Using Cognitive Load Theory to Understand and Improve Patient Handoffs

Abstract

Introduction Transfers of patients from one physician to another (handovers) are ubiquitous and occur with increasing frequency. Handovers are a common source of communication failures, which lead to medical errors and harm to patients. Considerable attention has focused on interventions to improve patient safety during handovers. While interventions that bundle current best practices have resulted in reduced errors and improved educational outcomes, handovers remain a common source of error. Addressing the cognitive complexity of a patient handover can help develop new strategies that support learning and further improvements in patient safety. Informed by Cognitive Load Theory (CLT), we examine the cognitive challenges that trainees face when learning how to perform a handover. This work advances understanding of how cognitive load can be measured and managed during a handover. Methods This thesis used multiple methodologies. An AMEE guide and subsequent analysis of handovers through the lens of CLT yielded a conceptual model of sign-out between the sending and receiving clinician. This conceptual model then informed two studies that described the development of a cognitive load inventory for handovers and collected evidence for validity. An experiment with early and advanced medical students explored the relative influence of learner knowledge and patient complexity on information loss and distortion during simulated handovers. A final study reported on a CLT-informed intervention to balance the mental workload associated with outpatient panels during the academic year-end handover of psychiatric continuity clinic patients. Results This work explored the implications of CLT for handovers education. Hypothesized drivers of intrinsic, extraneous, and germane load during a handover were identified. We demonstrated how current best practices primarily focus on reducing extraneous load. Managing intrinsic load and optimizing germane load are relatively under-addressed. Validity evidence obtained for the two versions of a cognitive load inventory for handovers yielded mixed results, with consistent support for the intrinsic load items, conflicting results on germane load, and poor performance by the extraneous load items. The experiment with early and advanced medical students suggested that learner knowledge influenced information loss and distortion more so than patient complexity. Our CLT-informed intervention showed that, compared to the traditional method, the workload-balancing method generated lower inter-caseload variation for each mental workload factor. The method reduced overall inter-caseload variation by 50%–61% in each of the four intervention years.