Evaluation of PIMA analyzer detecting CD4 cell count of venous and capillary blood in HIV-infected individuals

Abstract

This study is aimed at evaluating the utility of the portable CD4 analyzers (PIMA). The paired finger prick blood (25 µl) and 5 ml venous blood samples were collected from 196 HIV infected patients, who came to Yunnan CDC voluntary counseling and testing (VCT) clinic for CD4 test services, from May to August, 2012. The absolute CD4 cell counts were measured by PIMA (using venous and finger-prick blood) and by Calibur (using venous blood) as the reference. The PIMA and Calibur CD4 results were compared using the Wilcoxon matched-pairs test, and the Spearman's rank correlation coefficients were estimated. The Bland-Altman plots were used to assess the consistency of the two methods. The median absolute CD4 counts of 196 venous blood samples obtained by PIMA and by Calibur were 268 (range:169-403) cells/µl and 302 (range:181-474) cells/µl respectively, which showed significant difference (Z = -7.31, P < 0.01). The median absolute CD4 counts measured by PIMA and by Calibur (using 188 finger-prick and venous blood samples respectively) were 271 (range: 165-450) cells/µl and 304 (range:188-476) cells/µl, which also showed significant difference (Z = -7.60, P < 0.01). The CD4 counts obtained by PIMA CD4 analyzer (using venous and finger-prick blood) showed strong positive correlation with the CD4 counts obtained by the reference method (using venous blood), and the r values were 0.94 and 0.92 respectively (P < 0.01) . The mean biases (limit of agreement) were -38.7 (-210.9-133.5)cells/µl and -45.4 (-221.8-131.0) cells/µl, respectively.Using 350 CD4 counts as the threshold for ART treatment initiation, the sensitivity and specificity of PIMA were 99.1% and 79.3% for venous blood samples, and 97.2%and 78.5% for finger-prick blood samples, respectively. The CD4 counts obtained by PIMA are lower than that obtained by Calibur, while the sensitivity is high.