Validation of death certificate diagnosis of out-of-hospital coronary heart disease deaths in Olmsted County, Minnesota

Citation
Ty. Goraya et al., Validation of death certificate diagnosis of out-of-hospital coronary heart disease deaths in Olmsted County, Minnesota, MAYO CLIN P, 75(7), 2000, pp. 681-687
Citations number
29
Categorie Soggetti
General & Internal Medicine","Medical Research General Topics
Journal title
MAYO CLINIC PROCEEDINGS
ISSN journal
00256196 → ACNP
Volume
75
Issue
7
Year of publication
2000
Pages
681 - 687
Database
ISI
SICI code
0025-6196(200007)75:7<681:VODCDO>2.0.ZU;2-5
Abstract
Objective: To evaluate the validity of death certificate diagnosis of out-o f-hospital (OOH) coronary heart disease (CHD) and sudden cardiac death (SCD ) in Olmsted County, Minnesota, between 1981 and 1994. Methods: In this review of the medical records, autopsy reports, and corone r's files, OOH deaths with heart disease as the underlying cause of death o n the death certificate were classified into CHD (International Classificat ion of Diseases, 9th Revision, Clinical Modification [ICD-9-CM] codes 410-4 14) and non-CHD (other ICD-9-CM heart disease codes) deaths. A 10% random s ample (n=174) of these death certificates was reviewed by physicians, and p ublished validation criteria were applied to classify these deaths into val idated CHD or non-CHD categories. Sudden cardiac death was defined as valid ated CHI) that occurred at an OOH location with less than 24 hours between symptom onset and death. Results: The death certificate definition of OOH CHD death (ICD-9-CM codes 410-414) had high sensitivity and positive predictive value of 91% and 96%, respectively. The specificity and the negative predictive value were sligh tly lower at 86% and 72%, respectively. The sensitivity of death certificat e diagnosis of CHD for validated SCD was 89%, and the positive predictive v alue was 77%, Using a more restrictive definition of SCD, that is, less tha n 1 hour between the onset of symptoms and death, the positive predictive v alue of CHD codes for SCD was lower at 52%, Conclusions: In Olmsted County, the positive predictive values of death cer tificate diagnosis for OOH CHD and SCD are high. Relying on death certifica te diagnoses results in about 5% underestimation of the true CHD rates, whe reas their use as a surrogate for SCD yields a 16% overestimation of the tr ue SCD rates.