Thursday, May 22, 2008

Adverse Events of Chiropractic Care: Transparency Now!
So I've been binging on literature lately....One of my first courses at Texas Tech's doctoral program is Advanced Clinical Practice for the Cervicothoracic Junction and Thoracic Outlet. Let me be the first to say I had no idea this part of the body would be quite so involved and interesting at the same time.

Our weekly assignments involve reading research ranging from randomized clinical trials to anatomical reviews. After only one week in this program, I've already picked up some good information I can use in the clinic.

Today's article came from reading some anatomical studies on the vertebral artery. In reading more about the anatomy of this area, my mind wandered to the topic of potential risks for cervical manipulation. The most dramatic adverse events seem to focus on dire occurrances such as stroke. However, we've all seen folks come to our clinics saying they had a range of responses to chiropractic care ranging from pain and stiffness to a worsening of radicular symtoms.

I wondered in particular if the chiropractic profession was any more transparent with reporting adverse events pertaining to cervical manipulation. We've all heard the chiros tell us a manipulation has fewer complications associated with it than taking an NSAID, but we've heard this boy cry wolf before and a little hard data would be nice.

Sure enough, a trial spawned out of the UCLA Neck Pain study attempts to help us learn more. In a randomized clinical trial, a total of 336 patients with neck pain were randomized into three groups of chiropractic care:

  • Manipulation with or without heat
  • Manipulation with or without electrical stimulation
  • Mobilization with or without heat or electrical stimulation

The adverse event in this trial was "discomfort or unpleasant reactions from chiropractic care" assessed at 2 weeks from the baseline assessment. Of the 280 patients who responded, 85 patients (~30%) reported having one or more adverse symptoms as a result of chiropractic care. A total of 212 adverse reactions were reported from the 85 patients who had complications. Of the 212 adverse reactions there were:

  • 70 episodes of moderate to severe neck pain or stiffness
  • 44 episodes of moderate to severe headaches
  • 28 episodes of tiredness or fatigue

Other less common adverse events included dizziness, nausea, depression, tinnitus, arm or leg weakness, blurred vision, confusion or disorientation. Nearly 20 percent of the respondents reported the adverse events had a significant impact on their tolerance for ADL. There is plenty of good data within the results and discussion section to read over so I would encourage you to take a look through it.

The authors, one of whom is a chiropractor, concluded the following:

  • Adverse events from chiropractic manipulation are common
  • Adverse events are more likely to follow manipulation vs. mobilization
  • Chiropractors should consider mobilization over manipulation in the treatment of neck pain, particularly for those with severe pain.

I have bad news for the chiropractic profession... If you take the judicious application of manipulation based on a medical vs holistic model, you get what many well-trained manual physical therapists provide on a daily basis. This study does not bode well for a chiropractic profession that is very slow to let go of its roots in subluxation.

I don't take this study as an indictment of the chiropractic profession, as there are some very skilled manual practitioners out there. However, they dug their own hole in the form of subluxation-based care and are going to have to work hard to dig themselves out. It will be interesting to see if they can do it before the clock runs out on their profession....and believe me it is ticking. Chiropractic tuitions have never been higher and their revenues have never been lower...tick tock...tick tock....!

Hurwitz, E., Morgenstern, H., Vassilaki, M., Chiang, . (2005). Frequency and Clinical Predictors of Adverse Reactions to Chiropractic Care in the UCLA Neck Pain Study. Spine, 30(13), 1477-1484.


  1. I am curious why this trial didn't include a 'sham' manipulation as a control group. Wouldn't any study without control be limited in its application? Also did the levels of pain go up or down after manipulation, the information you provide merely indicates the pain levels after care not before. Something I would also be interested in is what type of diagnosis or complicating factors were present in the individuals and what technique was employed in treating the patients.

  2. I'll agree with the "Sham" manip point, that would have been helpful. However, as those that were included were randomized, it could be relatively safe to assume radomization blunted the effect the Dx or complicating factors will have.

    Finally, the technique used (as studies have shown in regards to the nonspecific nature of manipulation) shouldn't have an impact on outcomes. I'll admit it may on pt comfort, but why would a painful technique be chosen?

    A much more important question (IMO) does PT treatment have lesser reports of "Adverse effects". I'm not sure. I have a lot of people complain of "soreness" of "fatigue" after treatment.

  3. 91

    You bring up some good points. A sham manip would have added strength to the study. One of the limitations of this study was indeed the limited follow up of two weeks, but the stated purpose of the article was to report acute adverse events following chiropractic care. In this case the data reflects these adverse events.

    Factors associated with adverse events included: Pain >8/10 during presentation, history of trauma, pain less than one year,NDI >16, lack of confidence in treatment, nausea, and moderate to severe HA.
    This would indicate the need for more discretion with choosing the best circumstances for applying a manipulation.

    The manipulation protocol was as follows: "Subjects assigned to one
    of the spinal manipulation groups received at least one controlled
    dynamic thrust, applied with high velocity and low amplitude
    force, with minimal extension and rotation directed at
    one or more restricted upper thoracic and/or cervical spine
    joint segments within patient tolerance."

    The mobilization was similarly standardized. I hope that answers your questions. I wish I could share the entire article, but I'm not sophisticated enough to know how to do that at this point in my blogging career.

    Jason your question is addressed in the study. Reported adverse event rates for most non manipulative physical therapy interventions are lower than for manipulation. However, the adverse events associated with either manipulative or nonmanipulative manual interventions remain lower than for pharmacologic or surgical interventions according to the authors (one chiro, one allopathic physician, Ph.D.)
    My take home is that chiros often stand on the reported adverse events for medically based treatments to support their interventions. Articles such as these bring some balance to this debate.

    The author's essential message is for chiros to choose appropriate patients for manipulation. Unfortuantely if they do this, they loose a tremendous revenue stream from patients they shouldn't manipulate.

    Well trained PT's tend to be much more discriminating as to who we choose to manipulate. Thanks for both of your questions and comments. I hope I could answer them.

  4. Roderick Henderson, MPT, OCS, MA, CSCS,

    While I have no direct way to prove this I think that DCs often make note of negative effects in their SOAP notes. The questions are how should these be tabulated and reported and to whom? The ACA, ICA, WCA? I don't think they get along well enough to trust each other with the information. Even if insurance companies were the compliers of the data I don't know that these results would find their way into literature.

    "The author's essential message is for chiros to choose appropriate patients for manipulation. Unfortuantely if they do this, they loose a tremendous revenue stream from patients they shouldn't manipulate."

    I think you may have some bias with this line of thinking. Since I worked in a DCs office for over a year and am now in my second year at a prominent chiropractic college
    I may be biased as well, but I do have first hand experience with the subject at hand. And as we are told in class "the most important thing in adjusting is knowing when not to adjust vs when to adjust." I personally prefer the term manipulate but I think the idea is true.

    Is there the mentality of "when all you have is a hammer, everything looks like a nail" in practice but given what I have seen personally it isn't as common as some would guess.

    Plenty of times I have seen DCs refer out for things they either didn't feel equipped to handle or something they felt that manipulation and the small number of physiotherapy would be ineffective for.

    Perhaps my personal experience is not the norm but unfortunately it is all I have to base my opinion on.

    Jason L. Harris, PT, DPT said...

    I would disagree with your statement that "Finally, the technique used (as studies have shown in regards to the nonspecific nature of manipulation) shouldn't have an impact on outcomes." I have personally recieved several different types of chiropractic treatment (from diversified, Gonstead, Activator, SOT, LBT, and Proadjuster) and have had differing post treatment effects from each. Some I felt no better or worse, some I felt much worse, and some I truely felt better post treatment.

    While 2 of these are not HVLA treatments they did produce differing results from each other, despite having a similar theoretical mode of action. The other HVLA techniques have also produced differing results and have definitely lead to my adoption of a preference to learn and use myself, as well as which techniques I prefer to have performed on me.

  5. 91

    I can't argue that I don't have bias. It sounds as if you might be a new breed of chiropractor in choosing your manipulations based on clinical presentations. Like I said in my editorial comment, chiropractors like yourself may be the reason your professon remains solvent. However based on the manipulations performed in this study there is still an undercurrent of poor judgement in the use of manipulations regardless of presentations. Many who were manipulated in this particular study would not meet many practitioners manipulate (high pain, headache,nausea, high disability, acuity). Yet these patients were manipulated.

    You are probably going to be one of the chiropractors fighting the good fight, but I'm not sure if it will be enough to counter the ACA, WCA, ICA's reluctance to let go of Palmer's initial theory.

  6. And to get back to chosen technique and area for manipulation. I found that article (done by chiropractors) showing that whether a DC choosing the type and area of Cx manip by endfeel or a computer randomly assigning the technique and area, outcomes were the same.

    This just adds to the pile of evidence that manipulation is NON-SPECIFIC and no special super powers and 45 techniques are needed to effectively deliver and get optimal outcomes with manipulation.

    Here is the reference:

    Haas M, Groupp E, Panzer D, Partna L, Lumsden S, Aickin M.

    Efficacy of cervical endplay assessment as an indicator for spinal manipulation.

    Spine. 2003 Jun 1;28(11):1091-6; discussion 1096.

  7. Jason L. Harris, PT, DPT,

    While I do see the article does indicate that end range motion feel/motion palpation shouldn't be used as a specific indicator of cervical spine dysfunction, I don't see any mention about the effects of differing adjusting/manipulation methods. Although this is a small sample size study (N=104) and more research is probably needed before further acceptance is warranted.

    What it also indicates is that manipulating 1 segment isn't the only way to help a person (which makes sense since the tract of Lissauer asends/descends multiple levels before entering Lambda 1&2 and the mechanoreceptive mediated release of GABA initidated by manipulation/mobilization will reduce the centrally excited cord level). But I don't see how that indicates that all techniques are the same.

  8. -91

    These are obviously speculative comments here, but I don't think all the techniques are the same. However many of the physiologic effects are similar. In my mind this leaves room to determine if the effects are similar, how do we choose the best method? Safety and rate of adverse events would definitely be yardsticks to use in determining this.

    If cervical mobilization and/or thoracic manipulation are found to be equally effective at managing the outcome, it makes it difficult to consider using cervical manipulation to achieve the same goal.

    I agree with you that there are times when a specific mobilization may be more clinically useful than the generalized techniques. I think studies like this one are attempting to sift through the advantages and disadvantages of these approaches. As always I apprecaite your comments.

  9. 91z4me,

    I said it was another piece to the PILE of evidence that shows manipulation is non-specific. While I'm all for attempting to be as specific as possible, aiming for, say, C5 with a manipulation likely means at the very best you will get C5. In all likelihood (again from RESEARCH) you'll be manipulating C4 and C6 and maybe not even C5 by only above that or below that.

    Here's the great thing, doesn't matter. Because that PILE of research is also showing that NON-SPECIFIC manipulation still gets great results in the right population.

    Heck, we even know the cavitation isn't important in getting optimal results. In fact, I'll have to dig around, but another study showed that the cavitation most frequently occurred at a distant site from the level attempting to be manipulated.
    Like last time I'll have to dig for it so you don't have to take my word on it.

    Finally, how did you come to the conclusion the an N of 104 is a "small sample size"? Obviously in the end, it's the power not the number that is important. The sample size is just 1 component to determining the power of the study (thus decreasing the chance of a type II error). One can make a generalization that the higher the n the higher the potential power. However, this does not need to be the case, especially if there is a large effect size.

    In fact the authors address sample size:
    "The sample size (n  52/group)
    was selected a priori to detect a 10-point difference between
    groups in the primary outcome with 80% power for a twosided
    test set at the 0.05 level of significance"

    So, they (like good researcher do) determined beforehand what sample size would be needed to get a large power. Therefore, in this case, n of 104 IS NOT too small of sample size.

  10. Here are those reference regarding manipulation and cavitation level:

    Beffa R, et al. "Does the adjustment cavitate the targeted joint? an investigation into the location of cavitation sounds" Journal of Manipulative and Physiological Therapeutics Volume 27, Issue 2, February 2004, Pages 118-122

    Conclusion: Location of cavitation sounds does not appear to have a relationship with type of manipulative technique selected.


    Ross JK, et al. "Determining cavitation location during lumbar and thoracic spinal manipulation: is spinal manipulation accurate and specific?" Spine. 29(13):1452-1457, July 1, 2004

    Conclusion: In the lumbar spine, SMT was accurate about half the time. However, because most procedures were associated with multiple cavitations, in most cases, at least one cavitation emanated from the target joints. Inthe thoracic spine, SMT appears to be more accurate.


    I'm a little more pessimistic in the last article than the authors. I'd say that you are lucky to get cavitation at the selected site of manipulation only 50% of the time, and even then you are very likely to get cavitation at other non-targeted levels.

  11. Jason L. Harris, PT, DPT,

    Thanks for the articles, I will read them later when I have more time.

    I am really enjoying this discussion we have going on here guys.

    By any chance do either of you have an opinion on manipulation and its effects on Asthma? I have to do a collaborative paper on the effects of chiropractic on pediatric asthma and while I have seen some results indicating positive (though less than MMI under traditional asthma treatment) results I don't have much research at this point. Any info you have would be greatly appreciated, also what would the standard of care by a PT or DPT be for this condition?

  12. Interesting post, thanks for sharing. I do believe in chiropractic care but yet there are some side effects too, which needs to be taken care of.