Statistical process monitoring of WBC-reduced blood components assessed bytwo types of software

BACKGROUND: Statistical process control is required for monitoring of the W BC-reduction process. This study focused on some factors that may influence the outcomes of statistical process monitoring, such as WBC-reduction tech nologies, the anticoagulant used, and WBC-counting technologies, by using t wo types of software. STUDY DESIGN AND METHODS: Data were collected from January to September 199 9, before the implementation of universal WBC reduction. The effects of thr ee major factors were investigated: methods of preparation, the addition of EDTA to the sample, and the WBC-counting technologies used (microvolume fl uorimetry, flow cytometry, and Nageotte chamber). The WBC-reduction process capability was assessed by two types of software, EZQC (Gambro BCT) and NW A (Northwest Analytical). In addition, the differences between various sets of results were compared by the t test or ANOVA. RESULTS: There was no statistical difference (at the 0.05 level of signific ance) in WBC content when the three types of platelets in citrate samples w ere compared with EDTA samples. In general, the Nageotte chamber appeared t o count the lowest, and microvolume fluorimetry appeared to count lower tha n flow cytometry There were minor but significant methodologic differences between the software packages. However, these differences had negligible ef fects on the percentage of conforming components at both <1 x 10(6) and <5 x 10(6) WBCs per unit. CONCLUSION: Only the counting technologies were sufficiently different to w arrant consideration. This difference may make unacceptable the interchange of results obtained from various counting methods.