Abstract

Purpose: Selecting the optimal treatment regimen for antiviral-naïve patients may be difficult, given the concern about the antiviral activity, the development of drug resistance, and the increase in drug costs. This study evaluates the costs and effectiveness of using lopinavir/ritonavir (LPV/r) vs. nelfinavir (NFV), both coadministered with stavudine and lamivudine, as the first HAART regimen in treating HIV patients, based on the results from the published clinical trial M98-863. Method: A Markov model was developed using a combination of viral load (VL) and CD4 count as surrogate markers to define health states. VL and CD4 count data from the 48-week analysis of the clinical trial were used as measures of effect. The impact of resistance difference between NFV and LPV/r was also examined. Results: Over the first 5 years, the model estimated that LPV/r could save $3,461 per patient in total HIV care costs compared with NFV. If the resistance advantage of LPV/r was taken into account, the cost savings by LPV/r increased to $5,546. For longer term projection, without considering the resistance difference, the incremental cost-effectiveness ratio (CER) for LPV/r vs. NFV was $6,653 per quality-adjusted life-year (QALY). This CER compares favorably to therapies for HIV disease and for common drug treatments for other conditions and is well within accepted thresholds for health policy makers. Conclusion: When treatment options are being considered, this study suggests that use of LPV/r in the first antiretroviral regimen, as compared to NFV, is cost-effective based on improved efficacy and resistance.