Résumé :
|
[BDSP. Notice produite par INIST-CNRS 8zZtxR0x. Diffusion soumise à autorisation]. Objective. Data from death certificates are often used in research ; however, little has been published on the processing of vague or incomplete information reported on certificates. The goal of this study was to examine the querying efforts in the United States used to clarify such records. Methods. The authors obtained data on the querying efforts of the 50 states, New York City, and the District of Columbia. Descriptive statistics are presented for two units of analysis : registration area and death record. Using data from a single registration area, Washington State, the authors compared the percent change in age-adjusted death rates for data from before and after querying to analyze the effect of querying on selected causes of death. Results. Fifty-one of the 52 registration areas queried either demographic or cause-of-death information. Almost 90% of queries were returned ; the underlying cause of death changed in approximately 68% of these records. This data translates into about 3% of total U.S. death records, given that 4% of total U.S. death records were queried about cause of death. The impact of queries on age-adjusted death rates varied by cause of death. Generally, the effect is most obvious for cause-of-death categories that are specific and relatively homogenous. Conclusion. Querying continues to be widely practiced. In the case of cause-of-death queries, this method refines the assigned underlying cause of death for records reported with vague or incomplete information.
|