January 22, 1997

The Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS)

Author Affiliations

University of Frankfurt Medical School Frankfurt, Germany

JAMA. 1997;277(4):297. doi:10.1001/jama.1997.03540280035024

To the Editor.  —The article by Dr Borhani et al1 contains judgments of adverse event frequencies that call for some comments regarding the multiplicity2,3 aspects of testing events (listed in their Table 4). Although the authors mention possible chance occurrences of significances, they suggest that the results are "of concern."Concentrating on the 6 "major vascular events" (ie, stroke, myocardial infarction, sudden death, congestive heart failure, angina pectoris, and other cardiovascular disease deaths) the problem of multiple testing may be circumvented. Although the authors state that only 1 major event was counted per patient (leading to 6 mutually independent groups of patients each with 1 major event), the total number of events is 28 and not 25 (as indicated in their Table 4). Ignoring this deviation as well as the fact that the 6 test results are based on randomized patients, the 6 tests may be considered as essentially independent of