Multisite Evaluation of a Data Quality Tool for Patient-Level Clinical Data Sets.

Area of Science:

• Health Informatics
• Data Science
• Observational Health Data Sciences and Informatics (OHDSI)

Background:

• Data quality is essential for reliable analysis of electronic health record (EHR) and administrative claims data.
• Ensuring data integrity is critical for public and research community trust in health data outputs.

Purpose of the Study:

• To introduce and evaluate the Achilles Heel data quality analysis tool.
• To compare the performance of Achilles Heel across diverse healthcare datasets.

Main Methods:

• The Achilles Heel tool, developed by OHDSI, was applied to 24 large healthcare datasets from seven organizations.
• A structured interview was conducted with participating sites to gather qualitative feedback on the tool's utility.

Main Results:

• The study identified 71 data quality rules, with 12 rules flagging issues in at least 10 of the 24 datasets.
• Achilles Heel is a free, extensible software providing a foundational set of data quality rules.
• Qualitative feedback indicated the tool's value in data quality evaluation.

Conclusions:

• This analysis is the first to compare outputs from a data quality tool with a fixed, yet extensible, rule set.
• The use of a common data model facilitated rapid comparison of international datasets.
• The findings underscore the importance of standardized data quality assessment for multi-site research.