A Single Index of Mortality and Morbidity

Abstract

A CONTINUING interest of the National Center for Health Statistics is the development and evaluation of new health indices suited to diverse specific purposes. No one index can reflect all aspects of health, but there is considerable agreement that an index which measures some aspects of nonfatal illness as well as mortality would be desirable. A rationale for using both mortality and disability rates as the components of such an index has already been published (1). One technique for combining mortality and morbidity rates into a single index was devised and reported by Chiang in conjunction with his development of mathematical models of illness frequency, illness duration, and mortality (2). Moriyama has discussed criteria desired in an index of health and, in view of these, reviewed some approaches proposed in the literature (3). A deseription and evaluation of disability concepts and measures being considered as the basis of the morbidity component of a mortality-morbidity index appeared in a recent report (4). Another technique for merging death rates with illness rates, and some illustrative results are described in this paper. A primary objective of these studies is development of a summary measure which reflects changes over time in the health status of the nation's population. Too little is known as yet about these techniques, and in some cases about the data they employ, to permit thorough evaluation of alternative approaches to the construction of such indices. Results of studies of such measures are presented as they become available by the Center to stimulate consideration of the issues and, possibly, to stimulate further studies by those in a position to conduct related research. Some preliminary index values based upon the techniques presented in this paper have already been published for fiscal years 1958-66 (5). The estimates in this article are also preliminary. Although they relate to only a single year, they provide previously unpublished information on whites and other persons and on sex differences. These estimates are considered more accurate than earlier computations of such values.