Patients' attitudes towards increasing the offer of HIV testing in primary and secondary care

BACKGROUND

Recent reports have shown that 28% of the estimated 77,400 people living with HIV in the UK are unaware of their infection. ¹ By the time they are diagnosed, the CD4 count has dropped below 200 in a quarter of patients. ¹ This late diagnosis leaves patients at the risk of developing AIDS-defining conditions and leads to 24% of all deaths in HIV-positive adults, ² higher direct patient care costs ³ and increased risk of onward transmission of infection. ⁴

Many patients seek medical care in the year preceding their HIV diagnosis. ⁵ This represents a missed opportunity for preventing morbidity and mortality. In response to these data, the 2008 UK National Guidelines for HIV testing ⁶ outline situations where HIV testing should be performed. These guidelines recommend that testing should be routinely offered and recommended to all patients presenting for health care with HIV-associated conditions, the so-called ‘clinical indicator’ conditions. Additionally, they recommend that an HIV test should be considered in all patients registering in general practice or being admitted to hospital in areas where the local prevalence exceeds two per 1000 population. This opt-out form of testing has been shown to be acceptable to patients in an antenatal setting. ⁷ However, the acceptability of routinely offering and recommending HIV testing to patients in other settings is unknown.

In Cornwall, the population HIV sero-prevalence is unknown but the prevalence of diagnosed infection is the lowest in the south-west region, reported as 32 per 100,000 population aged 15 or over in 2008. ⁸ Routine testing for HIV in Cornwall is presently offered only in genitourinary medicine and antenatal clinics, with an uptake of 72% and 94%, respectively. ^9,8 We were interested in establishing potential barriers towards HIV testing as recommended in the national guidelines. We wished to obtain the views of patients attending primary care and hospital outpatients settings and feed this information back to health-care professionals from disciplines that may encounter ‘clinical indicator’ conditions.

METHODS

Six primary care practices in Cornwall were approached and all agreed to participate, as did the following hospital outpatient settings: neurology, respiratory medicine, gastroenterology, dermatology, eye casualty, gynaecology and the unplanned pregnancy unit. Two other hospital specialties were invited to participate, but they declined. Patients who attended participating services between November 2008 and January 2009 were given a short information sheet regarding HIV (Appendix 1) and then asked to respond to a series of statements by circling one of the five options (strongly agree, agree, neither agree nor disagree, disagree and strongly disagree). Participation was voluntary and anonymous. The questionnaires were distributed by the reception staff working in primary care and hospital outpatients settings, as funding to employ additional staff to recruit patients was not available. When staffing levels did not allow for direct distribution of the questionnaire, they were made available in the outpatient waiting area. Patients were asked to place completed surveys in boxes provided. Due to the public setting in which patients were asked to complete the survey, specific questions regarding risk behaviours for HIV were not asked. Instead, patients were asked if they would identify themselves as being in a high-risk group for HIV.

The local ethics committee deemed that the project did not require formal ethical approval. The P values were calculated using chi-squared or Fisher's exact test.

RESULTS

Six hundred and sixteen patients participated from the following settings: 339 (55%) primary care, 44 (7%) dermatology, 48 (8%) gastroenterology, 22 (4%) gynaecology, 52 (8%) neurology, 53 (9%) ophthalmology, 21 (3%) respiratory medicine and 37 (6%) the unplanned pregnancy clinic. The men to women ratio was 1:2.2 and the median age group was 45–55 (range from under 25 to over 75). There was no significant difference between the gender (P = 0.26) or age group (P = 0.15) of respondents from primary or secondary care. Responses to the statements are shown in Table 1.

Patients attending primary care were significantly more likely than those attending secondary care to be unwilling to test if presenting with a medical problem known to be associated with HIV (11/339 [3.2%] cf. 1/277 [0.4%], P = 0.01). However, the actual number of patients who responded negatively to this statement in either setting was small and the relevance of this difference is uncertain. There were no other statistically significant differences between the responses from primary and secondary care to any of the other statements.

Fifteen out of 616 (2.4%) patients identified as being from a high-risk group for HIV. All of these patients stated that they would be willing to be tested if presenting with a medical problem associated with HIV. None stated that they would not be willing to be tested routinely, although one patient circled ‘neither agree nor disagree’ for this statement. Two patients from this group stated that they would not be willing to have the result of their HIV test in their main medical notes.

Patients were asked to give reasons why people might decline HIV testing. The responses were as follows: do not feel at risk (n = 102), concerns regarding insurance (n = 27), concerns regarding confidentiality (n = 23), do not want to know HIV status (n = 2).

DISCUSSION

Ninety-four percent of patients responded that they would be willing to have an HIV test if presenting with a problem associated with HIV, in accordance with the national guidelines, while 71% stated that they would be willing to be tested as part of their routine care. This figure is very similar to the 72% of patients accepting routine HIV testing in genitourinary medicine clinics in Cornwall. ⁹

Patients who are at higher risk of HIV have previously been shown to be less likely to accept an HIV test. ¹⁰ Data on ethnic origin were not collected in this study, but the local population is not ethnically diverse ¹¹ and has a low prevalence of HIV. This may explain the apparent high acceptability of HIV testing in our study sample. However, all patients who identified as being from a high-risk group agreed that they would be willing to be tested for HIV in line with the national guidelines, although why these patients self-identified as being in a high-risk group is unknown.

The questionnaire survey was completed by a self-selected sample of patients. Each of the involved outpatient clinics has approximately 500 patient attendances per month. Due to the method of distributing the survey, it is not possible to determine the number of these patients who were offered the survey. However, with only 616 responses, only a small proportion of patients attending during the study period completed the survey and this introduces an element of bias. The demographics of patients who were either not offered or declined to complete the survey is unknown, and those patients who completed the survey may not be representative of the whole patient population. There were a disproportionate number of women respondents, which may be a consequence of the settings where the survey was distributed, e.g. gynaecology, unplanned pregnancy clinic.

Only 8% of patients stated that they would not be willing to have their HIV test results in their main medical notes. A greater proportion (13%) of patients who identified themselves as being from a high-risk group stated they would not be willing to have their HIV results in their main medical notes. The numbers are too small to demonstrate whether this was a significant difference (P = 0.44). It appears that many of the patients in this survey were only willing to have the result of their HIV test in their medical notes because they expected that result to be negative.

Patients responses regarding the extent of information and counselling required prior to testing are incongruent. While only 7.8% stated that the information on the sheet would be insufficient prior to taking a test, three times more (28%) stated they would require further one-to-one counselling prior to taking an HIV test. This may be a result of bias originating from the layout of the response options. This discrepancy was not revealed in the initial pilot of the survey. Further research into this area is required.

We were surprised by the resistance to the project we encountered from various health-care and administrative colleagues in the hospital setting. Of the 10 secondary care specialties invited to participate, two declined. One surgical specialty initially agreed to be included but then withdrew when a consultant was asked a question about HIV from a patient that he felt unable to answer. A medical specialty declined to participate on the basis that the consultants deemed it unnecessary as they felt they already knew when to offer HIV testing appropriately and did not want to upset their patient cohort. In another medical specialty, the questionnaire return boxes repeatedly disappeared from a prominent position in the waiting room. In yet another discipline, a receptionist stated that he was not willing to hand out the questionnaire as he did not agree with the purpose of the project.

The response rates between the participating primary care practices were very variable with some practices contributing as many as 187 completed questionnaires while others returned 10 or less. This may be due to issues similar to those that arose in the hospital setting or to other reasons, such as apathy towards participation. However, our experiences suggest that, while the patients who responded to this survey were very receptive to the increased offer of HIV testing, the same may not be true of health-care colleagues. More work is needed to explore health professionals' attitudes to increasing HIV testing in order to identify potential barriers. This may be a particular issue in areas of relatively low HIV prevalence, where substantial education of health-care workers and support staff across a number of disciplines may be required if we are to achieve the recommendations of the UK National Guidelines for HIV testing.