Pediatric septic shock is a leading cause of childhood death, in which morbidity-free survival depends on timely diagnosis and treatment. Although the medical community, general public and state legislatures have increasingly recognized that delays in diagnosing septic shock in children are an important cause of preventable morbidity, there are no validated predictive models to improve early sepsis diagnosis in children. The candidate, Dr. Halden Scott, is a pediatric emergency physician at Children's Hospital Colorado and the University of Colorado School of Medicine. She has critically evaluated early diagnostic strategies in an Electronic Health Record (EHR)-derived pediatric sepsis registry that she designed that will support the proposed research. She is seeking to develop into an independent health services researcher, through the proposed mentored research with the following goals: 1) Derive and test models to predict septic shock in an existing EHR-derived dataset, comparing models derived with two modern modeling techniques; 2) create an EHR decision support tool for the early recognition of septic shock; 3) conduct a pilot implementation trial of decision support in community and academic pediatric emergency care settings within the Children's Hospital Colorado network. This implementation and feasibility data will inform a future multicenter pragmatic trial of computerized decision support in diagnosis of pediatric septic shock, which Dr. Scott will propose in an R01 application before the end of the K award. Dr. Scott's research trajectory is facilitated and enriched by strong institutional clinical, informatics and administrative support, and her involvement in the Pediatric Emergency Care Applied Research Network and international sepsis guidelines committees. This award would support Dr. Scott's career development goals: acquire skills in predictive modeling, the use of health information technology in learning health systems, and pragmatic trials. Dr. Scott's nationally renowned mentors will support these goals: Dr. Kempe, expert in pragmatic trials, program evaluation and implementation research; Dr. Fairclough, expert in statistics and modeling; Dr. Kahn, expert in research and clinical informatics. Additional advisors contribute expertise in machine learning, clinical pediatric emergency research, critical care and sepsis. This proposal responds to AHRQ priorities: improving health care quality through information systems, improving safety and quality by synthesizing evidence, and focuses on AHRQ priority populations including children and those with chronic care needs. The research, mentorship and training plan proposed are ideally suited to provide Dr. Scott with the skills needed to develop into an independent health services researcher focused on improving the quality of care for pediatric emergency conditions, particularly sepsis, through the use of health information technology in learning health systems, and conduct of pragmatic trials to test implementation of evidence-based practice in emergency settings.
/ Public Health Relevance Pediatric septic shock, a leading cause of childhood death, is often recognized late, causing preventable morbidity, and improved early diagnostic approaches would improve the timeliness of sepsis diagnosis, accelerate treatment and improve outcomes. This proposal will provide mentored research training to use predictive modeling to improve early diagnosis of pediatric sepsis, and test the delivery of clinical decision support to clinicians through the electronic health record in an implementation trial. This proposal responds to the AHRQ priorities of improving health care quality through use of information systems, improving safety and quality of care by synthesizing and communicating evidence to health care professionals, advances knowledge generation in a learning health system and focuses on AHRQ priority populations including children and those with chronic care needs.
