Quantification of physician workload for radiotherapy planning, and possible associations with performance.

Abstract

245 Background: Workload is a hypothetical construct representing the overall human cost incurred during a task. Across disciplines, the NASA Task Load Index (TLX) has been used as a subjective measure of workload. It quantifies six dimensions of work (mental, physical, and temporal demands, frustration, effort, and performance) on a scale of 1-100 scale, with scores >55 (i.e. overworked) associated with lesser performance. We herein quantify differences in physician workload for planning a relatively “simple” (palliative 2 field opposed lateral brain) versus relatively “complex” (curative 4 field pancreas) case. Further, we explore the association between workload and performance to define TLX levels where performance could be expected to decline. Methods: Nine physicians planned two cases, each case involving three tasks (i.e., task 1 - review written patient medical records; task 2 - review diagnostic images and design of treatment fields; and task 3 - review and approval of the treatment plan). After completing each case, TLX scores of each case/task was collected. Differences in workload between cases/tasks were assessed via ANOVA. The association between workload and performance (assessed using time-to-task completion and self-assessments) was assessed via Pearson’s correlation test. Results: 54 TLX workload and performance assessments were obtained. Workloads for the simple brain case/tasks averaged 41-48, versus 62-69 for the complex pancreas case/tasks (p<0.001 via ANOVA). There were no differences between tasks for the individual cases (p>0.1). There was a correlation between TLX scores and time-to-complete (r=0.54, p<0.001), and with physician self-assessments (r=-0.74, p<0.001). There was a trend towards reduced performance with TLX scores > 55. Conclusions: Physician workload levels are markedly lower for “simple” versus “complex” cases, indicating that TLX is a reliable tool to quantify workload. Performance appears to decline at TLX levels > 55, which is consistent with findings from other industries. Thus, we propose workload assessment (via TLX) to be considered as an independent quality measure to assess the quality assurance (QA) of processes used to deliver radiation therapy.