EACS proposes list of ‘indicator diseases’ to prompt the offer of an HIV test

At its recent conference in Madrid, the European AIDS Clinical Society (EACS) proposed drawing up a list of ‘indicator diseases’ that would be disseminated to all doctors in Europe that indicate people who are at high risk of also having HIV infection, with the idea that the diagnosis of one or more of these would trigger the offer of an HIV test.

This proposal will be debated at an invitation-only conference of HIV healthcare workers, advocates and policy makers on 26-27 November this year in Brussels called HIV in Europe - Working Together for Optimal Testing and Earlier Care. This conference has been convened by AIDS Action Europe in partnership with the University of Copenhagen, with support from EACS, the European AIDS Treatment Group (EATG), HIV Europe and a number of other organisations including the British HIV Association.

The Indicator Disease proposal is one way of trying to reduce the number of people with HIV in Europe who remain undiagnosed. Professor Jens Lundgren of the Copenhagen HIV Programme, who will co-chair the HIV in Europe conference, told the EACS conference that 50-60% of people with HIV in Europe as whole remain undiagnosed. “This is the biggest unresolved public health problem in Europe,” he said. “We may think HIV is out of control in Africa, but we haven’t cleaned it up in our own back yard.”

Nikos Dedes of EATG added that this split into about 30% undiagnosed in western Europe and about 70% undiagnosed in the countries of eastern Europe, with some estimating even higher rates. Western countries also have populations that have high rates of non-diagnosis: for instance, the Health Protection Agency in the UK has found that 47% of heterosexuals with HIV are not diagnosed till their CD4 counts fall below 200 cells/mm³.

Dedes added: “You can improve the number of diagnoses by better targeting, but you need enabling factors that do not frighten people off taking tests, which is why it’s a subtle public health issue.”

For this reason, Jens Lundgren said, the EACS had decided against advocating for universal opt-out testing in the same way as the US Centers for Disease Control and Prevention (CDC) had done in September 2006. However there has been concern that this may be a coercive form of testing, and some research has suggested that carefully-targeted testing would be more efficient than effectively screening the entire sexually-active population.

“There has been a push that everyone who presents to healthcare settings should have an HIV test,” said Lundgren. “I would characterise that as not compatible with European philosophy. But we can identify patients with early signs of disease that should prompt testing.”

Professor Brian Gazzard of London's Chelsea and Westminster Hospital said it had been estimated that universal screening in a population with 0.1% HIV prevalence, the average amongst the general public in the USA, would cost US $113,000 (€78,500, £55,000) per Quality-Adjusted Life Year (QALY) saved. This was clearly not cost-effective. But in a group with 1% prevalence, targeted screening would cost $38,000 (€26,400, £18,500) per QALY saved, which would come within the generally-accepted limit of £30,000 for a cost-effective intervention.

Professor Nathan Clumeck of the St Pierre University Hospital, Brussels, said that the indicator diseases should therefore be ones with ‘comorbid HIV incidence’ of more than one per cent: in other words, where more than one per cent of people diagnosed with the illness also turn out to have HIV.

Clearly this is the case with AIDS-defining illnesses like TB and cervical cancer. But there were a number of other conditions strongly indicative of HIV ranging from STIs like LGV (comorbid HIV incidence 76%) and gonorrhoea (6%) to cancers like Hodgkin’s Lymphoma (comorbid HIV incidence, 9-18%). There were other diseases where the comorbid HIV incidence was much less certain, for instance clinical candidiasis or disseminated shingles, and many where it was not known at all, such as certain digestive and fungal infections. These were often suggestive either of impaired immunity or HIV risk behaviour but, commented Professor Gazzard, “Though we often know the disease prevalence in people with HIV, the HIV prevalence in people with the disease is often much less clear.”

Clumeck commented that “We don’t want to put GPs into a position where they have to HIV-test almost everyone who has almost every disease.” For this reason the list of indicator diseases had to be carefully drawn up.

Members of the audience questioned whether a list of new indicator diseases was necessary when rates of HIV testing in people presenting with AIDS-defining illnesses were low. For instance, a 2006 UK study found that only 49% of patients attending one UK TB clinic were offered an HIV test and of those who were, 31% refused a test, meaning that only a third of outpatients with active TB received an HIV test. HIV prevalence in outpatients tested was 2%.

Lundgren agreed and said that “the challenge will be to get colleagues to understand that when they see one of a particular list of HIV indicator diseases, they should offer a test.”

Gideon Hirsch of the Israeli AIDS Task Force commented that as a lot of the Indicator Diseases were suggestive of advanced immune deficiency, patients in early infection would be missed.

Professor Gazzard commented that this was “the dilemma: I agree it would make more sense to test high-risk groups, but it is very difficult for GPs and non-HIV practitioners to identify them. So we thought a list of indicator disease would be better.” He agreed however that routine HIV testing policy needed to be harmonised in Europe, too; he said he had been shocked to learn that pregnant women in Belgium are still not routinely offered HIV testing, for instance.

Audience members commented that it might also make sense to publicise the symptoms associated with acute HIV infection, but Prof. Gazzard commented that there was no consensus definition of these symptoms.

A study from Denmark also presented at the EACS conference supported the idea that pre-HIV-diagnosis morbidity from other diseases is significantly correlated with HIV diagnosis and subsequent mortality – see separate story.