Health systems throughout the world are increasingly concerned with the evaluation of the effectiveness and cost-effectiveness of pharmaceuticals (and health technologies more generally), in order to enhance the extent to which scarce health resources are used efficiently. Regulatory processes designed to demonstrate efficacy, quality and safety are added to the concept of efficiency or value for money. These regulatory processes serve to identify and ensure that a product does indeed have a therapeutic effect; that it is manufactured to acceptable standards; and that it does not bring undue hazard to patients with use. A subsequent appraisal process considers if and where society might rationally bring the intervention into usage. It may be efficient for a health system to adopt a less effective but less costly treatment where overall limited resources may be allocated in such a way as to achieve greater health benefit. In this article, I consider aspects of the statistical synthesis of evidence intended to support decisions on pharmaceutical reimbursement.