ABSTRACT 

Objective: No country can afford to provide all necessary healthcare for its citizens, so prioritization among interventions must feature in all health systems. Resources in health should be allocated among interventions/facilities/patients in such a way as to be in line with the objectives of the health system. To achieve this, resource allocation decisions must be informed by the relative contributions that prospective interventions will make to societal health and to costs. Internationally, the EQ-5D based quality adjusted life year (QALY) now dominates this kind of analysis. This paper reports on a pilot study to develop an EQ-5D-3L value set for Trinidad and Tobago based on a protocol that avoids some of the issues that are associated with other approaches to developing such value sets such as the complex elicitation tasks that respondents must carry out, and the large respondent samples required for collecting multiple valuation subset values using blocked designs.

Methods: An orthogonal discrete choice experiment design was used to elicit a set of choices from a sample of respondents.

Results: The choice data were analysed using mixed multinomial logistic regression to produce an internally valid model that predicts well.

Conclusion: This paper marks an important milestone in the development of health resource allocation in the Caribbean. It sets out the importance of incorporating the impact of health interventions to inform health resource allocation decisions, describes the elicitation and analysis methods used in the pilot and provides an illustration of the use of the EQ-5D value set.