A novel approach to the assessment of variations in the human platelet count.

This is the first report of a method to assess the significance of numerical changes in the platelet count based upon a result exceeding the normal intra-individual variation in platelet numbers. Serial platelet counts from 3,789 subjects were analysed to determine the intra-individual variation in platelet numbers. A platelet count difference of 98 x 10(9)/L in males was found to represent a change that would occur by chance in less than 1 in 1,000 platelet count determinations. Tables to determine the significance of platelet number variations, given N previous observations, are provided at two probability levels. The repeatability of the platelet count was calculated as 0.871 (males) and 0.849 (females) indicating that the heritability of platelet count is high and that the platelet count is predominantly genetically determined. A seasonal variation in platelet count was found with a 'winter' versus 'summer' difference of 5.10 X 10(9)/L (males) and 5.82 x 10(9)/L (females).