From: Dr. Daniel Redwood
Subject: Methodological Issues Raised by NEJM Studies
Date: Thursday, October 08, 1998 8:19 PM
The NEJM articles are quite thought-provoking. I have appreciated the initial thoughts expressed by list members thus far, and I'd like to add a few of my own.
In the asthma article (which I have only seen in abstract form thus far), the biggest question from my point of view is: what constitutes a true "sham" adjustment? The findings, with 34% improvement in symptom severity and decreased medication use in both sham and "real" adjusting groups -- which I assume is far beyond what would usually be expected as the natural course of the illness for patients who had already undergone extensive medical treatment -- are strongly reminiscent of some things I've been reading recently in the acupuncture literature. For example:
"Sham acupuncture was initially assumed by most investigators to be ineffective and, therefore, ideal as a placebo. However, in 1983 Lewith and Machin (1) pointed out that sham acupuncture appeared to have an analgesic effect in 40-50% of patients
compared with 60% for real acupuncture."(2)
The assumption that the sham manipulation in the NEJM study has no effect is just that, an assumption. If minimal-force chiropractic techniques like Toftness or DNFT have some therapeutic effect, then it is entirely possible that the "sham"
manipulation in the NEJM study was an accidentally effective treatment, no matter how surprising this may seem. Was any of
this discussed by the author of the NEJM study in his analysis of his data? If not, it's a significant oversight.
In a roundtable discussion at an NIH conference on acupuncture (published in 1996 in the Journal of Alternative and
Complementary Medicine, of which I am now associate editor), one of the participants said, "Some true points are not on the
meridians. If any sham point is more than 50% effective, then that is a new real point."(3) Similarly, whatever presumably
minimal physical contact the chiropractor applied as a sham in the NEJM study cannot be presumed to have no effect. Obviously, if the DC never touched the patient's back or neck, my point is moot. My assumption, however, is that there was contact applied.
Because chiropractors differ in the techniques they use, and (perhaps more significantly) in their technical mastery of their manual art, we are going to have to expect inconsistency from one study to the next. For the sake of discussion, consider the 1985 Cassidy-Kirkaldy-Willis study in Canadian Family Physician(4). The truly spectacular results from chiropractic treatment (approximately 80 percent of patients 'totally disabled' for an average of 7 years back on the job and doing well after 3 weeks of daily adjustments) reflect, in my view, not only the positive value of low back adjustments but also the exceptional skill of the particular chiropractor (I believe it was Dr. Cassidy). With chiropractic (as with other non-drug approaches including acupuncture, physical therapy, and massage) we are not dealing with a substance of standardized potency, but with a physical modality that varies from one adjustment to the next, and from one adjuster to the next.
These issues (sham vs. real, and individual variation) pose a genuine challenge in CAM research circles today, one with which some very good minds in other disciplines continue to struggle. In addressing the challenge of the NEJM studies, I think they are both worth pondering further.
Lewith Gt, Machin D:
Acupuncture compared with placebo in post-herpetic pain.
Lewith GT, Vincent C:
On the evaluation of the clinical effects of acupuncture: a problem reassessed and a framework for
J Altern Complement Med 1996;2(1):70-90
Hammerschlag R, Lewith GT, Bushar H et al:
Discussion of methodological considerations in human trials.
J Altern Complement Med 1996;2(1):91-100
Kirkaldy-Willis W, Cassidy J:
Spinal manipulation in the treatment of low back pain.
Can Family Phys 1985;31:535-540
One last thing. The Journal of Alternative and Complementary Medicine: Research on Practice, Paradigm and Policy is now
indexed on Medline and Index Medicus. Any DCs interested in submitting research papers, commentaries, or book reviews are
welcome to contact me.
Daniel Redwood, D.C.
Virginia Beach, VA
Return to the RESPONSES TO NEJM ARTICLES