Challenges for Data Quality in the Clinical Data Life Cycle: Systematic Review

Background: Electronic health record (EHR) data are anticipated to inform the development of health policy systems across countries and furnish valuable insights for the advancement of health and medical technology. As the current paradigm of clinical research is shifting toward data centricity, the utilization of health care data is increasingly emphasized. Objective: We aimed to review the literature on clinical data quality management and define a process for ensuring the quality management of clinical data, especially in the secondary utilization of data. Methods: A systematic review of PubMed articles from 2010 to October 2023 was conducted. A total of 82,346 articles were retrieved and screened based on the inclusion and exclusion criteria, narrowing the number of articles to 851 after title and abstract review. Articles focusing on clinical data quality management life cycles, assessment methods, and tools were selected. Results: We reviewed 105 papers describing the clinical data quality management process. This process is based on a 4-stage life cycle: planning, construction, operation, and utilization. The most frequently used dimensions were completeness, plausibility, concordance, security, currency, and interoperability. Conclusions: Given the importance of the secondary use of EHR data, standardized quality control methods and automation are necessary. This study proposes a process to standardize data quality management and develop a data quality assessment system.