The British HIV Association currently (July 2005) recommends that anti-HIV treatment should start before the CD4 count has fallen below 200, if the patient is ready for the commitment of treatment.

Around 33% of people who test positive every year in the UK already have CD4 counts below 200, the level at which immediate prophylaxis against Pneumocystis pneumonia is recommended, and the AIDS-defining limit in the USA.

Among gay men, the average CD4 count when diagnosed HIV–positive has remained almost constant since 1990, at around 350, but the proportion whose CD4 count was below 200 at diagnosis declined steadily, from 38% in 1993 to 25.4% in 2002. However the average CD4 count in those testing late was a life-threatening 60, compared with an average figure of 450 in the early testers, and the proportion of all deaths in gay men with HIV that happened in the year after diagnosis went up, from 8% of all deaths in 1996 to 27.5% in 2001.

Amongst those who have contracted HIV through injecting drug use, the average CD4 count at diagnosis is 315 cells. The proportion with CD4 counts under 200 at diagnosis was 37% in 2002.

In contrast, heterosexuals (the majority of whom are African) are being diagnosed with HIV at much lower median CD4 counts (around 200 to 260), with African men being diagnosed at an extremely late stage of infection (200 cells). Forty per cent of female heterosexuals and 49% of male heterosexuals (70% of whom are African) were diagnosed when their immune systems were already severely immunosuppressed.

Amongst Africans in the UK, fewer than 25% know their HIV status at least two months before developing an AIDS–defining illness. An increasing proportion of these `short interval' AIDS diagnoses amongst heterosexuals are cases of tuberculosis, which may occur at higher CD4 counts than other opportunistic infections (in other words, earlier in the course of HIV infection).

A survey from north London’s Royal Free Hospital HIV clinic found that half of all cases of AIDS diagnoses at the clinic occurred in the 15% of patients who did not come for an HIV test till their CD4 count was dangerously low (under 50).

Altogether 110 patients out of the 719 treated at the Royal Free between January 1996 (when combination therapy became available) and December 2002 were in this group of late testers.

Late diagnosis among those at low risk of infection

While most late HIV diagnoses in England, Wales and Northern Ireland are currently seen in black African heterosexuals, late diagnosis is also a problem for a smaller number of mainly UK nationals who may have low risk for HIV, may not be aware of their partners´ risks or may not acknowledge their own high risk of infection.

Newly-diagnosed and reported HIV infections in the UK are followed up if their exposure route requires clarification. Indepth confidential interviews are conducted if both the clinician and patient agree. In many cases HIV diagnosis is only made after frequent attendances with a range of symptoms to primary care physicians (GPs).

During 2001, the Communicable Disease Surveillance Centre (now the HPA) interviewed 265 people (155 men and 110 women) who had been diagnosed with HIV despite fitting into none of the known risk groups.

More than half of these individuals had not been diagnosed till the appearance of AIDS-related symptoms, and 60% had been infected in the UK or a low-prevalence country.

Of these 84 had been reported with an undetermined risk for acquiring HIV infection, 134 with a heterosexual risk, but without additional information and 47 were believed to have acquired their infection in the UK from partners not known to have a risk factor for HIV infection. After interview, 20% of the group turned out to have had high-risk sex abroad, but in low-prevalence countries, and 16% had had sex with a ‘high risk’ partner, generally a bisexual man.

Just under one quarter (23% or 62 individuals) were diagnosed because of the identification of an HIV-positive child or partner. In total 14% (37) were diagnosed because they were screened as a blood donor, at an antenatal or STI clinic, for insurance or for a visa. 8% (21) were tested of their own volition.

The majority of individuals reported frequent visits to their GP with a range of health concerns prior to diagnosis. For 213 of those interviewed, no risk factors for HIV were identified other than heterosexual contact, the majority of which occurred in the UK with partners believed not to be at any risk of HIV infection. Of the individuals diagnosed late (after the appearance of HIV-related symptoms, which were often misdiagnosed), 71% were men, compared with 42% of individuals diagnosed before symptoms appeared.

Almost all did not perceive themselves to be at risk of HIV infection and over half claimed little knowledge of HIV. A large proportion had been in a relationship for five years or more and this may have reinforced their sense that HIV was not of concern to them.

All those diagnosed late due to illness stated that no partners had contacted them to inform them of their possible HIV risk. In response to their own diagnosis, almost one third (of the 49 individuals asked) chose not to inform any of their partners. It is likely that some of the partners of those diagnosed late will also themselves be diagnosed late, as they will not perceive themselves to be at risk of HIV infection.

Primary care physicians should be encouraged to consider HIV testing as a possibility when patients “without risk” for HIV infection present with symptoms not responding to established treatments. The fact that these individuals presented to their GP on multiple occasions is particularly worrisome in light of Government plans to devolve HIV testing to Primary Care Trusts.

Undiagnosed HIV infections

About one-third of people in the UK with HIV are thought to be unaware of their infection. This undiagnosed proportion varies both between and within risk groups.

The HPA estimate that about 26% of gay men, 22% of injecting drug users, 22% of heterosexual women and 39% of heterosexual men with HIV are unaware of their infection, leading to the overall estimate that just over one in four HIV infections remain undiagnosed. The higher proportion of female heterosexuals with HIV infection diagnosed compared to males may be attributed to antenatal screening.

The proportion of undiagnosed infections also varies within risk groups. The annual HIV Prevalence and Sexual Behaviour Survey of gay men on the commercial scene in London, Manchester and Brighton, which gave anonymous saliva HIV tests to participants, found in 2003 that 33% of those with HIV were unaware of their infection and no less than 44% in London. In contrast, the HPA found that in 2003 2% of gay men nationally and 5% in London attending GUM clinics had undiagnosed HIV infection – suggesting that it may be people without STI symptoms or who think they have not been at risk that form the majority of the undiagnosed.

References

Gilbart VL. Late diagnosis of HIV infection amongst individuals with low, unrecognised or unacknowledged risks in England, Wales and Northern Ireland. abstract WePeC6078 Fourteenth World AIDS conference.

Gupta SW et al. CD4 cell counts in adults with newly diagnosed HIV infection: results of surveillance in England and Wales, 1990-1998. AIDS 14: 853-861, 2000.