Background

Reducing undiagnosed HIV infection in the UK remains a public health priority and there has been much discussion as to whether there should be universal offer of testing for women attending abortion services. In 2008, the British HIV Association (BHIVA) recommended this in their National Guidelines for HIV Testing.1 More recently, the National Institute for Health and Clinical Excellence (NICE) have published guidance on increasing HIV testing in Black African communities, in which they support the offer of a test to all women attending abortion services in accordance with the BHIVA guidelines.2

However, there remains a paucity of good research on testing in this particular setting. In 2004, evidence-based guidelines for women requesting induced abortion published by the Royal College of Obstetricians and Gynaecologists (RCOG) recommended selected offer of HIV testing, leaving the decision to local services based on population prevalence and available resources.3 These guidelines are soon to be updated and it will be of interest to see if the RCOG change their stance on this in light of the above mentioned publications.

Living with HIV

In the UK there were an estimated 86 500 people living with HIV infection at the end of 2009, and an estimated 26% of these were unaware of their infection.4 Of particular note, there has been an increase in infections that are acquired within the UK in heterosexual people, from an estimated 740 in 2004 to 1130 in 2009.4

With the advent of effective antiretroviral therapy, HIV infection has become a manageable, long-term condition, and the estimated survival for a person diagnosed at the age of 20 years is approximately two-thirds that of the general population.5 Yet despite this, over half of patients are not diagnosed until their CD4 count has fallen below 350, the level at which initiation of treatment would usually be recommended.4 A BHIVA audit in 2006 demonstrated that 24% of deaths in HIV-positive people in the UK occurred because the diagnosis was made too late for effective treatment.6 The message here is clear: diagnose people earlier and we will see fewer deaths from HIV.

Achieving a reduction in the number of people who remain unaware of their HIV infection will also reduce the number of new cases of HIV. For example, the aim of antiretroviral therapy is to suppress viral replication, and it is very rare for individuals with a low viral load to infect others.7 There is also evidence to suggest that knowledge of HIV status can reduce high-risk sexual activity.8 It is estimated that onward transmission from those unaware of their infection is 3.5 times more likely than from those known to have HIV.8

Cost effectiveness of HIV testing

Given the clear benefits of HIV testing, why is it that we still do not universally offer testing to women requesting abortion? The primary aim of an abortion clinic is to terminate an unwanted pregnancy and it could be considered that offering HIV testing is not directly related to patient care in this setting. This is somewhat different to the argument for chlamydia testing, the presence of which can result in increased complications as a result of the procedure. In general though, my perception is that colleagues in abortion services are not actually opposed to testing. More fundamentally the issue appears to be that in these times of increased financial constraints there remains no robust evidence of the cost effectiveness of universal testing within this particular patient group.

One of the major factors when considering the cost effectiveness of HIV testing is the population prevalence. BHIVA, the Health Protection Agency (HPA) and NICE all refer to a population prevalence of greater than 2:1000 (0.2%) as a level at which enhanced screening should be considered in the UK, and in this setting recommend universal testing for all medical admissions and new registrants with general practitioners.1 2 4 Thirty-seven English primary care trusts now have a general population prevalence above this level.4 In America, a lower population prevalence of 1:1000 is generally suggested for universal screening, although some reports suggest that this could be cost effective even with a prevalence as low as 0.05% (1:2000).9

Prevalence of HIV

What is the prevalence of HIV in women attending for abortion? There are some recent studies published which adopted an ‘opt out’ approach to testing, and rates of detection have been reported up to 0.77%.10,–,12 Two of these studies report on previously undiagnosed prevalence rates of 0.52–0.6%, indicating that the total population prevalence will be higher than this and in excess of the 2:1000 (0.2%) level suggested for screening.11 12 That said, large numbers of participants are required to accurately establish a low prevalence within a population, and some of these studies have wide confidence intervals. In addition, all were undertaken in or around London, an area known to have a higher prevalence of HIV within the general population, and as such the findings are unlikely to be transferable to the rest of the UK.

Further evidence of prevalence can be derived from unlinked anonymous screening programmes, where the trend would appear to be towards a higher prevalence of HIV infection in women attending for abortion than in women attending antenatal clinics, a setting where universal offer of testing has been in place since the late 1990s. Unlinked anonymous testing for HIV in six London abortion clinics in 2006 found a population prevalence nearly three times that seen in London antenatal clinics (10.8:1000 vs 4.2:1000).13 The advantages and hence cost savings of diagnosing HIV in pregnancy are, of course, greater in terms of prevention of neonatal infection, and hence offering testing at a lower population prevalence is more acceptable. However, it does seem somewhat inequitable to be offering antenatal testing but not to offer testing to a population with high rates of other sexually transmitted infections (STIs), increased likelihood of recent partner change and a known higher prevalence of HIV infection.14

Universal testing has already been introduced in some clinics and the anticipated obstacles explored. Concerns about patient acceptability have generally been unsubstantiated. One recent study looking at attitudes to testing in an abortion clinic found that only 2% of women thought offering HIV testing was unacceptable.15

Uptake of HIV testing

Although seemingly acceptable to women, actual uptake of HIV testing is generally lower in abortion clinics than that seen in antenatal clinics and this will impact adversely on the cost effectiveness of screening. Uptake rates of between 37% and 96.4% have been reported.10,–,12 It would appear that uptake is related to the experience of the staff offering testing, and in general increases over time when testing is first introduced. With experience though, high levels of uptake can be achieved and data from one service that has been offering testing for over 5 years now reported an uptake of over 96%.10 In the non-specialist setting, the time of the last HIV test, self-perception of low HIV risk, age of the patient, and the approach of individuals or practices to offering testing have all influenced uptake rates.16 Many of these factors have also been observed in abortion clinics.12 17 It is of course possible that those at higher risk of infection, including women who know they have risk factors for HIV, may self-select out of testing. In the setting of HIV testing in abortion clinics though, there does not seem to be any significant variation in uptake across demographic characteristics such as ethnicity and country of origin.12 17

Barriers to HIV testing

In practical terms, issues such as additional workload and staff knowledge and confidence in offering testing may represent barriers. There may be concerns regarding the requirement for pre- and post-test counselling, legal aspects of HIV care, or confidentiality issues. However, many of these issues can be easily overcome with adequate training and the development of robust care pathways. As part of normalisation of testing, there has been a significant move away from the idea that lengthy pre-test discussion, delivered only by appropriately trained individuals, is required before HIV testing can be undertaken, and this move has been actively supported in the BHIVA guidelines.1 Most genitourinary medicine (GUM) clinics now issue leaflets on HIV testing and only offer additional pre-test discussion if high-risk factors are identified during the consultation. Insurance companies are no longer allowed to ask if an individual has been HIV tested, only if they are known to have HIV infection.18

The view held by BHIVA is that it should be within the capabilities of all health care workers, including both nursing and medical staff, to offer HIV testing1 and perhaps we should remember that many patients actually believe that testing is routine anyway.

In terms of issuing positive results, care pathways have been successfully developed in other areas of HIV testing to involve GUM staff as soon as a positive or equivocal result is received. GUM staff can offer advice to the requesting clinician on delivery of the result and in some areas may even attend the consultation to offer immediate patient support. Confirmatory testing can then be undertaken without delay and follow-up arranged in the GUM clinic.