Automatic Second Check of RT Treatment Record Using Event Triggered Audit
S Hadley*, X Chen, G Weyburne W Keranen, The University of Michigan, Ann Arbor, MISU-E-T-228 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To enhance patient safety and radiation therapy quality assurance by implementing an immediate computer-based audit of radiation therapy treatment record. Provide an automatic second check of the RT record in a paperless environment. Replace some aspects of weekly physics chart checks with a computer-based audit of the treatment.
Method: A software agent receives signals from a system we have developed, EventNet, when sentinel events occur in the TMS. Audits are triggered when a "treatment canceled", and one of the many "treatment complete" signals are received. The treatment canceled audit assumes the patient treatment was canceled and verifies that no treatment happened. The treatment completed audit assumes that a treatment occurred and checks the treatment history parameters and dose summary against the scheduled plans, session, daily and total doses. The agent extracts the plan(s) and the day's treatment history from the TMS with a DICOM query and sends that data to a web service that compares the history against the plan. If the audit succeeds, the agent quietly adds an entry to a log file. When it fails the agent alerts staff via pager, email, or SMS message.
Result: The agent is able to catch events from the TMS and trigger one of two audit processes. Basic audits of the treatment histories works as designed. Intentional variation in treatment parameters in test cases are caught and trigger the immediate alert.
Conclusion: Immediate audit of the RT record is better than waiting for weekly physics chart check to verify plan parameters. Rather than have a physicist seek out variations in treatments by checking every treated parameter, a software agent can provide a basic audit and alert the physicist if needed. This brings more focus on patient treatments that require oversight and could free up valuable time for other QA tasks.