The effect of acupuncture utilization on healthcare utilization.

OBJECTIVE To determine whether acupuncture is a complement to or substitute for various medical services. DATA SOURCE This study used managed care claims data from a midsize metropolitan insurance company from 2002. Zip code level data from the 2000 US Census was also incorporated. The original dataset contained medical and drug claims data for every eligible acupuncture user (n = 1688) and every 18th eligible nonacupuncture user (n = 16,282) covered by the data provider. STUDY DESIGN Simultaneous equations models with an exclusion restriction were used in this cross-sectional study. The influence of acupuncture utilization was assessed independently on each conventional service of interest, controlling for numerous clinical and demographic characteristics. Bivariate probit models were estimated using distance to the nearest acupuncturist as the exclusion restriction. RESULTS Acupuncture was a statistically significant (P < 0.05) substitute for primary care, all outpatient services, pathology services, all surgery, and gastrointestinal medications. Acupuncture seemed to complement numerous therapies, particularly chiropractic and physical therapy; however, acupuncture did not statistically significantly complement any therapies after controlling for unobservable characteristics that influence the use of acupuncture and/or conventional medicine. CONCLUSIONS Acupuncture is an economic substitute for some medical services and pharmaceuticals, a finding of some importance to insurers, healthcare practitioners, and policy makers. The fact that acupuncture has an effect on other medical services needs to be explored more fully with an emphasis on how this substitution impacts patient health.