Optimal Bayesian sequential sampling rules for the economic evaluation of health technologies

We present a Bayesian sequential sampling model in which a researcher has flexibility over the timing of the decision to recommend adoption of a new health technology or continuation with an existing one and which accounts for the financial costs and benefits of research and adoption. We apply the model in the field of cardiovascular disease, deriving dynamic thresholds defining the researcher's optimal policies as a function of sample size. Failure to account for the dynamic nature of experimentation and its economic parameters can lead to the misallocation of resources within health care systems