J Korean Soc Emerg Med.
2007 Jun;18(3):256-262.
The Study of Accuracy of Death Statistics
- Affiliations
-
- 1Department of Emergency Medicine, Hanyang University College of Medicine, Seoul, Korea. ardoc@medimail.co.kr
Abstract
-
PURPOSE: In order to examine the accuracy of death statistics, the present study compared the underlying causes of death on death certificates with the underlying causes of death published by the National Statistical Office.
METHODS
A retrospective survey was performed comparing death certificates issued by a university hospital for one year in 2003, the dead patients' medical records, and death statistics for 2003 published by the National Statistical Office. We compared the underlying causes of death on death certificates, the underlying cause of death as classified by the National Statistical Office, and the underlying causes of death in medical records, in order to analyze their degree of coincidence and the causes of any inconsiste
RESULTS
The inconsistency rate between the underlying causes of death as listed on death certificates and the underlying causes of death in the published statistics was 26%. The most frequent reason for discrepancies was a switch from one diagnosis name to another (58.7%), and the next most frequent was a change from the general categories of death due to old age or unknown cause to a diagnosis name (41.3%). The inconsistency rate between the actual underlying causes of death and the underlying causes of death on death certificates was 18.9%, with the most frequent reason for inconsistency being the recording of an uncertain cause of death such as old age or unknown cause (53.3%), and next most frequent being the recording of an interim result as the underlying cause of death (38.7%). The inconsistency rate between the underlying causes of death in medical records and the underlying causes of death in statistics was 8.1%: the most frequent reason for inconsistency was a change to a diagnosis name irrelevant to the patient's underlying disease (60.9%). and next most frequent was a change of the patients' underlying disease to one that was not related to the patient's death (34.4%). The proportion of cases with concordance between the underlying cause of death on the death certificate and the actual underlying cause of death, but with discrepancy between the actual underlying cause of death and the underlying cause of death in statistics, was 6.2%. Inconsistency both between the underlying cause of death listed on the death certificate and the actual underlying cause of death and between actual underlying cause of death and the underlying cause of death published in statistics occurred in 11.3% of cases, marking a significant difference.
CONCLUSION
The overall accuracy rate of statistics on the causes of death was 91.9%, and the concordance rate between the actual underlying causes of death and the underlying causes of death in published statistics was high whenever death certificates were issued listing the actual underlying causes of death.