Modeling economic evaluations of pharmaceuticals: manipulation or valuable tool?

Controversy surrounds the use of models in economic evaluations of pharmaceuticals. Many believe that modeling is a way of manipulating results and is not credible, whereas others consider modeling a valuable tool in economic evaluations. The purpose of this article is to provide a historical perspective on modeling, focus on the controversy and policy implications of using models, and review the suggested framework and guidelines for modeling practices. Models can be used to extrapolate beyond intermediate end points, predict costs and consequences of alternative therapies, generalize data to other settings, pose questions instead of providing answers when no data exist, design an evaluation to reduce uncertainty, and perform direct comparisons that are not currently available. We believe that a useful model should document the detailed inner workings, assumptions, and inherent bias during production (and at publication time), so that its reviewers and users can evaluate the appropriateness of the model's outcomes. The acceptability of models in the future rests with the researchers constructing them. If constructed appropriately, modeling economic evaluations is not a manipulation but rather a valuable tool.