[Skip to Navigation]
March 1963

Myocardial Infarction Prognosis by Discriminant Analysis

Author Affiliations


Clinical Assistant in Medicine (Dr. Hughes); Clinical Assistant in Medicine (Dr. Kalbfleisch); Instructor of Medical Biomathematics, Resident in Medicine, and Director, Biostatistical Unit and Medical Research Computer Center (Dr. Brandt); Research Associate in Medicine and Trainee in Biostatistics (Mr. Costiloe).; Currently on active duty, U.S. Public Health Service (Dr. Kalbfleisch).; Department of Internal Medicine and Biostatistical Unit and Medical Research Computer Center in the Department of Preventive Medicine and Public Health, University of Oklahoma Medical Center.

Arch Intern Med. 1963;111(3):338-345. doi:10.1001/archinte.1963.03620270064010

Several large series of patients with acute myocardial infarction have been analyzed for factors which seemed important to prognosis. However, conclusions from these reports have been varied and at times conflicting. These results prompted us to study the hospital records of a group of patients with myocardial infarction by linear discriminant analysis utilizing high-speed computer equipment, seeking a key to the riddle of prognosis. This report presents the analysis of the study. It is a report on data accumulated from past clinical records. Only the method of analysis is the result of a carefully designed statistical experiment.

Material and Methods  Records of all patients with a diagnosis of acute myocardial infarction admitted between January, 1953, and January, 1960, to the University of Oklahoma Medical Center (including University Hospital, Veterans Administration Hospital, and Wesley Hospital) were reviewed. University Hospital is a state-owned teaching hospital which serves the State of Oklahoma on

Add or change institution