MATHEMATICAL-MODELS IN DECISION-ANALYSIS

Decision analysis offers powerful techniques to understand and evaluat e uncertain clinical situations better. Decision analytic models are a ppearing with increasing frequency in health policy planning, clinical information and decision-support computer systems, evaluations of cli nical pathways, development of clinical practice or utilization review guidelines, and epidemiologic research. This article describes the st ructure, application, and limitations of the more popular decision ana lytic methods, including decision trees, Markov models, Monte Carlo si mulation, survival and hazard functions, fuzzy logic, and sensitivity analysis. Understanding the nature of these methods will help readers to assess better the appropriateness of their use in published reports .