A Structural Solution for Task Management in a Resident-Directed, Team-Based Radiotherapy Clinic

Abstract

<h3>Purpose/Objective(s)</h3> The radiation oncology service at our urban safety net hospital is high volume and organized into a series of resident-directed Teams covering various disease sites. As configured in our OIS, task management was difficult, and OIS-related miscommunication led to rushed/excess work and risked patient impact. An initiative was started to reduce the incidence of such miscommunication and associated sequelae. <h3>Materials/Methods</h3> A root cause analysis was performed. Clinical and OIS processes were mapped and failure modes assessed. While the clinic was organized into Teams, the OIS was not, and manual assignment of users to tasks was required. This was time consuming and error prone, with errors most common after residents rotated, during cross coverage, and with long intervals between consultation and simulation. Scheduling therapists often scheduled the incorrect resident to at least one task. Multiple attendings had patients on any given Team, but each was unable to see the entire Team's tasks. Tasks associated to specific residents had to be accounted for after they switched rotations. OIS functionality was such that only attendings could be automatically assigned to tasks and that individual tasks could not be assigned to multiple user groups. An alternative solution was developed involving a specific naming schema for tasks, creation of virtual resources for task assignment representing Teams and key user groups, and procurement of Team-specific VOIP phones. <h3>Results</h3> The Team concept was integrated into the OIS, and missed tasks and other OIS-related miscommunication are no longer a common reason for process failure. A minimum set of actions necessary to meet needs was codified into a set of Team-based clinical pathways. Task naming used the convention "<responsible group> - <task>." Tasks assigned to Team resources allowed rapid Team-specific, user group-specific identification of rate-limiting actions when sorted alphabetically. This also allowed transparency into the status of key activities performed by all user groups. Attendings maintained an accurate personal list. The updated system was reliable, required fewer actions, and reduced cognitive demand for all users. For schedulers, the minimum number of clicks required went from 25 to 5 and the minimum number of choices went from 8 to 1. Work to determine which resident to schedule was eliminated as therapists could rely on the less variable Team concept. A forcing function detected likely failure modes of wrong attending and wrong Team. Team-specific VOIP phones allowed the correct resident to be contacted no matter their physical location while minimizing interruptions to co-residents. <h3>Conclusion</h3> A novel implementation of virtual task lists improved quality by eliminating systematic sources of error and excess work.