A test of the National Death Index using the Coronary Artery Surgery Study (CASS)

Abstract

The National Death Index (NDI) is a central, computerized index of death record information compiled from records submitted by the Vital Statistics Office of each state to the National Center for Health Statistics. A selection of 1180 cases, including 370 decedents, from the Coronary Artery Surgery Study was sent to the NDI to test the reliability of the index in identifying decedents. Five possible matching rules are considered: an NDI match with soundex included, an NDI match without soundex matches, an NDI match on social security number, an NDI match on personal data (first name, last name, sex, race, and month, day, and year of birth), and an NDI match on social security number or personal data. In determining death, the greatest positive predictive value was given by a match on social security number, while the greatest negative predictive value was given by an NDI match with soundex included. Further comparisons of sensitivity and specificity, and figures that relate predictive values to the population mortality rate, are given to assist potential users in determining an optimal matching rule.