COST-EFFECTIVENESS OF POSTEXPOSURE PROPHYLAXIS FOLLOWING SEXUAL EXPOSURE TO HIV

Objectives: To assess the cost-effectiveness, relative to other health -related interventions in the U.S., of post-exposure prophylaxis (PEP) following potential HIV exposure through sexual contact with a partne r who may or may not be infected, and to compare the relative cost-eff ectiveness of dual- and triple-combination PEP. Methods: Standard tech niques of cost-utility analysis were used to assess the cost effective ness of PEP with a four-week regimen of zidovudine and lamivudine, or zidovudine, lamivudine, and indinavir. Due to a lack of empirical data on the effectiveness of PEP with combination drug regimens, the analy sis assumed that combination PEP was no more effective than PEP with z idovudine alone. The main outcome variable is the cost per quality-adj usted life year (QALY) saved by the program. Results: Providing PEP to a cohort of 10 000 patients who report receptive anal intercourse wit h a partner of unknown HIV status (who is assumed to be infected with probability equal to 0.18) would prevent about 20 infections, at an av erage net cost of about US$ 70 000 per infection averted. The cost-uti lity ratio, US$ 6316 per QALY saved, indicates that PEP is highly cost -effective in this instance. Moreover, triple-combination PEP would ne ed to be about 9% more effective than dual-combination PEP for the add ition of indinavir to the regimen to be considered cost-effective. Pro phylaxis following receptive vaginal exposure is cost-effective only w hen it is nearly certain that the partner is infected; PEP for inserti ve anal and vaginal intercourse does not appear to be cost-effective. Conclusions: From a purely economic standpoint, PEP should be restrict ed to partners of infected persons (e.g., serodiscordant couples), to patients reporting unprotected receptive anal intercourse (including c ondom breakage), and possibly to cases where there is a substantial li kelihood that the partner is infected. Providing PEP to all who reques t it does not appear to be an economically efficient use of limited HI V prevention and treatment resources. (C) 1998 Lippincott-Raven Publis hers.