM. A randomized controlled trial comparing manipulation with mobilization for recent onset neck pain. Arch Phys Med Rehabil. 2010 Sep;91(9):1313-8.
Competing interests: No competing interests