ABSTRACT
BACKGROUND: [corrected] As part of the effort to control HIV/AIDS, the number of HlV voluntarycounselling and testingcentres (VCTCs) is increasing rapidly in the public health system of the Indian state of Andhra Pradesh, which is estimated to have one of the highest rates of HIV infection in India. However, systematic data on the cost and efficiency of providing VCT services in India are not available to help guide efficient use of resources for these services. METHODS: We used standardized methods to obtain detailed cost and output data for the 2002-03 fiscal year from written records and interviews in 17 VCTCs in the public health system in Andhra Pradesh. We calculated the economic cost per client receiving VCT services, and analysed the variation and determinants of total and unit costs across VCTCs. We used multivariate regression techniques to estimate incremental unit costs. We assessed hurdles towards serving an optimal number of clients by VCTCs. RESULTS: In the 2002-03 fiscal year, 32 413 clients received the complete sequence of services at the 17 VCTCs, including post-HIV test counselling. The number of clients served by each VCTC ranged from 334 to 7802 (median 979). The overall HIV-positive rate in post-test counselled clients was 20.5% (range 5.4%-52.6%). The cost per client for the complete VCT sequence varied 6-fold between VCTCs (range Rs 141.5-829.6 [US 2.92-17.14 dollars], median Rs 363.5 [US 7.51 dollars]). The cost per client was significantly lower at VCTCs with more clients (p < 0.001, R2 = 0.83; power function) due to substantial fixed costs. Personnel made up the largest component of cost (53.7%). The cost per client had a significant direct relation with percent personnel cost for VCTCs (p < 0.001, R2 = 0.58; exponential function). A multiple regression model revealed that the incremental cost of providing complete VCT services to each HIV-positive and -negative client was Rs 123.5 (US 2.54 dollars) and Rs 59.2 (US 1.22 dollars), respectively. Fourteen VCTCs (82.4%) reported that they could serve more clients with the available personnel and infrastructure, and that inadequate demand for their services was the main hurdle towards achieving this. CONCLUSION: These data suggest that the efforts of the National AIDS Control Organisation of India and the Andhra Pradesh State AIDS Control Society in increasing VCTCs could yield even higher benefit if the demand for these services was enhanced, as this would increase the number of clients served and reduce the cost per client. Ongoing systematic cost-efficiency analysis is necessary to help guide efficient use of HIV-control resources in India.