Use of Inter-Hospital Transfer Services in Critical Illness and Acute Respiratory Failure - PROJECT SUMMARY/ABSTRACT This K08 award will establish Christopher Worsham, MD, MPH as a clinician-investigator health services researcher examining critical care services and their value using advanced methods of causal inference. Dr. Worsham is an Instructor in Medicine, pulmonologist, and critical care physician at Harvard Medical School and Massachusetts General Hospital. Building on his clinical training, prior research experience, and coursework in quantitative methods, this award will allow him to develop expertise in (1) advanced methods of causal inference using large administrative datasets, (2) inter-hospital critical care transport services with an emphasis on acute respiratory failure (ARF), and (3) decision science and cost- effectiveness analysis. To achieve these goals, Dr. Worsham has assembled a team with expertise in: health services research and advanced methods of causal inference (Drs. Anupam Jena, Mary Beth Landrum, Hannah Wunsch, and Bruce Landon); critical care medicine and ARF (Dr. Taylor Thompson); critical care transport in ARF (Dr. Susan Wilcox); and decision science and cost-effectiveness analysis (Dr. Ankur Pandya). The inter-hospital transfer of critically ill patients is a common practice in the U.S. critical care system, generally done to move patients to hospitals with intensive care unit (ICU) resources commensurate with the severity of their acute illness. However, we do not fully understand current national practice patterns of inter- hospital ICU transfer across diverse patient populations. We also lack robust estimates of clinical benefits or value of using particularly costly air ambulance services to rapidly move ICU patients between hospitals, including those with ARF, the most common reason for ICU admission. To address these important questions in the critical care system, Dr. Worsham will analyze data from two large insurance claims databases, Medicare and the MarketScan commercial insurance claims database. In Aim 1, he will characterize current practices and risk factors associated with inter-hospital ICU transfer in the overall adult ICU population. In Aim 2, focusing on patients with ARF, he will estimate the effect of air compared to ground inter-hospital transport using a novel instrumental variables approach that accounts for unmeasured confounding by taking advantage of the random weather conditions that push patients to travel by ground rather than air. In Aim 3, he will use methods of decision science to perform a cost-effectiveness analysis of air compared to ground inter-hospital transport of patients with ARF to inform high-value care. This work and training will provide him the basis for a future R01 proposal that examines common and costly practices that impact patients as they move through the U.S. critical care system using advanced methods of causal inference and cost-effectiveness analysis. This K08 and future R01 will inform bedside decision making and policy surrounding resource allocation and high-value care in the U.S. critical care system.
