Using the CER Hub to ensure data quality in a multi-institution smoking cessation study

Area of Science:

• Health Informatics
• Comparative Effectiveness Research
• Public Health

Background:

• Comparative effectiveness research (CER) across multiple institutions relies on high-quality data from diverse electronic health records (EHRs).
• Ensuring data uniformity from varied EHR systems is crucial for reliable CER findings.
• The CER Hub informatics platform addresses this by developing standardized data quality assurance (QA) processes.

Purpose of the Study:

• To implement and evaluate a QA process for ensuring data uniformity and quality in a multi-institution CER study.
• To assess the effectiveness of the CER Hub's informatics platform and emrAdapter tool in improving EHR data quality.

Main Methods:

• A QA process was developed using the CER Hub informatics platform and its emrAdapter tool.
• The emrAdapter tool, programmed with quality checks, queried primary care encounter records standardized to the CER Hub common data framework.
• The distributed QA process generated error reports for local data correction and aggregate data for central review.

Main Results:

• Data completeness and correctness issues were prevalent initially across six health systems.
• Three iterations of the QA process led to considerable improvements in data quality.
• Incomplete mapping of local EHR data to the common data framework was a common challenge.

Conclusions:

• A highly automated and distributed QA process is effective in ensuring the correctness and completeness of patient care data extracted from EHRs.
• This approach is vital for multi-institution CER studies, particularly in areas like smoking cessation services.
• Standardized data quality measures are essential for robust comparative effectiveness research.