Assessing death data quality through skepticism about death data accuracy

Research Square (Research Square)(2022)

引用 0|浏览0
暂无评分
摘要
Abstract Background Death is a crucial measure in electronic medical record (EMR) studies, where it has significance as a criterion for analyzing mortality in the database. This study aimed to assess extracted death data quality and investigate the potential of a final administered medication variable as an indicator to quantify the accuracy of a newly extracted control group’s death data. Methods Data were collected through Asan Biomedical Research Environment, which comprised data from both the Asan Medical Center and The Korean Central Cancer Registry. The gold standard was established by examining differences according to death information sources through a chart review. Cosine similarity was employed to quantify the final administered medication similarities between the gold standard and other cohorts using the Anatomical Therapeutic Chemical classification system code. Results The gold standard was determined as patients who died in hospital after 2006, when the final hospital visit/discharge date and death date differed by 0 or 1. For all three criteria, a) Seer Stage, b) cancer type c) type of final visit; as the mortality rate increased, the final administered medication cosine similarity with the golden standard increased. Conclusion This study introduced an indicator that can provide additional accurate death information and differentiate reliability. In the future, variables other than EMR could be used to further determine death information quality, in addition to the final administered medication.
更多
查看译文
关键词
death data quality,death data accuracy
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要