OBJECTIVES: To enumerate CD4 and CD8 T-cells using the simple and cheap immuno-alkaline phosphatase (IA) method and to compare it with flow cytometry (FC); and to study the effects of duration of sample storage on the IA method results.

DESIGN: Method comparison study.

SETTING: Blair Research Laboratory, Harare, Zimbabwe.

SUBJECTS: 41 HIV positive and 11 HIV negative men and women from Harare participating in HIV studies at Blair Research Laboratory, Zimbabwe.

MAIN OUTCOME MEASURES: CD4 and CD8 T-cell counts by FC and the IA method.

RESULTS: The IA method and FC were highly correlated for CD4 counts (Spearman rs = 0.91), CD4 percentage (rs = 0.84), CD8 count (rs = 0.83), CD8 percentage (rs = 0.96) and CD4/CD8 ratio (rs = 0.89). However, CD4 cell counts and percentage measured by the IA method were (mean difference +/- SE) 133 +/- 24 cells/microL [corrected] and 6.7 +/- 1.1% higher than those measured by the FC method (p < 0.0001) respectively. CD8 counts and percentages by the IA method were lower than those by the FC method (p < 0.01). Accordingly, the IA method gave a higher CD4/CD8 ratio (p < 0.01). IA method CD4 counts < 300/mL best predicted FC CD4 counts < 200/mL while IA CD4% < 25 best predicted FC CD4% < 14%. IA method CD4/CD8 ratio < 0.8 best predicted FC CD4/CD8 ratio < 0.5. Smears stored for up to 18 months gave results similar to fresh smears.

CONCLUSION: The IA method correlates well with but gives CD4 counts and percentages that are higher than those determined by FC. On the contrary, the IA method gives CD8 counts and percentage that are lower than FC values. The method is a cheap and reliable alternative to FC and allows storage of samples for extended periods before analysis, making it an appropriate technology for resource poor countries.