MULTI-SITE EVALUATION OF A DATA QUALITY TOOL FOR BIG DATA IN HEALTHCARE

Evaluation of data quality in large healthcare datasets.

abstract:

Data quality and fitness for analysis are crucial if outputs of big data analyses should be trusted by the public and the research community. Here we analyze the output from a data quality tool called Achilles Heel as it was applied to 24 datasets across seven different organizations. We highlight 12 data quality rules that identified issues in at least 10 of the 24 datasets and provide a full set of 71 rules identified in at least one dataset. Achilles Heel is developed by Observational Health Data Sciences and Informatics (OHDSI) community and is a freely available software that provides a useful starter set of data quality rules. Our analysis represents the first data quality comparison of multiple datasets across several countries in America, Europe and Asia.

大型医疗数据集的数据质量评估

摘要：
若要让大数据分析的成果获得公众与科研界的认可与信任，数据质量与分析适配性便至关重要。本研究针对一款名为Achilles Heel的数据质量工具展开分析，该工具已被应用于7家不同机构的24个数据集，并对其输出结果进行了系统研究。本研究筛选出12条数据质量规则，这些规则可在至少10个数据集内检测出数据质量问题；同时汇总了至少在1个数据集内被检出问题的全部71条数据质量规则。该工具由观察性医疗数据科学与信息学（Observational Health Data Sciences and Informatics, OHDSI）社区开发，是一款免费可用的软件，可提供一套实用的基础数据质量规则集。本研究是首个针对美洲、欧洲及亚洲多个国家的多数据集开展的数据质量对比分析。