Pharmaceutical Feedback Loop - A Concept to Improve Prescription Safety and Data Quality

Area of Science:

• Health Informatics
• Data Quality Management
• Pharmacovigilance

Background:

• Real-world data (RWD) quality is critical for scientific research and clinical decision-making.
• Hospital Information Systems (HIS) generate valuable RWD but often contain unstructured data.
• Ensuring data quality in HIS is a significant challenge for accurate analysis.

Purpose of the Study:

• To develop and evaluate a feedback loop framework for improving drug prescription data quality in HIS.
• To demonstrate a method for mapping unstructured drug data to Anatomical Therapeutic Chemical (ATC) codes.
• To identify limitations and potential improvements in data quality for scientific utilization.

Main Methods:

• Algorithms were developed to map unstructured drug prescriptions from HIS to ATC codes.
• Data distributions were visualized to identify prescription data limitations, using proton pump inhibitors as a case study.
• A four-step feedback loop framework was inductively created, involving data processing, analysis, expert evaluation, and stakeholder feedback.

Main Results:

• A novel feedback loop framework with four crucial steps was established for continuous data quality improvement.
• The approach successfully visualized the distribution of structured versus unstructured drug data based on ATC codes.
• The study identified limitations in current drug prescription data, particularly with free-text entries not mapped to ATC codes.

Conclusions:

• The developed framework enables a continuously learning system for enhancing RWD quality from HIS.
• This approach has the potential to improve prescription data quality and patient safety.
• Future work will focus on evaluating the framework's impact on data quality and patient safety outcomes.