Some Demographic Characteristics of an Autopsied Population.

THE demographic characteristics of that special type of population that is made up of individuals who died in a hospital and were autopsied are necessarily of great interest to the many investigators whose studies employ postmortem data. Examination of certain of the characteristics of a series of autopsied individuals may aid the pathologist in determining the extent to which he may, with a reasonable degree of firmness, make generalizations beyond his immediate postmortem observations. Research utilizing autopsy data has proven itself to be of great importance, particularly when the condition investigated has crucial aspects that are detectable only at autopsy. When an institution is conducting investigations based upon autopsy findings, in general it would seem desirable, at regular intervals, to analyze available records regarding the demographic characteristics of those patients who come to autopsy. Only limited efforts, however, have been made either in this direction or in attempting to reach an understanding of autopsied populations generally. Almost 40 years ago, PEARL. and BACON described the age, sex, and racial distribution of patients autopsied at Johns Hopkins Hospital during the preceding 35 years. Those autopsied were compared with the living population of Baltimore, Maryland, and with all deaths in Maryland. These comparisons amply demonstrated “the special nature and mode of formation” of the autopsied population [l]. More recently, MCMAHIAN analyzed national mortality data in an effort to describe both the nature of the data as to whether an autopsy was or was not performed, and also selected demographic characteristics of those individuals reported as having been autopsied. He found that autopsy status was not reported on almost 20 per cent of the death certi&ates for whites and almost 30 per cent of those for nonwhites. Because of this lack of information, he concluded that death certificate evidence regarding the characteristics of individuals autopsied in the United States was “not sufficient to justify even tentative conclusions,” though the data did “lend statistical support” to several hypotheses regarding selection by sex, race, and state of residence [2]. WALLACH and BORGAITA