Quality assessment of pharmacoeconomic abstracts of original research articles in selected journals

OBJECTIVE: To assess and compare the quality of pharmacoeconomic abstracts of cost-minimization analyses, cost-effectiveness analyses, cost-utility analyses, and cost-benefit analyses of original research articles in selected medical, pharmacy, and health economics journals, METHODS: MEDLINE was used to identify articles in selected medical, pharmacy, and health economics journals using the MeSH word ''economic'' and text words ''cost'' and ''pharmacoeconomic''; the journal PharmacoEconomics was searched manually, All retrieved abstracts were evaluated, Original, comparative (at least one drug comparator) research articles (1990-1994) reporting both costs and clinical outcomes were included in the quality analysis. Abstract quality was assessed as a percentage by using a checklist with 29 objective criteria. Group consensus produced interrater reliability greater than 0.8. RESULTS: One thousand two published abstracts labeled with the above key words were identified, Of these, 951 were excluded from quality assessment because they were not original research (18%), were not pharmacoeconomic research (47%), lacked a drug comparator (35%), or did not report a clinical outcome (0.5%). Thus, the quality of 51 (5% of the total) remaining abstracts was assessed. Overall scores were 56% in 1990 and 58% in 1994 (p = 0.094). Medical articles scored highest (61.5%; n = 25), pharmacy articles were next (54.3%; n = 5), and health economics articles were lowest (53.4%; n = 21) (p = 0.091); structured abstracts scored significantly higher (62.5%; n = 20) than unstructured (53.3%; n = 31) (p = 0.003). CONCLUSIONS: Abstract quality was generally poor, with no significant change in quality over time. Medical journals scored highest, probably because they use structured abstracts. Guidelines for structured pharmacoeconomic abstracts may assist in improving quality.