Azathioprine and Prednisone in Polymyositis

Excerpt

To the editor: A recent paper in your journal (1) illustrates a common error in the use of probability methods. The authors conclude, on the basis of a test group and a placebo control group of eight patients each, that "the findings clearly indicate that azothioprine does not afford any therapeutic advantage when used in addition to accepted prednisone dosages in the initial management of polymyositis." Presumably, this conclusion was reached because the probability value associated with the statistical test used was greater than the traditional (although not divinely prescribed) 5%.

The main difficulty here is that no power values