Economic and epidemiologic impact of guidelines for early ART initiation irrespective of CD4 count in Spain

Abstract

Introduction: Emerging data suggest that early antiretroviral therapy (ART) could reduce serious AIDS and non-AIDS events and deaths but could also increase costs. In January 2016, the Spanish guidelines were updated to recommend ART at any CD4 count. However, the epidemiologic and economic impacts of early ART initiation in Spain remain unclear.

Methods: The Johns Hopkins HIV Economic-Epidemiologic Mathematical Model (JHEEM) was utilized to estimate costs, transmissions, and outcomes in Spain over 20 years. We compared implementation of guidelines for early ART initiation to a counterfactual scenario deferring ART until CD4-counts fall below 350 cells/mm3. We additionally studied the impact of early ART initiation in combination with improvements to HIV screening, care linkage and engagement.

Results: Early ART initiation (irrespective of CD4-count) is expected to avert 20,100 [95% Uncertainty Range (UR) 11,100-83,000] new HIV cases over the next two decades compared to delayed ART (28% reduction), at an incremental health system cost of €1.05 billion [€0.66 - €1.63] billion, and an incremental cost-effectiveness ratio (ICER) of €29,700 [€13,700 - €41,200] per QALY gained. Projected ICERs declined further over longer time horizon; e.g., an ICER of €12,691 over 30 years. Furthermore, the impact of early ART initiation was potentiated by improved HIV screening among high-risk individuals, averting an estimated 41,600 [23,200-172,200] HIV infections (a 58% decline) compared to delayed ART.

Conclusions: Recommendations for ART initiation irrespective of CD4-counts are cost-effective and could avert > 30% of new cases in Spain. Improving HIV diagnosis can amplify this impact.

Fig 1

Projected new HIV infections (A), cumulative HIV infections averted (B) HIV deaths (C), and cumulative deaths averted (D) in Spain from 2017 to 2037 under early versus delayed ART. All lines represent the median values of simulations, comparing scenarios representing delayed ART initiation at CD4 counts of <350 cells/mm³(solid red line in Panels A&C) against early ART initiation, both without (solid green line) and with additional improvements to the HIV care continuum (dashed blue lines). The shaded areas represent the corresponding interquartile uncertainty ranges (not shown for the additional improvement scenarios due to overlap).

Fig 2. Cost-effectiveness acceptability curve for early (versus delayed) ART initiation.

The y-axis represents the proportion of all simulations that were shown to be cost-effective in the probabilistic uncertainty analysis under different willingness-to-pay thresholds, expressed as 2017 euros per QALY gained (on the x-axis). At current proposed WTP thresholds of €30,000 per QALY averted, 62% of simulations are cost-effective (dashed blue lines). Increasing the WTP to €80,000 increases the proportion of cost-effective simulations to 99%.

Fig 3. One-way sensitivity analysis of the incremental impact of early ART initiation compared to delayed ART initiation (CD4<350).

Shown on the x-axis are the range of values for incremental HIV incidence (A), HIV deaths (B), costs (C) and cost-effectiveness (D) in the early ART initiation scenario compared to the delayed ART scenario. Shown on the y-axis are the five parameters [range of values] for which variation over their stated ranges resulted in the greater effect on each outcome. Yellow bars correspond to the median outcome value among the simulations containing the highest 10% of the parameter value, whereas green bars depict the median projected outcome among simulations containing the lowest 10% of the parameter value.