To the Editor.—
The clinical value of a laboratory test depends on more than just a statistically significant difference between two test groups. Grimm et al1 may have overemphasized the prognostic importance of the white blood cell count for coronary heart disease. The unwary may be misled into believing that a blood cell count can actually identify persons at risk.When the test groups are large, even minor differences may be statistically significant. For example, in one study of about 1,300 healthy subjects,2 there was a small but highly significant sex difference in the platelet count (P<.0001). Even so, there are better ways than a platelet count to separate the girls from the boys!