Issues in the Design of Trials of Therapies for Subjects with Human Immunodeficiency Virus Infection that Use Plasma RNA Level as an Outcome

Longitudinal plasma RNA levels were examined in 326 persons infected with human immunodeficiency virus (HIV), who were participating in 4 clinical trials and 1 observational study. Variations in plasma RNA levels among individuals were analyzed, and correlations were made of replicate and longitudinal measurements, plasma RNA levels, CD4 cell counts, and plasma RNA levels at enrollment. Sample sizes were calculated for detection of differences between treatment groups in mean plasma RNA level at one time point, the average change over time, and the area under the curve. On the basis of study findings, study eligibility criteria should specify a minimum plasma RNA level at least as high as the expected size of the treatment effect, and all longitudinal data should be used when determining the effect of treatment on plasma RNA level. Duplicating plasma RNA levels resulted in a decrease in the required sample size of 14%-41%.