Understanding multiple task coordination in a complex healthcare environment
Barg-Walkow, Laura Hillary
MetadataShow full item record
Understanding multiple task coordination is important in complex life-critical environments. In healthcare, for example, many situations occur in which there are multiple tasks and limited resources for addressing all tasks at the same time. Emergency departments in particular are complex, interruption-driven environments. In many cases, physicians in emergency departments do not complete a single task in isolation. Decisions regarding what tasks to do, and when to do them, can affect performance (e.g., time, accuracy, patient safety). Additionally, some task factors (e.g., priority, difficulty) can drive task coordination behaviors. Characteristics of interruptions, such as frequencies and types, in emergency departments have been studied, but there has been little research on how physicians schedule and manage multiple tasks. The purpose of this research was to investigate multiple task coordination by emergency physicians to understand strategies for task completion, strategies for task scheduling, and management of interruptions. I conducted two studies to understand how emergency physicians coordinate multiple tasks. The goal of the first study was to understand task scheduling decisions by physicians in emergency departments through a modeling approach. This study consisted of an online questionnaire conducted with 170 emergency physicians (120 attending and 50 resident physicians). There were two primary research aims: to understand (1) task scheduling decisions in a multiple task context, and (2) how task scheduling decisions varied across experience level. Attending physicians’ task scheduling decisions aligned more with a parsimonious one-reason rule, where priority was the only factor that influenced decisions. Alternatively, resident physicians’ decisions were not driven by priority, but rather were influenced by difficulty, salience, and engagement. This indicates that physicians may be differentially weighting different cues as they make decisions about how to order tasks, and provides insights for how to support decision making as these strategies are learned. The goal of the second study was to understand how multiple task demands are managed and coordinated by physicians in emergency departments. This study consisted of questionnaires and interviews with 30 emergency physicians (15 attending and 15 resident physicians). There were three primary research aims: to understand (1) strategies used for multiple task coordination, including both completion and scheduling strategies; (2) how interruptions were conceptualized and coordinated, and (3) how multiple task coordination varied across experience level. I identified and hierarchically categorized a broad set of strategies for task completion, and determined that these strategies did not change with experience. For task scheduling, I confirmed that previously-identified factors drove task scheduling. I also better defined factors (e.g., splitting priority into urgency and criticality) and identified additional factors (e.g., time and its subcomponents, interpersonal skills). Although there were common task scheduling factors mentioned by all 30 participants (e.g., priority, time), other factors were identified more often by attending physicians than resident physicians (e.g., interpersonal skills). I also found that conceptualizations of interruptions in this environment did not significantly differ from existing definitions; however, participants discussed the need to clarify between positive and negative interruptions. Overall, this research provided insights into task coordination in a complex, interruption-driven healthcare context. In this work, I investigated strategies for task scheduling, including further evaluating known factors (e.g., priority) and identifying additional factors (e.g., time) that drive task scheduling decisions. I combined insights from both quantitative and qualitative methods to evaluate hypothesis-driven models for task scheduling. In this case, findings from Study 1 indicated that a one-reason priority-only model best captured attending physicians’ task scheduling decisions and a multi-attribute model best captured resident physicians’ task scheduling decisions; however, findings from Study 2 indicated a rich set of factors that are used by emergency physicians beyond those factors in the models. This indicates more parameters should be included in modeling studies to better evaluate task scheduling decisions. The results of this dissertation have implications for improving training and evaluation of physicians as well as designing tools to support multiple task coordination.