Growth and quality of the cost-utility literature, 1976-2001

Abstract

PURPOSE: Cost-utility analyses (CUAs) have become increasingly popular, although questions persist about their comparability and credibility. Our objectives were to: 1) describe the growth and characteristics of CUAs published in the peer-reviewed literature through 2001; 2) investigate whether CUA quality has improved over time; 3) examine whether quality varies by the experience of journals in publishing CUAs, or the source of external funding for study investigators; and 4) examine changes in practices in US-based studies following recommendations of the US Panel on Cost-Effectiveness in Health and Medicine (USPCEHM). This study updates and expands our previous work, which examined CUAs through 1997. METHODS: We conducted a systematic search of the English-language medical literature for original CUAs published from 1976 through 2001, using Medline and other databases. Each study was audited independently by two trained readers, who recorded the methodological and reporting practices used. RESULTS: Our review identified 533 original CUAs. Comparing articles published in 1998 to 2001 (n = 305) with those published in 1976 to 1997 (n = 228), studies improved in almost all categories, including: clearly presenting the study perspective (73% vs. 52%, P < 0.001); discounting both costs and quality-adjusted life-years (82% vs. 73%, P = 0.0115); and reporting incremental cost-utility ratios (69% vs. 46%, P < 0.001). The proportion of studies disclosing funding sources did not change (65% vs. 65%, P = 0.939). Adherence to recommended practices was greater in more experienced journals, and roughly equal in industry versus non-industry-funded analyses. The data suggest an impact in methodological practices used in US-based CUAs in accordance with recommendations of the USPCEHM. CONCLUSIONS: Adherence to methodological and reporting practices in published CUAs is improving, although many studies still omit basic elements. Medical journals, particularly those with little experience publishing cost-effectiveness analyses, should adopt and enforce standard protocols for conducting and reporting CUAs.