关键词: 机采血小板, 影响因素, 纵向研究, 线性混合效应模型

Abstract: Objective To explore the effect of the blood donation interval adjustment of plateletpheresis donor released by Whole Blood and Component Donor Selection Requirements (GB18467-2011) on the peripheral platelet count level. Methods The linear mixed effect model in the longitudinal data analysis method was selected to model the plateletpheresis donors who donated plateletpheresis more than 10 times from September 30, 2016 to September 30, 2018, using R 3.5.1 statistical software. Periphery platelet count level and possible confounding factors within 23 plateletpheresis donations were selected for modeling analysis. Mixed effect model with a random intercept (model1) and with both a random intercept and a random slope for interval (model2) was established. Results AIC, BIC and the significance of distributions of the log (restricted) likelihood difference showed that model 2 was superior to model 1. In model 2, the parameter estimates of covariate intercept, periphery platelet baseline count, blood donation interval, collection volume, age, gender, blood donation times and the interaction effect of blood donation interval and blood donation times were 59.708, 0.795, -0.023, 3.798, -0.234, 7.173, 0.385 and -0.004, respectively. All of the above parameters were of statistical significance. Conclusion The current minimum interval rule does not cause a large fluctuation of periphery platelet count for plateletpheresis donors with a few blood donations. Long-term blood donation in a shorter interval may cause a certain fluctuation in periphery platelet count, but the fluctuation magnitude is still small. For frequently repeated blood donors with short intervals, less platelet volume or extended interval properly should be considered to reduce the fluctuation of periphery platelet count.

Key words: Plateletpheresis, Influencing factors, Longitudinal study, Linear mixed-effectives model