These letters illustrate the strong, emotional, often vitriolic responses any discussion of chiropractic evokes in certain physicians. The charge of bias cuts in many directions, as Dr. Ernst is well aware. We stand by our specific statement that overall, there is no evidence of biased reporting of randomized trials of spinal manipulation. We agree that the basic science of manipulation and of the concept of subluxation is controversial and has not yet been fully developed. However, scientists in chiropractic institutions and major universities around the world, supported by major public and private funding agencies, are making progress. Greater understanding of the theories on which manipulation is based will become available. We agree that many of the randomized trials we described were on spinal manipulation rather than specifically on chiropractic manipulation itself, but we believe that this is not a significant point. Chiropractors use all forms of manipulation. In the United States, more than 90% of all spinal manipulation services are provided by chiropractors, and research on spinal manipulation, like that on any other treatment method, is equally of value regardless of the practitioner providing it.

We elected to include only the most recent and comprehensive English-language reviews of manipulation trials out of at least 60 published in the past 25 years. The paper by Brox and colleagues to which Dr. Ernst refers is in Norwegian and does not include all trials of manipulation for chronic back pain. van Tulder, in addition, has published inconsistent conclusions (2–4). Nevertheless, 10 of 11 nationally developed practice guidelines for back pain have recommended manipulation as a treatment option (4). The trial by Olafsdottir and colleagues on infantile colic, also mentioned by Dr. Ernst, was published too late to be included in our list and does not change our already cautious conclusions.

Minor reactions to spinal manipulation are known to all practitioners and most patients of chiropractors and tend to be no more than a nuisance in practice. Patients do not seem to be concerned about these side effects. Patient satisfaction remains higher for chiropractic care than for any treatment with which chiropractic has been compared (5), and the dropout rate due to side effects in clinical trials on manipulation and chiropractic is negligible.

Chiropractic as a profession has made many advances in education, ethics, practice, and theory that give it the trappings of a mainstream health care profession, but there are still a number of claims, as pointed out by these letters, that have not been adequately evaluated. It was for this reason that we described chiropractic as a profession at the crossroads between mainstream and alternative medicine.