The principal investigator is an established researcher responding to topic number 2 of the Program Announcement: Racial and ethnic differences: studies on causes of differences among ethnic and racial groups in susceptibility to age-related disease and dysfunction. National Multiple Cause of Death data provide a wealth of information concerning mortality among the elderly, as well as other age groups, for all major ethnic classifications. These data contain disease patterns present at time of death that permit researchers to analyze the complex nature of these conditions. These data have been subjected to numerous analyses since they were made available in 1968, but researchers have concentrated their efforts on subsets of the data, usually involving a single geographic region, a single disease, or a particular year. There have, to date, been few attempts to combine the annual datasets into one large, inclusive set to subject them to comprehensive analyses. This is probably because of the magnitude and cumbersome nature of the annual sets. The 1988 multiple cause data alone consume six 9-track, 6,250 bpi tapes, containing 2,171,196 records. If copied to disk, these records require 955,326,240 bytes of disk space. Present technology in the form of large capacity fast access disks combined with data reformatting techniques, however, allow us to analyze these combined data. We propose to prepare the National Multiple Cause of Death datasets for seven selected years (1981 and 1986 to 1993) to test the feasibility of conducting future large-scale, comprehensive analyses. These forthcoming analyses will include all of the datasets from 1968 to 1993 and will examine differences in disease-related mortality patterns among the elderly from the major ethic groups within the U.S. population. We will then compare these findings to those involving younger subgroups from the same ethnic populations.
The specific aims of this pilot study are: 1. To reorganize the National Multiple Cause of Death data into one large comprehensive dataset for efficient analysis using a personal computer with moderate power. 2) To make available both the documentation and the reconstituted data to other investigators over the World Wide Web so that this wealth of information can be utilized for the public good. 3) To summarize conditions listed on death certificates using several methods from the literature. 4) To conduct preliminary analyses involving ethnic differences in the patterns of death certification for various age groups, with special emphasis on the elderly. These preliminary analyses are to test the feasibility of the large-scale, comprehensive analyses discussed above. This pilot study will establish the methods to be used to analyze all of the U.S. Multiple Cause of Death Data from 1968 to 1993.
