Evaluation of PIMA(TM) CD4 System for Decentralization of Immunological Monitoring of HIV-Infected Patients in Senegal

BACKGROUND: HIV infection is a concern in the army troupes because of the risk behaviour of the military population. In order to allow regular access to CD4(+) T cell enumeration of military personnel as well as their dependents and civilians living with HIV, the Senegalese Army AIDS program is implementing PIMA(TM) Alere technology in urban and semi-urban military medical centres. Validation such device is therefore required prior their wide implementation. The purpose of this study was to compare CD4(+) T cell count measurements between the PIMA(TM) Alere to the BD FACSCount(TM). METHODOLOGY: We selected a total of 200 subjects including 50 patients with CD4(+) T-cells below 200/mm(3), 50 between 200 and 350/mm(3), 50 between 351 and 500/mm(3), and 50 above 500/mm(3). CD4(+) T-cell count was performed on venous blood using the BD FASCount(TM) as reference method and the PIMA(TM) Point of Care technology. The mean biases and limits of agreement between the PIMA(TM) Alere and BD FACSCount(TM) were assessed with the Bland-Altman analysis, the linear regression performed using the Passing-Bablok regression analysis, and the percent similarity calculated using the Scott method. RESULTS: Our data have shown a mean difference of 22.3 cells/mm(3) [95%CI:9.1–35.5] between the BD FACSCount(TM) and PIMA(TM) Alere CD4 measurements. However, the mean differences of the two methods was not significantly different to zero when CD4(+) T-cell count was below 350/mm(3) (P = 0.76). The Passing-Bablok regression in categorized CD4 counts has also showed concordance correlation coefficient of 0.89 for CD4(+) T cell counts below 350/mm(3) whilst it was 0.5 when CD4 was above 350/mm(3). CONCLUSION: Overall, our data have shown that for low CD4 counts, the results from the PIMA(TM) Alere provided accurate CD4(+) T cell counts with a good agreement compared to the FACSCount(TM).