Thesis (B.A.)--Haverford College, Dept. of Economics, 2008. / Includes bibliographical references.

This thesis evaluates the cost-effectiveness of eight policies expanding human immunodeficiency virus (HIV) testing in South Africa. All policies entail provider-initiated test offers for primary healthcare users and one of two options across three policy components: (i) consent method, opt-in or opt-out; (ii) test protocol, rapid only or rapid plus acute infection testing; and (iii) linkage-to-care, standard or enhanced. This thesis highlights four methodological issues. First is the challenge of conducting a population-level analysis, projecting the cost-effectiveness of expanded testing for each member of South Africa’s adult African population. To this end, I conducted a retrospective, descriptive study to measure current population-level testing rates and epidemic descriptors in an African community near Cape Town, South Africa. Second, the effects of testing expansion on current testing uptake were estimated by distinguishing testing in the study community likely to cease after testing expansion (baseline testing) from that likely to continue (background testing). Third, because testing alone is an outcome of less interest than health benefits following treatment, study community linkage-to-care probabilities were estimated and models utilized to estimate the efficacy of treatment. Fourth, the methods to convert the study community testing data into inputs for these models’ parameters are outlined. The enhanced linkage-to-care policies proved the most cost-effective, with opt-in testing and a rapid-only test protocol the least expensive cost-effective option at $848 per life year gained (LYG). Adding an opt-out consent method or acute infection test protocol to this policy increased the LYGs, but at higher cost-effectiveness ratios.