KEY POINTS
- Development and evaluation study of an automated AURA report-based chart checking tool built in the RayStation Python scripting environment for external beam radiotherapy.
- The tool compared treatment planning system data with Varian ARIA radiation oncology system data captured through advanced AURA reports, avoiding routine access to the live clinical database.
- Testing included 463 plans before physics review and 159 plans after manual physics review, across volumetric modulated arc therapy, three-dimensional conformal radiotherapy, and electron treatments.
- Before physics review, 48.6% of plans contained at least one detected error; after manual physics review, residual errors were still detected in 37.1% of reviewed plans.
- Common error categories included imaging setup (24%), written directive (6.3%), prescription (5.6%), patient shifts (5.4%), and contours (2.6%).
CLINICAL TAKEAWAY
This study supports report-based automated chart checking as a practical safety layer for external beam radiotherapy workflows, especially for catching structured errors that can persist after manual review. Its strongest operational advantage is avoiding routine live database access while still enabling near-real-time checks. The evidence is local and technical: the tool depends on RayStation, ARIA, AURA report structure, timely use, and continued manual review for items not yet automatable.