Data quality in the outpatient setting: impact on clinical decision support systems

Area of Science:

• Health Informatics
• Clinical Pharmacology
• Medical Informatics

Background:

• Clinical decision support systems (CDSS) are crucial for safe medication prescribing.
• The accuracy of CDSS relies heavily on the quality of electronic health record (EHR) data.
• Incomplete or inaccurate data can compromise CDSS performance and patient safety.

Purpose of the Study:

• To evaluate the impact of medical record completeness and accuracy on a CDSS.
• To assess how data quality affects a CDSS for gastrointestinal bleeding risk and NSAID therapy recommendations.

Main Methods:

• Reviewed 178 standardized patient encounters.
• Examined the documentation of six key data elements in medical records.
• Input available data into the CDSS to analyze recommendation accuracy.

Main Results:

• Mean completeness score for medical record data was 0.34.
• Mean correctness score for present data elements was 0.94.
• Missing data led to inappropriate and unsafe CDSS recommendations in nearly 77% of encounters.

Conclusions:

• Significant gaps in medical record completeness critically impact CDSS accuracy.
• Ensuring high-quality EHR data is essential for reliable CDSS performance.
• Improving data completeness is vital for safe prescribing and patient safety.