Modification to the Lampariello approach to evaluate reactive oxygen species production by flow cytometry

The aim of this article is to perform a statistical analysis of reactive oxygen species (ROS) cytometric data. It is demonstrated that the classical parametric and nonparametric statistical tests are not suitable to examine these data; the Kolmogorov-Smirnov test and the modification proposed by Lampariello are shown to be too sensitive with respect to the experimental bias (due to procedure or to the instrument) and variability in the ROS production within the repeated samples. Several approaches are examined and discussed. Modifications of the Lampariello's procedure are proposed to include the variability within samples. The validity of the proposed approach is verified by analyzing repeated measurements of ROS formation in cultured human lymphocytes untreated or treated with ferrous sulfate. The proposed approach is successful in considering the "intersample" variability in the ROS data analysis and keeps a good level of validity. Nevertheless, this procedure is not user-friendly and needs to be handled by an expert operator.