Background Our outreach HIV Point of care testing (POCT) programme changed from 4^th generation testing to 3^rd generation POCT kits in August 2014, which led to a significantly quicker turnaround time for results and greater convenience for both outreach staff and patients. We continued to confirm all POCT serology by conventional laboratory testing.

Aims To compare 3^rd and 4^th generation POCT in clinical practice and review the need for laboratory confirmation of all samples.

Methods The INSTI™ HIV-1/HIV-2 Antibody Test was used for POCT testing at a city centre outreach service from August 2014 until July 2015. All samples were also tested in parallel, in real-time, by standard laboratory tests for HIV. Results were compared retrospectively.

Results POCT was provided for 399 patients. 31 patients were excluded. Of the remaining 368 patients, there were 6 true positive results (1.6%) and no false-negatives or false-positives. By contrast, our previous evaluation of Alere Determine™ 4^th generation testing, with a sample size of 367, found 3 true positives (0.8%); 2 false positives (0.6%); and 3 false negatives (0.8%), leading to negative predictive value 99.2%; positive predictive value 60%; sensitivity 50%; specificity 99.4%. This was a significant underperformance in clinical practice compared with advertised values.

Discussion INSTI™ is outperforming Alere Determine™ in our local experience. We intend to continue using 3^rd generation POCT in our outreach programme. Given INSTI™’s performance, the question now raised is can we consider moving away from carrying out backup serology in all cases?