J Manipulative Physiol Ther. 1993 Mar-Apr; 16(3): 161-8.

Does the goose really lay golden eggs? A methodological review of Workmen's Compensation studies.

Assendelft WJ, Bouter LM.

Department of Epidemiology, University of Limburg, Maastricht, The Netherlands.

OBJECTIVE: To assess the value of workmen's compensation (WC) studies to determine the effectiveness of chiropractic. Therefore, the results of the available WC studies are summarized and the methodological quality of WC studies is discussed. DATA SOURCES: All studies were eligible without time restriction. Studies were identified by a Medline search from 1966 to 1990 (key words: chiropractic, and manipulation-orthopedic in combination with comparative studies, follow-up studies, evaluation studies) by manual examination of the most important chiropractic reference systems (CRAC and Index to Chiropractic Literature), by tracking the reference lists of identified (reviews of) WC studies and by correspondence with researchers. STUDY SELECTION: Studies were regarded as WC studies if by means of databases of WC boards, a comparison was made between claimants treated by chiropractors and those treated by other health care professionals. At least one of the following outcomes should be reported: compensated days, compensation paid or treatment costs. DATA EXTRACTION: Relevant data (authors, year, state, study population, number of patients, days compensated, compensation paid, number of treatments, consultation costs, additional treatment costs and total cost per case) were extracted by one nonblinded observer. The methodological value was reviewed narratively. DATA SYNTHESIS: The retrospective character of WC studies and the use of large WC databases harbor severe methodological problems like incomparability of study groups, absence of information on prognostic indicators, insufficient outcome measures and missing data. The results of older WC studies (before 1980) and the more recent WC studies, which were of better methodological quality, are presented separately. The older studies are in favor of chiropractic. Two of the six more recent WC studies challenge chiropractic effectiveness. CONCLUSIONS: WC studies in general report positive results for chiropractic. Recent results are more ambiguous. Because of the methodological drawbacks identified, WC studies are insufficient to enable a valid study made of chiropractic effectiveness. Therefore, chiropractic (cost-) effectiveness is not yet convincingly proven. More effort should be directed at establishing randomized clinical trials including the question of (cost-) effectiveness.