The Current Status of Death Certificate Written in an Academic Hospital and the Degree of Agreement in Interpretation: A Single Center Observational Study

Baek, Daehyun; Cho, Hanjin; Moon, Sungwoo; Park, Jonghak; Song, Juhyun; Kim, Jooyoung; Jeon, Seoungho; Ahn, Eusang

Abstract

PURPOSE
This study aims to review the appropriateness of the issued death certificates and autopsy reports and to evaluate the improvement points of these documents in accordance with the guidelines of the Korean Medical Association and the National Statistical Office. Moreover, this study also examines why the guideline is necessary for the credibility of these documents.
METHODS
The death certificates and autopsy reports written by a training hospital were analyzed for a 12-month period, between December 2014 and November 2015. The reference to analysis was the "guidelines to medical certificate 2015" written by the Korean Medical Association, "World Health Organization (WHO) death certificate principle", and "guideline leaflet," as provided by the National Statistical Office. Two researchers analyzed the documents that were against the guidelines, and suggested improvement points. The analyzed variables were age, sex, issued date, direct cause of death, manner of death, location of death, and types of accident. The primary goal was to see the rate of issued documents written correctly according to the guidelines and to suggest possible improvement points. The secondary goal was to analyze the reason for accordance and discordance between researchers.
RESULTS
There were a total of 603 death certificates and autopsy reports issued during the research period; 562 (93.2%) and 41 (6.8%) cases, respectively. As for the manner of death, 521 cases were "death from disease," 64 were "external causes," and 18 were "others or unknown" (86.4%, 10.6%, and 3.0%, respectively). As for the issued department, internal medicine and emergency medicine issued 301 (49.9%) and 126 (20.9%) documents, respectively. Of these, 139 (23.1%) cases were regarded to be in accordance with the guidelines, while 304 (50.4%) were considered to be discordant cases. Among the discordant cases, there were 177 (29.4%) cases that were the mode of death directly written to cause of death. As for the records of "period of occurrence to death" were recorded only 70 (11.7%) cases (including "unknown" 65 cases) and the others were blank. The Kappa number of analysis regarding the evaluation correspondence of the two researchers was 0.44 (95% confidence interval, 0.38 to 0.51).
CONCLUSION
The most frequent error was "˜the condition of death to direct cause of death' with the ratio of 29.4%. This may have been because the rate of concordance between the researchers based on the guidelines was not high enough. There is a need to provide specific guidelines for each case, and also promote and educate regarding significant errors.