diabetes claims, cost-effectiveness modeling, simulation, credibility, replication, imaginary worlds, scientific method, pseudoscience


Over the past 20 years a number of simulations or models have been developed as a basis for tracking and evaluating the impact of pharmacological and other interventions in type 1 and type 2 diabetes mellitus. These models have typically tracked the natural course of these diseases generating long-term composite claims for cost-effectiveness. These claims can extend over the lifetime of the modeled patient cohort. Set against the standards of normal science, however, these claims lack credibility. The claims presented are all too often either immune to failure or are presented in a form that is non-testable. As such they fail to meet the key experimental requirements of falsification and replication. Unfortunately, there is a continuing belief that long-term or lifetime models are essential to decision-making. This is misplaced. The purpose of this review is to argue that there is a pressing need to reconsider the needs of health system decision makers and focus on modeled or simulated claims that are meaningful, testable, reportable and replicable in evaluating interventions in diabetes mellitus.

Conflict of Interest