A social cost–benefit criterion for evaluating Voluntary Counseling and Testing with an application to Tanzania

Rationale: There are many interventions for HIV/AIDS that require that people know their status and hence require a HIV test. Testing that is driven by a desire to prevent the spread of the disease often has an indirect effect on others. These external effects need to be identified, quantified and included as part of the benefits and costs of testing. Pioneering analyses of HIV testing by Philipson and Posner have introduced the economic calculus of individual expected benefits and costs of activities into an understanding of the HIV epidemic. What is required for social evaluations is an extension of the analysis to ensure that external effects are included. Objectives: The objective of this paper is two‐fold. First we seek to formulate cost–benefit criteria that incorporate fully the external effects in the evaluation of Voluntary Counseling and Testing (VCT). We achieve this by recasting the individual calculus of benefits and costs to a couple setting. We can then compare an individual's cost–benefit analysis of being tested with social criteria that look at outcomes from a couple's perspective for both separate and dual/joint testing. Second we aim to apply our social criteria to VCT programs as they currently operate in Tanzania and how these programs might operate in the future when they are scaled up to relate to the general population. Methodology: We develop social criteria for evaluating separate and dual VCT using a couple's perspective with and without altruism. Therefore, the welfare function is based on two individual expected utility functions viewed as a couple, either married or regular partners. The benefits are the averted lives lost whenever discordant couples are revealed. The costs of VCT are the benefits of unprotected sex that the couple foregoes and the costs of the testing and counseling. The cost–benefit criteria are applied to VCT programs in Tanzania. The four main ingredients estimated are: the foregone benefit of unprotected sex (measured by the compensated wage differentials charged by commercial sex workers); the probability of infection; the cost of an infection (measured by both the value of a statistical life and the human capital approaches) and the cost of a single test (which includes behavior‐modifying counseling). Conclusions: We find separate testing in existing VCT programs to be only marginally worthwhile. However, in scaled‐up programs the benefit–cost ratio is over three. Dual testing is always more beneficial than separate testing. However, this advantage is reduced in scaled‐up programs. VCT should be greatly expanded throughout Tanzania as future returns would be even higher for both separate and joint counseling and HIV testing. Copyright © 2009 John Wiley & Sons, Ltd.