Patient Safety Improvement with a Software Tool to Prevent Isocenter Errors
E Ford*, M Nyflot, L Jordan, J Carlson, University of Washington, Seattle, WASU-E-T-237 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: Based on one year of departmental experience with near-miss incident learning, issues with isocenter placement were identified as a significant potential risk. Motivated by this, we initiated a patient safety improvement intervention which employs modern concepts in human factors engineering, error prevention, and software design.
Methods: In February 2012 a departmental electronic incident learning system was launched that is unique in three respects: the high volume of near-miss/no-harm reports (20 per week or ~1 per patient), the ability to tag each event by category type, and ranking of events by potential severity (0-to-4 point scale). Data over ten months (774 reports) indicate that near-miss events related to isocenter placement have a significantly higher potential severity compared to other events: 3.0+-1.3 vs. 1.5+-1.0 (p<0.001). We therefore developed a custom software interface to manage the placement of isocenters through the clinical workflow. The approach draws from concepts in the psychology of human factors design and usability, including forcing functions and automatic error checking.
Results: The custom software interface is designed to accommodate all common workflow scenarios for isocenter placement. The user is presented with a single screen containing all relevant information and is guided with a color-coded status scheme. The software checks for common error pathways (e.g. isocenter point inadvertently moved during planning) and changes the interface status accordingly. Further evaluation is ongoing, after which wider distribution may be possible.
Conclusions: Incident learning is a valuable method for objectively identifying areas for safety improvement. The critical safety issue identified here, isocenter placement, was addressed with a custom software interface that utilizes approaches not commonly employed by the vendor community such as user-centered design. The design bridges the gap of execution and evaluation which lies at the root of most errors.