the following is an actual transcript of the chirosecure live event … · 2017-03-22 · dr. clum:...

The following is an actual transcript of the ChiroSecure Live Event with Dr. Stu Hoffman and Dr. Gerald Clum. We

do our best to make sure the transcript is as accurate as possible; however, it may contain spelling or grammatical

errors. If you have any questions about the transcript or would like to request any transcripts from our previous

shows, feel free to contact us at 866-802-4476

Dr. Hoffman: Hello everybody. Dr. Stu Hoffman, President of ChiroSecure. The intent of this show is to offer up some information to you on what happens when the stroke issue rears its ugly head. What is the proper way to respond? Utilizing the literature that is available to us and not to just let our emotions get in the way of good, sound judgment. In fact, when we see an article appear or a news show appear that's attacking what we do as chiropractors and we know the information is false, do we just go and respond? I don't know about that. I think that we need to respond, but it needs to be based on sound information.

Here we have an article published at Baylor University Medical Center Dallas, Vertebral Artery Dissection after a Chiropractic Neck Manipulation, and they reference an estimated 1 in 20,000 spinal manipulations results in a vertebral artery aneurysm or dissection and ischemic infarct. We all know this is not true, and the facts are absolutely wrong, so I asked Dr. Clum to come on again for us and walk us through the right way to go about letting the university and the author know that their article referenced statistics that are incorrect and not at all based on current research and literature that we have so readily available.

I want to bring Dr. Clum on, and our goal is to get inside your head, Dr. Clum, and understand how you approach these situations because you did respond to this on numerous occasions, and I know that's what you're going to bring us up to speed on, and how do you decide when to respond and ultimately what to expect the outcome to be? Can you tell us first, how did you even learn about this article from Baylor?

Dr. Clum: Thanks, Stu. I appreciate it very much, and it's great to be with you again today. The Baylor article came to light in a very odd fashion. There was a promotional piece that was produced by the Vanderbilt University Medical Center last year. It was talking about their neurology department and their helicopter air flight service and that sort of thing, and they used the example of a case of a woman who had a vertebral artery dissection. She was in the chiropractor's office. She had the vertebral artery dissection, and it was recognized. The chiropractor took the appropriate steps, made the appropriate emergency calls.

The woman wound up going to the Vanderbilt University Medical Center by way of their ambulance service. The neurology folks did their job. It had a very good outcome, and the woman returned to her family and to her life and, essentially, it was a feel-good story for everybody from Vanderbilt's perspective.

In that article, they made reference to this 1 in 20,000 number that you just spoke of. I thought I had an understanding of where they came up with that

number, I also understood that number that appeared in the literature to have been an opinion piece. I contacted Vanderbilt and, sure enough, Vanderbilt said, "Yeah, we got that from this article in Baylor that you referenced at the start of the discussion today," Stu. I said thank you, went back to Baylor and asked them about it.

First, reviewed the article carefully. The first step in the process was, in reviewing the article, if you understand the literature and so on and you see the references that they're making in the article, are the references being applied properly? In this case, they weren't, and the references in the article that they had tied to their assertion that an adjustment resulted in a dissection in 1 in 20,000 cases was completely inaccurate.

At first step, I wrote them and pointed out the fact that the evidence they offered, i.e, the reference, didn't support the statement they were making. They wrote back, and they said, "Thank you. We appreciate you pointing out the error. We'll change the reference. We did, in fact, make a mistake on it, and this is what we based it on." They based it on a 1989 article from the British Medical Journal by an author by the name of Vickers. The unfortunate part of the Vickers article is that it's an entirely opinion-based statement. There's no data. There's no study. There's no control group. It's just an opinion.

It's kind of like, Stu, how often do people move to Phoenix? A hundred a day. You don't know that, and you don't have any basis to say that. There are people that do know that and can tell you that number, probably down to the family, but you don't know when you say that. This number was tossed out, this 1 in 20,000, as an opinion piece. All of a sudden, it gained a life of its own.

Baylor came back, and they said, "By the way, there's two other pieces of literature that we'd also like to add to this discussion," and they came out of Australia. One was by a fellow by the name of Dunn that was in 1990, and then there was another by a fellow by the name of Mann. When you looked at all three of them, you soon realized that they got into this circular argument where each used the other to reinforce its opinion.

Mann relied on Vickers, and Dunn relied on Mann and Vickers, so this thing starts to have a life of its own where this number of 1 in 20,000 pops up, and then it gets reinforced by the next person using it and the next person using it, and nobody goes back and says, "Hey, wait a minute. That's just an opinion piece." In 1989, that may have been the best available at the moment, but the bottom line was that, certainly, in 2016 when this was written, let alone 2017, it's certainly not the best available today.

The next level of the discussion with Baylor was to point out the literature that's evolved since then, going on and starting with Rothwell in 2001 in Stroke, moving on to Cassidy in 2007 in Spine, moving on to Kosloff in 2015, and then moving on to Church in 2016, and how absolutely inconsistent their data-driven

numbers were with this 1 in 20,000 piece. Bottom line was they really weren't interested in the discussion.

Dr. Hoffman: Dr. Clum, let me ask you first, when did you even make a decision to even approach them on this and what were we actually looking for as a result of that?

Dr. Clum: The decision to pursue it was based upon the fact that Vanderbilt was running with this number, as I said before, in a promo piece for their own facility. It was members of the chiropractor community in Tennessee who were really being at risk and being disadvantaged by the fact that a very prominent and major university in their area was promoting inaccurate data. In an attempt, naively thinking on some level, that if we put the right numbers in front of them, obviously, they'll default to the right numbers. That wasn't the case. They backed up, and they said, "Well, this came from Baylor."

Once I got on the track of it, it kind of became, I don't want to go as far as to say an obsession, but I became irritated about the way they were approaching it and just stayed on it to try to run it to ground to get it corrected. Quite honestly, I talked with a number of people that are active in this area in and around the discipline, and they all laughed and said, "Well, don't expect it to change." I said, "Well, it's a Don Quixote moment. We'll [inaudible 00:09:25] at the window a little bit and hope that we can change the course of the culture in that regard."

In the final analysis, we didn't, but the point of this discussion today is for our colleagues and your listeners to understand how a bad number that's based on nothing but opinion can grow legs, get embedded into the literature, and move forward and, if they know how and where that number came from, they can undermine conclusions that are drawn on that number, or based on that number, and set the record straight and put the potential risk, potential risk, not verified risk, but potential risk in context in relationship to other activities in healthcare.

This became a process of, if we could change the hearts and the minds of the authors, great. If we could change the record, even better. If we could come to an agreement, fantastic. None of those things happened, but it's a great opportunity to point out to the practicing chiropractor how this stuff gets started, how it gets sustained, and what they can do to deal with it.

Dr. Hoffman: I think that something that you said is really important, but I want to shift it a little bit. I know that you've gone back and forth with them, on numerous occasions at this point and, to date, nothing has mattered. You won't ever say it, but I will. We have to assume that's a professional/political agenda because why would you ignore all of the overwhelming documentation that you recited in terms of what's been going on in this stroke awareness over the last 15-20 years.

The references that you cited are so in contrast to these numbers, how could they just simply ignore it if they didn't have their own motivation. I'm not asking you to even respond to that. That's just my point of view on it. Is there more to this that we're waiting on? Is there more to it that we can do because, as I said, I know you've gone back and forth, not just with Baylor, but with Vanderbilt also and, so far, really more of a stonewall.

Dr. Clum: That's true. In terms of next steps, the only thing that we can do at this point is be diligent. Early on in these discussions relative to this whole subject area, Allan Terrett in Australia produced a magnificent monograph where he went back, and he researched every single piece of literature that had made reference to a chiropractor related to a stroke, and he found, right off the bat, that more than half of them, a chiropractor was never involved. The first thing we need to do is, we need to study the literature that comes out on this subject and hold the authors accountable to what they say.

More recently than Allan Terrett's work, Adrian Wenban in Europe at the Barcelona College which, by the way, was accredited over the weekend, and we're very happy for Adrian and his colleagues in Barcelona. Adrian has run to ground a number of misstatements from different authors, particularly in Europe, and has been able to get letters of correction into the literature.

The whole goal of this process is to get more information, more correct information, into the literature to be able to offset and to begin to erode the garbage literature from decades ago that still pops up, like the Vickers piece. The Vickers assumption, again, purely a guess but, as you noted in the literature that I talked about from Rothwell to Cassidy to Kosloff and on to Church, they all have good, solid case-controlled, case-crossover data upon which epidemiological conclusions can be drawn.

When you have that data and you look at those conclusions, they're light years apart from that 1 in 20,000 discussion in Vickers. The goal in this process is to bring a focused point of view to the author that wrote the piece that you have concerns with, to get it into the literature, to get a correction statement in there that, whenever that's used, we have something else to make reference to and not let a misstatement perpetuate on its own without response. If you're not the editor of that journal, you don't control what's going to go into that journal but, if we don't speak up, we know for a fact it'll never get corrected.

The process that, from my side, that I've been involved in and many others have been, is how can we keep the authors honest in relationship to what they write about us and the incidents data and so on. From the practitioners' side, the value is that, when they're confronted with a number, they need to understand that there's a way to deal with it and put it in context that is logical, rational, data-driven, and evidence-based as opposed to being opinion-based.

Dr. Hoffman: When I'm out, including this last weekend, talking to a number of doctors about the whole stroke blame-game, some of them will say to me, "Well, I'm an

activator doctor. It's irrelevant," which is interesting because, this last week, we had an article come out documenting an activator adjustment causing a vertebral body dissection and stroke. In fact, I've been in touch with Arlan Fuhr about that, and he's doing exactly what you did in this situation. He's starting that process of responding to them with information that is relevant utilizing research, some that he's done and some that we've even talked about on here.

I think it's very important that the doctors know that this information that comes out will get in the hands of every attorney in the country. When that happens, it's another thing that we have to deal with to overcome whenever you are accused wrongly of causing an actual dissection and stroke. Some of the doctors are very appreciative of that because they either had their own experience and/or know someone that has, but a lot of people don't realize how close this hits to home.

One of the examples that I've given a number of times is, one of the claims, unfortunately, that we've taken in was an older woman that came in to the doctor's office, just filling out paperwork, and they see that she's not okay. They call 9-1-1, she winds up going to the hospital, they're saying it was a stroke, and she died. No one in the family knows what was going on. They called it a chiropractic stroke but, in the end, she never saw the chiropractor. She never even finished filling out her paperwork, but the fact that she merely came from the chiropractor's office just like this led to a lot of assumptions.

Based on those assumptions, the family sues the chiropractor for causing her stroke and death. We have to understand that this is an attack, whether we like it or not, and we need to utilize information that is credible and just responding, which is what I want to get to, just responding, "No, that doesn't happen. No, this. No, that" and having it be an emotional response does not work.

We need to use some of the references like, "You signed it earlier," and we have almost all of those in our informed consent packet already for the doctors to familiarize themselves with and be able to educate rather than defend. That's what my concern is with the doctors being able to understand that this information, when it comes out, it can be very far-reaching. It may not hit home for them personally or for them today, but it's something that we, as a profession, will have to deal with if it's not corrected for a long time to come.

Dr. Clum: That's correct. One of the great things of the moment that we're living is the capacity that the internet provides us, and the problem is that there's no filter on it. There's no truth panel on the internet that says, "This is true. This is false." It's just up there. That's one part of it on the public side. On the more scientific side, when you go to something like PubMed, and that's where that attorney that you're talking about is going to go, they're going to find these case studies.

The reality is that case studies cannot be used to establish causation nor can they be used to cause rates of incidence of a given circumstance or problem. That would be like you and me saying, "Stu, the last three people that I had

come through the office had headaches, and they were completely resolved after one adjustment. Therefore, 100% of headaches resolve with one adjustment." That's asinine. We wish that were true, but that's not the reality of the world.

The same thing happens in the literature out there is that, when people run with a case study, positively or negatively, if we, as chiropractors, assert because we have a patient that had a great resolution of a problem of a given nature, and we run with that and say, "Chiropractic can address this," that's as wrong as them running with it and saying, "Chiropractic can cause this" in the other direction, in the negative direction. We need to be careful about how we're behaving on our side of the aisle and then, at the same time, we need to hold them accountable on their side of the aisle.

The articles that I made reference to before, the Vickers, the Dunn, and the Mann, none of them, none of them have the epidemiological power to assert causation or incidence data or incidence of occurrence of dissection associated with, let alone caused by, a chiropractic adjustment. This house of cards get built up with the improper use of substandard data that has to be displaced by proper data that's properly used, and that's the discussion that we're talking about today.

You're absolutely right that, once things get into the literature, they do take on a life of their own, and we have to deal with it. For example, the case that you made relevance to relevant to activator care, I had a chance to review that article, which was really quite interesting but, if I'm understanding I properly, they're suggesting that an elderly woman had a hemorrhage in her occipital lobe as a result of an activator thrust at the cervical/cranial junction. It's not even a vertebral artery dissection. They're saying that the activator, through the skull, caused a hemorrhage in the occipital lobe of the brain.

If it's a stretch, no pun intended, for an activator to cause a vertebral artery, how far do you have to stretch to think about an activator being applied to the craniocervical junction causing a hemorrhage inside the brain. It's crazy-making in that regard. It's not biologically possible in any shape, form, or fashion. They did, in the article, say that they've never seen anything like this in history.

Well, yeah, that's true because it's Haley's Comet. It's going to come around once every 482 years or whatever, and it hasn't got anything to do with what you're wearing that day or who you went to or what you drank or what you ate. It's going to happen. Those things happen, and they need to be put in the proper context.

As our colleagues grapple with these things, I come across two attitudes. One, people that are overwrought with concern about this, to the point that it becomes crippling and gets in the way of them living their professional lives and doing the good they can do, and the other end of the spectrum that blows if off completely.

The truth of the matter is, like most things, the truth is in the middle, and both ends are wrong. That's the perspective I think you and I would like to see our colleagues take, is that every chiropractor, every single chiropractor should be intimately familiar with the literature that we've talked about today and that, when I talk about Rothwell or Cassidy or Kosloff or Church, they should know exactly what I'm talking about in detail. That may sound unreasonable to some people. They may say, "Well, that's just not part of my world."

Make it part of your world because the challenges that come to us are best responded to with the information that those articles carry and convey. You need to know what they say. You need to understand the methodology involved. You need to understand the implication relative to the epidemiology involved and the power of those studies in relationship to a case study or a case series or something of that nature.

The more we can get our colleagues to understand that, the old story that we were taught since we were kids, the best defense is a good offense, if we have the data, if we know what we have, we can use those tools in our play. If we don't, then we're left with nothing but defense. Lord knows, we don't want that at all.

Dr. Hoffman: I want to reiterate that this is about elevating our doctors' level of awareness, not elevating their fear. It's actually what I got from you, just a little different format. I think, on top of that, I get the same when I'm talking to doctors all the time. In fact, there was a doctor in the last week or so that contacted me because so many of our colleagues assume this has nothing to do with them or their practice, but the more and more publicity that is out there about this, it is affecting every single practice whether we are keeping our head in the sand, as you said, or not.

Some of the doctors don't realize until someone does ask them a question, and this doctor contacted me, do I have any information about this, this, and this? I sent the information that we had, but they wanted more specific. This particular situation was more [inaudible 00:26:25] that we talked about in great lengths. It was all because a patient's father, I believe it was, wanted more information, so this doctor, it's my opinion, felt that they needed more to prove to that one individual that we're okay.

We cite different things, even in our informed consent, things to demonstrate how ridiculously safe a chiropractic adjustment actually is, compared to any other healthcare service that's out there. Part of my devotion and passion for bringing this information to the profession is because I don't want this information to, one day, overcome our colleagues and their practices because it can if we don't do something about it and stay vigilant with each one of these things that come out.

Thank God someone like you does respond to Baylor and to Vanderbilt on the professions we have, but at least you did it based on the literature and clear,

concise information. That's what everyone should be utilizing and, if they don't have the tools for that, certainly bring to any of our attention things that you see out in the public.

You also said something about our own colleagues putting out information about, I'll just pick one, neuropathy. It's not to suggest that we shouldn't take care of patients suffering from neuropathy or diabetic neuropathy, but some of the promotion that we get to see simply says, "Where medicine has failed, we can definitely make you better." I'm exaggerating, but only a little bit. Those are the things that wind up going to our licensing boards, and then we complain that our boards are being unfair. I've written about our boards being activists and some of the things I think are unfair, but not everything.

We have to take responsibility, as you said, for what we're talking about, what we're writing, what we're putting out in the public because it affects the entire profession, not just you as an individual. I think that's where you touched on. I appreciate your words on that. I guess it's the long way of saying it. Where do we go from here? Is there any avenues left with this whole Baylor issue? Is there more back and forth? What's still available to us?

Dr. Clum: I, quite honestly, haven't thought that far into it. If you're suggesting can we bring litigation against them, can we complain to somebody, that sort of thing.

Dr. Hoffman: I was indicating that, yes.

Dr. Clum: I don't believe that's a viable option. I don't believe that's ... I think, as much as we want to do that, as much as we feel assaulted by their attitude and their perspective on some of these things, I'll be honest with you, the feedback came from them. I'm paraphrasing this, but it basically said, "Hey, listen. [inaudible 00:30:01] you want us to make to this matter is three times longer than the original article. We're not going to do that." It came down to the length of the response was too much in comparison to the original article.

I was dumbfounded by the idea that sometimes it takes a long time to explain a problem. When you put the problem on the table, then it takes a good bit of information to correct it so that it's understandable, why wouldn't we want to pursue this? I didn't understand there was a word limit to addressing these kinds of concerns.

Relative to this, the real bottom line is that Stu Hoffman can respond to the things that come onto the table, Jerry Clum can respond to the things that come onto the table, other people who have a good understanding of the literature can do the same thing, but we need to disseminate that knowledge base among the chiropractic community at large, and every chiropractor needs to take responsibility for being on top of this literature and understanding what it says and understanding, more importantly, what it means.

As you had said, one of the things that we do is that we want to get excited and agitated when we're confronted with something in our circle. I would ask that, as people do get angry and respond to those kinds of things, that they think about the logic and the rationale that they respond to concerns that they have in other disciplines and are you applying the same standard to yourself that you're expecting of other people. It leads to that circumstance you were talking about where you wind up with a half-page ad in the morning newspaper as you open it up that has a kernel of truth to it but, by the time you read and get through the entire ad, it's not a kernel, it's a whole ear of corn.

Extrapolating out and making inappropriate assertions and conclusions or allowing those to be made in the mind of the reader, it's troublesome. In time, it's eventually going to come back and bite people in the butt, and then they wind up at your doorstep wanting help or they wind up with me saying, "Will you come testify for me before the board," or whatever the case might be. We need to be careful. The formula for the day is, be diligent, pay attention to what comes across the table, and make sure that the information is countered and addressed.

Number two, pass it along to other people. Don't assume that because you saw it, everybody else saw it. Send it to you, send it to me. Make sure it gets into the discussion realm that we can begin to address it more formally. Is there an organizational response that you can make? Is there an institutional response that one of our educational institutions can make, and so on? Those are the things that we have to do. To the degree that we can do them effectively, we can impact this whole circumstance more efficiently.

Dr. Hoffman: Thanks, Dr. Clum. As always, I appreciate your insights. We share a lot of the same things as we cross paths around the country and on the telephone with doctors on a daily basis. For those of you that are watching that don't already utilize and have our informed consent packet, certainly go onto our Facebook and/or our website and sign up for our concierge service, request it. Please like us on Facebook and download our ChiroSecure app on your smartphone, but we will keep you up to date on any and everything that comes out.

This last weekend, I was thanked by so many people for the information that we've made available to them that they wouldn't have had had we not brought it to the forefront, so we have people like Dr. Clum that is intimately involved in protecting doctors from reading, understanding, and responding to issues like the Baylor article that came out. We're thankful for him, and we appreciate all of you for participating and staying in tune with what we have to offer.

As Dr. Clum said, please read the articles, understand it, and then you could pass along information to your patients and your community that's based on actual research rather than emotion. That's what we want to convey to you, so thank you and have a great day.

Sent: Thu, Feb 2, 2017 7:28 am Subject: RE: Requested correction to Vertebral artery dissection after a chiropractic neck manipulation, Proc (Bayl Univ Med Cent) 2015(28)(1)88-90

Dear Dr. Clum, Dr. Nugent has provided a response (see attached), and we can print your letter (or a slightly modified version of it) and his response in the April issue. I'll get the page proofs to you around February 18th. Regards, Cindy Orticio

-----Original Message----- From: Dr. Gerard Clum [mailto:[email protected]] Sent: Monday, January 23, 2017 3:26 PM To: Subject: {EXTERNAL} FW: Requested correction to Vertebral artery dissection after a chiropractic neck manipulation, Proc (Bayl Univ Med Cent) 2015(28)(1)88-90

Dear Ms.,

I am writing to you in your role as the Managing Editor of the Proceedings of the Baylor University Medical Center.

On December 9, 2016 I sent the attached letter to Dr. Catherine Jones, the corresponding author of "Vertebral artery dissection after a chiropractic neck manipulation" Proc (Bayl Univ Med Cent) 2015(28)(1)88-90 requesting a correction to this paper. The correction requested and the rationale for the correction were detailed in my December 9, 2016 letter.

On December 20, 2016 I followed up my letter of December 9, 2016 with a second request for response from Dr. Jones.

To date I have not heard anything from Dr Catherine Jones regarding this issue.

I now turn to you as the Managing Editor of the Proceedings to secure your help in having the inaccurate and very misleading information put forward in the article in question corrected.

I have reviewed the "Policy on Misconduct in Research" found on the Baylor University website. I would appreciate being provided with the e-mail address

mailto:[email protected]?

and name of the party at Baylor to whom a complaint under this policy should be directed in the event this becomes necessary.

My original letter of December 9, 2016 was copied to you. I am assuming that it would have been your expectation that Dr. Jones would have responded to this matter by now. In the absence of any response this message has been necessitated.

Thank you for your consideration and follow through.

Gerard W. Clum, D.C. [email protected]

________________________________________ From: Dr. Gerard Clum Subject: Requested correction to Vertebral artery dissection after a chiropractic neck manipulation, Proc (Bayl Univ Med Cent) 2015(28)(1)88-90

Dear Dr.,

Attached please find a letter detailing a correction request with respect to the publication noted above.

Thank you for your consideration.

Sincerely,

Gerard W. Clum, D.C. Life University Marietta, Georgia

mailto:[email protected]

1

February 2, 2017

Response to letter from Gerard W Clum, D.C. regarding case report: Vertebral artery dissection

after a chiropractic neck manipulation, Proceedings (Baylor University Medical Center) 2015; 28

(1): 88- 90.

Dear Dr. Clum,

We appreciate your letter alerting us to the discrepancy between the estimated frequency

of vertebral artery dissection following spinal manipulation and the reference cited in our

discussion. We did cite the wrong reference in this portion of our discussion of this case report.

The original reference for this number comes from a paper written by Andrew Vickers and

Catherine Zollman entitled “The manipulative therapies: osteopathy and chiropractic”. This was

published in the British Medical Journal in 1999; they provide estimates for severe adverse

effects ranging from 1 in 20,000 patients to 1 in 1,000,000 patients undergoing cervical spine

manipulation.1 These authors provided no reference for those numbers. Timothy Mann quotes

the same number range and includes an estimate as high as 1 in 4,500 based on abstract from JW

Dunne presented at a meeting in 2000.2,3 Rothwell and colleagues have provided an estimate

based on a population-based case-controlled study published in Stroke in 2001. These authors

suggest that 1.3 vertebral artery accidents occur per 100,000 persons aged <45 years within 1

week of manipulation. The 95% confidence interval for this estimate is 0.5-16.7 per 100,000.4

The upper boundary of this confidence interval translates into 1.7 accidents per 10,000

manipulations. Consequently, our second review of this literature indicates that these numbers

are difficult to obtain, and that there is a wide range in these estimates. The numbers 1 in 20,000

to 1 in million seem reasonable even though this range is extremely broad.

The important point in this case report is that clinicians need to think about cervical spine

trauma in young patients who present with vertebral artery dissections or aneurysms.

1. Vickers A, Zollman C. ABC of complementary medicine. The manipulative therapies:

osteopathy and chiropractic. BMJ (Clinical research ed) 1999; 319(7218): 1176-9.

2

2. Mann T, Refshauge KM. Causes of complications from cervical spine manipulation. Aust

J Physiother 2001; 47(4): 255-66.

3. Dunne JW HNaMD. Neurological complications after spinal manipulation: a regional

survey. Proceedings of the 7th Scientific Conference of the International Federation of

Orthopaedic ManipulativeTherapists 2000: 90.

4. Rothwell DM, Bondy SJ, Williams JI. Chiropractic manipulation and stroke: a

population-based case-control study. Stroke 2001; 32(5): 1054-60.

Respectfully yours,

Jeremy Jones, MD

Catherine Jones, MD

Kenneth Nugent, MD

SPINE Volume 33, Number 4S, pp S176–S183©2008, Lippincott Williams & Wilkins

Risk of Vertebrobasilar Stroke and Chiropractic CareResults of a Population-Based Case-Control and Case-Crossover Study

J. David Cassidy, DC, PhD, DrMedSc,*†‡ Eleanor Boyle, PhD,* Pierre Cote, DC, PhD,*†‡§Yaohua He, MD, PhD,* Sheilah Hogg-Johnson, PhD,†§ Frank L. Silver, MD, FRCPC,¶�

and Susan J. Bondy, PhD†

Study Design. Population-based, case-control andcase-crossover study.

Objective. To investigate associations between chiro-practic visits and vertebrobasilar artery (VBA) stroke andto contrast this with primary care physician (PCP) visitsand VBA stroke.

Summary of Background Data. Chiropractic care ispopular for neck pain and headache, but may increase therisk for VBA dissection and stroke. Neck pain and head-ache are common symptoms of VBA dissection, whichcommonly precedes VBA stroke.

Methods. Cases included eligible incident VBAstrokes admitted to Ontario hospitals from April 1, 1993to March 31, 2002. Four controls were age and gendermatched to each case. Case and control exposures tochiropractors and PCPs were determined from healthbilling records in the year before the stroke date. Inthe case-crossover analysis, cases acted as their owncontrols.

Results. There were 818 VBA strokes hospitalized in apopulation of more than 100 million person-years. Inthose aged �45 years, cases were about three times morelikely to see a chiropractor or a PCP before their strokethan controls. Results were similar in the case controland case crossover analyses. There was no increasedassociation between chiropractic visits and VBA strokein those older than 45 years. Positive associations werefound between PCP visits and VBA stroke in all agegroups. Practitioner visits billed for headache and neck

complaints were highly associated with subsequent VBAstroke.

Conclusion. VBA stroke is a very rare event in thepopulation. The increased risks of VBA stroke associatedwith chiropractic and PCP visits is likely due to patientswith headache and neck pain from VBA dissection seek-ing care before their stroke. We found no evidence ofexcess risk of VBA stroke associated chiropractic carecompared to primary care.

Key words: vertebrobasilar stroke, case control stud-ies, case crossover studies, chiropractic, primary care,complications, neck pain. Spine 2008;33:S176–S183

Neck pain is a common problem associated with consid-erable comorbidity, disability, and cost to society.1–5

In North America, the clinical management of backpain is provided mainly by medical physicians, physi-cal therapists and chiropractors.6 Approximately 12%of American and Canadian adults seek chiropracticcare annually and 80% of these visits result in spinalmanipulation.7,8 When compared to those seekingmedical care for back pain, Canadian chiropractic pa-tients tend to be younger and have higher socioeco-nomic status and fewer health problems.6,8 In On-tario, the average number of chiropractic visits perepisode of care was 10 (median 6) in 1985 through1991.7 Several systematic reviews and our best-evidence synthesis suggest that manual therapy canbenefit neck pain, but the trials are too small to eval-uate the risk of rare complications.9 –13

Two deaths in Canada from vertebral artery dissec-tion and stroke following chiropractic care in the 1990sattracted much media attention and a call by some neu-rologists to avoid neck manipulation for acute neckpain.14 There have been many published case reportslinking neck manipulation to vertebral artery dissectionand stroke.15 The prevailing theory is that extensionand/or rotation of the neck can damage the vertebralartery, particularly within the foramen transversarium atthe C1–C2 level. Activities leading to sudden or sus-tained rotation and extension of the neck have been im-plicated, included motor vehicle collision, shoulderchecking while driving, sports, lifting, working over-head, falls, sneezing, and coughing.16 However, mostcases of extracranial vertebral arterial dissection arethought to occur spontaneously, and other factorssuch as connective tissue disorders, migraine, hyper-tension, infection, levels of plasma homocysteine, ves-sel abnormalities, atherosclerosis, central venous

From the *Centre of Research Expertise for Improved DisabilityOutcomes (CREIDO), University Health Network RehabilitationSolutions, Toronto Western Hospital, and the Division of HeathCare and Outcomes Research, Toronto Western Research Institute,Toronto, ON, Canada; †Department of Public Health Sciences,Management and Evaluation, University of Toronto, Toronto, ON,Canada; ‡Department of Health Policy, Management and Evalua-tion, University of Toronto, Toronto, ON, Canada; §Institute forWork & Health, Toronto, ON, Canada; ¶University Health Net-work Stroke Program, Toronto Western Hospital, Toronto, ON,Canada; and �Division of Neurology, Department of Medicine, Fac-ulty of Medicine, University of Toronto, Toronto, ON, Canada.Supported by Ontario Ministry of Health and Long-term Care. P.C. issupported by the Canadian Institute of Health Research through a NewInvestigator Award. S.H.-J. is supported by the Institute for Work &Health and the Workplace Safety and Insurance Board of Ontario.The opinions, results, and conclusions are those of the authors and noendorsement by the Ministry is intended or should be inferred.The manuscript submitted does not contain information about medicaldevice(s)/drug(s).University Health Network Research Ethics Board Approval number05-0533-AE.Address correspondence and reprint requests to J. David Cassidy, DC,PhD, DrMedSc, Toronto Western Hospital, Fell 4-114, 399 BathurstStreet, Toronto, ON, Canada M5T 2S8; E-mail: [email protected]

S176

catherization, cervical spine surgery, cervical percuta-neous nerve blocks, radiation therapy and diagnosticcerebral angiography have been identified as possiblerisk factors.17–21

The true incidence of vertebrobasilar dissection is un-known, since many cases are probably asymptomatic, orthe dissection produces mild symptoms.22 Confirmingthe diagnosis requires a high index of suspicion and goodvascular imaging. The cases that are most likely to bediagnosed are those that result in stroke.19,22 Ischemicstroke occurs when a thrombus develops intraluminallyand embolizes to more distal arteries, or less commonly,when the dissection extends distally into the intracranialvertebral artery, obliterating branching vessels.22 Thebest incidence estimate comes from Olmstead county,where vertebral artery dissection causing stroke affected0.97 residents per 100,000 population between 1987and 2003.23

To date there have been two case-control studies ofstroke following neck manipulation. Rothwell et al usedOntario health data to compare 582 cases of VBA stroketo 2328 age and sex-matched controls.24 For those aged�45 years, cases were five times more likely than con-trols to have visited a chiropractor within 1 week of VBAstroke. Smith et al studied 51 patients with cervical ar-tery dissection and ischemic stroke or transient ischemicattack (TIA) and compared them to 100 control patientssuffering from other strokes not caused by dissections.25

Cases and controls came from two academic stroke cen-ters in the United States and were matched on age andsex. They found no significant association between neckmanipulation and ischemic stroke or TIA. However, asubgroup analysis showed that the 25 cases with verte-bral artery dissection were six times more likely to haveconsulted a chiropractor within 30 days before theirstroke than the controls.

Finally, because patients with vertebrobasilar arterydissection commonly present with headache and neckpain,23 it is possible that patients seek chiropracticcare for these symptoms and that the subsequent VBAstroke occurs spontaneously, implying that the associ-ation between chiropractic care and VBA stroke is notcausal.23,26 Since patients also seek medical care forheadache and neck pain, any association between pri-mary care physician (PCP) visits and VBA stroke couldbe attributed to seeking care for the symptoms of verte-bral artery dissection.

The purpose of this study is to investigate the asso-ciation between chiropractic care and VBA stroke andcompare it to the association between recent PCP careand VBA stroke using two epidemiological designs.Evidence that chiropractic care increases the risk ofVBA stroke would be present if the measured associa-tion between chiropractic visits and VBA stroke ex-ceeds the association between PCP visits and VBAstrokes.

Methods

Study DesignWe undertook population-based case-control and case-crossover studies. Both designs use the same cases. In the case-control design, we sampled independent control subjects fromthe same source population as the cases. In the case-crossoverdesign, cases served as their own controls, by sampling controlperiods before the study exposures.27 This design is most ap-propriate when a brief exposure (e.g., chiropractic care) causesa transient change in risk (i.e., hazard period) of a rare-onsetdisease (e.g., VBA stroke). It is well suited to our research ques-tions, since within person comparisons control for unmeasuredrisk factors by design, rather than by statistical modeling.28–30

Thus the advantage over the case control design is better con-trol of confounding.

Source PopulationThe source population included all residents of Ontario(109,020,875 person-years of observation over 9 years) cov-ered by the publicly funded Ontario Health Insurance Plan(OHIP). Available utilization data included hospitalizationswith diagnostic coding, and practitioner (physician and chiro-practic) utilization as documented by fee-for-service billingsaccompanied by diagnostic coding. We used two data sources:(1) the Discharge Abstract Database (DAD) from the CanadianInstitute for Health Information, which captures hospital sep-arations and ICD codes, and (2) the OHIP Databases for ser-vices provided by physicians and chiropractors. These data-bases can be linked from April 1992 onward.

CasesWe included all incident vertebrobasilar occlusion and stenosisstrokes (ICD-9433.0 and 433.2) resulting in an acute care hos-pital admission from April 1, 1993 to March 31, 2002. Codeswere chosen in consultation with stroke experts and an epide-miologist who participated in a similar past study (SB).24 Casesthat had an acute care hospital admission for any type of stroke(ICD-9433.0, 433.2, 434, 436, 433.1, 433.3, 433.8, 433.9,430, 431, 432, and 437.1), transient cerebral ischemia (ICD-9435) or late effects of cerebrovascular diseases (ICD-9438)before their VBA stroke admission or since April 1, 1991 wereexcluded. Cases residing in long-term care facilities were alsoexcluded. The index date was defined as the hospital admissiondate for the VBA stroke.

ControlsFor the case-control study, four age and sex-matched controlswere randomly selected from the Registered Persons Database,which contains a listing of all health card numbers for Ontario.Controls were excluded if they previously had a stroke or wereresiding in a long-term care facility.

For the case crossover study, four control periods were ran-domly chosen from the year before the VBA stroke date, usinga time-stratified approach.31 The year was divided into disjointstrata with 2 week periods between the strata. For the 1 monthhazard period, the disjoint strata were separated by 1 monthperiods and the five remaining control periods were used in theanalyses. We randomly sampled disjoint strata because chiro-practic care is often delivered in episodes, and this strategyeliminates overlap bias and bias associated with time trends inthe exposure.32

S177Risk of VBA Stroke and Chiropractic Care • Cassidy et al

ExposuresAll reimbursed ambulatory encounters with chiropractors andPCPs were extracted for the one-year period before the indexdate from the OHIP database. Neck-related chiropractic visitswere identified using diagnostic codes: C01–C06, cervical andcervicothoracic subluxation; C13–C15, multiple site subluxa-tion; C30, cervical sprain/strain; C40, cervical neuritis/neuralgia; C44, arm neuritis/neuralgia; C50, brachial radiculi-tis; C51, cervical radiculitis; and C60, headache. For PCP visits,we included community medicine physicians if they submittedambulatory fee codes to OHIP. Fee codes for group therapyand signing forms were excluded. Headache or neck pain-related PCP visits were identified using the diagnostic codes:ICD-9307, tension headaches; 346, migraine headaches; 722,intervertebral disc disorders; 780, headache, except tensionheadache and migraine; 729, fibrositis, myositis and muscularrheumatism; and 847, whiplash, sprain/strain and other trau-mas associated with neck (These codes include other diagnoses,and we list only those relevant to neck pain or headache). Thereis no limit on the number of reimbursed PCP visits per year.However, there are limits chiropractors, but less than 15% ofpatients surpass them.24

Statistical AnalysisConditional logistic regression was used to estimate the asso-ciation between VBA stroke after chiropractor and PCP visits.Separate models were built using different a priori specifiedhazard periods, stratified by age (�45 years and �45 years)and by visits with or without head and neck pain related diag-nostic codes. For the chiropractic analysis, the index date wasincluded in the hazard period, since chiropractic treatmentmight cause immediate stroke and patients would not normallyconsult a chiropractor after having a stroke. However, the in-dex day was excluded from the PCP analysis, since patientsmight consult these physicians after experiencing a stroke. Wetested different hazard periods, including 1 day, 3 days, 1 week,2 weeks, and 1 month before the index date. Exposure occurredif any chiropractic or PCP visits were recorded during the des-ignated hazard periods.

We also measured the effect of cumulative numbers of chi-ropractic and PCP visits in the month before the index date bycomputing the odds ratio for each incremental visit. These es-timates were similarly stratified by age and by diagnostic codesrelated to headache and/or neck pain. Finally, we conductedanalyses to determine if our results were sensitive to chiroprac-tic and PCP visits related to neck complaints and headaches.We report our results as odds ratios (OR) and 95% confidenceintervals. Confidence intervals were estimated by acceleratedbias corrected bootstraps with 2000 replications using the vari-ance-covariance method.33 All statistical analyses were per-formed using STATA/SE version 9.2.34

Results

A total of 818 VBA strokes met our inclusion/exclusioncriteria over the 9 year inception period. Of the 3272matched control subjects, 31 were excluded because ofprior stroke, one had died before the index date and 76were receiving long-term care. Thus, 3164 control sub-jects were matched to the cases. The mean age of casesand controls was 63 years at the index date and 63%were male. Cases had a higher proportion of comorbidconditions (Table 1). Of the 818 stroke cases, 337

(41.2%) were coded as basilar occlusion and stenosis,443 (54.2%) as vertebral occlusion and stenosis and 38(4.7%) had both codes.

Overall, 4% of cases and controls had visited a chiro-practor within 30 days of the index date, while 53% ofcases and 30% of controls had visited a PCP within thattime (Table 2). For those under 45 years of age, 8 cases(7.8%) had consulted a chiropractor within 7 days of theindex date, compared to 14 (3.4%) of controls. ForPCPs, 25 cases (24.5%) under 45 years of age had aconsultation within 7 days of the index date, com-pared to 27 (6.6%) of controls. With respect to thenumber of visits within 1 month of the index date,7.8% of cases under the age of 45 years had three ormore chiropractic visits, whereas 5.9% had three ormore PCP visits (Table 2).

The case control and case crossover analyses gave sim-ilar results. (Tables 3–7) Age modified the effect of chi-ropractic visits on the risk of VBA stroke. For those un-der 45 years of age, there was an increased associationbetween chiropractic visits and VBA stroke regardless ofthe hazard period. For those 45 years of age and older,there was no association. Each chiropractic visit in themonth before the index date was associated with an in-creased risk of VBA stroke in those under 45 years of age(OR 1.37; 95% CI 1.04–1.91 from the case crossoveranalysis) (Table 7). We were not able to estimate boot-strap confidence intervals in some cases because of sparsedata.

Similarly, we found that visiting a PCP in the monthbefore the index date was associated with an increasedrisk of VBA stroke regardless of the hazard period, or theage of the subject. Each PCP visit in the month before thestroke was associated with an increased risk of VBAstroke both in those under 45 years of age (OR 1.34;95% CI 0.94–1.87 from the case crossover analysis) and45 years and older (OR 1.52; 95% CI 1.36–1.67 fromthe case crossover analysis) (Table 7).

Our results were sensitive to chiropractic and PCPvisits related to neck complaints and headaches, and weobserved sharp increases in the associations when re-stricting the analyses to these visits (Tables 3–7). Overall,

Table 1. Age, Sex, and Comorbid* Condition of Casesand Controls

Variable Cases (n � 818) Controls (n � 3164)

Age: mean, median (SD)† 63.1, 66 (15.5) 62.6, 65 (15.4)Males: n (%) 518 (63.3) 2022 (63.9)Hypertension*: n (%) 276 (33.7) 738 (23.3)Heart Disease*: n (%) 275 (33.6) 506 (16.0)Diabetes*: n (%) 155 (19.0) 247 (7.8)High Cholesterol*: n (%) 62 (7.6) 200 (6.3)At least one comorbid

condition§: n (%)515 (63.0) 1294 (40.9)

*Comorbid conditions determined by ambulatory diagnostic codes from theOntario Health Insurance Plan (OHIP) during year prior to index date.†SD is standard deviation.§Indicates the presence of at least one of hypertension, heart disease, diabe-tes or high cholesterol.

S178 Spine • Volume 33 • Number 4S • 2008

these associations were more pronounced in the PCPanalyses. However, the data are sparse, and we wereunable to compute bootstrap confidence intervals inmany cases.

Discussion

Our study advances knowledge about the associationbetween chiropractic care and VBA stroke in two re-spects. First, our case control results agree with past casecontrol studies that found an association between chiro-practic care and vertebral artery dissection and VBAstroke.24,25 Second, our case crossover results confirmthese findings using a stronger research design with bet-

ter control of confounding variables. The case-crossoverdesign controls for time independent confounding fac-tors, both known and unknown, which could affect therisk of VBA stroke. This is important since smoking,obesity, undiagnosed hypertension, some connective tis-sue disorders and other important risk factors for dissec-tion and VBA stroke are unlikely to be recorded in ad-ministrative databases.

We also found strong associations between PCP visitsand subsequent VBA stroke. A plausible explanation forthis is that patients with head and neck pain due to ver-tebral artery dissection seek care for these symptoms,which precede more than 80% of VBA strokes.23 Since it

Table 2. No. (n) and Percentage (%) of Chiropractic (DC) and Primary Care Physician (PCP) Visits Before theIndex Date

Exposures

Entire Cohort Age �45 yr Age �45 yr

Cases (n � 818) Controls (n � 3164) Cases (n � 102) Controls (n � 408) Cases (n � 716) Controls (n � 2756)

Most recent DC visit0–1 day: n (%) 6 (0.7) 22 (0.7) * * * 21 (0.8)0–3 days: n (%) 9 (1.1) 40 (1.3) * 6 (1.5) * 34 (1.2)0–7 days: n (%) 14 (1.7) 56 (1.8) 8 (7.8) 14 (3.4) 6 (0.8) 42 (1.5)0–14 days: n (%) 27 (3.3) 88 (2.8) 12 (11.8) 17 (4.2) 15 (2.1) 71 (2.6)0–30 days: n (%) 36 (4.4) 125 (4.0) 13 (12.7) 18 (4.4) 23 (3.2) 107 (3.9)

Most recent PCP visit1–1 day: n (%) 63 (7.7) 41 (1.3) 12 (11.8) 6 (1.5) 51 (7.1) 35 (1.3)1–3 days: n (%) 111 (13.6) 130 (4.1) 18 (17.6) 10 (2.5) 93 (13.0) 120 (4.4)1–7 days: n (%) 205 (25.1) 290 (9.2) 25 (24.5) 27 (6.6) 180 (25.1) 263 (9.5)1–14 days: n (%) 311 (38.0) 517 (16.3) 38 (37.3) 46 (11.3) 273 (38.1) 471 (17.1)1–30 days: n (%) 437 (53.4) 945 (29.9) 46 (45.1) 83 (20.3) 391 (54.6) 862 (31.3)

No. of DC visitsNone in past month 782 (95.6) 3039 (96.0) 89 (87.3) 390 (95.6) 693 (96.8) 2649 (96.1)1 or 2 in past month 21 (2.6) 96 (3.0) * 13 (3.2) 16 (2.2) 83 (3.0)3 or more in past month 15 (1.8) 29 (0.9) 8 (7.8) * 7 (1.0) 24 (0.9)Mean (SD†) in past month 0.13 (0.82) 0.08 (0.52) 0.50 (1.65) 0.09 (0.49) 0.08 (0.60) 0.08 (0.53)

No. of PCP visitsNone in past month 381 (46.6) 2219 (70.1) 56 (54.9) 325 (79.7) 325 (45.4) 1894 (68.7)1 or 2 in past month 384 (46.9) 875 (27.7) 40 (39.2) 79 (19.4) 344 (48.0) 796 (28.9)3 or more in past month 53 (6.5) 70 (2.2) 6 (5.9) * 47 (6.6) 66 (2.4)Mean (SD) in past month 0.85 (1.09) 0.41 (0.73) 0.74 (1.01) 0.27 (0.61) 0.87 (1.10) 0.43 (0.74)

*Cell size �6 and cannot be reported.†SD is standard deviation.

Table 3. Odds Ratios and 95% Confidence Intervals (CI) and Accelerated and Bias Corrected Bootstrap 95% CI forCase-Control Estimates of the Association Between Chiropractic (DC) Visits and Vertebrobasilar Stroke

Exposures

Case-Control Age �45 yr Age �45 yr

Odds Ratio (95% CI) Bootstrap 95% CI Odds Ratio (95% CI) Bootstrap 95% CI Odds Ratio (95% CI) Bootstrap 95% CI

Any DC visit0–1 day 1.06 (0.43–2.62) 0.36–2.61 12.00 (1.25–115.36) * 0.55 (0.16–1.85) 0.14–1.930–3 days 0.87 (0.42–1.81) 0.40–1.78 3.33 (1.02–10.92) 0.80–14.00 0.44 (0.16–1.26) 0.12–1.250–7 days 0.98 (0.54–1.77) 0.51–1.79 2.41 (0.98–5.95) 0.80–6.29 0.55 (0.23–1.30) 0.21–1.230–14 days 1.22 (0.78–1.90) 0.77–1.92 3.07 (1.41–6.70) 1.25–7.31 0.82 (0.47–1.45) 0.45–1.470–30 days 1.14 (0.78–1.67) 0.75–1.62 3.13 (1.48–6.63) 1.34–7.21 0.83 (0.52–1.32) 0.51–1.30

Headache or cervicalDC visit

0–1 day 1.59 (0.56–4.54) 0.44–4.67 * * 0.63 (0.14–2.81) 0.00–3.370–3 days 1.22 (0.51–2.88) 0.48–3.01 5.00 (1.34–18.62) * 0.41 (0.09–1.77) 0.00–1.720–7 days 1.42 (0.71–2.86) 0.66–3.00 3.11 (1.16–8.35) 1.07–9.60 0.71 (0.24–2.10) 0.17–2.180–14 days 1.36 (0.79–2.33) 0.78–2.34 3.27 (1.36–7.90) 1.23–8.67 0.84 (0.41–1.75) 0.35–1.680–30 days 0.98 (0.60–1.61) 0.56–1.57 3.00 (1.26–7.12) 1.18–8.00 0.63 (0.33–1.19) 0.30–1.13

*Unable to compute due to small numbers.


is unlikely that PCPs cause stroke while caring for thesepatients, we can assume that the observed associationbetween recent PCP care and VBA stroke represents thebackground risk associated with patients seeking care fordissection-related symptoms leading to VBA stroke. Be-cause the association between chiropractic visits andVBA stroke is not greater than the association betweenPCP visits and VBA stroke, there is no excess risk of VBAstroke from chiropractic care.

Our study has several strengths and limitations. Thestudy base includes an entire population over a 9-yearperiod representing 109,020,875 person-years of obser-vation. Despite this, we found only 818 VBA strokes,which limited our ability to compute some estimates andbootstrap confidence intervals. In particular, our agestratified analyses are based on small numbers of ex-posed cases and controls (Table 2). Further stratificationby diagnostic codes for headache and neck pain relatedvisits imposed even greater difficulty with these esti-mates. However, there are few databases that can link

incident VBA strokes with chiropractic and PCP visits ina large enough population to undertake a study of such arare event.

A major limitation of using health administrative dataare misclassification bias, and the possibility of bias inassignment of VBA-related diagnoses, which has previ-ously been raised in this context.24 Liu et al have shownthat ICD-9 hospital discharge codes for stroke have apoor positive predictive value when compared to chartreview.35 Furthermore, not all VBA strokes are second-ary to vertebral artery dissection and administrative da-tabases do not provide the clinical detail to determine thespecific cause. To investigate this bias, we did a sensitiv-ity analysis using different positive predictive values forstroke diagnosis (ranging from 0.2 to 0.8). Assumingnondifferential misclassification of chiropractic and PCPcases, our analysis showed attenuation of the estimatestowards the null with lower positive predictive values,but the conclusions did not change (i.e., associations re-mained positive and significant—data not shown). The

Table 4. Odds Ratios and 95% Confidence Intervals (CI) and Accelerated and Bias Corrected Bootstrap 95% CI forCase-Control Estimates of the Association Between Primary Care Physician (PCP) Visits and Vertebrobasilar Stroke

Exposure

Case Control Age �45 yr Age �45 yr


Any PCP visit1–1 day 7.22 (4.70–11.08) 4.62–11.23 11.21 (3.59–35.03) 2.67–52.00 6.65 (4.18–10.58) 4.16–10.731–3 days 3.61 (2.76–4.73) 2.73–4.75 9.53 (3.96–22.97) 3.52–28.00 3.21 (2.41–4.27) 2.38–4.321–7 days 3.27 (2.67–4.00) 2.73–4.00 4.81 (2.57–9.02) 2.40–8.72 3.12 (2.52–3.87) 2.52–3.871–14 days 3.11 (2.61–3.69) 2.62–3.65 4.66 (2.78–7.84) 2.71–7.85 2.95 (2.46–3.54) 2.47–3.541–30 days 2.76 (2.35–3.24) 2.35–3.26 3.57 (2.17–5.86) 2.08–6.16 2.67 (2.25–3.17) 2.26–3.17

Headache or cervicalPCP visit

1–1 day 32.00 (7.36–139.17) * 12.00 (1.25–115.36) * 52.00 (6.80–397.50 *1–3 days 25.19 (8.78–72.24) 8.69–104.00 25.64 (3.13–209.78) * 25.04 (7.41–84.62) *1–7 days 16.72 (8.39–33.29 8.52–35.63 37.60 (4.80–294.70) * 14.39 (6.88–30.08) 6.88–35.311–14 days 10.89 (6.53–18.16) 6.59–18.76 37.60 (4.80–294.70) * 9.48 (5.56–16.19) 5.56–16.61)1–30 days 6.96 (4.66–10.41) 4.68–10.42 11.45 (3.68–35.62) 3.50–53.57 6.42 (4.17–9.89) 4.20–10.18


Table 5. Odds Ratios and 95% Confidence Intervals (CI) and Accelerated and Bias Corrected Bootstrap 95% CI forCase-Crossover Estimates of the Association Between Chiropractic (DC) Visits and Vertebrobasilar Stroke

Exposure

Case Crossover Age �45 yr Age �45 yr


Any DC visit0–1 day 1.77 (0.66–4.79) 0.49–5.60 5.04 (0.82–30.99) * 1.09 (0.30–4.02) 0.00–4.840–3 days 1.14 (0.52–2.51) 0.50–2.76 3.44 (0.96–12.28) * 0.61 (0.20–1.84) 0.18–2.110–7 days 0.80 (0.41–1.56) 0.35–1.85 12.19 (2.52–58.98) * 0.30 (0.12–0.77) *0–14 days 1.50 (0.89–2.52) 0.84–2.74 4.49 (1.60–12.63) * 0.98 (0.51–1.87) 0.47–2.010–30 days 1.25 (0.76–2.06) 0.74–2.13 3.60 (1.39–9.35) 1.46–10.84 0.86 (0.47–1.56) 0.45–1.57

Headache or cervicalDC visit

0–1 day 6.67 (1.59–27.90 * * * 2.67 (0.45–15.96) *0–3 days 2.42 (0.88–6.66) 0.70–8.00 17.70 (2.04–153.32) * 0.70 (0.14–3.40) 0.00–6.000–7 days 1.77 (0.80–3.94) 0.68–4.68 * * 5.18 (0.16–1.66) *0–14 days 3.16 (1.57–6.36) 1.43–7.35 33.61 (4.24–266.38) * 1.40 (0.58–3.34) 0.45–3.410–30 days 2.17 (1.09–4.31) 0.97–4.68 29.47 (3.60–241.54) * 1.03 (0.45–2.39) 0.37–2.42



reliability and validity of the codes to classify headacheand cervical visits to chiropractors and PCPs is notknown.

It is also possible that patients presenting to hospitalwith neurologic symptoms who have recently seen a chi-ropractor might be subjected to a more vigorous diag-nostic workup focused on VBA stroke (i.e., differentialmisclassification).36 In this case, the predictive values ofthe stroke codes would be greater for cases that had seena chiropractor and our results would underestimate theassociation between PCP care and VBA stroke.

A major strength of our study is that exposures weremeasured independently of case definition and handledidentically across cases and controls. However, there wassome overlap between chiropractic care and PCP care. Inthe month before their stroke, only 16 (2.0%) of ourcases had seen only a chiropractor, while 20 (2.4%) hadseen both a chiropractor and PCP, and 417 (51.0%) had

just seen only a PCP. We were not able to run a subgroupanalysis on the small number of cases that just saw achiropractor. However, subgroup analysis on the PCPcases (n � 782) that did not visit a chiropractors duringthe 1 month before their stroke did not change the con-clusions (data not shown).

Our results should be interpreted cautiously andplaced into clinical perspective. We have not ruled outneck manipulation as a potential cause of some VBAstrokes. On the other hand, it is unlikely to be a majorcause of these rare events. Our results suggest that theassociation between chiropractic care and VBA strokefound in previous studies is likely explained by present-ing symptoms attributable to vertebral artery dissection.It might also be possible that chiropractic manipulation,or even simple range of motion examination by any prac-titioner, could result in a thromboembolic event in apatient with a pre-existing vertebral dissection. Unfortu-

Table 6. Odds Ratios and 95% Confidence Intervals (CI) and Accelerated and Bias Corrected Bootstrap 95% CI forCase-Crossover Estimates of the Association Between Primary Care Physician (PCP) Visits and Vertebrobasilar Stroke

Exposure

Case Crossover Age �45 yr Age �45 yr


Any PCP visit1–1 day 4.36 (3.02–6.28) 3.09–6.54 15.24 (4.29–54.20) 3.70–68.00 3.72 (2.52–5.50) 2.53–5.641–3 days 2.91 (2.25–3.77) 2.24–3.71 5.62 (2.56–12.36) 2.10–14.60 2.68 (2.03–3.53) 2.02–3.551–7 days 2.36 (1.93–2.89) 1.92–2.95 2.90 (1.64–5.13) 1.56–5.09 2.30 (1.85–2.85) 1.85–2.931–14 days 2.38 (1.99–2.86) 1.96–2.87 3.53 (2.09–5.97) 1.96–6.49 2.26 (1.86–2.74) 1.85–2.801–30 days 2.42 (2.01–2.91) 2.01–2.96 2.99 (1.81–4.96) 1.69–5.09 2.34 (1.92–2.85) 1.91–2.96


1–1 day 16.00 (5.35–47.86) 5.20–72.00 * * 13.00 (4.24–39.87) 4.00–56.001–3 days 13.00 (5.89–28.71) 6.00–31.00 28.00 (3.44–227.58) * 10.86 (4.56–25.83) 4.57–32.001–7 days 8.28 (4.86–14.10) 4.87–14.82 20.00 (4.38–91.28) * 6.99 (3.93–12.44) 3.85–12.851–14 days 7.28 (4.60–11.52) 4.58–11.36 9.46 (2.95–30.31) 2.86–44.00 6.92 (4.20–11.40) 4.19–11.851–30 days 5.65 (3.88–8.22) 3.80–8.30 12.42 (3.95–38.99) 3.59–54.74 5.04 (3.37–7.54) 3.32–7.53


Table 7. Odds Ratios and 95% Confidence Intervals (CI) and Accelerated and Bias Corrected Bootstrap 95% CI forCase-Control and Case Crossover Estimates of the Association Between the Total Number of Chiropractic (DC) andPrimary Care Physician (PCP) Visits in the Month Prior to the Index Date

Exposures

All Cases Age �45 yr Age �45 yr


Case control estimatesAny DC visit 1.14 (1.02–1.27) 1.00–1.27 1.58 (1.19–2.10) 1.19–2.19 1.01 (0.86–1.17) 0.83–1.18Any PCP visit 1.77 (1.61–1.93) 1.61–1.94 2.12 (1.58–2.84) 1.53–3.01 1.73 (1.57–1.90) 1.58–1.91Headache or cervical

DC visit1.13 (0.98–1.30) 0.96–1.31 1.57 (1.17–2.11) 1.08–2.18 0.95 (0.77–1.18) 0.67–1.20


6.33 (4.27–9.37) 4.38–9.18 10.00 (3.26–30.63) * 5.87 (3.85–8.95) 4.04–9.04

Case crossover estimatesAny DC visit 1.10 (0.99–1.22) 0.97–1.24 1.37 (1.10–1.70) 1.04–1.91 0.98 (0.84–1.14) 0.80–1.15Any PCP visit 1.49 (1.36–1.63) 1.34–1.66 1.34 (1.05–1.70) 0.94–1.87 1.52 (1.38–1.67) 1.36–1.68Headache or cervical

DC visit1.18 (1.02–1.37) 0.99–1.45 2.80 (1.43–5.48) * 1.01 (0.83–1.22) 0.74–1.29


3.99 (2.88–5.53) 2.74–5.80 10.64 (3.45–32.78) 3.53–43.56 3.53 (2.51–4.98) 2.35–5.25



nately, there is no acceptable screening procedure toidentify patients with neck pain at risk of VBA stroke.37

These events are so rare and difficult to diagnose thatfuture studies would need to be multicentered and haveunbiased ascertainment of all potential exposures. Givenour current state of knowledge, the decision of how totreat patients with neck pain and/or headache should bedriven by effectiveness and patient preference.38

Conclusion

Our population-based case-control and case-crossoverstudy shows an association between chiropractic visitsand VBA strokes. However, we found a similar associa-tion between primary care physician visits and VBAstroke. This suggests that patients with undiagnosed ver-tebral artery dissection are seeking clinical care for head-ache and neck pain before having a VBA stroke.

Key Points

● Vertebrobasilar artery stroke is a rare event inthe population.● There is an association between vertebrobasilarartery stroke and chiropractic visits in those under45 years of age.● There is also an association between vertebro-basilar artery stroke and use of primary care phy-sician visits in all age groups.● We found no evidence of excess risk of VBAstroke associated chiropractic care.● The increased risks of vertebrobasilar arterystroke associated with chiropractic and physicianvisits is likely explained by patients with vertebro-basilar dissection-related neck pain and headacheconsulting both chiropractors and primary carephysicians before their VBA stroke.

AcknowledgmentsThe authors acknowledge the members of the Decadeof the Bone and Joint 2000 –2010 Task Force on NeckPain and its Associate Disorders for advice about de-signing this study. In particular, they acknowledge thehelp of Drs. Hal Morgenstern, Eric Hurwitz, ScottHaldeman, Linda Carroll, Gabrielle van der Velde,Lena Holm, Paul Peloso, Margareta Nordin, JaimeGuzman, Eugene Carragee, Rachid Salmi, AlexanderGrier, and Mr. Jon Schubert.

References

1. Borghouts JA, Koes BW, Vondeling H, et al. Cost-of-illness of neck pain inThe Netherlands in 1996. Pain 1999;80:629–36.

2. Cote P, Cassidy JD, Carroll L. The Saskatchewan Health and Back PainSurvey. The prevalence of neck pain and related disability in Saskatchewanadults. Spine 1998;23:1689–98.

3. Cote P, Cassidy JD, Carroll L. The factors associated with neck pain and itsrelated disability in the Saskatchewan population. Spine 2000;25:1109–17.

4. Cote P, Cassidy JD, Carroll L. Is a lifetime history of neck injury in a trafficcollision associated with prevalent neck pain, headache and depressive symp-tomatology? Accid Anal Prev 2000;32:151–9.

5. Cote P, Cassidy JD, Carroll LJ, et al. The annual incidence and course of neckpain in the general population: a population-based cohort study. Pain 2004;112:267–73.

6. Cote P, Cassidy JD, Carroll L. The treatment of neck and low back pain: whoseeks care? who goes where? Med Care 2001;39:956–67.

7. Hurwitz EL, Coulter ID, Adams AH, et al. Use of chiropractic services from1985 through 1991 in the United States and Canada. Am J Public Health1998;88:771–6.

8. Hurwitz EL, Chiang LM. A comparative analysis of chiropractic andgeneral practitioner patients in North America: findings from the jointCanada/United States Survey of Health, 2002– 03. BMC Health Serv Res2006;6:49.

9. Aker PD, Gross AR, Goldsmith CH, et al. Conservative management ofmechanical neck pain: systematic overview and meta-analysis. BMJ 1996;313:1291–6.

10. Gross AR, Kay T, Hondras M, et al. Manual therapy for mechanical neckdisorders: a systematic review. Man Ther 2002;7:131–49.

11. Hurwitz EL, Aker PD, Adams AH, et al. Manipulation and mobilization ofthe cervical spine. A systematic review of the literature. Spine 1996;21:1746–59.

12. McClune T, Burton AK, Waddell G. Whiplash associated disorders: a reviewof the literature to guide patient information and advice. Emerg Med J2002;19:499–506.

13. Peeters GG, Verhagen AP, de Bie RA, et al. The efficacy of conservativetreatment in patients with whiplash injury: a systematic review of clinicaltrials. Spine 2001;26:E64–E73.

14. Norris JW, Beletsky V, Nadareishvili ZG. Sudden neck movement and cer-vical artery dissection. The Canadian Stroke Consortium. CMAJ 2000;163:38–40.

15. Ernst E. Manipulation of the cervical spine: a systematic review of case reportsof serious adverse events, 1995–2001. Med J Aust 2002;176:376–80.

16. Haldeman S, Kohlbeck FJ, McGregor M. Risk factors and precipitating neckmovements causing vertebrobasilar artery dissection after cervical traumaand spinal manipulation. Spine 1999;24:785–94.

17. Rubinstein SM, Peerdeman SM, van Tulder MW, et al. A systematic reviewof the risk factors for cervical artery dissection. Stroke 2005;36:1575–80.

18. Inamasu J, Guiot BH. Iatrogenic vertebral artery injury. Acta Neurol Scand2005;112:349–57.

19. Schievink WI. Spontaneous dissection of the carotid and vertebral arteries.N Engl J Med 2001;344:898–906.

20. D’Anglejan-Chatillon J, Ribeiro V, Mas JL, et al. Migraine—a risk factor fordissection of cervical arteries. Headache 1989;29:560–1.

21. Pezzini A, Caso V, Zanferrari C, et al. Arterial hypertension as risk factor forspontaneous cervical artery dissection. A case-control study. J Neurol Neu-rosurg Psychiatry 2006;77:95–7.

22. Savitz SI, Caplan LR. Vertebrobasilar disease. N Engl J Med 2005;352:2618–26.

23. Lee VH, Brown RD Jr, Mandrekar JN, et al. Incidence and outcome ofcervical artery dissection: a population-based study. Neurology 2006;67:1809–12.

24. Rothwell DM, Bondy SJ, Williams JI. Chiropractic manipulation and stroke:a population-based case-control study. Stroke 2001;32:1054–60.

25. Smith WS, Johnston SC, Skalabrin EJ, et al. Spinal manipulative therapy is anindependent risk factor for vertebral artery dissection. Neurology 2003;60:1424–8.

26. Arnold M, Bousser MG, Fahrni G, et al. Vertebral artery dissection: present-ing findings and predictors of outcome. Stroke 2006;37:2499–503.

27. Maclure M. The case-crossover design: a method for studying transient ef-fects on the risk of acute events. Am J Epidemiol 1991;133:144–53.

28. Kelman CW, Kortt MA, Becker NG, et al. Deep vein thrombosis and airtravel: record linkage study. BMJ 2003;327:1072.

29. Mittleman MA, Maclure M, Tofler GH, et al. Triggering of acute myocardialinfarction by heavy physical exertion. Protection against triggering by regu-lar exertion. Determinants of Myocardial Infarction Onset Study Investiga-tors. N Engl J Med 1993;329:1677–83.

30. Redelmeier DA, Tibshirani RJ. Association between cellular-telephone callsand motor vehicle collisions. N Engl J Med 1997;336:453–8.

31. Janes H, Sheppard L, Lumley T. Overlap bias in the case-crossover design,with application to air pollution exposures. Stat Med 2005;24:285–300.

32. Janes H, Sheppard L, Lumley T. Case-crossover analyses of air pollutionexposure data: referent selection strategies and their implications for bias.Epidemiology 2005;16:717–26.

33. Efron B, Tibshirani RJ. An Introduction to the Bootstrap. New York: Chap-man and Hall/CRC, 1993.

34. STATA/SE [computer program]. College Station, Tex: Stata Corp, 2006.35. Liu L, Reeder B, Shuaib A, et al. Validity of stroke diagnosis on hospital


discharge records in Saskatchewan, Canada: implications for stroke surveil-lance. Cerebrovasc Dis 1999;9:224–30.

36. Boyle E, Cote P, Grier AR, et al. Examining vertebrobasilar artery stroke intwo Canadian provinces. Spine, in press.

37. Cote P, Kreitz BG, Cassidy JD, et al. The validity of the extension-rotation

test as a clinical screening procedure before neck manipulation: a secondaryanalysis. J Manip Physiol Therap 1996;159–64.

38. van der Velde G, Hogg-Johnson S, Bayoumi A, et al. Identifying the besttreatment among common non-surgical neck pain treatments: a decisionanalysis. Spine 2008;33(Suppl):S184–S191.


Received 01/26/2016 Review began 01/28/2016 Review ended 02/08/2016 Published 02/16/2016

© Copyright 2016Church et al. This is an open accessarticle distributed under the terms ofthe Creative Commons AttributionLicense CC-BY 3.0., which permitsunrestricted use, distribution, andreproduction in any medium,provided the original author andsource are credited.

Systematic Review and Meta-analysis ofChiropractic Care and Cervical ArteryDissection: No Evidence for CausationEphraim W. Church , Emily P. Sieg , Omar Zalatimo , Namath S. Hussain , Michael Glantz , Robert E. Harbaugh

1. Department of Neurosurgery, Penn State Hershey Medical Center 2. Department of Neurosurgery,Johns Hopkins University School of Medicine

Corresponding author: Ephraim W. Church, [email protected] Disclosures can be found in Additional Information at the end of the article

AbstractBackgroundCase reports and case control studies have suggested an association between chiropractic neckmanipulation and cervical artery dissection (CAD), but a causal relationship has not beenestablished. We evaluated the evidence related to this topic by performing a systematic reviewand meta-analysis of published data on chiropractic manipulation and CAD.

MethodsSearch terms were entered into standard search engines in a systematic fashion. The articleswere reviewed by study authors, graded independently for class of evidence, and combined in ameta-analysis. The total body of evidence was evaluated according to GRADE criteria.

ResultsOur search yielded 253 articles. We identified two class II and four class III studies. There wereno discrepancies among article ratings (i.e., kappa=1). The meta-analysis revealed a smallassociation between chiropractic care and dissection (OR 1.74, 95% CI 1.26-2.41). The qualityof the body of evidence according to GRADE criteria was "very low."

ConclusionsThe quality of the published literature on the relationship between chiropractic manipulationand CAD is very low. Our analysis shows a small association between chiropractic neckmanipulation and cervical artery dissection. This relationship may be explained by the high riskof bias and confounding in the available studies, and in particular by the known association ofneck pain with CAD and with chiropractic manipulation. There is no convincing evidence tosupport a causal link between chiropractic manipulation and CAD. Belief in a causal link mayhave significant negative consequences such as numerous episodes of litigation.

Categories: Neurology, Neurosurgery, Public HealthKeywords: vertebral atery dissection, cervical artery dissection, chiropractic manipulation, cervicalmanipulation, internal carotid artery dissection, cervical spine manipulative therapy

Introduction

1 1 1 1 2

1 1

Open Access OriginalArticle DOI: 10.7759/cureus.498

How to cite this articleChurch E W, Sieg E P, Zalatimo O, et al. (February 16, 2016) Systematic Review and Meta-analysis ofChiropractic Care and Cervical Artery Dissection: No Evidence for Causation. Cureus 8(2): e498. DOI10.7759/cureus.498

http://cureus.com/users/24556-ephraim-w-church

http://cureus.com/users/24557-emily-p-sieg

http://cureus.com/users/24558-omar-zalatimo

http://cureus.com/users/1054-namath-s-hussain

http://cureus.com/users/18125-michael-glantz

http://cureus.com/users/24560-robert-e-harbaugh

Neck pain is a common complaint in physicians’ and chiropractors’ offices. Data from theCenters for Disease Control and from national surveys document 10.2 million ambulatory carevisits for a neck problem in 2001 and 2002. By comparison, there were 11 million office-basedvisits for ischemic heart disease [1]. Many patients with neck pain seek chiropractic care andundergo cervical manipulation. As many as 12% of North Americans receive chiropractic careevery year, and a majority of these are treated with spinal manipulation [2].

In contrast to the frequency of neck pain and chiropractic treatments, spontaneous cervicalartery dissection (CAD) is rare. The annual incidence of internal carotid artery dissection hasbeen estimated at 2.5–3 per 100,000 patients and that of vertebral artery dissection at 1–1.5per 100,000 [3]. Stroke occurs in a small proportion of those with CAD, and its true incidence isdifficult to estimate. Overall, dissection accounts for two percent of all ischemic strokes [4].

Case reports and case series of cervical dissection following manipulation have been published.Despite their rarity, these cases are frequently publicized for several reasons. Patients are oftenyoung and otherwise in good health. Dissection accounts for 10–25% of ischemic strokes inyoung and middle aged patients [4]. If dissection is caused by cervical manipulation it ispotentially a preventable condition. Recent reports, including case control studies, havesuggested an association between chiropractic neck manipulation and cervical dissection [5-10]. Notably, a recent study from the American Heart Association evaluated the availableevidence and concluded such an association exists [11]. This report did not include a meta-analysis, nor did it seek to classify studies and grade the body of evidence. We sought toexamine the strength of evidence related to this question by performing a systematic review,meta-analysis, and evaluation of the body of evidence as a whole.

Materials And MethodsSearch terms “chiropract*,” “spinal manipulation,” “carotid artery dissection,” “vertebral arterydissection,” and “stroke” were included in the search. We used the Medline and Cochranedatabases. We additionally reviewed references of key articles for completeness. A librarianwith expertise in systematic review was consulted throughout the search process.

Two study authors independently reviewed all articles (EC, ES). They selected any applicablestudies for evaluation based on pre-specified inclusion and exclusion criteria. We included onlyhuman trials examining patients with carotid or vertebrobasilar artery dissection and recentchiropractic neck manipulation. We excluded non-English language studies. The articles wereindependently graded using the classification of evidence scheme adopted by the AmericanAcademy of Neurology [12-14]. A third author (MG) arbitrated any discrepancies in the class-of-evidence ratings for the included studies.

Data from all class II and III studies were included in a meta-analysis. A second meta-analysisexcluding class III studies was also performed. The inverse variance method and a fixed effectsmodel were employed. Additionally, we report results using a variable effects model. Theanalyses were performed using RevMan 5.3 software from the Cochrane Informatics andKnowledge Management Department. We did not compose a protocol for our review, althoughPRISMA and MOOSE methodologies were used throughout [15-16].

We evaluated the total body of evidence for quality using the GRADE system [17-20]. A finalGRADE designation was achieved by consensus after discussions involving all study authors asrecommended by GRADE guidelines. This system is designed to assess the total body ofevidence rather than individual studies. The criteria include study design, risk of bias,inconsistency, indirectness, imprecision, publication bias, effect size, dose response, and allplausible residual confounding. Four possible final designations are specified: high, moderate,

2016 Church et al. Cureus 8(2): e498. DOI 10.7759/cureus.498 2 of 11

low, and very low quality.

ResultsResults of the systematic reviewOur search strategy yielded 253 articles. Seventy-seven were judged by all reviewers to be non-relevant. Four articles were judged to be class III studies, and two were rated class II. There wereno discrepancies between the independent ratings (i.e., kappa=1). Studies rated class III orhigher are listed in Table 1. Figure 1 outlines our process of selecting studies for inclusion inthe meta-analysis.

Class II studies Design Patients Number of dissections/VBA strokes

Smith et al., 2003 Case control 151 51

Dittrich et al., 2007 Case control 94 47

Class III studies

Rothwell et al., 2001* Case control 2910 582

Cassidy et al., 2008 Case control 3982 818

Thomas et al., 2011 Case control 90 47

Engelter et al., 2013 Case control 1897 966

TABLE 1: Class II and III articles identified in the systematic review. *Cases overlap with Cassidy et al., 2008. VBA = Vertebrobasilar accidents


FIGURE 1: PRISMA flow diagram.[15]

Meta-analysisCombined data from class II and III studies suggests an association between dissection andchiropractic care, OR 1.74, 95% CI 1.26-2.41 (Figure 2). The result was similar using a randomeffects model, OR 4.05, 95% CI 1.27-12.91. We did not include the study by Rothwell et al.because it describes a subset of patients in the study by Cassidy et al. [5,8]. There was

considerable heterogeneity among the studies (I2=84%).

FIGURE 2: Meta-analysis of class II and III studies.


http://assets.cureus.com/uploads/figure/file/4734/lightbox_599b0390c3e011e5891c33d1242d14ae-Figure_1.png

http://assets.cureus.com/uploads/figure/file/4735/lightbox_c2ae0c50c3c811e580a4f9a708b4ae51-Figure_2.png

We repeated the meta-analysis excluding class III studies. The combined effect size was againindicative of a small association between dissection and chiropractic care, OR 3.17, 95% CI1.30-7.74). The result was identical when using a random effects model.

Class II studiesSmith et al. used a retrospective case control design, combining databases from two academicstroke centers to identify cases of arterial dissection [9]. They found 51 cases and 100 controls.Exposure to spinal manipulative therapy (SMT) was assessed by mail survey. The authorsreported an association between SMT and VBA (P = .032). In multivariate analysis, chiropractorcare within 30 days was associated with VBA, even when adjusting for neck pain or headache(OR 6.6, 95% CI 1.4-30). While this study controlled for possible confounders such as neck pain,there were several limitations. Head and neck pain as well as chiropractor visit were assessed ina retrospective fashion by mail survey, very possibly introducing both recall and survivor bias.The reason for reporting to the chiropractor (e.g., trauma) was not assessed. Further, there wassignificant variability among diagnostic procedures, which may reflect increased motivation byphysicians to rule out dissection in patients with a history of SMT. Such motivation could resultin interviewer bias.

Dittrich et al. compared 47 patients with CAD to a control group with stroke due to etiologiesother than dissection [6]. They assessed for risk factors using a face-to-face interview withblinding. These authors found no association between any individual risk factor and CAD,including cervical manipulative therapy. They blame the small sample size for the negativeresult, and they point out that cumulative analysis of all mechanical risk factors <24 hours priorto symptom onset showed an association (P = .01). This study is subject to recall bias.

Class III studiesRothwell et al. used a retrospective case control design to test for an association betweenchiropractic manipulation and vertebrobasilar accidents (VBA) [8]. They reviewed Ontariohospital records for admissions for VBA from 1993–1998. There were 582 cases and 2328matching controls. The authors report an association between VBA and visit to a chiropractorwithin one week (OR 5.03, 95% CI 1.32-43.87), but this was only true for young patients (<45years). This study represented the first attempt to delineate the association betweenchiropractic manipulation and extremely rare VBA with controls. Limitations included requisiteuse of ICD-9 codes to identify cases and associated classification bias, as well as potentialunmeasured confounders (e.g., neck pain).

In 2008, Cassidy et al. set out to address the problem of neck pain possibly confounding theassociation between chiropractic care and VBA [5]. Again using a retrospective case controldesign, they included all residents of Ontario over a period of 9 years (1993–2002, 109,020,875person years of observation). They identified 818 VBA strokes resulting in hospitalization andrandomly selected age and sex matched controls. Next, they examined ambulatory encounterswith chiropractors and primary care physicians (PCPs) in the one year preceding the stroke,limited to cervical manipulation, neck pain, and headache. Associations between chiropractorvisit and VBA versus PCP visits and VBA were compared. Indeed, there were associationsbetween both chiropractor visit and VBA (<45yrs OR 1.37, 95% CI 1.04-1.91), and PCP visit andVBA (<45 yrs OR 1.34, 95% CI .94-1.87; >45 yrs and OR 1.53, 95% CI 1.36-1.67). The associationfor chiropractor visit was not greater than for PCP visit. This data was interpreted as evidencethat a confounder such as neck pain may account for the association between chiropractor visitand VBA. This study was subject to many of the same limitations as previous efforts. Canadianhealth records would not reveal whether a patient with cervical complaints underwent cervicalmanipulation, and the researchers could not review each chart for imaging confirmingdissection. Additionally, the incidence of comorbidities (e.g., hypertension, heart disease,


diabetes) was significantly higher among cases as compared to controls, and we are concernedthat these differences were non-random.

In another case control study, Thomas et al. compared the records of 47 patients with confirmedor suspected vertebral or internal carotid artery dissection with 43 controls [10]. They limitedtheir analysis to young patients defined as <55 years. These authors report a significantassociation between dissection and recent head or neck trauma (OR 23.51, 95% CI 5.71-96.89)as well as neck manual therapy (OR 1.67, 95% CI 1.43-112.0). An inconsistent standard for caseascertainment (a significant number of patients lacked radiographic confirmation of dissection)and lack of blinding weaken this study.

Engelter et al. evaluated data from the Cervical Artery Dissection and Ischemic Stroke Patients(CADISP) consortium, identifying 966 patients with CAD, 651 with stroke attributable toanother cause, and 280 healthy controls [7]. The CADISP study involved both prospectively andretrospectively collected data at multiple centers in several countries. They assessed for priorcervical trauma within one month using questionnaires administered during clinic visits.Cervical manipulation therapy was more common for CAD versus stroke from another cause(OR 12.1, CI 4.37-33.2). The report notes that an association between any trauma and CAD waspresent even when restricting the analysis to prospectively recruited patients. However, inpatients to whom the questionnaire was administered after dissection, recall bias may havebeen at work whether or not the patient was enrolled prospectively. Indeed, the frequency ofprior cervical trauma in this study was substantially higher than previous reports (40% versus12-34%). Additional weaknesses include a highly heterogeneous standard for case definitionand no clear masking procedures.

Body of evidence quality (GRADE rating)Having performed a systematic review and rated articles according to their individual strengthsand weaknesses, we graded the overall body of evidence using the system proposed by Guyatt etal. [17-20]. The GRADE approach to rating quality of evidence proposes four categories that areapplied to a body of evidence: high, moderate, low, and very low. In the setting of systematicreview, a particular rating reflects the extent of confidence that the estimates of effect arecorrect. The GRADE approach begins with study design and sequentially examines featureswith the potential to enhance or diminish confidence in the meta-analytic estimate of effectsize.

Our final assessment of the quality of the body of evidence using these criteria was very low.The initial rating based on study design was low (observational studies). Given the controversialnature of this topic and the legal ramifications of results, there is certainly potential for bias (-1serious). However, blinding in the Class II studies mitigated this risk to some extent.Inconsistency and imprecision did not lower our rating. Because the body of evidence is derivedfrom measures of association, the rating was lowered for indirectness (-1 serious). Publicationbias is less likely because of the impact of a negative result in this case. The funnel plot fromour meta-analysis was inconclusive with regard to possible publication bias because of thesmall number of studies included but suggested a deficit in the publication of small negativetrials. There was not a large effect size, and currently there is no evidence for a dose responsegradient. Moreover, the most worrisome potential confounder (neck pain) would increase ratherthan reduce the hypothesized effect.

DiscussionThe results of our systematic review and meta-analysis suggest a small association betweenchiropractic care and CAD. There are no class I studies addressing this issue, and thisconclusion is based on five class II and III studies. Scrutiny of the quality of the body of data


using the GRADE criteria revealed that it fell within the “very low” category. We found noevidence for a causal link between chiropractic care and CAD. This is a significant findingbecause belief in a causal link is not uncommon, and such a belief may have significant adverseeffects such as numerous episodes of litigation.

The studies included in our meta-analysis share several common weaknesses. Two of the fivestudies used health administrative databases, and since conclusions depend on accurate ICDcoding, this technique for case ascertainment may introduce misclassification bias. It is notpossible to account for the type of spinal manipulation that may have been used. Retrospectivecollection of data is also a potential weakness and may introduce recall bias when a survey orinterview was used. Moreover, patients arriving at a hospital complaining of neck pain anddescribing a recent visit to a chiropractor may be subject to a more rigorous evaluation for CAD(interviewer bias). Another potential source of interviewer bias was lack of blinding in the classIII studies. Further, we noted substantial variability among diagnostic procedures performed.All of these weaknesses affect the reliability of the available evidence and are not “corrected”by performing a meta-analysis.

Perhaps the greatest threat to the reliability of any conclusions drawn from these data is thattogether they describe a correlation but not a causal relationship, and any unmeasured variableis a potential confounder. The most likely potential confounder in this case is neck pain.Patients with neck pain are more likely to have CAD (80% of patients with CAD report neck painor headache) [21], and they are more likely to visit a chiropractor than patients without neckpain (Figure 3). Several of the studies identified in our systematic review provide suggestiveevidence that neck pain is a confounder of the apparent association between chiropractic neckmanipulation and CAD. For example, in Engelter et al. patients with CAD and prior cervicaltrauma (e.g., cervical manipulation therapy) were more likely to present with neck pain but lessoften with stroke than those with CAD and no prior cervical trauma (58% vs. 43% for traumaand 61% vs. 69% for stroke) [7]. If patients with CAD without neurological symptoms came tomedical attention, it was probably because of pain. Patients with neck pain would also be morelikely to visit a chiropractor than those without neck pain.

FIGURE 3: The association between a chiropractor visit anddissection may be explained by headache/neck pain, a likely


http://assets.cureus.com/uploads/figure/file/4738/lightbox_5107a270c42f11e5a154ffedcdc315f0-Figure_3.png

confounder.

Cassidy et al. hypothesized that, although an association between chiropractor visits andvertebrobasilar artery stroke is present, it may be fully explained by neck pain and headache [5].These authors reviewed 818 patients with vertebrobasilar artery strokes hospitalized in apopulation of 100 million person-years. They compared chiropractor and PCP visits in thispopulation and reported no significant difference between these associations. For patientsunder 45 years of age, each chiropractor visit in the previous month increased the risk of stroke(OR 1.37, 95% CI 1.04-1.91), but each PCP visit in the previous month increased the risk in anearly identical manner (<45 yrs OR 1.34, 95% CI .94-1.87; >45 yrs and OR 1.53, 95% CI 1.36-1.67). The authors conclude that, since patients with vertebrobasilar stroke were as likely tovisit a PCP as they were to visit a chiropractor, these visits were likely due to pain from anexisting dissection.

Cervical artery dissection is a rare event, creating a significant challenge for those who wish tounderstand it. A prospective, randomized study design is best suited to control for confounders,but given the infrequency of dissection, performing such a study would be logistically and alsoethically challenging. Sir Austin Bradford Hill famously addressed the problem of assigningcausation to an association with the application of nine tests [22]. These criteria includestrength, consistency, specificity, temporality, biological gradient, plausibility, coherence,experimental evidence, and analogy. The specific tests and our assessment for the associationbetween cervical manipulation and CAD are summarized in Table 2. In our appraisal, thisassociation clearly passes only one test, it fails four, and the remaining four are equivocal due toabsence of relevant data [23]. Further, a 2013 assessment of the quality of reports of cervicalarterial dissection following cervical spinal manipulation similarly found lacking data tosupport a causal relationship [24].


Criterion Assessment for the Association Between Cervical Manipulation and CAD

1. Strengthof theassociation

The association is modest.

2.Consistencyof theassociation

Four of five class II and III studies demonstrate an association.

3. Outcomespecific toexposure

As seen in Cassidy et al., exposure to a primary care doctor and exposure to a chiropractor are equally likely toresult in CAD [5]. In this case the outcome is not specific.

4. Temporalrelationship

A temporal relationship is necessary but not sufficient to establish causation. In this case, the onset ofsymptoms following cervical manipulation is variable and often delayed.

5. Biologicalgradient

There are no data to support or refute a dose-response hypothesis.

6. PlausibilityAre there increased rates of CAD in regions with increased utilization of chiropractic manipulation? No relevantdata are available to address this criterion.

7. Coherence

Tests on human cadavers have revealed that vertebral artery strains during spinal manipulative therapy do notplace significant strain on the vertebral artery [23]. A review of data related to this topic sponsored by theAmerican Heart Association concluded that: “Current biomechanical evidence is insufficient to establish theclaim that spinal manipulation causes [CAD], including data from a canine model showing no significantchanges in [vertebral artery] lesions before and after cervical manipulation” [11].

8.Experimentalevidence

The available animal models do not support the association.

9. Analogousto provenassociation

While severe trauma most certainly causes dissection, it may be debated whether the situation in chiropracticcare is analogous.

TABLE 2: Hill’s criteria for assigning causation to association [22].CAD = Cervical artery dissection

In spite of the very weak data supporting an association between chiropractic neckmanipulation and CAD, and even more modest data supporting a causal association, such arelationship is assumed by many clinicians. In fact, this idea seems to enjoy the status ofmedical dogma. Excellent peer reviewed publications frequently contain statements asserting acausal relationship between cervical manipulation and CAD [4,25,26]. We suggest thatphysicians should exercise caution in ascribing causation to associations in the absence ofadequate and reliable data. Medical history offers many examples of relationships that wereinitially falsely assumed to be causal [27], and the relationship between CAD and chiropracticneck manipulation may need to be added to this list.


ConclusionsOur systematic review revealed that the quality of the published literature on the relationshipbetween chiropractic manipulation and CAD is very low. A meta-analysis of available datashows a small association between chiropractic neck manipulation and CAD. We uncoveredevidence for considerable risk of bias and confounding in the available studies. In particular, theknown association of neck pain both with cervical artery dissection and with chiropracticmanipulation may explain the relationship between manipulation and CAD. There is noconvincing evidence to support a causal link, and unfounded belief in causation may have direconsequences.

Additional InformationDisclosuresConflicts of interest: The authors have declared that no conflicts of interest exist.

AcknowledgementsThe authors wish to thank Elaine Dean, MLS, of the Penn State Hershey Medical Center GeorgeT. Harrell Health Sciences Library, for her assistance with the systematic review.

References1. Riddle DL, Schappert SM: Volume and characteristics of inpatient and ambulatory medical

care for neck pain in the United States: data from three national surveys. Spine. 2007, 32:132–140.

2. Hurwitz EL, Chiang LM: A comparative analysis of chiropractic and general practitionerpatients in North America: findings from the joint Canada/United States survey of health,2002-03. BMC Health Serv Res. 2006, 6:49. 10.1186/1472-6963-6-49

3. Micheli S, Paciaroni M, Corea F, et al.: Cervical artery dissection: emerging risk factors . OpenNeurol J. 2010, 4:50–55. 10.2174/1874205X01004010050

4. Schievink WI: Spontaneous dissection of the carotid and vertebral arteries . N Engl J Med.2001, 344:898–906. 10.1056/NEJM200103223441206

5. Cassidy JD, Boyle E, Côté PDC, et al.: Risk of vertebrobasilar stroke and chiropractic care:results of a population-based case-control and case-crossover study. Spine. 2008, 33:176–183.10.1097/BRS.0b013e3181644600

6. Dittrich R, Rohsbach D, Heidbreder A, et al.: Mild mechanical traumas are possible risk factorsfor cervical artery dissection. Cerebrovasc Dis. 2007, 23:275–281. 10.1159/000098327

7. Engelter ST, Grond-Ginsbach C, Metso TM, et al.: Cervical artery dissection: trauma and otherpotential mechanical trigger events. Neurology. 2013, 80:1950–1957.10.1212/WNL.0b013e318293e2eb

8. Rothwell DM, Bondy SJ, Williams JI: Chiropractic manipulation and stroke: a population-based case-control study. Stroke. 2001, 32:1054-1060.

9. Smith WS, Johnston SC, Skalabrin EJ, et al.: Spinal manipulative therapy is an independentrisk factor for vertebral artery dissection. Neurology. 2003, 60:1424-1428.

10. Thomas LC, Rivett DA, Attia JR, et al.: Risk factors and clinical features of craniocervicalarterial dissection. Man Ther. 2011, 16:351–356. 10.1016/j.math.2010.12.008

11. Biller J, Sacco RL, Albuquerque FC, et al.: Cervical arterial dissections and association withcervical manipulative therapy: a statement for healthcare professionals from the AmericanHeart Association/American Stroke Association. Stroke. 2014, 45:3155–3174.10.1161/STR.0000000000000016

12. AAN (American Academy of Neurology): Clinical Practice Guideline Process Manual.Gronseth GS, Woodroffe LM, Getchius TSD (ed): AAN (American Academy of Neurology), StPaul, MN; 2011.

13. French J, Gronseth G: Lost in a jungle of evidence: we need a compass . Neurology. 2008,71:1634–1638. 10.1212/01.wnl.0000336533.19610.1b


http://scholar.google.com/scholar?q=intitle:Volume and characteristics of inpatient and ambulatory medical care for neck pain in the United States: data from three national surveys

http://dx.doi.org/10.1186/1472-6963-6-49

http://dx.doi.org/10.1186/1472-6963-6-49

http://dx.doi.org/10.2174/1874205X01004010050

http://dx.doi.org/10.2174/1874205X01004010050

http://dx.doi.org/10.1056/NEJM200103223441206

http://dx.doi.org/10.1056/NEJM200103223441206

http://dx.doi.org/10.1097/BRS.0b013e3181644600

http://dx.doi.org/10.1097/BRS.0b013e3181644600

http://dx.doi.org/10.1159/000098327

http://dx.doi.org/10.1159/000098327

http://dx.doi.org/10.1212/WNL.0b013e318293e2eb

http://dx.doi.org/10.1212/WNL.0b013e318293e2eb

http://www.ncbi.nlm.nih.gov/pubmed/11340209

http://scholar.google.com/scholar?q=intitle:Spinal manipulative therapy is an independent risk factor for vertebral artery dissection

http://dx.doi.org/10.1016/j.math.2010.12.008

http://dx.doi.org/10.1016/j.math.2010.12.008

http://dx.doi.org/10.1161/STR.0000000000000016

http://dx.doi.org/10.1161/STR.0000000000000016

http://dx.doi.org/10.1212/01.wnl.0000336533.19610.1b

http://dx.doi.org/10.1212/01.wnl.0000336533.19610.1b

14. Gross RA, Johnston KC: Levels of evidence: taking Neurology® to the next level . Neurology.2009, 72:8–10. 10.1212/01.wnl.0000342200.58823.6a

15. Moher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews andmeta-analyses: the PRISMA statement. PLoS Med. 2009, 6:e1000097. Accessed: January 23,2016: http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000097.10.1371/journal.pmed.1000097

16. Stroup DF, Berlin JA, Morton SC, et al.: Meta-analysis of observational studies inepidemiology: a proposal for reporting. Meta-analysis of observational studies inepidemiology (MOOSE) group. JAMA. 2000, 283:2008-2012.

17. Guyatt G, Oxman AD, Akl EA, et al.: GRADE guidelines: 1. Introduction-GRADE evidenceprofiles and summary of findings tables. J Clin Epidemiol. 2011, 64:383–394.10.1016/j.jclinepi.2010.04.026

18. Guyatt GH, Oxman AD, Kunz R, et al.: GRADE guidelines: 2. Framing the question anddeciding on important outcomes. J Clin Epidemiol. 2011, 64:395–400.10.1016/j.jclinepi.2010.09.012

19. Balshem H, Helfand M, Schünemann HJ, et al.: GRADE guidelines: 3. Rating the quality ofevidence. J Clin Epidemiol. 2011, 64:401–406. 10.1016/j.jclinepi.2010.07.015

20. The Cochrane Collaboration: Cochrane Handbook for Systematic Reviews of Interventions(Version 5.1.0). Higgins JPT, Green S (ed): The Cochrane Collaboration, 2011.

21. Lee VH, Brown RD Jr, Mandrekar JN, et al.: Incidence and outcome of cervical arterydissection: a population-based study. Neurology. 2006, 67:1809-1812.

22. Hill AB: The environment and disease: association or causation?. Proc R Soc Med. 1965,58:295–300.

23. Herzog W, Leonard TR, Symons B, et al.: Vertebral artery strains during high-speed, lowamplitude cervical spinal manipulation. J Electromyogr Kinesiol. 2012, 22:740–746.10.1016/j.jelekin.2012.03.005

24. Wynd S, Estaway M, Vohra S, Kawchuk G: The quality of reports on cervical arterial dissectionfollowing cervical spinal manipulation. PLOS ONE. 2013, 8:e59170. Accessed: February 8,2016: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0059170.10.1371/journal.pone.0059170

25. Albuquerque FC, Hu YC, Dashti SR, et al.: Craniocervical arterial dissections as sequelae ofchiropractic manipulation: patterns of injury and management. J Neurosurg. 2011, 115:1197–1205. 10.3171/2011.8.JNS111212

26. Debette S, Leys D: Cervical-artery dissections: predisposing factors, diagnosis, and outcome .Lancet Neurol. 2009, 8:668–678. 10.1016/S1474-4422(09)70084-5

27. Artenstein AW: The discovery of viruses: advancing science and medicine by challengingdogma. Int J Infect Dis. 2012, 16:e470–e473. 10.1016/j.ijid.2012.03.005


http://dx.doi.org/10.1212/01.wnl.0000342200.58823.6a

http://dx.doi.org/10.1212/01.wnl.0000342200.58823.6a

http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000097

http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000097

http://dx.doi.org/10.1371/journal.pmed.1000097

http://scholar.google.com/scholar?q=intitle:Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of observational studies in epidemiology (MOOSE) group

http://dx.doi.org/10.1016/j.jclinepi.2010.04.026






http://scholar.google.com/scholar?q=intitle:Incidence and outcome of cervical artery dissection: a population-based study

http://scholar.google.com/scholar?q=intitle:The environment and disease: association or causation?

http://dx.doi.org/10.1016/j.jelekin.2012.03.005

http://dx.doi.org/10.1016/j.jelekin.2012.03.005

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0059170

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0059170

http://dx.doi.org/10.1371/journal.pone.0059170

http://dx.doi.org/10.3171/2011.8.JNS111212

http://dx.doi.org/10.3171/2011.8.JNS111212

http://dx.doi.org/10.1016/S1474-4422(09)70084-5

http://dx.doi.org/10.1016/S1474-4422(09)70084-5

http://dx.doi.org/10.1016/j.ijid.2012.03.005

http://dx.doi.org/10.1016/j.ijid.2012.03.005

RESEARCH Open Access

Chiropractic care and the risk ofvertebrobasilar stroke: results of acase–control study in U.S. commercialand Medicare Advantage populationsThomas M Kosloff1*†, David Elton1†, Jiang Tao2† and Wade M Bannister2†

Abstract

Background: There is controversy surrounding the risk of manipulation, which is often used by chiropractors, withrespect to its association with vertebrobasilar artery system (VBA) stroke. The objective of this study was to compare theassociations between chiropractic care and VBA stroke with recent primary care physician (PCP) care and VBA stroke.

Methods: The study design was a case–control study of commercially insured and Medicare Advantage (MA) healthplan members in the U.S. population between January 1, 2011 and December 31, 2013. Administrative data were usedto identify exposures to chiropractic and PCP care. Separate analyses using conditional logistic regression wereconducted for the commercially insured and the MA populations. The analysis of the commercial population wasfurther stratified by age (<45 years; ≥45 years). Odds ratios were calculated to measure associations for different hazardperiods. A secondary descriptive analysis was conducted to determine the relevance of using chiropractic visits as aproxy for exposure to manipulative treatment.

Results: There were a total of 1,829 VBA stroke cases (1,159 – commercial; 670 – MA). The findings showed nosignificant association between chiropractic visits and VBA stroke for either population or for samples stratified by age.In both commercial and MA populations, there was a significant association between PCP visits and VBA strokeincidence regardless of length of hazard period. The results were similar for age-stratified samples. The findings of thesecondary analysis showed that chiropractic visits did not report the inclusion of manipulation in almost one third ofstroke cases in the commercial population and in only 1 of 2 cases of the MA cohort.

Conclusions: We found no significant association between exposure to chiropractic care and the risk of VBA stroke. Weconclude that manipulation is an unlikely cause of VBA stroke. The positive association between PCP visits and VBAstroke is most likely due to patient decisions to seek care for the symptoms (headache and neck pain) of arterialdissection. We further conclude that using chiropractic visits as a measure of exposure to manipulation may result inunreliable estimates of the strength of association with the occurrence of VBA stroke.

Keywords: Chiropractic, Primary care, Cervical manipulation, Vertebrobasilar stroke, Adverse events

* Correspondence: [email protected]†Equal contributors1Optum Health – Clinical Programs at United Health Group, 11000 OptumCircle, Eden Prairie MN 55344, USAFull list of author information is available at the end of the article

CHIROPRACTIC & MANUAL THERAPIES

© 2015 Kosloff et al. This is an Open Access article distributed under the terms of the Creative Commons Attribution License(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium,provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Kosloff et al. Chiropractic & Manual Therapies (2015) 23:19 DOI 10.1186/s12998-015-0063-x

http://crossmark.crossref.org/dialog/?doi=10.1186/s12998-015-0063-x&domain=pdf

mailto:[email protected]

http://creativecommons.org/licenses/by/4.0

http://creativecommons.org/publicdomain/zero/1.0/

http://creativecommons.org/publicdomain/zero/1.0/

BackgroundThe burden of neck pain and headache or migraineamong adults in the United States is significant. Surveydata indicate 13% of adults reported neck pain in thepast 3 months [1]. In any given year, neck pain affects30% to 50% of adults in the general population [2].Prevalence rates were reportedly greater in more eco-nomically advantaged countries, such as the USA, with ahigher incidence of neck pain noted in office and com-puter workers [3]. Similar to neck pain, the prevalenceof headache is substantial. During any 3-month time-frame, severe headaches or migraines reportedly affectone in eight adults [1].Neck pain is a very common reason for seeking health

care services. “In 2004, 16.4 million patient visits or 1.5%of all health care visits to hospitals and physician offices,were for neck pain” [4]. Eighty percent (80%) of visitsoccurred as outpatient care in a physician’s office [4].The utilization of health care resources for the treatmentof headache is also significant. “In 2006, adults madenearly 11 million physician visits with a headache diagno-sis, over 1 million outpatient hospital visits, 3.3 millionemergency department visits, and 445 thousand inpatienthospitalizations” [1].In the United States, chiropractic care is frequently

utilized by individuals with neck and/or headache com-plaints. A national survey of chiropractors in 2003 re-ported that neck conditions and headache/facial painaccounted respectively for 18.7% and 12% of the patientchief complaints [5]. Chiropractors routinely employspinal manipulative treatment (SMT) in the manage-ment of patients presenting with neck and/or headache[6], either alone or combined with other treatment ap-proaches [7-10].While evidence syntheses suggest the benefits of SMT for

neck pain [7-9,11-13] and various types of headaches[10,12,14-16], the potential for rare but serious adverseevents (AE) following cervical SMT is a concern for re-searchers [17,18], practitioners [19,20], professional organi-zations [21-23], policymakers [24,25] and the public[26,27]. In particular, the occurrence of stroke affecting thevertebrobasilar artery system (VBA stroke) has been associ-ated with cervical manipulation. A recent publication [28]assessing the safety of chiropractic care reported, “…the fre-quency of serious adverse events varied between 5 strokes/100,000 manipulations to 1.46 serious adverse events/10,000,000 manipulations and 2.68 deaths/10,000,000 ma-nipulations”. These estimates were, however, derived fromretrospective anecdotal reports and liability claims data,and do not permit confident conclusions about the actualfrequency of neurological complications following spinalmanipulation.Several systematic reviews investigating the association

between stroke and chiropractic cervical manipulation

have reported the data are insufficient to producedefinitive conclusions about its safety [28-31]. Twocase–control studies [32,33] used visits to a chiropractoras a proxy for SMT in their analyses of standardizedhealth system databases for the population of Ontario(Canada). The more recent of these studies [32] also in-cluded a case-crossover methodology, which reduced therisk of bias from confounding variables. Both case–controlstudies reported an increased risk of VBA stroke in as-sociation with chiropractic visits for the populationunder age 45 years old. Cassidy, et al. [32] found, how-ever, the association was similar to visits to a primarycare physician (PCP). Consequently, the results of thisstudy suggested the association between chiropracticcare and stroke was noncausal. In contrast to thesestudies, which found a significant association betweenchiropractic visits and VBA stroke in younger patients(<45 yrs.), the analysis of a population-based case-seriessuggested that VBA stroke patients who consulted achiropractor the year before their stroke were older(mean age 57.6 yrs.) than previously documented [34].The work by Cassidy, et al. [32] has been qualitatively ap-

praised as one of the most robustly designed investigationsof the association between chiropractic manipulative treat-ment and VBA stroke [31]. To the best of our knowledge,this work has not been reproduced in the U.S. population.Thus, the main purpose of this study is to replicate thecase–control epidemiological design published by Cassidy,et al. [32] to investigate the association between chiroprac-tic care and VBA stroke; and compare it to the associationbetween recent PCP care and VBA stroke in samples of theU.S. commercial and Medicare Advantage (MA) popula-tions. A secondary aim of this study is to assess the utilityof employing chiropractic visits as a proxy measure for ex-posure to spinal manipulation.

MethodsStudy design and populationWe developed a case–control study based on the experi-ence of commercially insured and MA health plan mem-bers between January 1, 2011 and December 31, 2013.General criteria for membership in a commercial or MAhealth plan included either residing or working in a re-gion where health care coverage was offered by the in-surer. Individuals must have Medicare Part A and Part Bto join a MA plan. The data set included health planmembers located in 49 of 50 states. North Dakota wasthe only State not represented.Both case and control data were extracted from the

same source population, which encompassed nationalhealth plan data for 35,726,224 unique commercialand 3,188,825 unique MA members. Since membersmight be enrolled for more than one year, the average

Kosloff et al. Chiropractic & Manual Therapies (2015) 23:19 Page 2 of 10

annual commercial membership was 14.7 million mem-bers and the average annual MA membership was 1.4million members over the three year study period,which is comparable to ~5% of the total US populationbased on the data available from US Census Bureau [35].Administrative claims data were used to identify cases,as well as patient characteristics and health serviceutilization.The stroke cases included all patients admitted to an

acute care hospital with vertebrobasilar (VBA) occlusionand stenosis strokes as defined by ICD-9 codes of 433.0,433.01, 433.20, and 433.21 during the study period. Pa-tients with more than one admission for a VBA strokewere excluded from the study. For each stroke case, fourage and gender matched controls were randomly se-lected from sampled qualified members. Both cases andcontrols were randomly sorted prior to the matchingusing a greedy matching algorithm [36].

ExposuresThe index date was defined as the date of admission forthe VBA stroke. Any encounters with a chiropractor or aprimary care physician (PCP) prior to the index datewere considered as exposures. To evaluate the impact ofchiropractic and PCP treatment, the designated hazardperiod in this study was zero to 30 days prior to theindex date. For the PCP analysis, the index date was ex-cluded from the hazard period since patients might con-sult PCPs after having a stroke. The standard healthplan coverage included a limit of 20 chiropractic visits.In rare circumstances a small employer may have se-lected a 12-visit limit. An internal analysis (data notshown) revealed that 5% of the combined (commercialand MA) populations reached their chiropractic visitlimits. Instances of an employer not covering chiropracticcare were estimated to be so rare that it would have hadno measureable impact on the analysis. There were nolimits on the number of reimbursed PCP visits per year.

AnalysesTwo sets of similar analyses were performed, one for thecommercially insured population and one for the MApopulation. In each set of analyses, conditional logisticregression models were used to examine the associationbetween the exposures and VBA strokes. To measurethe association, we estimated the odds ratio of havingthe VBA stroke and the effect of total number of chiro-practic visits and PCP visits within the hazard period.The analyses were applied to different hazard periods,including one day, three days, seven days, 14 days and30 days for both chiropractic and PCP visits. The resultsof the chiropractic and PCP visit analyses were thencompared to find evidence of excess risk of havingstroke for patients with chiropractic visits during the

hazard period. Previous research has indicated that mostpatients who experience a vertebral artery dissection areunder the age of 45. Therefore, in order to investigatethe impact of exposure on the population at differentages, separate analyses were performed on patientsstratified by age (under 45 years and 45 years and up)for the study of the commercial population. The numberof visits within the hazard period was entered as a con-tinuous variable in the logistic model. The chi square testwas used to analyze the proportion of co-morbidities incases as compared to controls.A secondary analysis was performed to evaluate the

relevance of using chiropractic visits as a proxy forspinal manipulation. The commercial and MA databaseswere queried to identify the proportions of cases of VBAstroke and matched controls for which at least onechiropractic spinal manipulative treatment proceduralcode (CPT 98940 – 98942) was or was not recorded.The analysis also calculated the use of another manualtherapy code (CPT 97140), which may be employed bychiropractors as an alternative means of reporting spinalmanipulation.

EthicsThe New England Institutional Review Board (NEIRB) de-termined that this study was exempt from ethics review.

ResultsThe commercial study sample included 1,159 VBAstroke cases over the three year period and 4,633 ageand gender matched controls. The average age of the pa-tients was 65.1 years and 64.8% of the patients weremale (Table 1). The prevalence rate of VBA stroke in thecommercial population was 0.0032%.There were a total of 670 stroke cases and 2,680

matched controls included in the MA study. The aver-age patient age was 76.1 years and 58.6% of the patientswere male (Table 2). For the MA population, the preva-lence rate of VBA stroke was 0.021%.Claims during a one year period prior to the index

date were extracted to identify comorbid disorders. Boththe commercial and MA cases had a high percentage ofcomorbidities, with 71.5% of cases in the commercialstudy and 88.5% of the cases in the MA study reportingat least one of the comorbid conditions (Table 3). Six co-morbid conditions of particular interest were identified,including hypertensive disease (ICD-9 401–404), ischemic

Table 1 Age and gender of cases and controls(Commercial)

Variable Cases (n = 1159) Controls (n = 4633)

Age: mean (median) 65.1 (64.7) 65.1 (64.7)

Males: n (%) 751 (64.8) 3001 (64.8)


heart disease (ICD-9 410–414), disease of pulmonary cir-culation (ICD-9 415–417), other forms of heart disease(ICD-9 420–429), pure hypercholesterolemia (ICD-9272.0) and diseases of other endocrine glands (ICD-9249–250). There were statistically significant differences(p = <0.05) between groups for most comorbidities. Greaterproportions of comorbid disorders (p = <0.0001) werereported in the commercial and MA cases for hyper-tensive disease, heart disease and endocrine disorders(Table 3). The commercial cases also showed a largerproportion of diseases of pulmonary circulation, whichwas statistically significant (p = 0.0008). There were nosignificance differences in pure hypercholesterolemiafor either the commercial or MA populations. Overall,cases in both the commercial and MA populationswere more likely (p = <0.0001) to have at least one co-morbid condition.Among the commercially insured, 1.6% of stroke cases

had visited chiropractors within 30 days of being admit-ted to the hospital, as compared to 1.3% of controls visit-ing chiropractors within 30 days prior to their indexdate. Of the stroke cases, 18.9% had visited a PCP within30 days prior to their index date, while only 6.8% of con-trols had visited a PCP (Table 4). The proportion of ex-posures for chiropractic visits was lower in the MAsample within the 30-day hazard period (cases = 0.3%;controls = 0.9%). However, the proportion of exposuresfor PCP visits was higher, with 21.3% of cases havingPCP visits as compared to12.9% for controls (Table 5).The results from the analyses of both the commer-

cial population and the MA population were similar(Tables 6, 7 and 8). There was no association betweenchiropractic visits and VBA stroke found for the

overall sample, or for samples stratified by age. Noestimated odds ratio was significant at the 95% confi-dence level. MA data were insufficient to calculatestatistical measures of association for hazard periodsless than 0–14 days for chiropractic visits. Whenstratified by age, the data were too sparse to calculatemeasures of association for hazard periods less than0–30 days in the commercial population. The datawere too few to analyze associative risk by headacheand/or neck pain diagnoses (data not shown).These results showed there is an association existing

between PCP visits and VBA stroke incidence regardlessof age or length of hazard period. A strong associationwas found for those visits close to the index date (OR11.56; 95% CI 6.32-21.21) for all patients with a PCPvisit within 0–1 day hazard period in the commercialsample. There was an increased risk of VBA stroke asso-ciated with each PCP visit within 30-days prior to theindex date for MA patients (OR 1.51; 95% CI 1.32-1.73)and commercial patients (OR 2.01; 95% CI 1.77-2.29).The findings of the secondary analysis showed – that

of 1159 stroke cases from commercial population –there were a total of 19 stroke cases associated withchiropractic visits for which 13 (68%) had claims docu-mentation indicating chiropractic SMT was performed.For the control group of the commercial cohort, 62 of4633 controls had claims of any kind of chiropracticvisits and 47 of 4633 controls had claims of SMT. In thecommercial control group, 47 of 62 DC visits (76%) in-cluded SMT in the claims data. Only 1 of 2 stroke casesin the MA population included SMT in the claims data.For the MA cohort, 21 of 24 control chiropractic visits(88%) included SMT in the claims data (Table 9).None of the stroke cases in either population included

CPT 97140 as a substitute for the more conventionally re-ported chiropractic manipulative treatment proceduralcodes (98940 – 98942). For the control groups, there werethree instances where CPT 97140 was reported withoutCPT 98940 – 98942 in the commercial population. TheCPT code 97140 was not reported in MA control cohort.

Table 2 Age and gender of cases and controls (Medicare)

Variable Cases (n = 670) Controls (n = 2680)

Age: mean (median) 76.1 (76.2) 76.1 (76.2)

Males: n (%) 393 (58.6) 1572 (58.6)

Table 3 Comorbid conditions

Conditions n (%) Commercial Medicare

Cases (n = 1159) Controls (n = 4633) p-value Cases (n = 670) Controls (n = 2680) p-value

Hypertensive disease 767 (66.2) 2078 (44.9) <0.0001 554 (82.7) 1721 (64.2) <0.0001

Ischemic heart disease 300 (25.9) 638 (13.8) <0.0001 258 (38.5 ) 563 (21.0) <0.0001

Diseases of pulmonary circulation 29 (2.5) 55 (1.2) 0.0008 18 (2.7) 70 (2.6) 0.9140

Other forms of heart disease 357 (30.8) 800 (17.3) <0.0001 306 (45.7) 713 (26.6) <0.0001

Pure Hypercholesterolemia 9 (0.8) 24 (0.5) 0.2957 6 (0.9) 26 (1.0) 0.8590

Diseases of other endocrine glands 319 (27.5) 754 (16.3) <0.0001 285 (42.5) 740 (27.6) <0.0001

At least one of the conditions 829 (71.5) 2317 (50.0) <0.0001 593 (88.5) 1885 (70.3) <0.0001


DiscussionThe primary aim of the present study was to investigatethe association between chiropractic manipulative treat-ment and VBA stroke in a sample of the U.S. population.This study was modelled after a case–control design previ-ously conducted for a Canadian population [32]. Adminis-trative data for enrollees in a large national health careinsurer were analyzed to explore the occurrence of VBAstroke across different time periods of exposure to chiro-practic care in comparison with PCP care.Unlike Cassidy et al. [32] and most other case–control

studies [33,37,38], our results showed there was no sig-nificant association between VBA stroke and chiroprac-tic visits. This was the case for both the commercial andMA populations. In contrast to two earlier case–controlstudies [32,33], this lack of association was found to be

irrespective of age. Although, our results (Table 8) didlend credence to previous reports that VBA stroke oc-curs more frequently in patients under the age of45 years. Additionally, the results from the present studydid not identify a relevant temporal impact. There was nosignificant association, when the data were sufficient tocalculate estimates, between chiropractic visits and strokeregardless of the hazard period (timing of most recent visitto a chiropractor and the occurrence of stroke).There are several possible reasons for the variation in

results with previous similar case–control studies. Theyounger (<45 yrs.) commercial cohort that receivedchiropractic care in our study had noticeably fewercases. The 0–30 days hazard period included only 2VBA stroke cases. There were no stroke cases for otherhazard periods in this population. In contrast, earlierstudies reported sufficient cases to calculate risk esti-mates for most hazard periods [32,33].Another factor that potentially influenced the differ-

ence in results concerns the accuracy of hospital claimsdata in the U.S. vs. Ontario, Canada. The source popula-tion in the Province of Ontario was identified, in part,from the Discharge Abstract Database (DAD). The DADincludes hospital discharge and emergency visit diagno-ses that have undergone a standardized assessment by amedical records coder [39]. To the best of our know-ledge, similar quality management practices were notroutinely applied to hospital claims data used in sour-cing the population for our study.An additional reason for the disparity in results may

be due to differences in the proportions of chiropracticvisits where SMT was reportedly performed. Our studyshowed that SMT was not reported by chiropractors inmore than 30% of commercial cases. It is plausible that anumber of the cases in earlier studies also did not

Table 4 Chiropractic and PCP visits prior to the index date (Commercial)

Exposures All Age <45 yr Age ≥45 yr

Cases (n = 1159) Controls (n = 4633) Cases (n = 98) Controls (n = 392) Cases (n = 1061) Controls (n = 4241)

Most recent DC Visit

0-1 day: n (%) 3 (0.3) 11 (0.2) * * 3 (0.3) 11 (0.3)

0-3 days: n (%) 6 (0.5) 21 (0.5) * 1 (0.3) 6 (0.6) 20 (0.5)

0-7 days: n (%) 8 (0.7) 31 (0.7) * 1 (0.3) 8 (0.8) 30 (0.7)

0-14 days: n (%) 9 (0.8) 44 (0.9) * 3 (0.8) 9 (0.8) 41 (1.0)

0-30 days: n (%) 19 (1.6) 62 (1.3) 2 (2.0) 7 (1.8) 17 (1.6) 55 (1.3)

Most recent PCP Visit

1-1 day: n (%) 41 (3.5) 15 (0.3) 4 (4.1) 1 (0.3) 37 (3.5) 14 (0.3)

1-3 days: n (%) 78 (6.7) 41 (0.9) 8 (8.2) 2 (0.5) 70 (6.6) 39 (0.9)

1-7 days: n (%) 115 (9.9) 93 (2.0) 10 (10.2) 4 (1.0) 105 (9.9) 89 (2.1)

1-14 days: n (%) 157 (13.5) 165 (3.6) 12 (12.2) 15 (3.8) 145 (13.7) 150 (3.5)

1-30 days: n (%) 219 (18.9) 316 (6.8) 23 (23.5) 29 (7.4) 196 (18.5) 287 (6.8)

*Insufficient data to compute an estimate.

Table 5 Chiropractic and PCP visits prior the index date(Medicare)

Exposures Cases (n = 670) Controls (n = 2680)

Most recent DC Visit

0-1 day: n (%) * 4 (0.1)

0-3 days: n (%) * 8 (0.3)

0-7 days: n (%) * 9 (0.3)

0-14 days: n (%) 1 (0.1) 15 (0.6)

0-30 days: n (%) 2 (0.3) 24 (0.9)

Most recent PCP Visit

1-1 day: n (%) 16(2.4) 18 (0.7)

1-3 days: n (%) 30 (4.5) 36 (1.3)

1-7 days: n (%) 55 (8.2) 97 (3.6)

1-14 days: n (%) 90 (13.4) 183 (6.8)

1-30 days: n (%) 143 (21.3) 346 (12.9)



include SMT as an intervention. Differences betweenstudies in the proportion of cases reporting SMT mayhave affected the calculation of risk estimates.Also, there were an insufficient number of cases hav-

ing cervical and/or headache diagnoses in our study.Therefore, our sample population may have includedproportionally less cases where cervical manipulationwas performed.Our results were consistent with previous findings

[32,33] in showing a significant association between PCPvisits and VBA stroke. The odds ratios for any PCP visitincrease dramatically from 1–30 days to 1–1 day (Tables 6and 7). This finding is consistent with the hypothesis thatpatients are more likely to see a PCP for symptoms relatedto vertebral artery dissection closer to the index date oftheir actual stroke. Since it is unlikely that the servicesprovided by PCPs cause VBA strokes, the association

between recent PCP visits and VBA stroke is more likelyattributable to the background risk related to the naturalhistory of the condition [32].A secondary goal of our study was to assess the utility

of employing chiropractic visits as a surrogate for SMT.Our findings indicate there is a high risk of bias associ-ated with using this approach, which likely overesti-mated the strength of association. Less than 70% ofstroke cases (commercial and MA) associated withchiropractic care included SMT. A somewhat higherproportion of chiropractic visits included SMT for thecontrol groups (commercial = 76%; MA = 88%).There are plausible reasons that support these find-

ings. Internal analyses of claims data (not shown) con-sistently demonstrate that one visit is the most commonnumber associated with a chiropractic episode of care.The single visit may consist of an evaluation without treat-ment such as SMT. Further; SMT may have been viewedas contraindicated due to signs and symptoms of vertebralartery dissection (VAD) and/or stroke. This might explainthe greater proportion of SMT provided to control groupsin both the commercial and MA populations.Overall, our results increase confidence in the findings

of a previous study [32], which concluded there was noexcess risk of VBA stroke associated chiropractic carecompared to primary care. Further, our results indicatethere is no significant risk of VBA stroke associated withchiropractic care. Additionally, our findings highlight thepotential flaws in using a surrogate variable (chiropracticvisits) to estimate the risk of VBA stroke in associationwith a specific intervention (manipulation).Our study had a number of strengths and limitations.

Both case and control data were extracted fromthe same source population, which encompassednational health plan data for approximately 36 million

Table 6 Estimated odds ratios and 95% confidence interval (Commercial)

Exposures All Age < 45 yr Age > =45 yr

Odds ratio 95% CI Odds ratio 95% CI Odds ratio 95% CI

Any DC Visit

0-1 day 1.09 0.30-3.91 * * 1.09 0.30-3.91

0-3 days 1.14 0.46-2.83 * * 1.20 0.48-2.30

0-7 days 1.03 0.48-2.25 * * 1.07 0.49-2.33

0-14 days 0.82 0.40-1.68 * * 0.88 0.43-1.81

0-30 days 1.23 0.73-2.06 1.14 0.24-5.50 1.24 0.72-2.14

Any PCP Visit

1-1 day 11.56 6.30-21.21 16.00 1.79-143.2 11.22 5.96-21.11

1-3 days 7.75 5.29-11.35 16.00 3.40-75.35 7.31 4.93-10.86

1-7 days 5.23 3.95-6.93 10.00 3.14-31.88 5.00 3.73-6.68

1-14 days 4.24 3.36-5.35 3.72 1.62-8.53 4.29 3.37-5.46

1-30 days 3.22 2.66-3.89 4.08 2.17-7.68 3.14 2.58-3.83


Table 7 Estimated odds ratios and 95% CI (Medicare)

Exposures Odds ratio 95% CI

Any DC Visit

0-1 day * *

0-3 days * *

0-7 days * *

0-14 days 0.26 0.03-2.00

0-30 days 0.32 0.08-1.39

Any PCP Visit

1-1 day 3.66 1.85-7.26

1-3 days 3.38 2.07-5.51

1-7 days 2.37 1.68-3.34

1-14 days 2.09 1.60-2.73

1-30 days 1.81 1.46-2.25



commercial and 3 million MA members. A total of1,829 cases were identified, making this the largest case–control study to investigate the association betweenchiropractic manipulation and VBA stroke. Due to thenationwide setting and large sample size, our study likelyreduced the risk of bias related to geographic factors.However, there was a risk of selection bias – owing tothe data set being from a single health insurer – includ-ing income status, workforce participation, and links tohealth care providers and hospitals.Our study closely followed a methodological approach

that had previously been described [32], thus allowingfor more confident comparisons.The current investigation analyzed data for a number

of comorbid conditions that have been identified as po-tentially modifiable risk factors for a first ischemicstroke [40]. The differences between groups were statis-tically significant for most comorbidities. Informationwas not obtainable about behavioural comorbid factorse.g., smoking and body mass. With the exception ofhypertensive disease, there are reasons to question theclinical significance of these conditions in the occurrenceof ischemic stroke due to vertebral artery dissection. Alarge multinational case-referent study investigated the as-sociation between vascular risk factors (history of vasculardisease, hypertension, smoking, hypercholesterolemia, dia-betes mellitus, and obesity/overweight) for ischemic strokeand the occurrence of cervical artery dissection [41]. Onlyhypertension had a positive association (odds ratio 1.67;95% confidence interval, 1.32 to 2.1; P <0.0001) with cer-vical artery dissection.

While the effect of other unmeasured confounderscannot be discounted, there is reason to suspect the ab-sence of these data was not deleterious to the results.Cassidy, et al. found no significant differences in the re-sults their case-crossover design, which affords bettercontrol of unknown confounding variables, and the find-ings of their case–control study [32].Our results highlight just how unusual VBA stroke is

in the MA cohort (prevalence = 0.021%) and – evenmore so – for the commercial population (prevalence =0.0032%). As a result, some limitations of this study re-lated to the rarity of reporting VBA stroke events. Des-pite the larger number of cases, data were insufficient tocalculate estimates and confidence intervals for sevenmeasures of exposure (4 commercial and 3 MA) forchiropractic visits. Additionally, we were not able tocompute estimates specifically for headache and neckpain diagnoses due to small numbers. Confidence inter-vals associated with estimates tended to be wide makingthe results imprecise [42].There were limitations related to the use of adminis-

trative claims data. “Disadvantages of using secondarydata for research purposes include: variations in codingfrom hospital to hospital or from department to depart-ment, errors in coding and incomplete coding, for examplein the presence of comorbidities. Random errors in codingand registration of discharge diagnoses may dilute and at-tenuate estimates of statistical association” [43]. The recor-dings of unvalidated hospital discharge diagnostic codes forstroke have been shown to be less precise when comparedto chart review [44,45] and validated patient registries

Table 9 Chiropractic (DC) visits with spinal manipulative treatment (SMT)

Commercial Medicare

DC visit with SMT Any DC visit Total # in sample DC visit with SMT Any DC visit Total # in sample

Stroke cases 13 19 1159 1 2 670

Controls 47 62 4633 21 24 2680

All 60 81 5792 22 26 3350

Table 8 Odds ratio and 95% CI for association between # of exposures during 30-day hazard period

Exposures All cases Age <45 yr Age >45 yr

Odds ratio 95% CI Odds ratio 95% CI Odds ratio 95% CI

Commercial

Any DC* visit 1.03 0.86-1.26 1.32 0.64-2.71 1.01 0.81-1.25

Any PCP visit 2.01 1.77-2.29 2.38 1.55-3.66 1.97 1.72-2.26

Medicare

Any DC* visit 0.54 0.23-1.28

Any PCP visit 1.51 1.32-1.73

*DC = Chiropractic.


[43,46]. Cassidy, et al. [32] conducted a sensitivity analysisto determine the effect of diagnostic misclassification bias.Their conclusions did not change when the effects of mis-classification were assumed to be similarly distributed be-tween chiropractic and PCP cases.A particular limitation in using administrative claims

data is the paucity of contextual information surround-ing the clinical encounters between chiropractors/PCPsand their patients. Historical elements describing the oc-currence/absence of recent trauma or activities reportedin case studies [47-51] as potential risk factors for VBAstroke were not available in claims data. Confidence waslow concerning the ability of claims data to provide ac-curate and complete reporting of other health disorders,which have been described in case–control designs asbeing associated with the occurrence of VBA stroke e.g.,migraine [52] or recent infection [53]. Symptoms andphysical examination findings that would have permittedfurther stratification of cases were not reported in theclaims data.The reporting of clinical procedures using current pro-

cedural terminology (CPT) codes presented additionalshortcomings concerning the accuracy and interpretationof administrative data. One inherent constraint was thelack of anatomic specificity associated with the use ofstandardized procedural codes in claims data. Chiropracticmanipulative treatment codes (CPT 98940 – 98942) havebeen formatted to describe the number of spinal regionsreceiving manipulation. They do not identify the particularspinal regions manipulated.Also, treatment information describing the type(s) of

manipulation was not available. When SMT was re-ported, claims data could not discriminate among therange of techniques including thrust or rotational ma-nipulation, various non-thrust interventions e.g., mech-anical instruments, soft tissue mobilizations, muscleenergy techniques, manual cervical traction, etc. Manyof these techniques do not incorporate the same bio-mechanical stressors associated with the type of mani-pulation (high velocity low amplitude) that has beeninvestigated as a putative risk factor for VBA stroke[54-56]. It seems plausible that the utility of future VBAstroke research would benefit from explicit descriptions ofthe particular type of manipulation performed.Moreover, patient responses to care – including any

adverse events suggestive of vertebral artery dissectionor stroke-like symptoms – were not obtainable in thedata set used for the current study.In the absence of performing comprehensive clinical

chart audits, it is not possible to know from claims datawhat actually transpired in the clinical encounter. Fur-ther, chart notes may themselves be incomplete or other-wise fail to precisely describe the nature of interventions[57]. Therefore, manipulation codes represent surrogate

measures, albeit more direct surrogate measures, thansimply using the exposure to chiropractic visits.Our study was also limited to replication of the case–

control design described by Cassidy, et al. [32]. Forpragmatic reasons, we did not attempt to conduct acase-crossover design. While the addition of a case-crossover design would have provided better control ofconfounding variables, Cassidy, et al. [32] showed theresults were similar for both the case control and casecrossover studies.The findings of this case–control study and previous

retrospective research underscore the need to rethinkhow to better conduct future investigations. Researchersshould seek to avoid the use of surrogate measures oruse the least indirect measures available. Instead, thefocus should be on capturing data about the types of ser-vices and not the type of health care provider.In alignment with this approach, it is also important

for investigators to access contextual data (e.g., fromelectronic health records), which can be enabled byqualitative data analysis computer programs [58]. Theacquisition of the elements of clinical encounters – in-cluding history, diagnosis, intervention, and adverseevents – can provide the infrastructure for more action-able research. Because of the rarity of VBA stroke, largedata sets (e.g., registries) containing these elements willbe necessary to achieve adequate statistical power formaking confident conclusions.Until research efforts produce more definitive results,

health care policy and clinical practice judgments arebest informed by the evidence about the effectiveness ofmanipulation, plausible treatment options (includingnon-thrust manual techniques) and individual patientvalues [20].

ConclusionsOur findings should be viewed in the context of thebody of knowledge concerning the risk of VBA stroke.In contrast to several other case–control studies, wefound no significant association between exposure tochiropractic care and the risk of VBA stroke. Our sec-ondary analysis clearly showed that manipulation may ormay not have been reported at every chiropractic visit.Therefore, the use of chiropractic visits as a proxy formanipulation may not be reliable. Our results addweight to the view that chiropractic care is an unlikelycause of VBA strokes. However, the current study doesnot exclude cervical manipulation as a possible cause orcontributory factor in the occurrence of VBA stroke.

Competing interestsAll authors are employees of UnitedHealth Group – a U.S based commercialhealth care company. The authors declare that they have no othercompeting interests.


Authors' contributionsDE conceived of the study, and participated in its design and coordination.JT participated in the design of the study, performed the statistical analysisand helped to draft the manuscript. TMK participated in the design andcoordination of the study, and wrote the initial draft and revisions of themanuscript. WMB participated in the coordination of the study and thestatistical analysis, and helped to draft the manuscript. All authorscontributed to the interpretation of the data. All authors read and approvedthe final manuscript.

Author details1Optum Health – Clinical Programs at United Health Group, 11000 OptumCircle, Eden Prairie MN 55344, USA. 2Optum Health – Clinical Analytics atUnited Health Group, 11000 Optum Circle, Eden Prairie MN 55344, USA.

Received: 14 October 2014 Accepted: 28 April 2015

References1. Paulose R, Hertz R. The burden of pain among adults in the United States.

In Pfizer Facts. Edited by Pfizer Inc. 2008. [http://www.pfizer.com/files/products/PF_Pain.pdf] Accessed May 14, 2014.

2. Carroll L, Hogg-Johnson S, van der Velde G, Haldeman S, Holm L, CarrageeE, et al. Bone and Joint Decade 2000–2010 Task Force on Neck Pain and ItsAssociated Disorders: Course and prognostic factors for neck pain in thegeneral population: results of the Bone and Joint Decade 2000–2010 TaskForce on Neck Pain and Its Associated Disorders. Spine (Phila Pa 1976).2008;33(4 Suppl):S75–82.

3. Hoy D, Protani M, De R, Buchbinder R. The epidemiology of neck pain.Best Pract Res Clin Rheumatol. 2010;24(6):783–92.

4. Jacobs J, Andersson G, Bell J, Weinstein S, Dormans J, Gnatz S, et al.Spine: low back and neck pain. In The Burden of Musculoskeletal Diseasesin the United States. Chapter 2. Edited by Bone and Joint Decade USA2002–2011. Rosemont, IL: The American Academy of Orthopaedic Surgeons;2008:21–56.

5. Christensen M, Kollasch M, Hyland J, Rosner A. Chapter 8 – Patient Conditions.In Practice Analysis of Chiropractic: A Project Report, Survey Analysis, andSummary of the Practice of Chiropractic Within the United States. Greeley, CO:The National Board of Chiropractic Examiners. 2010:95–120.

6. Christensen M, Kollasch M, Hyland J, Rosner A. Chapter 9 – Professionalfunctions and treatment procedures. In Practice Analysis of Chiropractic:A Project Report, Survey Analysis, and Summary of the Practice ofChiropractic Within the United States. Greeley, CO: The National Board ofChiropractic Examiners. 2010:121–136.

7. D’Sylva J, Miller J, Gross A, Burnie S, Goldsmith G, Graham N, et al.Manual therapy with or without physical medicine modalities for neck pain:a systematic review. Man Ther. 2010;15(4):415–33.

8. Gross A, Miller J, D’Sylva J, Burnie S, Goldsmith G, Graham N, et al.Manipulation or mobilisation for neck pain: A Cochrane review. Man Ther.2010;15(4):315–33.

9. Bryans R, Decina P, Descarreaux M, Duranleau M, Marcoux H, Potter B, et al.Evidence-based guidelines for the chiropractic treatment of adults withneck pain. J Manipulative Physiol Ther. 2014;37(1):42–63.

10. Bryans R, Descarreaux M, Duranleau M, Marcoux H, Potter B, Ruegg R, et al.Evidence-based guidelines for the chiropractic treatment of adults withheadache. J Manipulative Physiol Ther. 2011;34(5):274–89.

11. Childs J, Cleland J, Elliott J, Teyhen D, Wainner R, Whitman J, et al.Neck pain: clinical practice guidelines linked to the InternationalClassification of Functioning, Disability, and Health from the OrthopaedicSection of the American Physical Therapy Association. J Orthop Sports PhysTher. 2008;38(9):A1–A34.

12. Clar C, Tsertsvadze A, Court R, Hundt G, Clarke A, Sutcliffe P. Clinicaleffectiveness of manual therapy for the management of musculoskeletaland non-musculoskeletal conditions: systematic review and update of UKevidence report. Chiropr Man Therap. 2014;22(1):12.

13. Vincent K, Maigne J, Fischhoff C, Lanlo O, Dagenais S. Systematic review ofmanual therapies for nonspecific neck pain. Joint Bone Spine.2013;80(5):508–15.

14. Bronfort G, Assendelft W, Evans R, Haas M, Bouter L. Efficacy of spinalmanipulation for chronic headache: a systematic review.J Manipulative Physiol Ther. 2001;24(7):457–66.

15. Chaibi A, Tuchin P, Russell M. Manual therapies for migraine: a systematicreview. J Headache Pain. 2011;12(2):127–33.

16. Racicki S, Gerwin S, Diclaudio S, Reinmann S, Donaldson M. Conservativephysical therapy management for the treatment of cervicogenic headache:a systematic review. J Man Manip Ther. 2013;21(2):113–24.

17. Cassidy J, Bronfort G, Hartvigsen J. Should we abandon cervical spinemanipulation for mechanical neck pain? No BMJ. 2012;344, e3680.

18. Wand B, Heine P, O’Connell N. Should we abandon cervical spinemanipulation for mechanical neck pain? Yes BMJ. 2012;344, e3679.

19. Moloo J. What's the Best Approach for Managing Neck Pain? NEJM JournalWatch 2012. [http://www.jwatch.org/jw201202090000004/2012/02/09/whats-best-approach-managing-neck-pain] Accessed May 14, 2014.

20. Schneider M, Weinstein S, Chimes G. Cervical manipulation for neck pain.PM&R. 2012;4(8):606–12.

21. Biller J, Sacco R, Albuquerque F, Demaerschalk B, Fayad P, Long P, et al.Cervical arterial dissections and association with cervical manipulative therapy:a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2014, Epub ahead of print.

22. American Chiropractic Association: ACA Response to AHA Statement onNeck Manipulation. 2014 (Aug 7). [http://www.acatoday.org/press_css.cfm?CID=5534] Accessed August 15, 2014.

23. American Physical Therapy Association: APTA responds to American HeartAssociation cervical manipulation paper. 2014 (Aug 7). [http://www.apta.org/Media/Releases/Consumer/2014/8/7/] Accessed August 15, 2014.

24. Kardys JA. Declaratory ruling regarding informed consent. Connecticut StateBoard of Chiropractic Examiners – State of Connecticut Department ofPublic Health. 2010. [http://www.ctchiro.com/upload/news/44_0.pdf]Accessed May 14, 2014.

25. Wangler M, Fujikawa R, Hestbæk L, Michielsen T, Raven T, Thiel H, et al.Creating European guidelines for Chiropractic Incident Reportingand Learning Systems (CIRLS): relevance and structure. Chiropr ManTherap. 2011;19:9.

26. Berger S: How safe are the vigorous neck manipulations done bychiropractors? Washington Post 2014 (Jan. 6). [http://www.washingtonpost.com/national/health-science/how-safe-are-the-vigorous-neck-manipulations-done-by-chiropractors/2014/01/06/26870726-5cf7-11e3-bc56-c6ca94801fac_story.html] Accessed January 10, 2014.

27. Group wants provincial ban on some neck manipulation by chiropractors.Winnipeg Free Press 2012 (Oct 4). [http://www.winnipegfreepress.com/local/Group-wants-provincial-ban-on-some-neck-manipulation-by-chiropractors-172692471.htm] Accessed May 14, 2014.

28. Gouveia L, Castanho P, Ferreira J. Safety of chiropractic interventions: asystematic review. Spine (Phila Pa 1976). 2009;34(11):E405–13.

29. Carlesso L, Gross A, Santaguida P, Burnie S, Voth S, Sadi J. Adverse events associatedwith the use of cervical manipulation and mobilization for the treatment of neckpain in adults: a systematic review. Man Ther. 2010;15(5):434–44.

30. Chung C, Côté P, Stern P, L'Espérance G. The association between cervicalspine manipulation and carotid artery dissection: a systematic review of theliterature. J Manipulative Physiol Ther 2014, [Epub ahead of print].

31. Haynes M, Vincent K, Fischhoff C, Bremner A, Lanlo O, Hankey G. Assessingthe risk of stroke from neck manipulation: a systematic review. Int J ClinPract. 2012;66(10):940–7.

32. Cassidy J, Boyle E, Cote P, He Y, Hogg-Johnson S, Silver F, et al. Risk ofvertebrobasilar stroke and chiropractic care: results of a population-basedcase–control and case-crossover study. Spine (Phila Pa 1976).2008;33 Suppl 4:S176–83.

33. Rothwell D, Bondy S, Williams J. Chiropractic manipulation and stroke:a population-based case–control study. Stroke. 2001;32(5):1054–60.

34. Choi S, Boyle E, Côté P, Cassidy JD. A population-based case-series ofOntario patients who develop a vertebrobasilar artery stroke after seeing achiropractor. J Manipulative Physiol Ther. 2011;34(1):15–22.

35. U.S. Census Bureau: State and County QuickFacts. Data derived fromPopulation Estimates, American Community Survey, Census of Populationand Housing, State and County Housing Unit Estimates, County BusinessPatterns, Nonemployer Statistics, Economic Census, Survey of BusinessOwners, Building Permits. 2014 (rev July 8). [http://quickfacts.census.gov/qfd/states/00000.html] Accessed August 19, 2014.

36. Kosanke J, Bergstralh E. GMatch Macro (SAS program): Mayo Clinic Collegeof Medicine. 2004. [http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biomedical-statistics-informatics/software/locally-written-sas-macros] Accessed June 6, 2014.


http://www.pfizer.com/files/products/PF_Pain.pdf

http://www.pfizer.com/files/products/PF_Pain.pdf

http://www.jwatch.org/jw201202090000004/2012/02/09/whats-best-approach-managing-neck-pain

http://www.jwatch.org/jw201202090000004/2012/02/09/whats-best-approach-managing-neck-pain

http://www.acatoday.org/press_css.cfm?CID=5534

http://www.acatoday.org/press_css.cfm?CID=5534

http://www.apta.org/Media/Releases/Consumer/2014/8/7/

http://www.apta.org/Media/Releases/Consumer/2014/8/7/

http://www.ctchiro.com/upload/news/44_0.pdf

http://www.washingtonpost.com/national/health-science/how-safe-are-the-vigorous-neck-manipulations-done-by-chiropractors/2014/01/06/26870726-5cf7-11e3-bc56-c6ca94801fac_story.html




http://www.winnipegfreepress.com/local/Group-wants-provincial-ban-on-some-neck-manipulation-by-chiropractors-172692471.htm



http://quickfacts.census.gov/qfd/states/00000.html

http://quickfacts.census.gov/qfd/states/00000.html

http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biomedical-statistics-informatics/software/locally-written-sas-macros



37. Smith W, Johnston S, Skalabrin E, Weaver M, Azari P, Albers G, et al. Spinalmanipulative therapy is an independent risk factor for vertebral arterydissection. Neurology. 2003;60(9):1424–8.

38. Engelter S, Grond-Ginsbach C, Metso T, Metso A, Kloss M, Debette S, et al.Cervical Artery Dissection and Ischemic Stroke Patients Study Group:Cervical artery dissection: trauma and other potential mechanical triggerevents. Neurology. 2013;80(21):1950–7.

39. Ardal S, Baigent L, Bains N, Hay C, Lee P, Loomer S: The health analyst’s toolkit.Ministry of Health and Long-Term Care Health Results Team - InformationManagement. Ontario (CA) 2006 (January) [http://www.health.gov.on.ca/transformation/providers/information/resources/analyst_toolkit.pdf]Accessed January 12, 2015.

40. Sacco RL, Benjamin EJ, Broderick JP, Dyken M, Easton JD, Feinberg WM, etal. American Heart Association Prevention Conference. IV. Prevention andrehabilitation of stroke. Risk factors. Stroke. 1997;28(7):1507–17.

41. Debette S, Metso T, Pezzini A, Abboud S, Metso A, Leys D, et al. CervicalArtery Dissection and Ischemic Stroke Patients (CADISP) Group: Associationof vascular risk factors with cervical artery dissection and ischemic stroke inyoung adults. Circulation. 2011;123(14):1537–44.

42. Guyatt G, Oxman A, Kunz R, Brozek J, Alonso-Coello P, Rind D, et al.GRADE guidelines 6. Rating the quality of evidence – imprecision.J Clin Epidemiol. 2011;64(12):1283–93.

43. Krarup L, Boysen G, Janjua H, Prescott E, Truelsen T. Validity of strokediagnoses in a National Register of Patients. Neuroepidemiology.2007;28(3):150–4.

44. Goldstein L. Accuracy of ICD-9-CM coding for the identification ofpatients with acute ischemic stroke: effect of modifier codes. Stroke.1998;29(8):1602–4.

45. Liu L, Reeder B, Shuaib A, Mazagri R. Validity of stroke diagnosis on hospitaldischarge records in Saskatchewan, Canada: implications for strokesurveillance. Cerebrovasc Dis. 1999;9(4):224–30.

46. Ellekjaer H, Holmen J, Krüger O, Terent A. Identification of incident stroke inNorway: hospital discharge data compared with a population-based strokeregister. Stroke. 1999;30(1):56–60.

47. Braksiak R, Roberts D. Amusement park injuries and deaths. An Emerg Med.2002;39(1):65–72.

48. Dittrich R, Rohsbach D, Heidbreder A, Heuschmann P, Nassenstein I,Bachmann R, et al. Mild mechanical traumas are possible risk factors forcervical artery dissection. Cerebrovasc Dis. 2007;23(4):275–81.

49. Mas J, Bousser M, Hasboun D, Laplane D. Extracranial vertebral arterydissection: a review of 13 cases. Stroke. 1987;18(6):1037–47.

50. Slankamenac P, Jesic A, Avramov P, Zivanovic Z, Covic S, Till V. Multiple cervicalartery dissection in a volleyball player. Arch Neuro. 2010;67(8):1024–5.

51. Weintraub M. Beauty parlor stroke syndrome: report of five cases. JAMA.1993;269(16):2085–6.

52. Tzourio C, Benslamia L, Guilllon B, Aïdi S, Bertrand M, Berthet K, et al.Migraine and the risk of cervical artery dissection: a case control study.Neurology. 2002;59(3):435–7.

53. Guillon B, Berthet K, Benslamia L, Bertrand M, Bousser M, Tzourio C.Infection and the risk of cervical artery dissection: a case–control study.Stroke. 2003;34(7):e79–81.

54. Symons B, Leonard TR, Herzog W. Internal forces sustained by the vertebralartery during spinal manipulative therapy. J Manip Physiol Ther.2002;25(8):504–10.

55. Wuest S, Symons B, Leonard T, Herzog W. Preliminary report: biomechanicsof vertebral artery segments C1-C6 during cervical spinal manipulation.J Manip Physiol Ther. 2010;33(4):273–8.

56. Herzog W, Leonard TR, Symons B, Tang C, Wuest S. Vertebral artery strainsduring high-speed, low amplitude cervical spinal manipulation.J Electromyogr Kinesiol. 2012;22(5):747–51.

57. Centers for Medicare & Medicaid: Comprehensive error rate testing (CERT).2015 (Jan. 15). [http://www.cms.gov/Research-Statistics-Data-and-Systems/Monitoring-Programs/Medicare-FFS-Compliance-Programs/CERT/index.html?redirect=/cert] Accessed February 4, 2015.

58. Welsh E: Dealing with data: using NVivo in the qualitative data analysisprocess. Forum: Qualitative Social Research 2002, 3(2): Art. 26 [http://nbn-resolving.de/urn:nbn:de:0114-fqs0202260] Accessed February 4, 2015.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit


http://www.health.gov.on.ca/transformation/providers/information/resources/analyst_toolkit.pdf

http://www.health.gov.on.ca/transformation/providers/information/resources/analyst_toolkit.pdf

http://www.cms.gov/Research-Statistics-Data-and-Systems/Monitoring-Programs/Medicare-FFS-Compliance-Programs/CERT/index.html?redirect=/cert



http://nbn-resolving.de/urn:nbn:de:0114-fqs0202260

http://nbn-resolving.de/urn:nbn:de:0114-fqs0202260

the following is an actual transcript of the chirosecure live event … · 2017-03-22 · dr. clum:...

Documents