Massive scale-up of HIV testing, plus HIV treatment on diagnosis, needed to curtail UK gay epidemic

Incidence could drop below 0.1%
This article is more than 10 years old. Click here for more recent articles on this topic

Increasing the proportion of HIV-positive gay men in the UK who have an undetectable viral load from the current figure of around 60% to 90% could result in a substantial drop in new HIV infections, a modelling study suggests. In order to achieve this, it wouldn’t be enough to change treatment guidelines – a substantial increase in HIV testing rates would also be required, Andrew Phillips of University College London told the 21st Conference on Retroviruses and Opportunistic Infections (CROI) in Boston last week.

However, Phillips noted that any decreases in current rates of retention in care, treatment adherence or condom use could have a negative impact on these predictions.

Reviewing the HIV epidemic among gay and other men who have sex with men (MSM) in the UK, Phillips said that there are approximately 600,000 MSM in the country, of whom around 45,000 are living with HIV. While approximately 20% are unaware of their infection, over 80% of diagnosed men are taking HIV treatment and 95% of those on treatment have an undetectable viral load. In all, approximately 60% of gay men living with HIV are virally suppressed.


retention in care

A patient’s regular and ongoing engagement with medical care at a health care facility. 

virological suppression

Halting of the function or replication of a virus. In HIV, optimal viral suppression is measured as the reduction of viral load (HIV RNA) to undetectable levels and is the goal of antiretroviral therapy.

mathematical models

A range of complex mathematical techniques which aim to simulate a sequence of likely future events, in order to estimate the impact of a health intervention or the spread of an infection.

primary infection

In HIV, usually defined as the first six months of infection.

person years

In a study “100 person years of follow-up” could mean that information was collected on 100 people for one year, or on 50 people for two years each, or on ten people over ten years. In practice, each person’s duration of follow-up is likely to be different.

It might therefore be expected that this high coverage of HIV treatment would have a substantial preventive effect, and that new infections would be falling. However, Phillips’ previously published modelling suggests that, in recent years, around 2500 men have newly acquired HIV each year, and this figure is either stable or increasing. Incidence is around 0.45% (in other words, each year around four in 1000 men become HIV positive).

His analyses suggest that while current provision of treatment has curbed infections, its impact has been undermined by an increase in the number of men having sex without a condom. Similar conclusions have been reached by researchers looking at the gay epidemics in the Netherlands and Switzerland. Moreover, epidemics of HIV in men who have sex with men continue to expand in most countries of the world.

Future scenarios

Andrew Phillips then shifted his attention to predictions of the potential impact of changes to treatment guidelines, or the uptake of HIV testing, or both. His estimations are based on mathematical models which take into account multiple sources of data on the UK epidemic and varying assumptions about sexual and testing behaviour.

In terms of policy changes, he considered the impact of immediately recommending HIV treatment after diagnosis with HIV, rather than the current treatment guideline of starting treatment with a CD4 cell count of 350 cells/mm3. The application of this policy should depend on the outcome of the START randomised trial, examining the impact of early treatment on the health of the person living with HIV. 

He also considered the possibility of changes to HIV testing, involving both an increase in the frequency of testing and a reduction in the number of men who have never tested. At present, approximately 40% of men are diagnosed within a year of acquiring their infection, but Phillips modelled the impact of two possible scenarios:

  • A scale-up of testing, so that 60% are diagnosed within a year of infection.
  • A larger scale-up, so that 90% are diagnosed within a year.

He did not know how such changes in testing behaviour could be achieved, but did say that they were "highly ambitious" and would likely require substantial investment.

The model forecast new infections up to the year 2030. It predicted that if there was no change in treatment policy or in testing, over 3000 men who have sex with men would acquire HIV during the year 2030. However, there is potential to reduce the number of infections through policy changes:

  • Immediate treatment could reduce infections by 32%.
  • A scale-up of testing could also reduce infections by 32%.
  • A larger scale-up of testing could reduce infections by 54%.
  • Immediate treatment and a scale-up of testing could reduce infections by 64%.
  • Immediate treatment and a large scale-up of testing could reduce infections by 80%.

In the latter scenario, around 600 gay and other men who have sex with men would acquire HIV in 2030. 

Preliminary analysis suggests that this scenario would be highly cost-effective. Although many more people would be taking HIV treatment in the first few years, the numbers on treatment would be falling by 2030 as fewer men acquire HIV.

Phillips noted a number of factors which could influence these outcomes. If retention in care and adherence deteriorated, so that 15% fewer men with HIV had an undetectable viral load, there would be around 1000 more infections each year.

Changes in sexual behaviour could also have an impact – a decrease in condom use is possible as the impact of treatment on transmission becomes better known. However, the effect of this would vary in the different scenarios.

If testing and treatment policy does not change, a relatively small increase of 10% more men having sex without a condom each year would result in a massive increase in infections – up to around 6000 in the year 2030.

In the scenario of immediate treatment and a large scale-up of testing, there would be a smaller pool of men with infectious HIV, so the impact of sexual behaviour change would be lesser. A 10% increase in sex without a condom would result in a few hundred more infections each year.

Phillips also suggested that in the most optimistic scenario, primary infection might have less impact on the epidemic than it does at present. He said that the effective treatment of men who have chronic infection would help prevent outbreaks of primary infection occurring. 

Summing up

Phillips’ analyses suggest that if the current situation is maintained – in which around 60% of men living with HIV have an undetectable viral load – over 3000 men will acquire HIV each year.

If fewer men are virally suppressed, infection rates would be dramatically higher. If 30% of men had an undetectable viral load, well over 10,000 men would be newly infected each year.

But an improvement of viral suppression rates to 90% could result in only around 600 annual infections, an incidence of less than 1 per 1000 person-years.

He concluded by noting that to achieve the 90% viral suppression rate, it will be necessary for 90% of men to be diagnosed within a year of infection; for treatment to begin immediately after diagnosis; for linkage, retention and adherence to remain good; and for levels of condomless sex not to increase.


Phillips A MSM in the UK: Prevention Effects of ART in Perspective. 21st Conference on Retroviruses and Opportunistic Infections, Boston, abstract 116, 2014.

A webcast of this session is available through the CROI website.