I found, in the March 25, 2002 edition of Dynamic Chiropractic, a letter to the editor written by Paul Hooper,DC, concerning a piece I had written for DC: "Late Whiplash: The Controversy of Organic vs.Biopsychosocial Model." My longtime colleague from LACC (now Southern California University of Health Sciences) took me to task over this issue, and I am inclined to offer a brief rebuttal. In the same issue, John Lowe,MA,DC, chastened me for promoting what he referred to as a "logical fallacy." This also warrants a brief comment.
Dr. Hooper began by pointing out that he had been invited to write a part of a chapter for our latest textbook (Whiplash Injuries: the Cervical Acceleration/Deceleration Syndrome, 3rd edition, Lippincott Williams & Wilkins, 2001); his piece was "Psychosocial Factors of Chronic Pain." Clearly it was the conclusion in my article that the biopsychosocial model (let's call it BM) holds very little weight in the context of whiplash, and Dr. Hooper wished to disagree in a collegiate way. Witty comments notwithstanding (and I can appreciate wit), the fact that he wrote something in our book seems hardly a valid argument to support his position.
Perhaps at the very heart of this nettling issue of BM is the question of its relevance, or value. If it can be shown to be useful, conceptually, for automotive safety engineers, trauma epidemiologists, and the thousands of physicians who treat victims of motor vehicle crashes, then perhaps we should put the model into the crucible of science and see what boils out. Allow me to make an example of my point about value or relevance. In nearly every clinical or epidemiological study of crash trauma, researchers ask questions about age, sex, and a number of other variables that are beyond our control to change.
Yes, females are injured twice as often as males, and older persons are at more risk than younger persons. But if the goal of research is to alter the risks for injury or death, we need to find out why there is a difference in risk between the sexes and age groups. And we need to concentrate on other variables that can be changed to reduce the risk for injury or death. We can't simply take away or restrict the driving privileges of females and older persons. What we can do is make cars safer for females and older drivers by understanding why they are at more risk. We can't always stop people from crashing, but we can make the car more crashworthy; likewise, with the case of the BM. We can't limit a person's driving privileges because of his or her low self-esteem or educational attainment, or because he or she is unhappy with a job. Also, we cannot expect to change these variables in most people.
So, what is the relevance of this model? If the BM is valid, why does it single out low back and neck pain, two of the biggest cost scourges big business finds itself saddled with? Why don't we expect the outcome in persons who break their legs in skiing accidents to be affected with the BM? Why is it a "nonissue" outside the context of the medicolegal (W/C and PI) world?
While proponents often argue that the BM can be helpful in getting the proper care for patients, there's no good evidence to support this. Even if someone with late whiplash has psychological problems, he or she still has neck pain, which can't be fixed with counseling.
Does counseling help in most cases? It is unknown. If you examine this issue objectively, it's clear that the model's real utility lies in the subtle art of dissimulation, in shifting the blame, and providing a plausible deniability and an excuse for looking in the wrong direction. Those of us in the research community suffer because funding agencies often are convinced that the "real" issues are not physical injuries, but some convoluted and complex elements of the human psyche: the need to be compensated, secondary gain issues, etc. Why spend money chasing phantoms? But when research is not funded, everybody potentially can suffer.
Of course, several scientists have skipped the questions of relevancy and value, and have tested the model anyway. The collective results, which are quite dismal, should be the requiem for the "little model that couldn't." But the politics of big business has kept its head above water, by encouraging through various means copious amounts of junk science that ostensibly lend it some credence.
Many scientists and authors also fail to fully understand the unfortunate confounding nature of the many studies, purporting to support this somewhat ethereal BM of pain. If one were to interview 100 randomly selected persons off the street, one would find many with low social or educational attainment, low self-esteem, unhappy marital relationships, and many of the other components of this variously defined BM, who also may be dissatisfied with their jobs, bosses, and coworkers. And only a few of them would have chronic pain. Who doesn't have at least one BM factor?
But if you were also to look at a large number of injured persons in car crashes and try to correlate outcome with these numerous factors, the mathematics of statistics predicts that you will eventually find a correlation between one or another, even when the correlation is due to chance. For example, if you count the number of blue jays and motorcycle crashes in Wisconsin, you'll find a correlation. The confounder is nice weather. Blue jays do not increase the risk of motorcycle crashes. If you compare shoe size to IQ, you'll also find a correlation. Children score lower than adults. Shoe size, however, is a confounder and not directly relevant to IQ. These are examples of what we call in epidemiology, "ecological fallacies." The BM literature is full of examples. While I don't doubt that some of these BM factors do adversely affect outcome, the question remains: How can this model help us to reduce risk for injury or death, and what utility does it hold for clinicians?
This literature is full of other problems that affect BM's external validity. For example, the supporting literature for this BM generally relies heavily on a variety of psychometric tools that have been validated in medically healthy populations, not persons with neuromusculoskeletal pain, for example. Persons with such injuries will score high on many subscales of such tests, such as the somatiform and hypochondriasis scales, because the tests presume that healthy people do not have headaches, neck pain, etc. In order to apply these tests to persons from these injured populations, the tests must first be validated in those populations to determine whether they are actually capable of measuring what they purport to find.
Moreover, quite a few studies within this whiplash population have clearly shown that these "psychological" measures are not independent and that psychological scores improve in parallel with reductions in somatic complaints. It has been irrefutably shown in several recent studies that these complaints cause the abnormal psychiatric scores: not the obverse. Thus, many of the assumptions of the BM are wrong-at least in the context of late whiplash.
Of the studies I have seen that have found a slightly higher incidence of one or more of these ill-defined factors in the groups that had poor outcomes, it is interesting that there is little agreement between studies, with some finding educational attainment the only factor; other factors included low income, unemployment, etc. This, to a scientist, is far from satisfying, since reproducibility and repeatability are crucial components of a valid hypothesis. And again, many of these can also be confounders.
For example, persons from lower income groups may have a more difficult time getting medical care, paying for it, etc. This can result in a greater tendency for legal adventure which, in turn, can result in their being selected for inclusion in some studies. And very few of the studies I'm referring to have carefully stratified the subjects according to severity of injury, further obfuscating the results. They are often examples of irrelevant questions begging the wrong answers.
I would estimate that at least half of the references to these numerous biopsychosocial factors are entirely editorial in nature, yet are then cited as authoritative by the next author, and the next after the next. Validity is gained neither by the magic wand of publication nor by repetition, except in certain circles outside of science.
The bottom line is that in this particular area, there is no room for a realistic debate. The literature simply fails to support the BM as an "important consideration" in the context of whiplash. Of course, I don't disagree with Dr. Hooper that there are persons with pre-existing psychological disorders. And I also don't doubt that they are likely to have something of a more difficult recovery than, say a Navy SEAL. The gripe I have is with those who want only to apply the fait accompli and find one of these factors, and then disregard these patients, or, worse yet, the condition, in general, out of hand. And, as one who is rather heavily invested in this research area of whiplash, and who frequently descends from the old ivory tower of academia and testifies in court, I can assure Dr. Hooper that this is a problem. Until there is formal scientific support to suggest that this phenomenon really is an important factor, and that it has a bona fide relevancy, let's just hold the notion in abeyance.
In his article, "Spine Lesions as the Most Common Source of Posttraumatic Pain: an Institutionalized Fallacy" (www.chiroweb.com/archives/20/07/13.html), Dr. Lowe takes exception to a statement I made in the same article we've been discussing. I wrote, "Rather elegant clinical research has also pointed to the facet joint as a key player in the genesis of neck pain following whiplash, being responsible for some 60-70 percent of it." My meaning in this was that 60-70 percent of late whiplash, the title of the article identifying this as the subject of discussion was the result of injury to the facet joint.
As I read it, I felt that Dr. Lowe's article was something of a catharsis to vent a frustration that's been long festering in him regarding the tendency for many writers and researchers to promote what he refers to as a "logical fallacy." So, I am in good company: My friends Nick Bogduk, Steve Troyanovich, Dan Murphy, and Don Harrison are all codefendants in this conspiracy, but using different tools. So I'll let them defend their own honor.
Dr. Lowe argues that this logical fallacy stems from making unfair generalizations from research. In this he is correct; it is inappropriate to make inferences from one population to the next, or to extrapolate the findings between groups that may differ in some important ways. I wouldn't refer to it as a logical fallacy, because the error does not prove that such extrapolation is necessarily invalid or illogical. But that's another issue.
He writes: "Like Drs. Bogduk, Harrison, Troyanovich, and Murphy, Dr. Croft failed to distinguish between the general population of whiplash patients and those who were studied invasively, because they failed to respond to conservative care." His meaning, I gather, was that I was suggesting, tacitly at least, that Bogduk's group could stand as representative of all whiplash patients. Their group were people who had intractable chronic pain from whiplash, and therefore were representative of that group of whiplash patients who suffer from that more commonly termed "late (or chronic)" whiplash. I gather that Dr. Lowe believes that somehow the Newcastle group's patients were not even representative of all late whiplash patients, but as far as I know, they were.
Moreover, although I did not do so in the article in question, it is a fairly reasonable logical step to suppose that the lesions identified in those suffering from a more severe form of a condition are likely to be a source of pain in those with a less severe form. In any case, a huge body of literature, from cadaver studies; to animal studies; to clinical studies (such as that of Bogduk, et al., and some more recent similar studies); to autopsy studies supports the z-joint as a major player in the genesis of neck pain. That provides a lot of face validity.
I gather that Dr. Lowe believes that a large(er) portion of late whiplash pain is the result of myofascial disorders. I note, however, that this literature is particularly murky, on a scientific basis, with more than a few rheumatologists questioning the validity of the very condition itself. As for myself, I believe that the typical muscular pain syndromes seen after whiplash are the secondary result of injuries to the discs and z-joints. That doesn't mean that treatment of the muscles isn't helpful, so long as we remember that they are primarily "end-organs" in this chain.
Arthur C. Croft, DC,MS,MPH,FACO
Director, Spine Research Institute San Diego, California
President, Center for Research into Automotive Safety and Health (CRASH)
"The location of the pain is usually not its source."
I'd like to comment on John Lowe's recent article on the sources of pain ("Spine Lesions as the Most Common Source of Posttraumatic Pain: An Institutionalized Fallacy" (DC, March 25, 2002, or on line at www.chiroweb.com/archives/20/07/13.html). He points out the logical fallacies in Bogduk's work, noting that pain blockage via anesthetic injection was done strictly on patients who failed previous conservative care. He then notes how this research has penetrated the minds of the chiropractic profession, making us comfortable working only with the joints, ignoring other structures, such as the myofascia.
I'd like to point out another problem with the needle-injection type of research. Injecting a joint and blocking pain may help prove that the pain generator is the joint, or its capsule, but the question remains: Is this the underlying cause of the pain?
Our profession has long prided itself in looking for the cause, and vilifying medicine for simply treating symptoms. In practice, I believe we tend to oversimplify, as many of us are satisfied to correct spinal subluxations, which may or may not be the cause.
What if the pain generator is better seen as the pain "summator?" What if the subluxated joint is just the final point where all of the dysfunction gathers to create pain? "Pain is a liar" is an oft-quoted statement. The location of the pain is usually not its source.
Daily, I find painful and tender joints that immediately change in their tenderness and degree of restriction when I correct restrictions that are distant to the painful area. The simple version of this is right-sided sacroiliac pain and tenderness, caused by left-sided sacroiliac restriction, which creates compensatory hypermobility in the right sacroiliac. A more complicated, but more common version might be chronic right-sided sacroiliac pain and tenderness, contributed to by a dropped first metatarsal; an anteromedial talus; restriction within the intraosseous structures of the tibia; hip joint dysfunction; fascial restrictions affecting the kidney; and left sacroiliac and fourth lumbar restriction. This may be accompanied by tensor fascia lata tightness, and weakness of the gluteus and transverse abdominal muscles. Either of these cases could have the pain temporarily blocked by right sacroiliac anesthetic injection, but is this the cause?
Rather than fight over chickens and eggs, we should all commit to further study of functional anatomy and a deeper understanding of how the body works, and how it breaks down. We must at least understand dysfunction of spinal and extremity joints and the myofascial structures. I would encourage all of us to broaden our understanding of fascia to include the visceral fascia and the cranial structures.