Abstract
The objective of this study is to audit the implementation and knowledge of the British HIV Association (BHIVA) UK National guidelines for HIV testing (2008) in key conditions at Basildon & Thurrock University Hospital. Basildon Hospital is a district general hospital, serving over 400,000 patients in south-west Essex. A total of 348 patients were assessed through electronic to pathology data and patients' notes to investigate if they had been tested for HIV when diagnosed with the following conditions: tuberculosis (TB), hepatitis B and C, cervical intraepithelial neoplasia grade II/III, lymphoma, anal cancer, seminoma or Castleman's disease. The physicians involved were questioned as to their knowledge of the HIV testing guidelines. Of the 348 patients who were identified as having the above mentioned conditions, only 13.8% of those with any of the key conditions had received an HIV test. Only one non-HIV physician was aware of the guidelines. Knowledge of the 2008 BHIVA HIV testing guidelines is scanty among non-HIV-trained physicians. Health-care professionals in the field, irrespective of their role, should work harder to disseminate information and reduce prejudice that decreases testing of at-risk individuals.
INTRODUCTION
While the availability of combination antiretroviral therapy (cART) has transformed the outcome for individuals with HIV infection, there continues to be significant and avoidable morbidity and mortality relating to HIV infection in the UK. 1 A British HIV Association (BHIVA) national audit showed that of the deaths occurring among HIV-positive adults in the UK in 2006, 24% were directly attributable to the diagnosis of HIV being made too late for effective treatment. 2 Furthermore, it has been shown that many of the ‘late presenters’ have been seen in the recent past by health-care professionals without diagnosis, with conditions that are associated with HIV. Due to the HIV/genitourinary (GU) medicine department residing in the outpatient hospital on another site, there is no permanent HIV health professional present at Basildon Hospital, except for inpatient consultations for HIV inpatients. As a result, apart from coverage in the national medical press and general medicine news sources, there was little dissemination of the 2008 HIV testing guidelines other than on an ad hoc basis, a situation mirrored in many other hospital trusts. We present data from a wide reaching audit and re-audit into HIV testing in key conditions, a practice advocated by BHIVA in its 2008 guidelines, which states that certain ‘key’ conditions should automatically trigger the offer of an HIV test. 3
METHODS
Sample
Data were collected retrospectively from 1 August 2008 to 31 July 2009. This period of time was chosen as it coincided with the new intake of junior doctors for a two-year period in the hospital as well as the appointment of a new consultant in HIV. The re-audit was carried out from 1 August 2009 to 30 June 2010.
We obtained the names of all patients diagnosed with the following conditions (between 1 August 2008 to 31 July 2009): tuberculosis (TB), hepatitis B, hepatitis C, cervical intraepithelial neoplasia (CIN) grade II or III, lymphoma, anal cancer, seminoma, aspergillosis or Castleman's disease. The choice of these conditions was based on the ability to obtain verifiable data on the numbers of new diagnoses and HIV testing. All positive tuberculosis, hepatitis B surface antigen-positive and hepatitis C antibody-positive patients are maintained in a database in the microbiology and virology departments. These were then compared with internal departmental databases maintained by the respective clinicians. In addition, all positive cellular pathology samples are maintained in a database by the cellular pathology department. Unfortunately, due to a lack of a suitable database, recalcitrant seborrhoeic dermatitis and psoriasis were two conditions that we were not able to audit. A previous audit determined a 98% HIV testing rate in the antenatal population and as such was not re-audited on this occasion.
Names of all patients were obtained from individual departmental electronic records and HIV testing was double checked using the electronic pathology records system and a separate database of HIV testing for the given period. For any patients where HIV testing histories were unclear, those patients' notes were then reviewed and the responsible clinician contacted. All physicians and clinical nurse specialists (CNSs) involved in the management of the key conditions were asked individually during the collection of data as to whether they were aware of the existence of and the details within the BHIVA 2008 HIV testing guidelines.
Due to the remit of the audit and the data sources used, the percentage of HIV-positive patients resulting from the testing carried out was not determinable.
RESULTS
A total of 348 patients were assessed in the initial audit, and in the re-audit 557 patients were assessed (see Table 1).
HIV testing rates per key condition for initial audit of 1 August 2008 to 31 July 2009 (2008–2009) and re-audit of 1 August 2009 to 20 June 2010 (2009–2010)
GU = genitourinary; CIN = cervical intraepithelial neoplasia
Tuberculosis
HIV tests were offered during the second or third consultation, and, if offered, were only offered by the CNS in TB. No reason was given as to why the remaining six patients had not been tested. In the re-audit, the rates were much lower due to risk stratification of age and the severity of illness precluding consent.
Hepatitis B and C
The CNS for hepatology co-managed the hepatitis C but not the hepatitis B patients. Barriers to testing from the physicians in hepatology included systemic responses, i.e. ‘The results are not available on the intranet pathology system hence it is difficult to determine if tests have been previously requested’, and personal views regarding mode of infection, ‘if a 20-something has hepatitis B with no risk factors it is going to be due to vertical transmission and hence a HIV test is not needed’. The CNS for hepatology considered testing primarily according to risk factors. None of the physicians or nurses were aware of the 2008 guidelines. In the re-audit, the rates were similar for hepatitis B but lower for hepatitis C.
Lymphoma
Very few patients were tested for HIV. Physicians primarily managed the patients. Only one of four physicians was aware of the guidelines and admitted that they were failing to achieve the standard. Other physicians had more personal barriers to testing: ‘I only test them if they are really camp’ and ‘yet another thing we have to do’. In the re-audit the numbers were similar.
CIN II/III and anal cancer
Again, very few patients had been tested and no reason was provided as to why they were not tested, neither from the physicians nor the notes. No physician or CNS was aware of the guidelines. In the re-audit the data were similar.
The data from the first audit were disseminated via an Audit Presentation Day and a Medical Grand Round presentation and follow-up education meetings for doctors and nurses of all grades were held. In addition, active input from the microbiology and virology department was provided regarding patients who should be offered testing based on positive results for key conditions. In addition, the pathology department is currently considering releasing HIV test results on the computer system. During meetings disseminating this information, the majority of hospital physicians still thought that GU medicine physicians and CNSs were required to give extensive counselling prior to HIV testing.
In the re-audit, the data were disappointing as the education and promotion of HIV testing had not appeared to effect change in testing practice (see Table 1). While the overall absolute numbers of HIV tests had increased in the hospital (due in part to the absolute increased numbers of antenatal testing); the increase was not reflected in the key conditions assessed in the re-audit.
CONCLUSIONS
Non-HIV and infectious diseases-trained physicians are still not aware of the 2008 UK HIV testing guidelines and still think that testing requires extensive pretest counselling by HIV-trained physicians or CNSs. The success of the guidelines in expanding HIV testing and reducing the numbers of late-presenters will depend on the engagement of other specialties and disciplines.
Our audit results were disappointing, reflecting the difficulties faced in changing general attitudes and perceptions of HIV testing. While HIV and GU medicine services have stopped insisting on extensive counselling prior to HIV testing for low-risk patients to aide normalization of the test in line with other conditions, this has not translated to the wider medical community. In particular, our audit showed that even with knowledge of the audit and an awareness of the need for testing, this does not necessarily translate into higher testing rates, and in our audit actually led to lower testing rates. Tradition and longstanding perceptions may be a difficult combination to overcome.
The specific reasons for the lower testing rates in our hospital were not readily discernible, further hampering future endeavours to determine the best methods to increase testing. However, one of the barriers identified in the re-audit was geography; Basildon is an Essex-based town, with ready access to London in the south-east of England, but one of the views is that HIV is mainly a problem ‘of London’ and not one really affecting Basildon. We suspect that this view is commonly held in non-major city-based hospitals all over the country.
Despite all the methods already identified and implemented, including grand round and audit meeting presentations, personal dissemination to individuals and departments, greater input by microbiology and virology departments and altering access to pathology information, we observed a drop in HIV testing. Our audit demonstrated that not only are we faced with a great challenge that might not be surmounted quickly, but also that a greater concerted effort with innovative ideas is needed if we are to identify those patients with HIV early enough to initiate life-saving treatment. We hope our audit will encourage colleagues to develop new ideas.
Footnotes
ACKNOWLEDGEMENTS
The authors would like to especially thank Dr Sheena Sikka and Dr John Williamson for the collection of data. The authors would also like to thank Maggy Stephens, Trevor Meacock, Eric Watts, Andy Hare and Sarah Tarff for their invaluable help in obtaining information.