Written by In Part 1, the relative risks and important screening areas prior to cervical manipulation were discussed. In this second installment, the effectiveness and a theoretical framework for utilizing cervical manipulation will be laid out.
According to the Neck Pain Guidelines published by Cleland et al, manual therapy (mobilization and manipulation) was graded both an ‘A’ (strong evidence) and a ‘1’ (evidence from high-quality RCT, prospective, or diagnostic studies) in the treatment of neck pain. These guidelines advocate both manipulative and lower grade mobilization in reducing neck pain and cervicogenic headache symptoms. Published in the same year as the clinical guidelines, Walker et al conducted an RCT looking into the effectiveness of manual therapy and exercise in the management of mechanical neck pain. Those randomized to the manual therapy and exercise group (MTE) received treatments that ranged from cervical manipulation to soft-tissue mobilization and every technique in between. The intervention period lasted for 3 weeks with a total of 6 treatment sessions during this timeframe. At the conclusion of the study, those patients in the MTE group demonstrated statistically superior improvements in NDI scores at all 3 follow-ups, pain reduction at 3 and 6 weeks, perceived patient improvement on the GRC at all follow-ups, and overall treatment success rates were almost two-times as large for the MTE group in comparison to the minimal intervention group. These results bode well for a comprehensive manual therapy approach, but is there evidence to support the use of cervical manipulation in isolation?
Saavedra-Hernández et al recently published a RCT investigating the short-term effects of spinal manipulation in individuals suffering from chronic neck pain. In this study, patients were either randomized to an isolated cervical manipulation program (CMP) or a comprehensive manipulation program (cervical spine, cervicothoracic junction, and thoracic spine). After treatment (7-day follow-up), those in the CMP program demonstrated a decrease in pain from 4.8 to 2.7, NDI from 23.7 to 16.8, and increased ROM in all planes. While these values did reach statistical significance and this does lend evidence in support of cervical manipulation, those in the comprehensive program did achieve a greater reduction in disability than the CMP group. This lends support to a thorough manual therapy approach but does not necessarily discredit the use of manipulation. With so many options available to the treating clinician, how do we know when to use manipulation?
Tseng et al investigated what predictors helped to identify those patients who would demonstrate an immediate reduction in pain intensity, significant perceived improvement, and/or a high satisfaction level following cervical manipulation. After analyzing the outcomes of the 100 patients included and grouping them into either the ‘responder’ or ‘non-responder’ group, 6 variables were determined to be predictors of a positive response. These variables included initial scores on Neck Disability Index < 11.5, having a bilateral involvement pattern, not performing sedentary work greater than 5 h/day, feeling better while moving the neck, did not feel worse when extending the neck, and the diagnosis of spondylosis without radiculopathy. If 4 of these variables were present, the probability of success increased from 60% to 89% following manipulation. More recently, Puentedura et al conducted a similar study attempting to develop a clinical prediction rule identifying those who would respond favorably to cervical manipulation. Eighty-two consecutive patients presenting to an outpatient physical therapy clinic received manipulation of the cervical spine. Of the patients in this study, only 32/82 (39%) reported a favorable outcome. However, of those 32 who met the criteria for a favorable response, there were 4 variables that proved to be predictive. These items included symptom duration less than 38 days, positive expectation that manipulation will help, the side-to-side difference in the cervical rotation range of motion of 10° or greater, and pain with posteroanterior spring testing of the middle cervical spine. When > 3 of the 4 variables were present, the likelihood of a positive response increased from 39% to 90%. These are overwhelmingly positive findings, but it must be noted that these are merely preliminary findings and further supplemental research needs to be conducted in order to validate Puentedura and colleagues’ findings.
Many clinicians will argue that cervical manipulation should be abandoned in favor of less ‘dangerous’ interventions. The justification is primarily due to a lack of overwhelmingly superior outcomes compared to thoracic manipulation and/or cervical mobilization. While theoretically, these may be considered more conservative, are the outcomes really the same? Puentedura et al evaluated the outcomes of cervical manipulation versus those of thoracic manipulation in a small RCT. At the conclusion of the study, patients randomized to the cervical manipulation + exercise (CME) group demonstrated superior improvement in pain and disability in comparison to the thoracic manipulation + exercise group (TME). Additionally, there were no serious adverse events reported for either group, however, the TME group reported significantly more side effects than the CME group (8 side effects versus 1). So, can we honestly say thoracic manipulation is safer and equally effective? This study seems to disagree. While this was a very small study (n = 24) and all subjects were treated by one clinician, it still offers sufficient evidence to compare the two interventions.
The second intervention that is typically used in place of manipulation is a lower grade joint mobilization (grades I-IV). Dunning et al set out to determine whether a combination of thoracic and cervical thrust manipulation was more effective than non-thrust techniques. At follow-up (48 hours post-intervention), those in the manipulation group demonstrated superior improvements in disability, pain, atlantoaxial ROM, and motor performance of the deep neck flexor musculature. While this study only included data immediately following treatment, the benefits of manipulation seem to outweigh those of joint mobilization (in the short-term at least). This study is in agreement with an older study (Cassidy et al) that directly compared the short-term effects of cervical manipulation to mobilization. Both groups demonstrated similar improvements in ROM, but the manipulation group yielded superior improvements in pain intensity. However, contrary to the findings of these two studies, there have also been several studies that have found a lack of significant improvements between manipulation and mobilization. Hurwitz et al, Boyles et al, and Leaver et al all found a lack of discernible improvement in patient outcomes between manipulation and mobilization. Additionally, Saavedra-Hernández et al conducted a RCT comparing the effects of cervical manipulation to those of Kinesiotape in patients with mechanical neck pain. Both groups had similar reductions in neck pain and cervical ROM… This does not lend supporting evidence to Kinesiotape, rather it provides evidence against manipulation. Understandably, it is easy to use the findings of these studies as evidence to support the avoidance of manipulation, but what key concept are all of these studies lacking?
Cervical manipulation, actually manipulation in general, is not an intervention that should be used with every patient presenting with a particular diagnosis. There are specific sub-groups where manipulation is highly beneficial and then there are patients that will show no improvement with its use. Look at the Cochrane Review published on spinal manipulation for low back pain (LBP)… It was determined that, when analyzing all of the data, manipulation was no better than other inert interventions. However, based on the most recent developments in sub-grouping and treatment-based classification, matched treatment with regards to manipulation results in a positive likelihood ratio of 13.2. This is why one of the caveats included within the Cochrane Review is that more research needs to be done looking into the benefits of sub-grouping within this patient population. This is the route that cervical manipulation and its subsequent research needs to take (recent studies by Puentedura et al are a step in the right direction). As seen by the early results (improving probability of benefit from 39% to 90%), those patients who are likely to respond deserve to be provided the appropriate intervention. Not all patients need to be manipulated. In fact, not very many will need cervical manipulation at all, but to eliminate the intervention from those who will likely benefit seems counterproductive to me.
Based on the risk (albeit very minimal) and similar therapeutic effects from other seemingly less provocative interventions, I propose that cervical manipulation should be a potential intervention, but only for specific patient presentations. In general, I believe if you have exhausted other typical treatment approaches and the patient has not responded adequately, it may be time to consider more aggressive methods. My treatments are determined based on patient presentation and response to treatment, but if possible I try to progress from distal to proximal and from low grade to high-velocity low amplitude in terms of manual therapy interventions. In addition, motor control, stabilization, and direction preference exercises must not be left out. As seen in several studies throughout this article, benefits are far superior when coupled with appropriate therapeutic exercise… This is further supported by a systematic review conducted by Gross et al who found combined treatment (manual therapy + exercise) to be superior to any treatment in isolation. It is our job as clinicians to develop and implement a comprehensive program utilizing manual therapy, exercise, pain modulation, and our patient’s beliefs and experiences. Use best evidence and clinical reasoning to determine an appropriate plan of care and your patient will demonstrate far superior outcomes compared to any intervention in isolation.
Thoracic Manipulation → Cervicothoracic Manipulation → Cervical Mobilization (Grade III/IV) → Meet CPR? → Cervical Manipulation (Grade V)
Continued Reading…
John Snyder, DPT 
Let’s not move too fast on concretely tagging sub-grouping and CPR’s at identifying individuals who respond to a SPECIFIC intervention per se. Instead, it appears that the CPR for manipulation is likely more prognostic in that it identifies a sub-group with an excellent prognosis regardless of treatment content.
http://ptjournal.apta.org/content/early/2012/08/08/ptj.20120216.abstract
“Meeting the CPR was prognostic for all outcomes measures and should be considered a universal prognostic predictor.”
Further, this study is profound not just in results but in design: http://forwardthinkingpt.com/2012/10/10/ground-breaking-new-study-mobilization-and-manipulation-are-equally-effective-and-produce-the-same-outcomes-in-mechanical-lbp/
http://www.manualtherapyjournal.com/article/S1356-689X(12)00189-0/abstract
LikeLike
I don’t know where to start. Perhaps congradulation for the blog and the time and effort invested in it. But I can’t help to feel you miss the target with some of the arguments. First, the recent Dunning study that compares manips and mobs 48hrs later… What kind of an outcome is that? And Did you look at the methodology – the way they administered the mobs is no way near what is done in practice. And, needless to say, with Dr Dunning’s affiliations, the risk of bias here is HUGE. When we take more pragmatic studies that compared mobs to manip for the C-spine and for all the spine for that matter, they both end up being the same outcome wise. http://www.ncbi.nlm.nih.gov/m/pubmed/20801246/
Sure, we could say it wasn’t the right subgroup but most of the studies looking into predictors of response to HVLA thrusts/manips – basicaaly CPRs – either were not validated yet and are only predicting if the pts responds to the manips, not if he will respond better to the manip than the mobs. In fact recent work by Chad Cook who did a group of studies for LBP and manips actually shows that these manips CPR might simply highlight who are the patients with a good pronostic rather than who are the patients who respond to manips. Sure, they respond well to manips, but they respond well to just about any knid of HPSG (human primate social grooming). If so, why risk it with a manip if they will fare just fine without it.
The risk issue is a potentially long discussion and I don’t want to dig into the trillion references for either opinions but I will only say that with an under-reporting rate neighboring 100% for the severe vascular adverse events, none of the studies looking at the relative risk of Cx manips reach any valid conclusions with regards to risk. In fact, they all will probably underestimate the risk by a mile. The comparison with NSAID is certainly fair but the differences in methodologies in studies looking at NSAIDS and severe events and cx manips and severe events precludes any real conclusion. BUT, we don’t need to throw the NSAID arguments here as we are not here to prescribe NSAID but to apply a physical therapy treatment in which NSAID isn’t really among our true choices. And NSAID can and should have its own debate on risk/benefice in another discussion. It is not a matter of either/or manips vs NSAID but rather of what is best for our patient and present the less risk. After all I’ve read about pain and manual therapies mechanisms of effectiveness to relieve pain (mostly non-specific) I really, and I mean really, can’t think of a situation where Cx manips would be the treatment oif choice over anything else I could provide in a comprehensive manual therapy approach, not with the state of the evidence as it is.
LikeLike
I think the obviously tongue-in-cheek image heading this blogpost where Rambo is in the process of rupturing the blood supply to the bad guy’s brain stem speaks volumes.
I’m all for light banter and joking around, but this is a critically important issue facing us as a profession. I don’t think using an image like this is appropriate within the context of a serious discussion, and it makes me wonder if the author has a deep enough understanding and appreciation of the ramifications of this issue.
Real patients might come upon this blog, right? Is that what you want them to see, John?
LikeLike
John, I commend you on writing a blog like this. However, what I see missing in these two reviews, is the underlying model of moving joints further by either manip or mob.
Both these techniques assume that the human neuromuscular system under the hands of the therapist is unable to increase its ROM by itself, without outside forces.
Furthermore, the perceived lack of motion in the C-spine is not a condition that can be clinically and reliably shown to be a CAUSE of the patient’s complaint. At the best of times we can say it is correlated with the complaint; it may even just be a defensive, adaptive and protective finding.
The above points make the extensive search for subgroups, proper technique application, better studies and so forth, rather an expensive search for the holy grail: restoring the patient’s pain free function with a specific technique.
Outcome studies and subgroupings are part of our evidence base, but there needs to be a review of deeper aspects such “defect versus defense” and correlation versus causation, and the long term outcomes of specific manual interactions.
LikeLike
In performing any manipulative therapeutic procedure of the C spine, I was taught to always perform a test for Adson’s sign prior to manipulative intervention. If negative then to proceed. Also, I would like to point out that in any contemplated manipulation of the axial skeleton, one should always be aware of the vertebral subluxation complex. (VSC). The VSC consists of the vertebrae of interest, the one manifesting abnormal signs and symptoms. And the one immediately Superior to it as well as the one immediately Inferior to it. All three should be treated with highly specific techniques.per the developed treatment plan for this patient.
LikeLike
Dr Cochin, Applying your rules would be really interesting and usefull if the subluxation only existed in the first place. Unfortunatly it doesn’t…
http://www.ncbi.nlm.nih.gov/pubmed/?term=hill's+criteria+manipulation
I could put many other references to support my affirmation, whereas there isn’t much proof for its existence. So manipulation don’t serve any specific purpose with regards to putting joints back into place or into proper alignment. And it was shown that a joint position doesn’t change after a manipulation procedure. http://www.ncbi.nlm.nih.gov/pubmed/9615363
LikeLike
“Dr” Cochin’s comment about the subluxation is instructive on at least one other level than the one Frederic already indicated, which is that such a thing doesn’t exist. Thank you for reminding us, Frederic.
The other is a less conspicuous though perhaps even more pernicious aspect of the spinal manipulation debate and what drives it. And that is the “turf war” between the PT and chiropractic professions, the latter of which I think it’s safe to assume “Dr” Cochin is a member. Is this what we want to spend our precious research resources and blog articles on? Fighting with chiros over the right to do something that has been documented to kill people, who were unfortunate enough to enter a chiro or PT clinic with neck pain?
I think it’s an utter waste of brainpower and bandwith. Let’s talk about something else and let the chiros kill and maim people with the “vertebral subluxation complex” or “VSC”.
LikeLike
I’m not sure where to begin but here I go. The idea of PT’s or Chiro’s “killing” thousands with cervical manipulations doesn’t hold up. In fact the first study was disproven and it actually noted that you are more likely to cause someone to have a VBA stroke by visiting their PCP. The study (Cassidy et al. 2008 ) found that the billed visits by PCP’s for headache and neck complaints were highly associated with VBA strokes. There was No Evidence to support the precious study. Now on to VBI or VA testing. One study (Magarey et al 2004) found that “premanipulative testing alone is clinically unlikely to distinguish between patients with varying degrees of flow impedance.” This study also suggests that it was clear that greater emphasis should be placed on detecting risk or indicators of risk in the patient history. Mitchell et al. (2004) ,Kerry (2005), and Kerry &Taylor (2005),(2006), Thiel and Rix (2005), and Rivett et al. (2005) have all suggested that we should reconsider using these tests at all. NHMRC, 2003 is one study that speaks of the limited evidence in regards to manipulation but also states that there is even less for alternative interventions such as lower grade mobilizations. Also, someone states Dunning et al. only looked at symptoms up to 48 hours post treatment. So what? at least it was a start. While this may be true, at present there are several studies looking to build on this study with longer term data being collected. At this time, there is really no significant evidence to suggest that cervical manipulation causes more risk to a patient as opposed to non-thrust, or hands off therapy of the cervical spine. Because of this, we should all look closer at the history as well as BP testing, cranial nerve testing, and the Canadian C-spine rules for trauma instead of bashing each other. If you do not feel comfortable treating patients with cervical manipulations by all means don’t do it because you will put your patient and yourself in jeapordy. I am not a researcher. I am a full-time practicing clinician with more manual therapy and manipulation continuing education than I care to speak about. As for myself, I by all means do not perform manipulations on every patient but the ones that are candidates do very well with the treatment. We can’t get stuck in the past with our thoughts on certain treatments. If we as a profession would approach treatment techniques with an open mind we may just start to see that some are very valuable and should be used. I encourage all of you to broaden your horizons whether it be manipulation, kinesiotaping, active release techniques, instrument assisted STM, McKenzie, etc. Remember, MCKenzies original theory was disproven as well but hey, it works on a lot of people.
LikeLike
Dr Conlin, The mere fact that VBI and VA testing can’t adequatly detect flow impedance alone should make us wonder about using cx manips at all. Obviously, GP will see patients with stroke risks, people with stroke symptoms often seek medical attention. They migth seek chiropractic attention too, true. But since we can’t really compare the incidence of people with stroke related symptoms (prior to the visit) that go see a GP vs a chiroprator, that argument isn’t really valid. It migth be that people who have strokes after a manip really did have a VA issue beforehand, but that’s not the reason for us to give them the coup de grâce.
On the effectiveness issue, there as been many studies since 2003 to say that manip are not really superior than mobs when the latter were done in a pragmatic manner. There just isn’t any solid evidence for superiority of manips(anywhere in the spine). BTW, who are the candidate for c spine manips? No sign of any validated CPR to my knowledge. And when there are manip CPR, these CPR generally also predict good pronostic, so why risking it with patients who will improve anyway, regardless of treatment? I was highly trained in manual and manip therapy myself, And I can’t think of a single condition (that is valid anyway- I don’t count fixations or segmental dysfonctions as valid entities btw) where a c spine manip is required. Broadening my horizon to pain neuroscience is what lead me to such constatations, among others.
Perhaps, as for Dr Dunning ( his 48 hr study has a high risk of bias yet he declared no conflict of interest in the paper…), you may lack the detachement required to be critic about C spine manips as you migth have invested a lot of time and money into it, or perhaps not. But, I really don’t comprehend the current literature the way you seem to.
http://www.ncbi.nlm.nih.gov/pubmed/20518945
this is an excellent perspective by Dr E. Ernst utilizing the Hill’s criteria of causation on manips and stroke. I share his views.
LikeLike
I find it interesting that practioners who are willing to open a debate about spinal manipulation with the thought of bias in the “pro” research, are often citing Ernst, when multiple journals (Spine, New Zealand Med Journal) have reported a portion of his research as biased stating “Rather than uncovering bias on our part, Ernst has created his own bias with poor scholarship at the core of his comments”. With the current fight that our profession is fighting to keep the rights to spinal manipulation in many states, I believe it is irresponsible to challenge its efficacy within our own profession. One may not be able to do it effectively and safely or may feel uncomfortable with it but this does not discount its efficacy and should not be grounds to challenge the procedure. Also, while the current evidence does not overwhelmingly provide that either mobilization or manipulation is better than the other, the fact that some evidence does on a continual bases show substantial benefits from spinal manipulation means that we need to continue the scientific process and research when/where/why/how exactly do we get the most out of this treatment when involved in a patients care. As a profession if we do not come together to fight these battles we will be left with little to no access, limited treatment options and even less respect than we already have.
LikeLike
Well, it is to be expected Dr Ernst comments will such reactions by the one being scrutinized. Bu regardless of the messanger, many of his comments are worthy of consideration and have been evoked by others recently (Neil.O’Connell for instance). Now, One thing I will definitively disagree on, is this :
«I believe it is irresponsible to challenge its efficacy within our own profession. One may not be able to do it effectively and safely or may feel uncomfortable with it but this does not discount its efficacy and should not be grounds to challenge the procedure»
Rather than being irresponsible to challenge the efficacy of manipulation within our profession, I find it perfectly normal as I would challenge any intervention for that matter. Further more, the more potentially risky the procedure the more its efficacy should be challenged. I am really not fearfull this will discredit our profession, rather, I think it will show our brilliance and clairvoyance and our hability to question ourselves as would any good health care professionnals putting the patient’s well being way ahead of our own personnal or professionnal interests.
LikeLike
Everyone here seems generally to agree that the kind of stroke under discussion is rare. It therefore follows that the small subset of strokes temporally related to cervical manipulation are even MORE rare – almost stratospherically so in fact. Further as Mr. Ridgeway has explained, there are at least three ways in which manipulation may be temporally related to stroke – only ONE of which is that the manipulation definitely caused a stroke that would otherwise NOT have happened [herein lies one of SEVERAL pitfalls for “expert witnesses” giving evidence ‘against’ a manipulating therapist – causation is a scientific, philosophical and logical minefield – one can be so easily made look like a complete fool on the witness stand by an appropriately prepared Barrister – the problem is that only a few physiotherapists are adequately armed to brief a counsel to this level of sophistication).
To return to my main point: a small and unknown fraction of a tiny and undetermined percentage of a rare stroke may be ASSUMED to have caused – but may in fact NOT have caused [AND perhaps may NEVER be proven to have caused] the rare strokes under discussion. I live in Europe, so perhaps I’m ‘too far from the Centre of Gravity of the Scientific Universe’ over here, but I wonder if US Physical Therapy is unique in having so MANY (to paraphrase the great Churchill), discuss SO LITTLE, for SO LONG to benefit SO FEW, while using for support a body of research whose methodology many scientists consider to be of mediocre quality at BEST [i.e. rehabilitation medicine/Physical Therapy]. One MUST wonder: Qui Bono?
To Mr. Wellens I would say this by way of caution – you have rightly pointed out that there is latitude to identify a potential for bias in Dr. Dunning’s research because of his ‘interests’ – the same can be said (but I note wasn’t) for nearly EVERY other article published (there is ample evidence for this in the literature – itself also subject to bias) and EVERY contributor HERE. Biases in the selection of which articles to read and in how one interprets and criticizes them are VERY influential – being hubristic, we humans like to describe ourselves, not as biased, but as ‘informed’. One or two of the contributors here have ‘interests’ that oppose those of Dr. Dunning – how can we possibly say that Dr. Dunning alone fails to resist temptation? I know I can’t! But I know few with my reverence for honesty.
LikeLike
I forgot to say how informative I found the blog – thank you for writing it. I wish you well in your DPT.
LikeLike
I found some typos in my comment – that’s what I get for typing ex tempore and not editing before pressing ‘return’ – sorry for that: the corrected AND FINAL version for comment (or not) is below.
Everyone here seems generally to agree that the kind of stroke under discussion is rare. It therefore follows that the small subset of strokes temporally related to cervical manipulation are even MORE rare – almost stratospherically so in fact. Further as Mr. Ridgeway has explained, there are at least three ways in which manipulation may be temporally related to stroke – only ONE of which is that the manipulation definitely caused a stroke that would otherwise NOT have happened [herein lies one of SEVERAL pitfalls for “expert witnesses” giving evidence ‘against’ a manipulating therapist – causation is a scientific, philosophical and logical minefield – one can be so easily made look like a complete fool on the witness stand by an appropriately prepared Barrister – the problem is that only a few physiotherapists are adequately armed to brief a counsel to this level of sophistication).
To return to my main point: a small and unknown fraction of a tiny and undetermined percentage of a rare stroke may be ASSUMED to have been caused by manipulation – but in fact, manipulation may in fact NOT have caused [AND perhaps may NEVER be proven to have caused] the rare strokes under discussion. I live in Europe, so perhaps I’m ‘too far from the Centre of Gravity of the Scientific Universe’ over here, but I wonder if US Physical Therapy is unique in having so MANY (to paraphrase the great Churchill), discuss SO LITTLE, for SO LONG to benefit SO FEW, while using for support a body of research whose methodology many scientists consider to be of mediocre quality at BEST [i.e. rehabilitation medicine/Physical Therapy]. One MUST wonder: Qui Bono?
To Mr. Wellens I would say this by way of caution – you have rightly pointed out that there is latitude to identify a potential for bias in Dr. Dunning’s research because of his ‘interests’ – the same can be said (but I note wasn’t) for nearly EVERY other article published (there is ample evidence for this in the literature – itself also subject to bias) and EVERY contributor HERE. Biases in the selection of which articles to read and in how one interprets and criticizes them are VERY influential – being hubristic, we humans like to describe ourselves, not as biased, but as ‘informed’. One or two of the contributors here have ‘interests’ that oppose those of Dr. Dunning – how can we possibly say that Dr. Dunning alone fails to resist temptation? I know I can’t! But I know few with my reverence for honesty.
LikeLike
Just for the record, here are my own biases :
I was trained and certified in manual therapy – that includes HVLA thrust to the Cx spine.
I don’t believe HVLA thrust to be necessary anywhere in the spine.
I believe the effects manips have on pain are mostly (if not only) non-specific.
I do use HVLA thrust in the thx and Lx-spines, seldomly. I wouldn’t care not doing it anymore.
I am technicaly proficient in delivering manipulations anywhere in the spine.
I own my own physiotherapy clinic and doing or not doing c-spine manips would’nt change anything in my incomes.
I don’t profit in any way from the promotion of evidence against the use of s-spine manips aside than my own intellectual satisfaction and pride associated with the reputation of the profession I am associated with.
If C-spine manips where totally proven to be safe (which will likely never happen – my bias) I wouldn’t mind using them then, but I would still see no real incentive to do so nonetheless.
LikeLike
I wouldn’t be so rude as to question your training, qualifications or technical proficiency – I was merely trying to point out the slippery nature of Bias. My main point rests on the observation that by the very nature of our biases, we are unaware of them, and unfortunately, the ones of which we are LEAST aware are our most pernicious. Fortunately for some, with age, experience and with years of painful self-examination and much study, comes wisdom.
I have a question or two, and I realize that your ‘first’ language may be French, so bear with me for being picky for I am a lowly monoglot. If you don’t believe HVLA is necessary anywhere in the spine: 1. why do you use it in thoracic and lumbar spine? 2. do you use it for peripheral joints? 3. If the effects of HVLA are ‘non-specific’ – [though I’m not sure I fully understand this] why do you perform them in the T and L spine – or indeed anywhere?
I don’t think neck HVLA will ever be shown to be 100% safe – but neither is exercise, neither is cycling to the clinic – or even driving there. The Holy Grail of 100% Safety is a red herring: ask any physician who injects Z-joints (a much more dangerous procedure than manipulation will ever be), or any GP who prescribes NSAIDS.
Finally: and again I admit that you may not have intended to say what you said, in precisely this way: “I don’t profit in any way from the promotion of evidence against the use of s-spine manips aside than my own intellectual satisfaction and pride associated with the reputation of the profession I am associated with” – what you are describing here is not an innocuous bias; it is the intentional promotion of EVIDENCE AGAINST X. This is not the way scientific evidence is supposed to be used; it leaves you open to very high levels of confirmation bias.
Thank you for your response M. Wellens.
LikeLike
Mr Shortall,
Je l’admets, je parle principalement français! Effectively, “promoting evidence against” was a bad choice of word. It would have been better to say that “I do not profit from my current understanding of the literature about the C spine manips which is biased towards the procedure being too risky, nor do I profit by arguing in favor of this view here in this blog.”
As for the rest of your question – as you numbered them
1-I said “necessary”, which is not to be mistaken with sufficient. The nuance is pristine here. If I feel the context and the patient’s system of beliefs are right, I might use it… seldomly. My approach is parcimonius – but never in the c spine.
2-For the same reasons, yes and really seldomly.
3-Really, I understand your question but to be good and have merit, my answer would be too long. I could direct you to this review I wrote : http://physioaxis.ca/The%20Traditionnal%20Mechanistic%20Paradigm%20of%20Manual%20Therapy%20-%20Time%20for%20a%20Reality%20Check.pdf.
Or I could add that I think a manipulation need not to be specific to a joint to be effective and its effects on pain are not specific to the fact a manipulation, as a technic, was employed per se but rather to innumerable factors among which patient expectation, context, the impression of mastery felt by the patient, novel stimulation, … But again, that is really too shallow of an explanation to really make my point. So I invite you to read the review!
Again, the point about other interventions being as risky is besides the point as I am not the one providing Facet blocks and I am not the one driving to the clinic. These are not factors I need to consider when I decide to provide a manual treatment. Anyway, Patients need to drive to the clinic both for a mob and a manip. So I argue my net risk is lower 🙂
LikeLike
If HVLA are in your own words ‘Too Risky’ at what stage do procedures become ‘Too Risky’ rather than being ‘Acceptably Risky’? Specifically, in the event that in the future it becomes evident that whereas HVLA result in (for example) one stroke in 100 million manipulations and ‘mobilizations’ or MWM with (for example) 2 strokes per 100 million ‘mobilization processes’ (as Mr. Ridgeway puts it), would you still reject the former and accept the latter?
Naturally I think you are wrong in arguing that the fact that other interventions are not so risky: after all, it is precisely to these other interventions that patients will be directed. And one could argue that ‘mobilization’ is less risky, but as I pointed out earlier, we are dealing with stratospherically rare incidences here – in either case.
Physiotherapists (and doctors) claim that it is important to explain ‘risk’ of procedures to patients but few actually do it in practice. One reason for this is that they do not understand the relative nature of ‘risk’ – thus, simply telling a patient that e.g. HVLA is a ‘risky procedure’ is to promote a lie – one should give the patient and idea of what this means in relative terms. Thus my example of ‘cycling to the clinic’ (though ‘tongue in cheek’) is highly relevant. Perhaps we should tell patients that there is (for example) one chance in a million that a HVLA will precipitate a stroke (these numbers used only for illustration) whereas if they opt for (again only for example) a facet joint injection, their chances of very serious side effects are 100 times greater and if they opt for a surgery (for example, fusion) perhaps 1000 times greater. I suspect may is a degree of irrationality and hysteria amongst the ‘anti-HVLA’ lobby – this is MY bias. I will read your paper as soon as I can.
In any case – I enjoyed talking with you. Thank you for your response. Merci.
LikeLike
On a final note, I will concede the risk of serious side effects after a c spine manip is low. Yet, the real risk is unknown, some say the under reporting rate neighbors 99%. While this may be exagerated – or not, it makes the stratospherically rare qualification a bit of an optimistic statement in the face of a serious side effect such as possible death, especially with the lack of pristine evidence in favor of manipulation or for the whole manual therapies for that matter.
I will concede that many other interventions share similar or perhaps higher risks – read for example surgeries – but I’m no fan of these either. I’m all for less medicalization.
Would the manips be shown less risky than mobs, well yes, I would consider using more of the former, although, I would argue that the way I do perform mobilizations that is really unlikely to happen.
Give me your thoughts on the paper, my email is on it.
Enjoyed the talk too. Regards,
LikeLike
Having witnessed dozens (probably hundreds) of unnecessary surgeries I too am in favour of less medicalization – particularly with respect to low back pain and chronic pain sufferers. Some surgeons seem to be unaware that there are profound ethical issues here.
Despite your mention of the possibility of a “99%” under-reporting rate, I stand by my rough estimate of strokes directly caused by HVLA as ‘stratospherically rare’ – mathematically, and from a clinical risk perspective, there is no significant difference between 1 in a million and 2 in a million (despite the 100% difference!). And irrespective of whether the “true” value is 1, 2, 3 or 43 in a million I still maintain that in many of those claiming to ‘count’ the serious side effects, the risk of bias is HUGE because of their vested interests. The number of supposed ‘strokes’ actually DUE to a HVLA has not been reliably counted; this is why I said discussions like this seem so much like “so MANY, talking about so LITTLE, to benefit so FEW, using for support mediocre research that is likely to be drenched in BIAS”.
LikeLike
M. Wellens, I actually had your article at work and read it a while ago, adding copious notes (mostly remarks of my approval!) of my own in the margins and on the reverse blank pages! I would not characterize it (as you do) as a ‘review’ in the sense that that term is used in modern science/clinical writing. It is an opinion piece – an argument in support of your stance on manual therapy and the biomechanical model with your point of view supported by selected (and well selected) articles and reviews. I was so impressed by it that I pinned it to the noticeboard at work for my other colleagues to peruse during breaks in the hope of sparking a debate there. Well written mostly!
However, it is an ‘opinion piece’ with articles selected to support your opinion – it is therefore a platform for confirmation bias (in addition to all the other sources of bias on which ‘confirmation bias’ rests: e.g. selection bias, interpretation bias and so on) – this is one of the reasons why, though they are much more gratifying to read AND write, this kind of ‘review’ has fallen out of favor since the 90’s.
As I said, I enjoyed your article and I actually agree with much that is in it, but I think like most (perhaps all) of the contributors to this debate, you are unaware of your own biases – though I noticed several references above to OTHER people’s biases so the notion of bias (per se) is not foreign to our contributors: just the notion of bias as it applies to themselves, research readers and Prime Tauro-therapists. My reason for entering into this comment stream was to attempt to highlight this very tendency – but self-examination takes courage and honesty. I feel the debate with respect to neck HVLA is always so emotional that histrionics, bias and personal insults (as seen in the rude treatment of a Chiropractor who contributed above – the poor deluded man probably thought he was contributing on a professional level) usually predominate. Strangely enough when I have such debates ‘in the real world’ (face to face) as it were, the aggressive Alpha Male behavior that one sees online is nowhere to be seen. I deduce from this that while internet communication has coarsened us all, it has made none of us more courageous.
With luck, my efforts to highlight the possibility of bias and unhelpful ’emotion’ in THIS debate will bear fruit in some future debates. I feel there is no possibility of a professional debate on HVLA of the neck because the same Alpha Males (or their proxies) will rampage on and trot out the same material each time.
LikeLike
M. Shortall, Thanks for your remark about the article. It is obviously more of an editorial than a review, I concede, although it’s done to review two models for the effect of manual therapy. About the biases. I obviously have my own biases and I am well aware of many of them. Opinions are fraught with them. Although, I choose to use reviews supporting my point of view to make an argument, I am well aware articles that seek to support the opposite position do exist. I also know that, as anyone else, I can be prone to confirmatory bias, I won’t negate that. I enjoy stumbling on articles supporting my opinions yet that doesn’t mean I am blinded to evidences against my own views. My article was written as it is for a single purpose and that is to be read and discussed about in an almost purposefully black and white dualectic way that would spark debates. I try to be more nuanced in my head about these things. But having your own biases doesn’t revoke one’s right to highlight the biases of others, especially when they are obvious and (willfully?) undisclosed in a peer reviewed publication where it is common practice to do so.
All in all, it seems we agree more than we disagree on many things. And surely the discussion on HLVA in the c-spine is far from settled and the jury is still out – so to speak. One thing though, I now know what you think of the debate about that but I still am unsure whether you think manips to the c-spine are really useful and should be used on a regular basis or not. Perhaps, you are comfortable with uncertainty as to both its effectiveness and safety and thus feel this debate has simply lasted too long!
LikeLike
I prefer articles written in the style you chose for yours – better reading and more likely to spark a debate; more provocative. And we agree on many things. The HVLA ‘debate’ just seems to bring out the worst – perhaps because it rattles the cages of fervent followers of the NAIOMT, the APTA and other ‘interested parties’ (parties who, like the patently superfluous and irrelevant Licensing Boards, always seem to have their hand out for OUR money. With a national exam, a Licensing Board is simply legalized extortion), and these formerly august bodies may have noticed that the world is moving along quite nicely (and disturbingly fast!) without their imprimatur. Where,( they may feel), is their rightful supplication and their comforting “jizya”? Perhaps they feel guilt, because some of our ‘professional bodies’ left individual physiotherapists alone and un-funded in the battlefield against for example chiropractors? No matter why… this ‘debate’ triggers more sub-par behavior than usual. Mention HVLA therefore, and the ‘Dogs of Debating’ are ‘let loose’.
I agree that having biases of one’s own does NOT negate the right to point out perceived biases in others – however it DOES leave one open to charges of hypocrisy, most especially if (as always happens on blogs like this) one does not balance the argument by giving at least equal attention to not only one’s own actual biases (and I don’t mean just ‘conflicts of interest’ – I mean scientific, cultural, clinical biases) but those of ALL the other authors whose research is cited. Clearly one cannot select a single author (with whose approach one does not happen to agree) and declare his bias to be “HUGE” while ignoring the fact that this off the cuff assessment is in itself an example of bias, and that the authors of ALL of the other articles cited may be biased in their research too – this would have elevated the comments here to a ‘debate’ I think.
In bias, as in other sins, ‘let he who has not sinned, cast the first stone’, and as this is not a woman buried up to her neck in the ground so she can’t defend herself, stones thrown may be thrown back, so beware.
And I repeat what I said a few comments ago: the biases of which we are unaware or that we will not admit to in public (or in private; or even to ourselves) are the most influential and pernicious ones.
For many reasons I have resisted the ‘lure’ of entering the debate about HVLA of the neck – I wanted to start a debate about ‘debating’ instead in the hope that what Mr. Reagan termed ‘trickle down’ would occur, not of wealth, health and handsome children this time, but trickle down of just enough self-examination to imbue future debates with less Digitally Expressed Testosterone, and a lot less sarcasm leveled against non-herd members like our Chiropractor contributor and the PT chap who ‘argued against’ the tide.
As for HVLA – I will do as I have always have done: I will read the relevant literature as it comes out; I will evaluate whether or not it is rubbish, relevant or just more mundane PhD/DPT ‘publish or perish’ resume padding (we’ve lots of that now), and I will either permit it to influence my clinical practice, or not, as I see fit. The reality is that Rehab/Physio research is simply NOT precise enough to justify the formation of distinct battle lines like those in evidence in this ‘debate’ – physiotherapy is not, and never will be Quantum Mechanics – hell, it’s not even Geology. Scientifically, we’re in the Penny Arcade – a very important fact for our chief blogger and new graduates like Kyle Ridgeway to learn at this early stage or their careers. The more ‘humble’ the science, the more important it is to portray scientific humility. The strident stances and aggressive posturing in evidence here (and elsewhere) are scientifically insupportable: they are expressions of the people involved, not representations of the quality of the science they espouse. What I see on most of these debates is not science, I see Alpha Male chest beating and ‘my fact is bigger than yours, so there’. ‘Fact Flashing’.
Each individual clinician must decide what is best. That seems to be what you do; and it is what I do – and I very much doubt that either of us could sway the other very far. I doubt anyone on any of these blogs reads research so much more expertly than I do (judging by some of the contributions – especially the sarcastic and gratuitously rude ones) that I would be left speechless by their erudition. One of them, I won’t embarrass him by saying his name, even challenged another contributor on a blog by telling him that HE had a DPT (I could almost hear the poor boy chanting:’so there, na na na na!’). I feel the ‘best of my profession’ is not on show in these debates. The main reason why I almost never take part in them. And they take such a lot of time! I’m off now to read an article in Nature on Myokines. Much more my thing than Alpha Male Fact Waving.
I did enjoy your article and your responses, thanks
LikeLike
Thanks mr Shortall, it was crystal clear!
LikeLike
Mais, bien sur. Merci beaucoup M. Wellens. Au revoir.
LikeLike
Post hoc ergo propter hoc? Black Cat – Bad Luck? Can’t we do better than this?
http://www.ncbi.nlm.nih.gov/pubmed/%2021923248
LikeLike
Taleb would call it a black swann. It could be just a correlation, taken in itself to argue against cx manips would be a post hoc fallacy I agree. Taleb also says many interesting things about randomness. Like for instance, if you play russian roulette with a stratospheric number of bullet chambers you will surely make a lot of money and think you figured out how to win at russian roulette. But, eventually, if you keep playing and playing, armed with such confidence, that black swann may show up. We all know the outcomes of that. You don’t get more chances to win back your earnings after your first and only loss. Despite the odds of losing being so small, we could say its relative weight against the odds of winning is much bigger. You can win all you want but the first loss ends the game.
That’s a bit how I see cx manips. There is surely a fair amount of randomness behind the vascular events post manips, patients might need to have a specific profile AND the practician migth need to have a bad day AND what else? Who knows? I don’t care, it’s enough for me not to take the risk. Not with an intervention with this low grade of evidence in its favor.
LikeLike
Sorry for the delay in responding – I was ‘away from my desk’ as it were – you make a good point. It’s been a long time since I read Nicholas Taleb’s book, but I believe he has a different concept of the Black Swan than you have. You seem to have characterized it as a highly rare event: but that’s just ONE characteristic of a true Black Swan.
Most of his examples (I can recall e.g. 9/11 and Market ‘Crashes’) are NOT like strokes ‘after’ HVLA. The former (and this was one of the central messages of Taleb’s concept) examples involve incredibly complex INTERCONNECTED (non-mutually independent) systems rather than just a large number of discrete events. Indeed given that 9/11 was an obvious expression of Blowback, even IT isn’t unrelated to the behavior of ‘Wall Street’! Thus (in a somewhat ‘fractal’ deconstruction) one could theorize that even both of these events are merely ‘data’ in a larger and more catastrophic future ‘Black Swan’. The reason there is practically no terrorism in Canada for instance is that you don’t kill people in other countries and you don’t manipulate the economies of other countries for your ‘interests’. INTERconnectedness (‘interference’) is key to the Black Swan thesis.
Unless one believes in a concept like ‘collective unconsciousness’ (?Jungian?) then one must view patients as distinct, non-interdependent entities – therefore lacking the primary driving mechanisms for a Talebian Black Swan: interconnectivity, interdependence, summation, interaction. No treatment of patient A will predispose patient B to a stroke ‘after’ a HVLA, in the same way that say US manipulation of the price of (dare I bring this up! LOL) Bananas in Central America would affect the feelings (and perhaps the livelihood) of a peasant in Iran toward the US. In fact even successive HVLA on the same patient may not interact as strongly as this. So I feel you are misinterpreting Taleb’s thesis.
You argue that there is no difference in efficacy between HVLA and mobilizations – I disagree but I don’t as yet have a ‘Fact’ to ‘Flash’ and my conclusion is that the jury is still ‘OUT’ and the research is simply not a good enough tool to demonstrate the effects I see clearly in my patients on a daily basis. I have a ‘bias’ for empiricism I suppose. Scientifically, as I said above, much rehab. research is ‘mickey mouse’ … so I’m waiting for improvements.
LikeLike
An excellent collection of resources and articles curated by Matthew Rupier: http://www.somasimple.com/forums/showpost.php?p=153286&postcount=1
LikeLike
Reblogged this on Chiroresearcher and commented:
And here’s part 2 of John Snyder’s blog post on cervical manipulation. Again, the comments are also worth a read.
LikeLike