A critical appraisal of review articles on the effectiveness of conservative treatment for neck pain

STUDY DESIGN: A criteria-based appraisal of review articles on neck pain. OBJECTIVES: To assess the methodologic quality, conclusions, and extent of concordance among reviews on the conservative treatment of neck disorders. SUMMARY OF BACKGROUND DATA: During the past decades there has been an increasing interest in summarizing and analyzing the available evidence on the effectiveness of conservative management of neck pain. Considering the growing number and quality of reviews, consumers may question which reviews to read and believe. METHODS: Computerized bibliographic databases were searched without language restriction. The reviews assessed had been published before January 1998, included neck pain and evaluated conservative therapies, and reported at least one controlled clinical trial. Identification, selection, and quality assessment were performed independently by two investigators. RESULTS: Of the 108 identified articles, 25 review articles were selected, of which 12 were systematic reviews. The reviews differed in their reporting of study population, interventions, and outcomes. Statistical pooling was performed in two high-quality systematic reviews, whereas in other reviews, the investigators explicitly decided not to pool data. The results of the current study show that the concordance among reviews varied. Regarding manipulation and traction, there is inconclusive evidence among reviews. Concordance regarding the effectiveness of other conservative interventions was absent. Many of the reviews displayed major methodologic flaws. CONCLUSIONS: Consumers should consider reports of reviews both carefully and critically, given the wide variety of review methodology, descriptive information, and final conclusions. There is a paucity of evidence from primary studies on neck pain. Therefore, more research is needed to allow systematic reviews to formulate stronger conclusions