Integrating and Evaluating the Data Quality and Utility of Smart Pump Information in Detecting Medication...

Area of Science:

• Clinical Informatics
• Health Data Integration
• Patient Safety

Background:

• Electronic health records (EHRs) are primary clinical data sources but rely on manual input, leading to errors.
• Smart pump records (SPRs) offer granular data on medication administration but face integration challenges with EHRs.
• The full potential of SPRs for medication error detection remains unrealized due to data quality and utility issues.

Purpose of the Study:

• To evaluate the quality and utility of integrated SPR and EHR data for detecting medication administration errors.
• To determine if SPRs provide unique information for identifying discrepancies in the medication use process.
• To hypothesize that combined data sources enable more comprehensive error detection.

Main Methods:

• Evaluated medication use for 9 high-risk medications in a neonatal intensive care unit over 1 year.
• Developed an algorithm to align SPRs with EHR medication orders using patient ID, medication name, and timestamp.
• Assessed SPR data quality and compared discrepancy detection between SPRs and EHR medication administration records (MARs).

Main Results:

• Approximately 70% of SPRs were feasible for integration with EHR data.
• Combined data revealed 321 EHR MAR discrepancies and 682 SPR discrepancies, with vasopressors, narcotics, and total parenteral nutrition showing high rates.
• SPRs detected more substantial dosing discrepancies than EHR MARs, with minimal overlap in identified errors.

Conclusions:

• Integration of SPR and EHR data enhances the analysis of medication lifecycle errors.
• SPRs emerge as a more reliable data source for detecting medication errors.
• Integrating SPR information with EHR data is crucial for comprehensive medication error detection and mitigation.