As well as two models of how PrEP/microbicides would impact on HIV prevalence and incidence in a southern African context (see Forecasters agree PrEP/microbicides could cut HIV infections in South Africa), two mathematical models looked at the cost-effectiveness of these prevention methods.
Costs of treatment, PrEP, and treatment-plus-PrEP
Rochelle Walensky from Harvard Medical School calculated the cost-effectiveness of topical and oral PrEP if used in South Africa by heterosexual women. The World Health Organization defines a measure as “very cost-effective” if it costs less per year than the average annual per capita GDP, which in the case of South Africa is US$5400.
She calculated the incremental cost-effectiveness – the cost of each added year of healthy life – of pre-exposure prophylaxis for women using three scenarios.
One drew the age of participants, background annual HIV incidence, observed efficacy and cost per dose from CAPRISA 004, one used the same figures from the iPrEX study, and one substituted South Africa’s national HIV incidence in young women (2.2% a year) instead of the very high baseline annual incidence of 9.1% seen in CAPRISA, but otherwised use the CAPRISA figures.
The annual cost of a microbicide plus a monthly HIV test and a six-monthly full blood screen would be $204 per person per year for the CAPRISA microbicide and $257 a year for the iPrEx pill.
She calculated that a microbicide used in the same high-incidence situation as CAPRISA would cut the lifetime risk of HIV infection in women from 85 to 70%, or – with the lower national incidence – from 40 to 27%. A pill used with the same background incidence as iPrEx would cut the lifetime risk from 56 to 38%. At these rates the incremental cost-effectiveness would be $1200 per healthy life-year for the CAPRISA scenario, $3600 for the iPrEx scenario, and $4600 for the national-incidence one.
Would there be any situations in which PrEP would be cost-saving? At present prices, the cost would only break even in situations where PrEP was 80% effective and the background HIV incidence more than 7%. If PrEP was less than 80% effective, it could not save money. If, however, the cost of PrEP could be halved, then it would be cost saving with an efficacy of 50% and an incidence of more than 5.5%, or 75% effective with an incidence of 3%.
Cost-effectiveness in serodiscordant couples
Timothy Hallett of Imperial College London looked at the cost-effectiveness of PrEP when used for serodiscordant couples. He commented that the question was often asked: "Wouldn’t it just be easier to give the pills to the HIV-positive partner?"
He agreed that this made sense if the only risk to the negative partner was from their spouse, but pointed out that in the Partners in Prevention study 25% of infections came from an outside partner. He added that at South African prices Truvada PrEP only cost 60% as much as first-line combination therapy.
He therefore devised a cost-effectiveness model looking at the most efficient way of delivering ARV prevention to a serodiscordant couple, using as his endpoint the cost of ensuring that the HIV-negative partner was alive and remained free of HIV by the time they reached 50.
He said that in most cases it was indeed cheaper to give treatment to the HIV-positive partner. He posed the question, is it more cost-effective to give PrEP to the HIV-negative partner until the HIV-positive partner reaches a CD4 count of 200 cells/mm3, and then switch to treating them; or is it cheaper to switch when they reach a CD4 count of 350 cells/mm3? The answer is that in general it is cheaper to start the HIV-positive partner at a CD4 count of 350 cells/mm3. It would only be cheaper to prolong PrEP, if it cost 60% of what combination ARVs cost or if it has an efficacy of more than 75%.
What about the cost of either stopping PrEP for the negative partner and starting treatment for the positive one when they reach a CD4 count of 350 cells/mm3, versus starting treatment for the positive partner at a count of 500 cells/mm3? In this case, PrEP and later treatment would be a lot more cost-effective. It would only have to be 30% effective to cost less than starting the positive partner’s treatment at 500 cells/mm3.
So the answer is that, for a serodiscordant couple, PrEP is generally more cost-effective than treatment if the HIV-positive partner would not normally be taking treatment according to WHO guidelines.
Abstracts and webcasts
You can view the abstracts from this research on the official conference website:
Abstract 37LB: www.retroconference.org/2011/Abstracts/42531.htm
Abstract 99LB: www.retroconference.org/2011/Abstracts/42617.htm
You can also watch webcasts of these presentations. Abstract 37LB was presented by Ji-Eun Park in the session HIV Prevention: HSV-2, Topical and Oral PrEP, and Circumcision. Abstract 99LB was presented in the session Advances in PrEP.
Rochelle Walensky presented Tuesday's plenary session: Cost-effectiveness in HIV Care: Understanding Value in a World of Limited Resources.
We rely on donations to continue our work to help people with HIV, support us today at: www.aidsmap.com/donate
Walensky R et al. Cost Effectiveness of PrEP for HIV Infection in South Africa. Eighteenth Conference on Retroviruses and Opportunistic Infections, Boston, abstract 37LB. 2011.
Hallett T et al. ART or PrEP for HIV Prevention in HIV Serodiscordant Partnerships: A Mathematical Modeling Comparison. Eighteenth Conference on Retroviruses and Opportunistic Infections, Boston, abstract 99LB. 2011.