CONCERN - Communicating Narrative Concerns Entered by RNs (CONCERN): Clinical Decision Support Communication for Risky Patient States

Sarah Rossetti, Assistant Professor of Biomedical Informatics and Nursing at the Columbia University Medical Center, School of Nursing 

The aim of this project is to test in a clinical trial a novel predictive early warning score for patients at risk of detection in the hospital using a SMARTapp on FHIR clinical decision support system across two large academic medical centers. Our CONCERN predictive model uses our healthcare process modeling approach to derive when a nurse is concerned about a patient from the patterns of EHR documentation. The model outperforms other early warning scores and has increased lead time - identifying at risk patients 5-24 hours earlier. We are extending and validating our CONCERN predictive model on the COVID-19 hospitalized population to provide clinical decision support for patients at risk of deterioration during a time of increased severity of illness and decreased resources. In addition we are leveraging our modeling approach and applying it to several resource modeling needs, including data-driven bed classification during the surge of creating new ICU beds and tracking ventilator status. These are challenging activities to track in the EHR, often requiring manual configuration and tracking.