I agree with Dr. Deyo that the AHCPR guidelines are limited to acute low back pain, found many conventional medical treatments to be ineffective, and are lengthy and complex. I also concur that more research is needed. I appreciate his point of clarification about the lack of any formal technical cost-effectiveness analysis reported in the guideline. On reviewing the guidelines, I noted that they state that spinal manipulation relieves symptoms and improves function (while none of the other nonsurgical interventions offer both). A typical practitioner or patient may conclude that a manual (nonsurgical) treatment that works is more cost-effective than any number of treatments that don't work.

Stano (1) provides data on the costs of spinal manipulative therapy specifically; this analysis compared health insurance payments for common lumbar and low back conditions initiated by chiropractic or medical treatment. These results complement a preponderance of studies from workers compensation data (2, 3) and a nationwide longitudinal survey of medical care utilization and costs for 6000 randomly selected households (4). The relatively well-known cost-effectiveness studies by Carey and colleagues and by Shekelle and associates, cited by Dr. Deyo, have been interpreted by Rosner (5) as having several limitations: Severity of illness was not adequately considered, degree of recovery was not adequately assessed, matching of services with provider type may have been irregular, compliance was disregarded, medications and side effects were not specified, calculation of charges was not sufficiently precise or balanced among provider types, and episodes were not adequately defined or contained.

Overall, I intended to provide a balanced view (not a "ringing endorsement"), stating that more than one fourth of spinal manual adjustment was found to be inappropriate and that efforts must be made to reduce the amount of inappropriate spinal manipulation. Dr. Deyo's letter is an important reminder of the need for balance in interpretation of these results. I take this opportunity to repeat the other major points of my editorial—that additional outcomes-based research and research funding is needed for AHCPR and others to conduct appropriate studies and that professional training and practice standards in complementary medicine should continue to be raised.