Assessing explicit error reporting in the narrative electronic medical record using keyword searching

Area of Science:

• Medical Informatics
• Patient Safety
• Healthcare Quality

Background:

• Medical errors, both in and out of hospitals, carry significant consequences and increase healthcare costs.
• Accurate identification of medical errors is crucial for effective management and prevention strategies.
• Electronic health records offer a potential avenue for systematically identifying reported medical errors.

Purpose of the Study:

• To assess the explicit reporting of medical errors within electronic health records.
• To evaluate the effectiveness of keyword searches combined with manual review for identifying reported medical errors.

Main Methods:

• Utilized five keywords: "mistake," "error," "incorrect," "inadvertent," and "iatrogenic."
• Surveyed narrative reports including discharge summaries, sign-out notes, and outpatient notes from 1991 to 2000.
• Manually reviewed all identified cases, with physician reporting serving as the basis for error identification.

Main Results:

• Identified 222 explicitly reported medical errors.
• Positive predictive value for keywords ranged from 3.4% to 24.4%, indicating variability.
• Therapeutic-related errors, predominantly medication errors, were the most commonly reported.

Conclusions:

• Keyword searches coupled with manual review successfully identified some reported medical errors in records.
• The methodology demonstrated low sensitivity but a moderate positive predictive value, influenced by the chosen search terms.
• Physicians most frequently documented errors in the Hospital Course and History of Present Illness sections of discharge summaries.