An optimized strategy for choosing the number of platelet concentrates to pool.

Abstract

The number of platelet concentrates (PCs) pooled per dose varies widely. The number should be based on quality control statistics. To determine the optimal number of PCs per pool that will achieve a target dose. The population statistics of PC cell counts were obtained from routine quality control records. A target value of 3.00x10(11) cells was chosen. A theoretical model to predict the probability of achieving this target cell count was developed. The calculated probability was tested in a Monte Carlo simulation. The mean of a sequential series of 100 PC counts (obtained from routine quality control records) was 9.08x10(10), and the standard deviation was 2.70x10(10). Platelet loss during pooling was measured at 8.3%, so that the pooling of 3.27x10(11) cells is required to deliver a dose of 3.00x10(11) cells. The pooling of 4 PCs from this data set predicts an average pool cell count of 3.63+/-2.30x10(11). Using standard tables for normal distribution, the probability of achieving a minimal pool dose of 3.00x10(11) cells is estimated to be greater than 0.76. Similarly, the probability of achieving the target value by pooling 5 PCs is greater than 0.99. In a Monte Carlo simulation selecting 4 PCs, 76.5% of the trials were successful, and for 5 PCs, 99.5% were successful (estimated P values of .77 and .99, respectively). Five PCs per pool are necessary and sufficient to achieve the chosen target value given the cell count characteristics of our PC supply.