Cost‐effectiveness of atazanavir+ritonavir (ATV+RTV) vs. lopinavir/ritonavir (LPV/r) in women of childbearing age in the United Kingdom

Purpose of the study In the UK, an increasing proportion of individuals with HIV are women of childbearing age (WOCBA). Literature on the potential effects of antiretroviral therapy (ART) on certain risk factors for coronary heart disease (CHD), such as total cholesterol (TC), has not differentiated the reported ART‐associated risk by other risk factors for CHD (e.g. age and gender). Given the different age‐specific CHD risk estimates for men and women, particularly the low risk of CHD in WOCBA, the cost‐effectiveness of LPV/r vs ATV+RTV specifically among WOCBA warrants examination. Therefore, the objective of this study was to perform a cost‐effectiveness and budget impact analysis of two first‐line protease inhibitor‐based regimens, LPV/r versus ATV+RTV, among HIV‐infected, ARV‐naïve WOCBA in the UK. Methods A modified version of a previously published Markov model was utilized [1]. CHD risk estimates were based on the Framingham risk score. Baseline assumptions were that 33% of women were smokers with a mean age of 25 years. Guidelines regarding therapeutic drug monitoring among pregnant women in the 3rd trimester who receive ATV+RTV were incorporated [2]. Age‐specific pregnancy rates were estimated in order to determine ARV utilization during gestation. The model employed a lifetime horizon under a NHS perspective and a discount rate of 3.5%. Costs were presented in 2011 GBP. Summary of results The model predicted no appreciable difference in quality‐adjusted survival (in terms of QALYs) between the two regimens over a lifetime (0.2 days in favor of ATV+RTV) and an increased cost of £3,003 per patient on the ATV+RTV regimen. Cost savings of £1,977 over 5 years and £2,916 over 10 years were predicted for patients who initiated LPV/r. Furthermore, for every 100 patients started on LPV/r, the savings accrued after one year allow for the treatment of 6 additional patients. After 10 years, the number of additional patients that can be treated accumulated to 42. The model predicted a mean of 1.4 pregnancies per woman, and an overall difference between the regimens of 0.3% in CHD events after 10 years. Conclusions Initiating HIV infected, ARV‐naïve WOCBA on a LPV/r‐based regimen compared to an ATV+RTV‐based regimen produces substantial short‐ and long‐term cost savings with similar life expectancy. These results warrant consideration, as selecting LPV/r over ATV+RTV may provide an opportunity for improving access to ART for WOCBA living with HIV in the UK.