Probabilistic analysis of cost-effectiveness models: choosing between treatment strategies for Gastro-Esophogeal Reflux Disease

When choosing between mutually exclusive treatment options, it is common to construct a cost-effectiveness frontier on the cost-effectiveness plane that represents efficient points from among the treatment choices. Treatment options internal to the frontier are considered inefficient and are excluded either by strict dominance or by appealing to the principle of extended dominance. However, when uncertainty is considered, options excluded under the baseline analysis may form part of the cost-effectiveness frontier. By adopting a Bayesian approach, where distributions for model parameters are specified, uncertainty in the decision concerning which treatment option should be implemented is addressed directly. The approach is illustrated using an example from a recently published cost-effectiveness analysis of different possible treatment strategies for gastro-esophageal reflux disease. It is argued that probabilistic analyses should be encouraged since they have potential to summarise the strength of evidence in favour of particular treatment choices.