PrEP would save the NHS money, in numerous scenarios

A pre-exposure prophylaxis (PrEP) programme for men who have sex with men in the UK would not only be cost-effective, but actually cost-saving (in other words, both improving health and lowering NHS spending), according to an economic evaluation published by Valentina Cambiano and colleagues in The Lancet Infectious Diseases today.

Modifying a wide range of assumptions – drug prices, event-based or daily PrEP dosing, changes in sexual behaviour and so on – did have an impact on the results, but in all the scenarios examined, PrEP remained cost-saving. In other words, those factors only decreased or increased the amount of money the NHS would save by introducing PrEP.

“There is no doubt about the effectiveness of PrEP," commented Valentina Cambiano of University College London. "In addition to delivering a substantial health benefit, our work suggests that introduction of PrEP will ultimately lead to a saving in costs, as a result of decreased numbers of men in need of lifelong HIV treatment.” 

Nonetheless, the medium-term budgetary impact of PrEP for the NHS is not trivial. During the first two decades of a PrEP programme, there will be extra costs associated with introducing PrEP. The researchers found that the savings in NHS costs will likely be seen within 40 years – but sooner if the availability of generic medications makes prices fall substantially.

The greater the reduction in drug costs, the sooner PrEP will be cost-saving. In a linked comment article, Paul Revill of the University of York says: “These findings therefore call for NHS policy makers to negotiate with manufacturers to get favourable deals on prices and to be far sighted: invest now and reap long-term gains. This approach is currently challenging when unmet needs across all areas of health care are so great and NHS resources are stretched so thinly. However, this study provides the definitive evidence to support such a decision.” 

The study

Especially as PrEP is more expensive than some other interventions, NHS decision makers need to know whether the benefits of providing PrEP to gay men are likely to be greater or lower than the benefits of other interventions that may no longer be provided if resources are committed to PrEP.

Building on a previous analysis, the researchers used a dynamic individual-based mathematical model to estimate costs, the number of HIV infections prevented and the additional years lived in good health (quality-adjusted life years or QALYs) in men who have sex with men, if PrEP was introduced in 2016. They compared results against projections for a scenario in which PrEP was not introduced.

Some of the key assumptions in the model were as follows:

• The eligibility criteria for PrEP in men who have sex with men are broad and similar to those used in the PROUD study – testing HIV negative and having had condomless anal sex in the last three months. If condomless sex only occurred with an HIV-positive partner who had an undetectable viral load, PrEP would not be provided.
• PrEP is provided on an event-based regimen (taking pills before and after sex), as in the French Ipergay study. In those prescribed PrEP, there are 86% fewer HIV infections.
• Men with a higher number of condomless partners or with a recent history of sexually transmitted infections are more likely to start PrEP. For example, 30% of those with one condomless partner and 80% of those with ten or more partners would start PrEP. Men remain on PrEP while they continue having condomless sex, except for 10% each year who stop it. Each year, half of those who had stopped PrEP re-start it.
• Among men eligible for PrEP but not taking it, the rate of new HIV infections is 2% per year (in 2016), similar to the rate seen in repeat testers in sexual health clinics.
• PrEP drugs cost £4331 and HIV treatment costs £6288 per person, per year. These prices remain constant over time, apart from an annual reduction (discounting) of 3.5%.
• PrEP users test for HIV every three months.
• Sexual behaviour, uptake of HIV testing in non-PrEP users, and uptake of HIV treatment remain at current levels.

Other scenarios are possible. For example, the introduction of generic drugs will almost certainly lower medication costs in the coming years, although it is not clear when and by how much. Moreover, HIV incidence in potential PrEP users may be higher than 2% (in the deferred arm of the PROUD study, incidence was 9%). To deal with the uncertainty and the range of credible scenarios, the researchers included a series of ‘sensitivity analyses’ which were based on different assumptions.

Key findings

According to the model, the uptake of PrEP would be gradual and only a minority of the estimated 725,000 men who have sex with men living in the UK would be PrEP users. During the first year of the PrEP programme, 4000 men would start PrEP and by the year 2030, around 15,000 men would be taking PrEP, with a little under 40,000 ever having taken PrEP. They would take PrEP for a mean of 4.5 years each.

Thanks to current trends in frequent HIV testing and rapid uptake of HIV treatment after diagnosis, the model predicts some decline in the annual number of HIV infections, regardless of PrEP. But introducing PrEP will prevent a further 25% of HIV infections over 80 years: instead of 179,000 additional infections in gay men, there will be around 135,000. Of the infections averted, a little under half are infections directly averted in PrEP users and the remainder are due to the prevention of onward transmission.

Each HIV infection averted is estimated to result in a gain of five additional healthy life years (QALYs). Across the population, introducing PrEP would result in a gain of 220,000 healthy life years (QALYs).

Without PrEP, annual spending on HIV programmes would rise from around £450 million in 2016 to reach a peak of around £850 million in thirty years’ time. Over the 80-year time period, total spending on HIV programmes would be £64 billion.

Adding in a PrEP programme would result in more spending initially, with annual costs reaching their peak sooner, in twenty years’ time. But over the 80-year period, total spending on HIV programmes would be lower at £56 billion.

As is standard in cost-effectiveness analyses, financial expenditure and QALYs gained that occur in the future are ‘discounted’ (at a rate of 3.5% per year). This is meant to reflect the tendency of policy makers and the public to put less value on long-term consequences than short-term consequences. In other words, financial commitments and health improvements that occur far in the future are seen as less important than those that occur immediately.

After discounting, the PrEP programme would result in 40,000 QALYs gained, with a discounted saving in costs of £1 billion. Therefore, over the 80-year time horizon, the introduction of PrEP is cost-saving – the money saved on future spending on HIV treatment and care is greater than the cost of PrEP itself.

Changing the assumptions

The researchers conducted a series of sensitivity analyses, examining different scenarios in which the assumptions underlying the model were altered. In all of them, PrEP remained cost-saving for the NHS, although the amount of money saved varied.

Several analyses altered the cost of tenofovir and emtricitabine, drugs used both for PrEP and for HIV treatment. The patent protection for the Truvada combination pill will expire soon, allowing cheaper generic versions to be used. A 50% reduction in drugs cost from 2019 would mean that the PrEP programme would save £1.4 billion rather than £1 billion. An 80% reduction in costs would lead to the programme saving £1.6 billion.

Similarly, PrEP could save the NHS even more money in several other scenarios. This includes an increase in HIV incidence due to a modest increase in condomless sex in the wider population of gay men. (If the risk of acquiring HIV is greater, a PrEP programme will prevent more infections.) Extra savings would also occur if PrEP was used by more men who meet the eligibility requirements – if more men use PrEP, more sexual partnerships are protected by PrEP, more HIV infections are avoided and less money needs to be spent on HIV treatment.

In contrast, the amount of money PrEP could save the NHS would be lower (but PrEP would remain cost-saving) in other scenarios that were examined. A daily PrEP regimen involves more pills so would be more expensive than the event-based regimen, but would still save £0.6 billion. Savings would also be reduced if the uptake of PrEP was 50% lower, if more PrEP users stopped using PrEP despite still needing it, if PrEP users had condomless sex for a longer period of time, and so on.

When will PrEP start to save the NHS money?

Evaluating interventions over a lifetime (i.e. 80 years) is recommended by NICE, the body which considers the cost-effectiveness of NHS interventions, because the health benefits (such as reduced onward transmission) of an intervention may build up over a long period of time.

Nonetheless, some decision makers may want to consider shorter time horizons. In the base case scenario, PrEP would be cost-saving over 40 years. If drug costs were 80% lower, PrEP would be cost-effective over 20 years (£6000 per QALY gained). If HIV incidence increased, PrEP would be cost-effective over 30 years (£13,000 per QALY gained). If drug costs fell and incidence rose, PrEP would be cost-saving over 20 years.

Conclusion

“The results of our modelling study and economic analysis suggest that the introduction of event-based PrEP among MSM [men who have sex with men] in the UK with the eligibility criteria proposed is cost-saving and leads to health benefits, caused by a substantial reduction in HIV incidence among MSM,” conclude the authors.

“Although introduction of PrEP is cost-saving when considering an appropriately long time horizon, there are increases in overall costs for 20 years in our main results and it takes 40 years for the incremental cost-effectiveness ratio to reach less than £13,000 per QALY gained… Commissioners will have to sustain an additional cost for the first 20 years, unless drug prices are substantially reduced.”