Routine HIV testing in the emergency department: tough lessons in sustainability.

OBJECTIVES: Routine HIV testing in nonspecialist settings has been shown to be acceptable to patients and staff in pilot studies. The question of how to embed routine HIV testing, and make it sustainable, remains to be answered. METHODS: We established a service of routine HIV testing in an emergency department (ED) in London, delivered by ED staff as part of routine clinical care. All patients aged 16 to 65 years were offered an HIV test (latterly the upper age limit was removed). Meetings were held weekly and two outcome measures examined: test offer rate (coverage) and test uptake. Sustainability methodology (process mapping; plan-do-study-act (PDSA) cycles) was applied to maximize these outcome measures. RESULTS: Over 30 months, 44,582 eligible patients attended the ED. The mean proportion offered an HIV test was 14%, varying from 6% to 54% per month over the testing period. The mean proportion accepting a test was 63% (range 33-100%). A total of 4327 HIV tests have been performed. Thirteen patients have been diagnosed with HIV infection (0.30%). PDSA cycles having the most positive and sustained effects on the outcome measures include the expansion to offer blood-based HIV tests in addition to the original oral fluid tests, and the engagement of ED nursing staff in the programme. CONCLUSIONS: HIV testing can be delivered in the ED, but constant innovation and attention have been required to maintain it over 30 months. Patient uptake remains high, suggesting acceptability, but time will be required before true embedding in routine clinical practice is achieved.

Automatic oral fluid-based HIV testing in HIV screening programmes: automatic for the people.

OBJECTIVES: UK guidelines recommend routine HIV testing in general clinical settings when the local HIV prevalence is > 0.2%. During pilot programmes evaluating the guidelines, we used laboratory-based testing of oral fluid from patients accepting tests. Samples (n = 3721) were tested manually using the Bio-Rad Genscreen Ultra HIV Ag-Ab test (Bio-Rad Laboratories Ltd, Hemel Hempstead, UK). This was a methodologically robust method, but handling of samples was labour intensive. We performed a validation study to ascertain whether automation of oral fluid HIV testing using the fourth-generation HIV test on the Abbott Architect (Abbott Diagnostics, Maidenhead, UK) platform was possible. METHODS: Oral fluid was collected from 143 patients (56 known HIV-positive volunteers and 87 others having contemporaneous HIV serological tests) using the Oracol+ device (Malvern Medicals, Worcester, UK). Samples were tested concurrently: manually using the Genscreen Ultra test and automatically on the Abbott Architect. RESULTS: For oral fluid, the level of agreement of results between the platforms was 100%. All results agreed with HIV serology. The use of the Oracol+ device produced high-quality samples. Subsequent field use of the test has shown a specificity of 99.97% after nearly 3000 tests. CONCLUSIONS: Laboratory-based HIV testing of oral fluid requires less training of local staff, with fewer demands on clinical time and space than near-patient testing. It is acceptable to patients. The validation exercise and subsequent clinical experience support automation, with test performance preserved. Automation reduces laboratory workload and speeds up the release of results. Automated oral fluid testing is thus a viable option for large-scale HIV screening programmes.