Study to evaluate the performance of a point-of-care whole blood HIV viral load test (SAMBA II HIV-1 Semi-Q Whole Blood).

ABSTRACT

Remote areas of many low and middle income (LMI) countries have poor access to HIV viral load (HIV VL) testing. The SAMBA-II (Simple Amplification-based Assay) Semi-Q Whole Blood Test (Diagnostics for the Real World (DRW), Cambridge, UK) is a point of care assay which uses leucodepletion technology to allow whole blood testing in remote settings. 1540 consecutive HIV-positive clinic patients in Cameroon (250), UK (633), Ukraine (412) and Zimbabwe (245) donated venous blood (all countries) and finger-prick blood (all except UK) for testing on SAMBA-II. SAMBA II results were compared with simultaneous plasma results on the Abbott RealTime HIV-1 (Abbott Molecular, Des Plaines, IL) viral load assay and interpreted as either <1000 RNA copies/ml or ≥1000 RNA copies/ml. For 1528 venous whole-blood samples tested on SAMBA II, overall percent agreement with the reference test at a cut-off of HIV VL ≥1000 cps/ml was 96.9% (1480/1528 95% CI 95.9-97.7), negative percent agreement 97.7% (1259/1289 95% CI 96.7-98.4), positive percent agreement 92.5% (221/239 95% CI 88.4-95.5). For 854 finger-prick samples there was 95.0% (811/854 95% CI 93.3-96.3) overall percent agreement; negative percent agreement 98.0% (625/638, 95% CI 96.5-98.9); positive percent agreement 86.1% (186/216 95% CI 80.8-90.4). These rose to 93.5% (82.1, 98.6), 97.6% (95.6, 98.8) and 95.6% (93.3, 97.3) after exclusion of aberrant results from the Ukraine centre. These results show a high level of agreement between SAMBA-II and a laboratory-based assay. SAMBA-II has a performance that is suitable to use as a VL point of care assay in remote settings