RESEARCH NOTE Use of pharmacoeconomics in prescribing research. Part 2: cost-minimization analysis – when are two therapies equal? D. Newby BPharm PhD and S. Hill BMed PhD FAFPHM Clinical Pharmacology, School of Medical Practice and Population Health, Faculty of Health, University of Newcastle NSW, Australia SUMMARY This is the second paper in a series looking at pharmacoeconomic methods. The first paper dis- cussed how costs are identified for pharmaco- economic studies. This article will examine one of the four main evaluation methods in health eco- nomics, cost-minimization analysis (CMA). The remaining three methods (cost-effectiveness, cost- utility and cost-benefit analysis) will be discussed in later papers. Key messages • Cost-minimization is the appropriate form of economic analysis to carry out whenever two drugs have the same clinical effect. • True equivalence studies are uncommon; a more useful approach is to assess the size of the confidence interval around the difference between treatments and determine whether it excludes clinically relevant effects. • The critical issue for cost-minimization analy- ses is determining equi-effective doses. Keywords: cost-minimization analysis, equiva- lance studies, pharmacoeconomics INTRODUCTION Cost-minimization analysis (CMA) is the simplest of the four evaluation methods (1) (see Table 1). A CMA should be performed and is appropriate when two interventions have been shown to produce the same, or similar, effects. If two therapies are con- sidered clinically equivalent, then only the costs of the interventions need to be considered. Economic efficiency (see Part 1 of this series [Robertson et al. Journal of Clinical Pharmacy and Therapeutics (2003) 28, 73–79]) then dictates selection of the least costly intervention. Thus, the critical first step before conducting a CMA is determining the therapeutic equivalence of the interventions. STATISTICAL SIGNIFICANCE, CLINICAL SIGNIFICANCE AND EQUIVALENCE The randomized controlled trial (RCT) is consid- ered to be the ‘gold standard’ for comparing two interventions. A well conducted RCT provides the best estimate of any difference in efficacy by pro- tecting against uncontrolled confounding variables (through randomization) and any potential sub- jectivity of outcome assessments (through blind- ing) (2). Any potential differences between two therapies should be assessed from both a clinical and statis- tical perspective. A clinically significant difference is the smallest difference in an outcome that would have a clinical impact on the patient. However, deciding what a clinically important difference is can be problematic. Unlike bioequivalence studies, where there are standards for defining ‘equival- ence’ that have been established by drug regulatory authorities, when it comes to assessing clinical equivalence there are few clear guides, and hence there is often considerable difficulty in deciding what is a clinically important difference (3). The usual approach in assessing studies is to consider both statistical and clinical significance, based on the summary estimate of the size of the Series Editor: Paramjit Gill, University of Birmingham Received 27 September 2002, Accepted 27 November 2002 Correspondence: David Newby, Clinical Pharmacology, Newcastle Mater Hospital, Edith St, Waratah, NSW 2298, Australia. Tel.: 2 +61 2 49211292; fax: +61 2 49602088; e-mail: david.newby@newcastle.edu.au Journal of Clinical Pharmacy and Therapeutics (2003) 28, 145–150 Ó 2003 Blackwell Publishing Ltd 145