HIV diagnoses in English gay men have been falling since 2014

A new analysis by Public Health England of testing rates and HIV diagnoses from all of England’s sexual health clinics shows that the decline in diagnoses is England-wide, started at least a year before the decline was first noticed at London's 56 Dean Street clinic, and is not restricted to gay men who test frequently.

One of the big stories in HIV prevention this year has been the dramatic drop in HIV diagnoses being seen among gay men in HIV clinics in London. This started when, less than a year ago, Dean Street clinic announced a 40% drop in infections in 2016. This trend, there and at other inner London clinics, has been maintained to the extent that there were only four HIV diagnoses at Dean Street in October 2017 compared with 60 in October 2015 – a drop of 93%.

However Noel Gill of Public Health England, presenting data from English sexually transmitted infection (STI) clinics at the recent 16th European AIDS Conference (EACS 2017) in Milan, said that there was evidence that the falls in HIV diagnoses had started earlier and were national.

In summary, HIV diagnoses have fallen by 65% in London and by 48% outside London from their peak in 2014. In the third quarter of 2014 there were approximately 370 HIV diagnoses in gay men in London and 250 outside. They then started to fall, and in the second quarter of 2017 the total of new HIV diagnoses both in and outside London was 130. The decline was already well underway in London by mid-2016, which is when Dean Street noticed it. London HIV diagnoses in the second quarter of 2016 had already fallen by 46% relative to their peak in 2014.

What has caused the decline in diagnoses? Noel Gill did not ascribe it to any one cause, but said that the “lower threshold of access to HIV testing” with results available within an hour and the use of self-sampling and, increasingly, self-testing as options were all helping. He also said it was important that STI clinics in the UK were generally “held in high regard” by their users.

An increase in attendances and in the frequency of HIV testing among gay men are associated with the declines in diagnosis.

Starting in 2014, there was a large increase in the number of attendances at STI clinics by gay men and in the frequency of HIV testing. There were 14,500 attendances by gay men at high risk of HIV in London clinics in the first quarter of 2014. This expanded, within 18 months, to about 19,000 attendances and has stayed at that level since. Out of London, the increase took longer, but there were over 14,000 attendances per quarter by the third quarter of 2016, up from 10,000 in early 2014.

The number of attendees who were classed as “frequent testers” increased in London from 6400 in the first quarter of 2014 to 9700 in the first quarter of 2015 and then stayed at that level. Again, the increase was slower to take hold outside London but increased from 4000 in the first quarter of 2014 to 6000 in the first quarter of 2016.

The quarterly number of HIV tests among frequently testing gay men increased from roughly 3500 per quarter in 2012 to over 8200 in 2017 and in men who tested less frequently from about 4500 to about 7000. However, the number of tests in the less-frequent testers has stayed at that level since early 2015 and may even have fallen slightly. This may indicate that there needs to be new encouragement for frequent testing among gay men who are not aware of the importance of testing frequently, or who are not at high risk or do not see themselves as ‘high risk’.

This is especially important as there were more HIV diagnoses among the less-frequent testers than among the frequent testers. This may seem counter-intuitive, but frequent testers are presumably more health-conscious, and may already belong to networks where most of the HIV-positive people are diagnosed and on HIV treatment. Or, because tests are spaced more widely in infrequent testers, the proportion of results that are positive are higher, simply because more time has elapsed. And finally, ‘infrequent testers’ includes ‘first testers’ – a proportion of the diagnoses are late diagnoses in people infected years ago.

Whatever the reason, the quarterly number of diagnoses in frequently testing gay men averaged about 40 in 2012, increased to 70 in 2014-15 and then decreased again to 40 in 2016.

In infrequent testers there were more diagnoses, but the decline in diagnoses was also more marked. They averaged 150 per quarter in 2012-13, increased to about 200 a quarter in 2014, but started to decline from mid-2015 onwards and in the second half of 2016 were only averaging 80 per quarter.

The public health significance of frequent testing is only fully realised if people who are diagnosed start antiretroviral therapy (ART) as soon as possible, as reducing the time spent between infection and viral suppression is the most critical factor in bringing down HIV transmission.

In 2012, the time taken until half of those diagnosed were on ART was 1.2 years and until 90% were on ART was 3.6 years. By 2014 this had been reduced to 3 months and 1.9 years respectively. The latest figures, from 2016, show that 50% of people diagnosed are on ART within a month and 90% within 3 months.

So far, Gill said, the three most important factors contributing to the observed decline were:

• A 50% increase in STI clinic attendance in gay men since 2011
• An increase in the frequency of gay men’s HIV testing, with the average now 2.5 tests a year
• Putting 90% of those diagnosed with HIV on treatment within less than a year.

He did not speculate on whether there was an additional role for pre-exposure prophylaxis (PrEP), as the decline in diagnoses started happening before its more widespread availability. However, PrEP may have subsequently accelerated the decline in the inner-London clinics, as the rate of decline in clinics like Dean Street and Mortimer Market is both more recent and steeper than that seen in the Public Health England data. San Francisco is another example of a city where HIV diagnoses were already declining but where the rate of decline increased after clinic attendees started using PrEP.