Epidemiological impact and cost-effectiveness of providing long-acting pre-exposure prophylaxis to injectable contraceptive users for HIV prevention in South Africa: a modelling study

Abstract

Introduction: Although pre-exposure prophylaxis (PrEP.) is an efficacious HIV prevention strategy, its preventive benefit has not been shown among young women in sub-Saharan Africa, likely due to non-adherence. Adherence may be improved with the use of injectable long-acting PrEP methods currently being developed. We hypothesize that providing long-acting PrEP to women using injectable contraceptives, the most frequently used contraceptive method in South Africa, could improve adherence to PrEP, result in a reduction of new HIV infections, and be a relatively easy-to-reach target population. In this modelling study, we assessed the epidemiological impact and cost-effectiveness of providing long-acting PrEP to injectable contraceptive users in Limpopo, South Africa.

Methods: We developed a deterministic mathematical model calibrated to the HIV epidemic in Limpopo. Long-acting PrEP was provided to 50% of HIV negative injectable contraceptive users in 2018 and scaled-up over two years. We estimated the number of HIV infections that could be averted by 2030 and the drug price of long-acting PrEP for which this intervention would be cost-effective over a time horizon of 40 years, from a healthcare payer perspective. In the base-case scenario we assumed long-acting PrEP is 75% effective in preventing HIV infections and 85% of infected individuals are on antiretroviral drug therapy (ART) by 2030. In sensitivity analyses we adjusted PrEP effectiveness and ART coverage. Costs between $519 and $1119 per disability-adjusted life-year (DALY) averted were considered potentially cost-effective, and <$519 as cost-effective.

Results: Without long-acting injectable PrEP, 224,000 (interquartile range 176,000 to 271,000) new infections will occur by 2030; use of long-acting injectable PrEP could prevent 21,000 (17,000 to 26,000) or 9.8% (8.9% to 10.6%) new HIV infections by 2030 (including 6000 (4000 to 7000) in men). Long-acting PrEP would prevent 34,000 (29,000 to 39,000) or 12,000 (8000 to 15,000) at 75% and 95% ART coverage by 2030 respectively. To be considered potentially cost-effective the annual long-acting PrEP drug price should be <$16, and/or ART coverage remains at <85% in 2030.

Conclusions: Providing long-acting PrEP to injectable contraceptive users in Limpopo is only potentially cost-effective when long-acting PrEP drug prices are low. If low prices are not feasible, providing long-acting PrEP only to women at high risk of HIV infection will become important.

Keywords: HIV; South Africa; cost-effectiveness; injectable contraceptives; long-acting pre-exposure prophylaxis.

Figure 1

Short‐term epidemiological impact of long‐acting pre‐exposure prophylaxis (PrEP) on the HIV epidemic in Limpopo (2018 to 2030), assuming 50% of injectable contraceptives users use long‐acting PrEP. In the analysis, we assumed an effectiveness of long‐acting PrEP ranging between 50% and 100%. Depicted are the median and interquartile ranges of all accepted simulations.

Figure 2

Effects of the coverage with antiretroviral therapy (ART) in the population and long‐acting pre‐exposure prophylaxis (PrEP) provided to half of injectable contraceptive users in Limpopo on the number of new HIV infections in the period 2018 to 2030. Baseline scenario assumes no long‐acting PrEP and 85% of infected individuals using ART by 2030. Effects of lower (75%) and higher (95%) ART coverage in 2030 are depicted as well as the effect of long‐acting PrEP for all three different ART scenarios. Depicted are the median and interquartile ranges of all accepted simulations.

Figure 3

Cost‐effectiveness of providing half of HIV negative injectable contraceptive users in Limpopo with long‐acting pre‐exposure prophylaxis (PrEP) if the proportion of HIV infected individuals using antiretroviral therapy (ART) (a) increases to 75% by 2030; (b) increases as predicted to 85% by 2030; (c) increases to 95% by 2030. A time horizon of 40 years is used. Red represents scenarios not cost‐effective (costs over $1119/DALY), light green represents potentially cost‐effective scenarios (cost between $519–$1119 per DALY) and dark green represents cost‐effective scenarios (cost <$519/DALY). Depicted are the median incremental cost‐effectiveness ratios of all accepted simulations.cting PrEP for all three different ART scenarios. Depicted are the median and interquartile ranges of all accepted simulations.

Figure 4

One‐way sensitivity analysis of the incremental cost‐effectiveness of long‐acting pre‐exposure prophylaxis (PrEP) provided to half of HIV negative injectable contraceptive users in Limpopo compared to no long‐acting PrEP over 40 years. At baseline, the proportion of infected individuals using antiretroviral therapy (ART) will rise as predicted to 85% in 2030, long‐acting PrEP is 75% effective, long‐acting PrEP drug costs $50 per year and non‐drug related costs for long‐acting PrEP are $66.91. Bars show the change in cost‐effectiveness if the value of the corresponding parameter is replaced by the value in parentheses. Red represents scenarios not cost‐effective (costs over $1119/DALY), light green represents potentially cost‐effective scenarios (cost between $519 and $1119 per DALY) and dark green represents cost‐effective scenarios (cost <$519/DALY). Depicted are the median incremental cost‐effectiveness ratios of all accepted simulations. Baseline scenario cost $1657 per DALY. DALY, disability‐adjusted life‐year, LA‐PrEP, long‐acting pre‐exposure prophylaxis.