Using name overlap analysis to understand medication name search safety.

Area of Science:

• Pharmacology
• Health Informatics
• Human-Computer Interaction

Background:

• Medication name ambiguity in electronic health records poses a risk for medication errors.
• Computerized systems often rely on text-based searches for medication identification.

Purpose of the Study:

• To quantify the extent of left-to-right character overlap in medication names within real-world computerized systems.
• To assess the implications of this overlap for medication search functionality and safety.

Main Methods:

• A computational approach was used to analyze character overlap in 20,020 lists of medication names.
• Lists were generated by random sampling from a database of 2,249 medication names.
• Analysis focused on lists ranging from 100 to 500 medication names.

Main Results:

• Maximum left-to-right character overlap ranged from 4 to 29 characters (mode: 14) in general medication lists.
• For high-alert medications, overlap ranged from 3 to 10 characters (mode: 6).
• Initial keystrokes are most critical for disambiguating medication names during searches.

Conclusions:

• Significant variation in medication name overlap necessitates adaptive search strategies.
• Implementing incremental dynamic search, where systems respond to each keystroke, can enhance medication search safety.
• This approach may allow users to identify medications with fewer keystrokes, reducing the potential for errors.