PROJECT SUMMARY/ABSTRACT Septic shock is a common, often fatal, dysregulated host response to infection that results in acute circulatory failure. Almost all patients with septic shock either have anemia (i.e., low blood counts) on presentation or develop anemia during hospitalization and anemia in septic shock is strongly associated with morbidity and mortality. Almost 60% of patients with septic shock receive autologous red blood cell transfusions to treat anemia once hemoglobin levels drop below 7.0 g/dL (threshold-based transfusion strategy). However, there is conflicting guidance and evidence for whether all patients with septic shock and hemoglobin levels <7.0 g/dL should receive transfusion or whether blood transfusion should be guided by additional clinical criteria beyond hemoglobin level alone. In a recent study, we found that routine threshold-based transfusion strategies for patients with sepsis were, on average, associated with harm. In the setting of conflicting guidance and limited evidence of effectiveness, practice patterns for blood transfusion in septic shock are also unclear. Guided by a Learning Health System framework that is consistent with the National Heart, Lung, and Blood Institute's clinical trial and implementation research priorities, this proposal's objective is to improve the effectiveness of blood transfusions during septic shock. We will (1) benchmark hemoglobin threshold- based blood transfusion practices in patients with septic shock across United States hospitals using a novel econometric-based approach, (2) use cutting-edge causal inference techniques to identify blood transfusion heterogeneity of treatment effect (i.e., identify patients most, and least likely to benefit from blood transfusion), and (3) pilot the implementation of an evidence-informed transfusion decision aid in the Boston Medical Center Medical Intensive Care Unit. Through his comprehensive career development plan, interactions with his excellent mentoring and advisory team, and the excellent training environment of Boston University Chobanian & Avedisian School of Medicine and Boston Medical Center, Dr. Bosch will achieve training objectives in (1) econometrics, (2) observational comparative effectiveness and causal inference, (3) mixed methodologies (4) hybrid implementation-effectiveness clinical trials, and (5) research team leadership that are aligned with core competencies to train the new generation of Learning Health System researchers. Results from this mentored proposal will fill large knowledge gaps in septic shock and transfusion medicine, will yield clinically actionable information regarding the optimal approach to blood transfusion in critically ill patients, and will inform Dr. Bosch's R01 application for a multicenter, pragmatic Hybrid Type 1 randomized clinical trial studying the effectiveness of transfusion guided by decision aids.