Assessing Imaging Metadata Quality in Radiology: Investigating Discrepancies Between PACS and RIS Data

Area of Science:

• Medical Informatics
• Radiology Data Management

Background:

• Data reliability is crucial for clinical decisions and research.
• Radiology Information Systems (RIS) and Picture Archiving and Communications Systems (PACS) often operate independently, leading to data inconsistencies.
• Uncertainty in clinical practice can arise from discrepancies between RIS and PACS data.

Purpose of the Study:

• To investigate the coherence and integrity of radiological data between RIS and PACS.
• To identify discrepancies in contrast agent documentation and anatomical region information.
• To assess the need for data quality improvements in radiological metadata.

Main Methods:

• Extracted data on contrast agents and anatomical regions from RIS and PACS at a German University Hospital.
• Linked and aggregated data at the study level for comparison.
• Analyzed coherence and identified discrepancies between the two systems.

Main Results:

• Moderate coherence for contrast agent documentation (61% identical records, 39% discrepancies).
• Discrepancies in contrast agents were often due to missing PACS entries.
• Variable agreement for anatomical regions: high for chest (81%), lower for abdomen (52%) and brain (51%).

Conclusions:

• Significant data inconsistencies exist in radiological metadata between RIS and PACS.
• Standardized documentation and cross-system data quality checks are essential.
• Addressing these inconsistencies is vital for reliable clinical decision-making and research integrity.