Validity of resource-use data in economic evaluations

Abstract

Aims: To analyse issues associated with resource-use data collection in economic evaluations, aiming at improvement in the design and conduct of future studies.Methods: Paper I: A health-economic study in the Swedish sub-set of patients in the Heart Outcomes Prevention Evaluation (HOPE) trial. The objective was to assess if ramipril, an angiotensin-converting-enzyme inhibitor, was cost-effective as compared to placebo. HOPE was a clinical multinational study (n=9 297) in patients at high risk of a cardiovascular event. Primary end-point was a composite of cardiovascular events. Paper II: A cost-effectiveness analysis of two alternative programmes for the eradication of EMRSA-16 based on decision analytical modelling. The background was an outbreak of EMRSA-16, which from a single patient assumed epidemic proportions despite preventive actions. Paper III: A cost-analysis of laparoscopic (LR) versus open resection (OR) of colon cancer after 12 weeks in a Swedish sub-set of patients in the Colon Cancer Laparoscopic or Open Resection (COLOR) study. Paper IV: Design of a new health-economic study based on COLOR II, a study comparing laparoscopic and open surgery for rectal cancer. Experience gained in previous studies was used to improve the principles for the design of the study.Results: Paper 1: The cost/life year gained was 1 940. Paper II: The forceful eradication programme was found to be cost-saving shortly after 2 years of implementation. Modelling is a suitable method in complex situations with scarce information. Paper III: From the perspective of the health care system costs were significantly higher in the LR group (mean difference 2 244, p=0.018). Societal costs did not differ significantly between the groups (mean difference 1 846). The analyses of paper I and III were associated with methodological problems. Paper IV: It was possible to increase the power and to avoid other methodological problems previously encountered by designing the study differently.Conclusions: Health-economic evaluations alongside clinical trials are associated with numerous methodological issues. The framing of the research question and the design of the study have a great influence on the validity of the result, and thus the basis for health-care decisions.