Integrating real-time clinical activity and behavioral responses for characterizing cognitive load and errors (IGNITE)

整合实时临床活动和行为反应来表征认知负荷和错误（IGNITE）

• 批准号: 10504867
• 负责人: Thomas George Kannampallil
• 金额: $ 40万
• 依托单位: WASHINGTON UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-30 至 2027-07-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10504867
• 关键词: Integrating; real; time; clinical; activity

PROJECT SUMMARY Cognitive load during the delivery of clinical care is affected by a number of factors including specialty, practice setting, patient complexity, electronic health record (EHR) use, and clinician expertise. In the United States, clinical care is primarily documented using EHRs: documentation burden, poor usability, and unnecessary navigation contribute to increased cognitive load. Such increases in cognitive load, in turn, contribute to more work hours, dissatisfaction with work, poor patient outcomes (e.g., errors), burnout and poor clinician outcomes. Much of the prior research characterizing clinician cognitive load or its impact on errors has relied on retrospective approaches including self-reports, time-motion studies, and focus groups. Similarly, burnout also has been measured exclusively using surveys. EHR-based audit logs have shown considerable promise as a viable resource for tracking and measuring clinical activities without the incremental survey burden on clinicians. Our research team has demonstrated that workload measures based on audit logs can be used to assess cognitive load, burnout, and errors. Based on this promising pilot work, the primary focus of the IGNITE (Integrating real-time clinical activity and behavioral responses for characterizing cognitive load and errors) study is to utilize EHR-based audit logs and decision support tools to objectively determine the direct relationships between (a) cognitive load and errors, and (b) the mediating role of clinician burnout in explaining the relationship between cognitive load and errors. We will accomplish this through a large-scale multi-site study conducted at three large academic medical centers associated with Washington University/BJC HealthCare, Stanford University, and University of Colorado. For the first aim, we will utilize EHR-based audit logs collected across non-surgical, inpatient settings across three sites over a 3-year period (1/1/2019 to 12/31/2021) to develop measures of cognitive load—both intrinsic and extraneous—and assess the effect of cognitive load on objectively measured wrong-patient errors (using the retract-and-reorder alerts). For the second aim, we will prospectively collect data on a cohort of 300 trainees (residents, fellows) from Medicine and Pediatrics from each study site over a 5-month period. Monthly burnout surveys, along with cognitive load measures from the EHR-based audit logs, and wrong-patient errors during the study period will be used to determine the mediating relationships between cognitive load, burnout and errors. In addition, for both of the proposed aims, we will develop advanced machine learning algorithms to predict errors and burnout from EHR-based activity sequences. Insights from this study will help in designing targeted interventions aligned with the contextual work practices of physicians, designing clinical trials for evaluating such interventions, and in developing informed policy guidelines for the safety and well-being of physicians.

项目总结