Consistency of routine measurements of CD4+, CD8+ peripheral blood lymphocytes.

In order to evaluate the reliability of CD4 and CD8 T lymphocyte counts in large scale studies, a quality control study was performed in 12 French laboratories. CD4 and CD8 counts, assessed by various haematological and immunological techniques, were compared in order to assess possible differences between the laboratories and the techniques used. Our data showed that (a) the consistency of CD4 measurements was satisfactory since the between-laboratory coefficient of variation for absolute CD4 cell numbers above 200/mm3 was around 15% instead of 5-10% for all laboratories but one; (b) the major sources of variability arose from the use of automatic devices in the two-step measurement procedure: immunophenotyping and haematological counting. These data suggest that multicentre assays of CD4 and CD8 counts result in some increase in their variability. Nevertheless the results of large multicentric trials can be extrapolated with confidence in the routine care of HIV+ patients. Together, the results justified the involvement of several experienced laboratories in a clinical trial of HIV-related disease.