Related Experiment Videos
Enhancing medical data quality through data curation: a case study in primary Sjögren's syndrome
Vasileios C Pezoulas1, Konstantina D Kourou2, Fanis Kalatzis1
1Unit of Medical Technology and Intelligent Information Systems, Department of Materials Science and Engineering, University of Ioannina, Greece.
Clinical and Experimental Rheumatology
|July 10, 2019
Summary
This study introduces an automated method to improve clinical data quality by detecting errors and standardizing terms. The approach enhances data accuracy and relevance for better research outcomes.
Area of Science:
- Clinical Informatics
- Data Science
- Medical Data Quality
Background:
- Assessing clinical data quality is crucial for reliable research.
- Manual data curation is time-consuming and prone to errors.
- Automated methods are needed to efficiently evaluate data accuracy, relevance, conformity, and completeness.
Purpose of the Study:
- To develop and apply an automated method for assessing clinical data quality.
- To automatically detect problematic data fields and match clinical terms within a specific domain.
- To improve the accuracy, relevance, conformity, and completeness of clinical datasets.
Main Methods:
- Automated construction of diagnostic reports summarizing data characteristics (types, ranges).
- Identification of outliers, inconsistencies, and missing values within datasets.
- Utilizing a reference model of domain-specific knowledge for term matching and data standardization.
Main Results:
- A case study involving 250 primary Sjögren's syndrome (pSS) patients demonstrated reliable outcomes.
- The method identified 28 features with issues like outliers and unknown data types.
- Data standardization successfully matched 89.41% of pSS-related terms against a clinician-defined reference model.
Conclusions:
- The data curation method significantly improves dataset quality.
- Automated identification of outliers, missing values, and inconsistencies enhances data reliability.
- Automated detection and standardization of pSS-related terms support clinical research and data harmonization.