An Automated, Dynamic Radiation Oncology Prescription Checking System

Area of Science:

• Radiation Oncology
• Medical Physics
• Health Informatics

Background:

• Radiation oncology prescriptions are critical for treatment delivery.
• Manual prescription entry is susceptible to errors, potentially impacting patient safety and resource allocation.

Purpose of the Study:

• To develop and implement an automated system for checking radiation oncology prescriptions.
• To reduce errors in prescription entry and minimize their propagation.

Main Methods:

• Generated rules for clinically appropriate doses per fraction based on disease site and treatment technique.
• Implemented an automated report running every 30 minutes to check new/approved prescriptions.
• Sent automated email alerts to prescribers for rule violations, with continuous monitoring and adjustment of criteria.

Main Results:

• Checked 24,047 prescriptions from June 2021 to November 2022, with a 1% alert rate (241 alerts).
• Identified 82.2% of alerts as unapproved doses per fraction, 12% as too high for nonstereotactic technique, and 5.8% as too low.
• 31% of alerts resulted in prescription amendments, indicating true errors. Trainee entry and low doses (<150 cGy) were associated with errors.

Conclusions:

• Automated prescription checking systems offer significant value in radiation oncology.
• Implementation can mitigate risks associated with erroneous prescription entry, including treatment delays and misadministration.
• Such systems enhance patient safety and optimize resource utilization.