This paper introduces a framework with which to conceptualise the decision-making process in health technology assessment for new interventions with high budgetary impacts. In such circumstances, the use of a single threshold based on the marginal productivity of the health care system is inappropriate. The implications of this for potential partial implementation, horizontal equity and pharmaceutical pricing are illustrated using this framework.

Under the condition of perfect divisibility and given an objective of maximising health, a large budgetary impact of a new treatment may imply that optimal implementation is partial rather than full, even at a given incremental costeffectiveness ratio that would nevertheless mean the decision to accept the treatment in full would not lead to a net reduction in health. In a one-shot price-setting game, this seems to give rise to potential horizontal equity concerns. When the assumption of fixity of the ICER (arising from the assumed exogeneity of the manufacturer’s price) is relaxed, it can be shown that the threat of partial implementation may be sufficient to give rise to an ICER at which cost the entire potential population is treated, maximising health at an increased level, and with no contravention of the horizontal equity principle.