Abstract
Sir: The use of HIV point-of-contact-tests (POCTs) is increasing within clinics and outreach settings to overcome some of the traditional barriers to testing. The increasing use of POCTs, and the expansion of their use in ‘low-risk’ populations, means that a greater proportion of reactive tests are likely to be false-positive results. In our service, the INSTI™ (Biolytical Labs, Inc., Richmond, BC, Canada) HIV POCT test has been used routinely in high-risk populations since October 2007. The sensitivity and specificity of this POCT, as per the manufacturer's data, is 99.6% and 99.3% respectively (INSTI HIV-1/HIV-2 antibody test. Biolytical laboratories product insert). In a high HIV prevalence population such as gay men (5%), the calculated positive predictive value of the POCT is 88%. If used in low-risk populations, such as the general population (0.2%), the positive predictive value would be expected to fall to 22% (i.e. nearly four in five reactive results would be expected to be false reactive). If HIV-POCT testing was offered to all individuals requesting an HIV test, there is concern that this could result in an unacceptable proportion of false-reactive results. There would be a clear advantage if ‘false reactives’ could be identified during the initial clinic visit.
We looked at all patients with a reactive INSTI™ POCT to see if performing a Determine™ (Inverness Medical, Priory Business Park, Bedford, UK) POCT on these patients could identify the false-reactive tests. Between September 2007 and February 2008 sequential high-risk patients (defined as unprotected anal intercourse between men, sharing needles, unprotected vaginal intercourse with a partner from a region with HIV prevalence or oral intercourse with an HIV positive partner) were tested with an INSTI™ HIV POCT and a parallel fourth generation laboratory test (BIORAD) (Bio-Rad Laboratories Ltd, Hemel Hempstead, UK). Paired stored serum samples of those patients with a reactive INSTI test were then retrospectively tested blindly with a Determine™ HIV POCT. One hundred and eighteen patients tested reactive with the INSTI™ POCT. Of these 91 stored serum samples were available, 12 of these were false-reactive results (as identified by the initial BIORAD test). When tested with the Determine™ POCT all 12 were non reactive. Of the 79 true reactive INSTI tests, all were reactive when tested with the Determine™ POCT.
As a result of this study our clinic policy changed. All patients requesting an instant HIV test are tested using the INSTI™ POCT; all individuals with a reactive result have a laboratory HIV test sent (4th generation HIV test-BIORAD) and also have a Determine™ POCT performed. If both POCTs are reactive, patients are told that it is highly likely that this represents a true HIV-positive result; however if the Determine™ test is negative patients are informed it is likely that the INSTI™ result is a false reactive one. With either circumstance patients are made aware of the laboratory result being the definitive result. Since this policy was introduced, the second HIV POCT has correctly predicted the final laboratory result in 100% of cases. Allowing false-reactive tests to be identified at ‘point of contact’ makes testing low-risk populations using HIV POCT more practical by reducing ‘false-positive’ associated patient anxiety.