A clinical data warehouse-based process for refining medication orders alerts.

Area of Science:

• Clinical Informatics
• Health Data Analytics

Background:

• Medication order alerts are crucial for patient safety but can be burdensome if not optimized.
• Clinical information systems generate numerous alerts, necessitating refinement before full implementation.

Purpose of the Study:

• To evaluate the use of a clinical data warehouse (CDW) and clinical information system (CIS) for refining medication order alerts.
• To test and optimize alert parameters prior to widespread clinical deployment.

Main Methods:

• A clinical decision rule refinement process was employed to assess alert frequencies.
• Retrospective analysis of physicians' medication orders from a CDW was conducted.
• Alerts for 10 renal function-dependent medications were evaluated in multiple iterations.

Main Results:

• Initial alert frequencies for 'exceeds maximum daily dose' were 7.10% and 'under dose' were 3.14%.
• Across iterations, 16.07% of rules were removed, 37.5% were modified, and 37.5% were newly introduced.
• Extensive data analysis facilitated significant alert optimization.

Conclusions:

• Coupling a CDW with a CIS enables effective testing and refinement of medication alerts.
• Retrospective analysis of medication orders is key to optimizing alert systems for enhanced patient safety.
• Optimized alerts minimize overridden alerts, improving clinical workflow efficiency.