HIV screening outside specialist settings is feasible, acceptable and effective in UK

The universal offer of a HIV test in UK general practice and hospital settings is highly feasible and acceptable to patients, researchers have concluded. Moreover, when conducted in areas of relatively high HIV prevalence, the interventions are effective in identifying a substantial number of people with undiagnosed HIV.

Staff who took part also found testing acceptable and reported few problems with it. Nonetheless many projects found that the number of tests offered varied considerably from doctor to doctor and from team to team. As one member of the audience at the Time to Test for HIV conference in London last week commented, “The greatest barrier to testing is doctors”.

The conference presented the findings from a series of pilot projects conducted in recent months in parts of England where the adult prevalence of diagnosed HIV is above 0.2% (2 in 1000). Whilst testing guidelines published in 2008 recommend the universal offer of an HIV test in medical settings in such areas, the implementation of this has been limited. Some pilots were conducted with general practioners (GPs or family doctors), some in acute hospital units and others in other hospital departments.

In addition, three outreach projects were conducted in community settings by voluntary sector agencies. Whilst relatively few individuals were tested within these programmes, the proportion who were diagnosed with HIV was extremely high.

The cost-effectiveness of HIV testing in these pilot projects still needs to be established (work is ongoing), but in the absence of any UK cost-effectiveness studies, it is often assumed that calculations made by the CDC in the United States provide a useful guide. The CDC judge that testing is cost-effective as long as at least 0.1% of results are HIV-positive.

To put the new findings in context, in English sexual health clinics, the proportion whose results are positive (the positivity rate) far exceeds this threshold. Looking at individuals living in areas with a relatively high HIV prevalence, the HIV-positive rate is 3.55% in tests of gay or bisexual men, 0.53% in other men and 0.45% in women. Even in heterosexual men and women in low-prevalence areas, the rate is 0.25%.

In English antenatal clinics, the average positivity rate is 0.18%, rising to 0.4% in London. In only two regions is the rate below 0.1%.

So to turn to the pilot projects, in general practices in Brighton and Lewisham, the positivity rates were 0.3% and 0.7% respectively. In acute hospital departments, the rates ranged from 0.14% (Brighton) to 1% (both in London and Leicester). However in a non-acute outpatient setting, no diagnoses were made at all.

But the highest positivity rates were seen in community outreach settings in London. Rates varied from 2.5% in a gay men's project and 1.9% in an African project to a lower 0.25% in an 'assertive outreach' programme with black Africans.

General practitioners (GPs) can offer HIV testing in a number of different circumstances. It may occur as part of the diagnostic process, when symptoms could be associated with HIV infection (but many symptoms are commonly seen and non-specific). Alternatively, a doctor may opportunistically suggest HIV testing to a patient thought to be at risk at HIV infection (but this requires a discussion of sexual practice and risk assesment, which many GPs find challenging).

In contrast, the two pilot projects described here follow a screening model - offering HIV testing to all adult patients newly registering with the practice, regardless of lifestyle or symptoms. In both cases, financial incentives were offered to GPs to participate. Moreover point-of-care INSTI tests, which provide results in one minute, were used and this considerably reduced the administrative burden of providing results.

Ruth Hutt presented the results from Lewisham. This inner-city London borough has a high HIV prevalence (0.7% pregnant women test positive) and a large black African population, but no sexual health clinic within its borders. The fact that there is a high turnover in patients at GP practices here means that a large number of people are new patients who can be offered an all-round health check, including HIV testing.

The primary care trust was pleasantly surprised that over the half the GP practices in the area expressed an interest in participating in the project, of whom 18 finally took part. Ruth Hutt emphasised that training was provided to all members of staff, including receptionists, healthcare assistants and nurses. In order to achieve this, each practice was visited and training provided onsite.

During the pilot period, 6183 new adult patients registered, of whom 4342 were offered a test and 2713 took the test. This amounts to an overall patient acceptance rate of 62%, but both the acceptance rate and the number of tests performed varied considerably from practice to practice. The researchers say that the biggest barriers to testing are staff attitudes and anxieties.

During the pilot 19 individuals were diagnosed HIV-positive, giving a high positivity rate of 0.7%. (An additional three people’s initial test results were reactive, but confirmatory testing procedures identified these as false positives.)

However, the researchers are concerned that four of the 19 people who received an HIV-diagnosis failed to attend their appointment at the specialist HIV clinic. Moreover, many of those who tested positive subsequently stopped attending the GP surgery where they had tested positive. This could be because individuals went to an HIV clinic for all their healthcare needs, but alternatively could suggest concerns around stigma and confidentiality.

The second project was conducted in Brighton, also an area with a high HIV prevalence, but with infections concentrated in gay men. People registering with a new GP were offered an HIV test, but uptake was hampered by the fact that only a third of people attend health check appointments when they register.

During the first four months of the pilot, 596 of 799 people offered a test took up the offer (74.6%). Those accepting the offer were more likely to be female, aged 20 to 29 and not to have tested in the last few months. Gay men were no more or less likely to accept the offer than other people. Once again, the number of tests offered and accepted varied considerably from practice to practice.

Questionnaires about the experience of being offered an HIV test suggested that even those who declined the offer thought that it was appropriate for HIV testing to be provided by GPs. For example, 93% of those who refused the offer agreed with the statement “It was a good idea to offer me an HIV test today during my new patient health check”. Only 10% of those who refused said they would prefer to test at a specialist sexual health clinic.

Commenting on these results, Phillipa Matthews, a London GP, noted that GPs’ interest and confidence in HIV testing tends to increase when they have had the experience of identifying previously undiagnosed infection and providing the result.

But she said that, at present, HIV testing only tends to be provided by “self-selected enthusiasts”. She pointed to some of the organisational barriers that need to be overcome. Neither HIV testing nor patient health checks at registration are part of the GP contract, so they aren’t specifically paid to do them. Moreover, there is no geographical flexibility in contract schemes such as the Quality and Outcomes Framework, meaning that it’s impossible to provide incentives only to practices in high-prevalence areas.

In hospital, the offer of an HIV test needs to be genuinely universal if it is to be effective, one of the pilot projects concluded. Although all new patients were meant to be offered the test, it became clear that clinicians targeted the offer to certain people. Within the project, two of 1413 patients were diagnosed with HIV (0.14%). But an anonymised testing programme, run in the same setting at the same time, found that a further six individuals had HIV infection that remained undiagnosed.

The project took place at the Royal Sussex County Hospital in Brighton, for all patients aged 16 to 79 admitted to acute general medicine. Doctors were asked to offer an HIV test to all patients not known to be HIV-positive.

In fact clinicians offered the test to less than half the patients, but the vast majority of patients accepted the offer. Of 3913 patients admitted, 1553 were offered the test (40%) and 1413 accepted the offer (91%).

In 92% of cases where the clinician did not offer testing, they did not record a reason for not doing so. While younger patients and those with a clinical indicator disease were more likely to be offered a test, there was once again great variability between doctors in the number of HIV tests offered. Doctors at more junior grades offered more tests, but there was also great variability between clinical teams, suggesting to the researchers the importance of clinical leadership.

The two individuals diagnosed HIV-positive as part of the programme were both black African and presented with clinical indicator diseases (one male, one female).

Looking at the extra six individuals identified by the anonymised programme but who remained undiagnosed, all were men and five were white British (the ethnicity of the sixth man was unrecorded). Their age ranged from 34 to 68, with three being over the age of 60 - doctors were significantly less likely to offer testing to those in this age group. In all cases, their sexuality was unrecorded (in fact the sexuality of 98.5% of all patients was not recorded and presumably not discussed, highlighting one challenge in attempting to target testing rather than offer it universally).

Had all eight cases of undiagnosed HIV been identified, the positivity rate would have been 0.2%.

The researchers suggest that in order to genuinely achieve routine opt-out testing, clincians’ barriers to HIV testing need to be better understood and tackled. But they point out that similar challenges were overcome in the implementation of opt-out antenatal testing, and suggest that there are lessons to be learnt from this experience.

Following the study, the hospital’s trust in Brighton has built HIV testing in to Commissioning for Quality and Innovation (CQUIN) payment framework. This provides financial incentives for the hospital to offer the test to at least 60% of newly admitted patients and for at least 90% of them to accept the offer.

While the pilot project at the acute medical unit in Leicester had a high overall positivity rate (1%), the proportion of eligible patients tested was low and varied month by month. At the beginning of the pilot 22% of patients were tested but this dropped at one point to 6%. There was considerable variation from one doctor to another.

In London, the HINTS project evaluated offering testing in three hospital settings - the emergency department at the Chelsea & Westminster (positivity rate 0.2%), the acute care unit at Homerton (1%) and the outpatient dermatology clinic at King's College (none).

Of those who were offered a test, two-thirds took up the offer. Those who declined to do so most commonly cited having tested recently, not being at risk and having other health concerns at the time as their reasons.

Staff surveys conducted before the project began showed that more than three-quarters of staff agreed with the principle of HIV testing being offered routinely to everyone and with this occurring in their own hospital department. Nonetheless, only 42% of acute care unit staff and 57% of emergency department staff agreed that they would feel comfortable offering tests themselves.

Staff concerns included needing additional training, there not being sufficient privacy, not having enough time and being unable to answer all the patients’ questions. However, at the end of the pilot, focus groups showed that many of these fears had been allayed, especially the latter two.

Commenting on these studies, Andy Ustianowski of North Manchester General Hospital noted that making ongoing testing sustainable could be challenging. It needs to become embedded in routine practice but it remains unclear what will facilitate this. Mentioning HIV screening in general patient information leaflets and on admission forms, making testing a performance indicator and providing resources for staff training may all help. He also suggested that HIV specialists should jointly develop procedures with their colleagues in other specialities and avoid attempting to impose systems from outside.

The conference also heard results from outreach activities in community settings. A Terrence Higgins Trust project targeted black African communities in London, working in partnership with 22 community groups and community centres. Over eleven months, 407 people tested (11.5% of those who were approached). Three individuals tested positive but in fact two had previously been diagnosed, yielding a new diagnosis rate of 0.25%.

A partnership of Positive East, the Metro Centre and the West London Gay Men’s Project also offered testing to black African communities in London. Despite testing being offered several times a week at three separate sites, only 106 people tested in a six-month period (the number who were approached was not recorded). With two individuals testing positive, the positivity rate was 1.9%.

The same partnership offered testing to gay and bisexual men at three different sites, out of office hours. In a six-month period 161 men were tested, of whom four had HIV. This amounts to a positivity rate of 2.5%, the highest seen in any pilot project.

Acceptability analysis showed that 'while-you-wait' results, opening outside work hours and, in the case of Africans, operating clinics in African-specific settings were all cited as reasons for using community testing sites rather than a clinic. Half the people in the Terrence Higgins Trust project had never taken an HIV test before.

These projects revealed a number of challenges, particularly forming partnerships to host and promote services, finding and training staff prepared to work irregular hours, and building up trust and awareness of the service in the community. These challenges were exacerbated by the short-term provision of funding, meaning projects were just bedding in when their funding came to an end.

Summing up the community results, Yusef Azad of the National AIDS Trust commented that community testing yielded more diagnoses if programmes were funded to continue beyond a pilot stage and if they were carefully targeted to underserved populations.

He pointed to a project run by Body Positive North West in Manchester that has identified 21 new diagnoses in 863 tests (2.4%) in a two-year programme. Moreover a programme run by Trade Sexual Health in Leicester, performing tests in a gay sauna, had produced the very high figure of seven new diagnoses in 91 tests (7.7%). This latter project identified a population of largely Asian and married men who have sex with men who had never been to an STI clinic but who were willing to test in a sauna environment.

Powerpoint slides for all presentations are available on the BHIVA website. Moreover, abstracts for all studies are included in the Health Protection Agency report Time to test for HIV: Expanded healthcare and community HIV testing in England.

Hutt R and Bell S Primary care HIV screening in a high prevalence area (abstract AB5).

Bryce G et al. A study to assess the acceptability, feasibility and cost-effectiveness of universal HIV testing with newly registering patients (aged 16-59) in primary care (abstract AB4).

Perry N et al. HIV testing in acute general medical admissions must be universally offered to reduce undiagnosed HIV (abstract AB2).

Palfreeman A et al. Leicester acute medical admissions unit HIV testing pilot (abstract AB3).

Rayment M et al. HIV Testing in Non-Traditional Settings – the HINTS Study (abstract AB1).

Brady M et al. Community HIV testing: the feasibility and acceptability of assertive outreach and community testing to reduce the late diagnosis of HIV (abstract AB7).

Ussher G et al. The MSM Comparative Community HIV testing pilot (abstract AB6a).

Ussher G et al. African Community Testing (ACT) Pilot (abstract AB6b).