Wrong-side imaging orders: automated detection using electronic health record data - a retrospective cohort study

Area of Science:

• Medical Informatics
• Patient Safety
• Diagnostic Imaging

Background:

• Wrong-side diagnostic imaging errors are preventable but underdetected due to reporting system limitations.
• These errors can delay diagnosis and lead to patient harm.

Purpose of the Study:

• To develop and validate an automated electronic health record (EHR)-based method for detecting potential wrong-side diagnostic imaging order errors.
• To identify risk factors associated with these errors using an adapted Retract-and-Reorder (RAR) approach.

Main Methods:

• Retrospective cohort study across six healthcare facilities.
• Adapted RAR methodology to screen 355,000 imaging orders from 2021, extending detection windows and analyzing side-switching events.
• Validated the method via chart review and used multivariate logistic regression to identify risk factors.

Main Results:

• Identified 1667 Retract-and-Reorder (RAR) events (4.70 per 1000 orders), with an 87% positive predictive value for confirmed errors (4.09 per 1000 orders).
• Outpatient settings (OR 4.53) and administrative staff (OR 2.08) had higher odds of RAR events.
• CT scans (OR 1.79) showed higher odds compared to X-rays.

Conclusions:

• A validated, scalable EHR-based method can automatically detect potential wrong-side diagnostic imaging order errors.
• This approach utilizes readily available EHR data for continuous surveillance and intervention evaluation.
• The findings support improved diagnostic safety by addressing a critical, yet often undetected, patient harm event.