Variability in leukocyte subset measurements among five laboratories in Vancouver

Cell counts and percentages of CD4 are widely used in the prognostic and clinical management of HIV-infected patients, and as surrogate outcomes in clinical trials involving HIV-infected individuals. Considerable variability in CD4 counts has been documented due to physiologic and methodologic factors. While studies of variability of CD4 counts among American and French laboratories have been reported in the literature, no published data are available for Canadian laboratories. This paper describes the results of a study to determine the variability of leukocyte subsets among 5 laboratories in Vancouver, British Columbia. Samples were collected in a prospective fashion from 52 HIV-negative patients from July 1991 to November 1993. Coefficients of variation (CV) were calculated for leukocyte subset percentages and absolute cell counts among laboratories. Our results demonstrate that the variability in leukocyte subsets among 5 Vancouver laboratories was lower than or comparable with published findings. The variability remained stable over the time period of the study, although 4 of the 5 laboratories participated in quality assurance programs. This suggests a plateau in the impact of this program. Since the variability among laboratories is less than the variability attributable to physiologic factors, further research efforts to reduce this variability should focus on physiologic sources.