Abstract 4021: Economic Evaluations within Randomized Controlled Trials of Cardiovascular Treatments
Background: Adoption and diffusion of new cardiovascular disease (CVD) treatments depends increasingly on evidence of costs and cost effectiveness, which in turn is increasingly being derived from randomized clinical trials (RCTs) in which patient-level economic information is collected simultaneously with clinical information. Yet the quality of the statistical methods used for the analysis of the economic data, and hence the validity of economic results, is unknown.
Methods: We performed a systematic review of economic evaluations based on patient-level cost or resource use data collected in RCTs of CVD interventions published in 2003. A total of 28 articles identified from Medline were evaluated for their use of statistical methods for:
joint comparison of costs and effects and assessment of sampling uncertainty;
incremental cost estimation; and
handling of incomplete or censored cost data.
Results: 11 (39%) of the 28 studies compared costs and effects by use of an incremental cost-effectiveness ratio or net benefit and 5 (45%) of these reported the uncertainty of this outcome. Depending on the strictness of the criteria, either 7 (25%) or 15 (54%) additional studies should have compared costs and effects, but did not do so. Comparison of mean costs alone was more common, yet 5 (18%) of the 28 studies made claims about cost-effectiveness or cost savings without statistically comparing costs between treatment groups. Of the remaining 23, 19 (83%) used at least one statistical test appropriate for drawing inferences for mean cost differences. 9 (32%) of the studies reported some incomplete cost data; 6 (21%) reported that data were complete; the remaining 45% did not report whether cost data were incomplete or not. 3 (33%) of the 9 studies that reported incomplete cost data conducted a complete-case analysis. None of the studies used published statistical approaches for handling incomplete cost data.
Conclusion: The quality of statistical methods used in economic evaluations conducted alongside randomized trials of CVD treatments was poor in the majority of studies published in 2003. Adoption of appropriate statistical methods is required before the results from such studies can consistently provide valid information to decision-makers.