Evaluation of accuracy of the CD4+ T-cell counts using PIMA CD4 as compared to BD Facscaliber flow cytometry

Abstract

Introduction: Pima CD4 (Inverness Medical group 2009), a new method for enumerating CD4+ T-cells is affordable, technically simple, and economical in that it can use either electricity or battery. It is fully automated and thus, useful in remote settings. However, limited information on its performance exists in current literature. Thus, this study aimed at assessing the accuracy of Pima CD4 to BD FASCaliber flow cytometer (Becton Dickinson, Franklin Lakes and NJ) and precision of CD4+ T-cell counting using freshly collected capillary and venous whole blood samples, and establishing the types and frequency of errors incurred when using Pima CD4. This study reported the performance of the relatively new Pima CD4 laboratory test as compared to the more established FASCaliber flow cytometer in determination of the CD4 counts in HIV-infected individuals. The study addressed an important effort in up-scaling rapid, efficient and reliable means of management of HIV in resource limited setting such as in rural Uganda.

Materials and methods Compared results of absolute CD4 counts obtained on replicate samples from 206 HIV-infected individuals (adult men and women) by Pima CD4 with those generated by BD FASCaliber flow cytometry at the Infectious Diseases Institute in Kampala, Uganda.

Results Using venous whole blood, the mean CD4 counts were higher for BD FACSCaliber (422±220 cells/µL) but not significantly different (P=0.1289) from that for Pima CD4 (391±201 cells/µL). Results from the two machines were highly positively correlated (r = 0.96). The mean CD4 counts for BD FASCaliber using venous whole blood (422±220 cells/µL) were significantly higher (P=0.0116 at p=0.05 level) than for Pima CD4 using capillary blood (371±185 cells//µL). Further, the mean CD4 counts enumerated by Pima CD4 for venous whole blood (391±201 cells//µL) were higher than that in capillary blood (371±185 cells//µL) but not significantly different (P=0.3142). Further, within run precision demonstrated that the inherent imprecision of the Pima CD4 instrument is within the manufacturer’s claims and clinically acceptable limits. Also, between run precision demonstrated that the overall inherent imprecision of Pima CD4 instrument and due to other external variants is within manufacturer’s claims and clinically acceptable limits. Using venous whole blood the operator committed 24.3% errors with channel filling and reagent quality control being the most prominent. Likewise, when using capillary blood, the operator committed 13.6% errors with image and reagent quality control being the most prominent.

Conclusions/recommendations Despite the few short comings, Pima CD4 maybe currently one of the suitable instruments for health centers and remote areas with limited access to CD4 testing centers. However, there is still need for proper and thorough training of the operators. Lastly, this study was absolutely done in laboratory environment. Thus, there is a need to evaluate the Pima CD4 in field conditions particularly in remote areas.