PROJECT SUMMARY High nursing workload is a threat to care quality, patient safety, and nurses’ well-being and job satisfaction. Workload – which lacks a universally accepted definition - is a complex multi-dimensional construct that is affected by external task demands and environmental, organizational, and psychological factors. The importance of managing high workload is nowhere more evident than in neonatal intensive care units (NICUs). Critically ill neonates are highly vulnerable to iatrogenic events due to their immaturity and fragility, and high workload has been directly associated with increased incidence of adverse neonatal safety outcomes. Despite the evidence and need, patient safety researchers have been slow to develop multi-level models, scalable workload measurement systems, or other health information technology interventions to improve workload management and patient safety. Conventional nursing workload management tools predominantly measure and predict workload using unit-level (e.g., staffing ratios) or patient-level (e.g., acuity) data rather than data collected across the four levels of workload recommended by human factors engineers (HFEs) - unit, job, patient, and situation. As a result, current tools under-measure the workload experienced by nurses and are not designed to identify mutable microsystem factors that contribute most to nursing workload. A promising development in nursing workload research is the increasing emphasis on measuring situational workload which best explains the workload experienced by nurses due to healthcare microsystem design. Situational workload is most affected by performance obstacles (i.e., delays, interruptions, etc.) in the local work environment and can be applied at the unit, job, or patient-levels. Most importantly, it is diagnostic of underlying contributory factors and therefore actionable for improvement. To date, situational workload has been measured using subjective surveys which are work-interrupting, thus difficult to integrate into practice. Vanderbilt University Medical Center (VUMC), in collaboration Johns Hopkins University (JHU), will employ a systems engineering human-centered design process to design, develop, and validate new multi-level models of NICU nursing workload derived from readily accessible electronic health record (EHR) data. The validated models will be the foundation for a future EHR-based clinical decision support (CDS) tool that will track the real-time workload of registered nurses, predict near- term future unit workload, and guide workload reduction and balancing interventions. The project’s three Specific Aims are: Aim 1. To conduct a comprehensive HFE-based analysis of NICU nursing workload; Aim 2. To design and develop real-time multivariable workload models and Aim 3. To validate the real-time workload models at VUMC (A) and to determine the generalizability of the models at an external hospital (B).

PROJECT NARRATIVE This project is relevant to public health because it aims to improve the safety of nursing care delivered in neonatal intensive care units (NICUs). The research is significant because it will improve neonatal patient safety by changing the way nursing workload in the NICU is measured and used for workload management, organizational learning, and improvement. The project is innovative because it will use data readily accessible in electronic health records to develop a new multi-level model of workload for registered nurses and use this model to guide the design and development a future clinical decision support tool for workload management and safety improvement.