Thank you for the opportunity to respond to the critiques by Drs.
We understand that our findings may appear inconsistent with the experiences of chiropractors and their patients who have felt benefited by chiropractic care.
Something that many clinicians, spinal surgeons and patients do not appreciate is that studies consistently have found that low back pain patients improve substantially over time even if they receive little or no treatment. At least for patients with acute, low back pain (who comprised the majority of patients in our study), the passage of time may be the most potent "therapy." This is why studies of back pain treatments that lack a comparison group invariably report high success rates for the treatments under study.
Understandably, both patients and providers often attribute this improvement to the treatment being administered at the time. In addition, positive patient and provider expectations that a treatment will be effective may contribute to healing, independent of the actual effectiveness of the specific treatment being used.
I have attempted to address the specific concerns (bolded) raised by the three reviewers in my responses below.
1) The study "is an inaccurate and unfortunate representation of the patients who normally seek chiropractic care for low back pain." The patients included in this study were not meant to represent patients who normally seek chiropractic care for low back pain. Our study focused on patients who sought medical care for low back pain, but who had not benefited substantially from one week of medical treatment. We wanted to know if this group of patients would benefit from referral for chiropractic manipulation or McKenzie physical therapy.
2. "In this trial, only one high-velocity technique (side-posture) was applied." We intentionally restricted the study to the chiropractic manipulative technique that our chiropractic advisers indicated was used most commonly by chiropractors, at least in Washington state. Discussions with the chiropractors invited to participate in the study confirmed that the side-posture high velocity thrust technique was in fact the one they used most often for low back pain.
3) Doctors of chiropractic were limited in the types of treatment they could provide while McKenzie practitioners were only allowed full scope of therapy options and specially trained for the purpose of the project by Robin McKenzie. The chiropractors in our study were selected because they were experienced providers (6-14 years in practice) who relied heavily on side posture, high velocity thrusts as their primary method of treatment. While it is true that the study chiropractors were unable to use physical therapy modalities, this constraint was imposed by Washington state law and not by the study design. We tried to make it clear to readers that this was a study of chiropractic manipulation and not of chiropractic care.
The chiropractors in the study were allowed to treat the subjects just as they treated their regular patients, except the number of visits was limited to nine over a one month period. The chiropractors (in consultation with their patients) chose to give the full 9 treatments to 25 percent of their patients. All four study chiropractors, who normally gave patients a brief exercise handout that did not emphasize extension exercises, were permitted by the study protocol to continue to give their patients an exercise handout consistent with their usual practices.
Although we designed the treatment protocols to deviate as little as possible from usual practice, we could not be certain that the study would not impose significant constraints on the providers. To evaluate this, at the end of the study we asked all the participating chiropractors (and McKenzie therapists) if they had "felt significantly constrained by the study protocol in providing high quality care to study subjects." Only the chiropractor with a masters degree in exercise physiology felt constrained, indicating that he would have recommended more intensive exercise programs for some of his patients. As noted in our article, however, patient outcomes were similar for all four chiropractors.
The McKenzie providers were also limited to nine visits over a one month period and, per state law, were proscribed from using spinal manipulation, which is used by McKenzie therapists outside Washington state for a small fraction of their back pain patients. In addition, the study protocol prohibited the McKenzie therapists from using physical therapy modalities, back classes and educational materials other than McKenzie's Treat Your Own Back book. Prior to their training by McKenzie faculty (which did not include training by Robin McKensie himself), the physical therapists in our study had received only a few weeks of training in the McKenzie method and had been using it for only about two years. Furthermore, there was real doubt prior to the study that the therapists were correctly employing the McKenzie Method. In fact, when the first McKenzie trainer observed the therapists prior to the start of the study, he said he would not have known they were using the McKenzie method if he had not been told.
4) "Baseline values regarding severity among the three groups tested appear to create a bias in the outcomes." While it is true that several of the 33 baseline measures were less favorable in the chiropractic group, it is also true that several of these measures were more favorable in the chiropractic group. The three treatment groups differed significantly on only four baseline measures; two of these favored the chiropractic group (less severe symptoms than the McKenzie group and greater expectations of improvement in back pain than the McKenzie and booklet groups).
The important thing is that the analyses were "controlled" for baseline differences in all variables found to predict outcome. Thus, any baseline differences among the groups were adjusted for in the final analyses.
5) "The initial bothersome (symptoms) and Roland-Morris disability scores of 4 and 7-8 are substantially below the respective values of 6-7 and 10 which are more frequently observed in trials involving significant low back pain." Dr. Rosner appears to have confused the one week outcomes in Figures 1 and 2 with the baseline values, which were included in Table 1. In fact, the subjects in our study were fairly similar to those in other studies: baseline symptoms of almost 6 on a 0-10 scale, and baseline Roland-Morris disability scores of about 12 on a 0-23 scale.
6) Clinical significance of observed treatment effects. How large a difference needs to be to be considered clinically significant is admittedly somewhat arbitrary. However, the literature suggests that 1.5 points on a 0-10 symptom scale and 2.5 points on the Roland-Morris disability scale should be considered clinically significant. These criteria are equally applicable to studies of all types of back pain treatments, from invasive procedures (e.g., spine surgery) to more conservative treatments (e.g., chiropractic manipulation). Even at the four week follow-up, where there was the greatest evidence for a statistically significant difference among the treatments, the difference between the chiropractic manipulation group and the booklet group was only one point on the symptom scale. We therefore concluded that the benefits of these treatments were modest. Moreover, decisions about the comparative value of various therapies need to take cost, as well as extent of effectiveness, into account.
The total two year costs of care for back pain in the McKenzie and chiropractic treatment groups were approximately $280 higher than those for the booklet group. We therefore concluded that "Whether the small benefits of these treatments are worth their additional costs is open to question."
7) "Treatment was given to a heterogeneous population without regard to standard assessment". It is true that we considered all patients with persistent pain or dysfunction seven days after a physician visit to be potentially eligible for this study. However, we then excluded patients with sciatica, previous back surgery, systemic or visceral causes of pain, osteoporosis, vertebral fractures or dislocations, severe neurologic signs, spondylolisthesis, pregnancy, involvement with claims for compensation or litigation or significant co-morbid conditions. Thus, we did not include all comers. More importantly, however, was that every chiropractor (and physical therapist) was allowed to use his or her standard methods of assessment. As noted above, the chiropractors did not feel particularly constrained by the protocol.
We agree that it would be inappropriate to draw sweeping conclusions about chiropractic care based on the results of this one study. Our results are most applicable to persons with uncomplicated low back pain who do not respond well to conventional medical care and who receive a common form of chiropractic manipulation. We recognize that other chiropractic treatments may be more effective than the one we studied and that chiropractic care may be more effective for persons who have already concluded that this type of treatment is their best option. We respect the patient-oriented goals of the chiropractic profession and are aware that there are millions of Americans who feel they have benefited from chiropractor care.
Thank you for providing an open forum for discussing the methods and findings of our study. Such interchanges are valuable for educating all parties involved and will ultimately contribute to improvements in the quality and relevance of future research.
Reference
1. Cherkin D, Deyo R, Battie M, Street J, Barlow W. A comparison of physical therapy, chiropractic manipulation, and provision of an educational booklet for the treatment of patients with low back pain. N Engl J Med 1998;339:1021-9.
Dynamic Chiropractic editorial staff members research, investigate and write articles for the publication on an ongoing basis. To contact the Editorial Department or submit an article of your own for consideration, email
.