# Data and Medical Decision-Making Data and computation are reshaping medicine. Here, we focus on their impact on clinical decision-making. We define “decision-making” as cognitive and other process that directly drive clinical action. Historically, the clinical judgment that underlies decision-making was the sole province of individual providers. Increasingly, however, data and feedback are being incorporated into decision-making processes to improve outcomes and healthcare value. We also note that decisions are not only made within the context of the doctor-patient relationship but also at various organizational levels within the healthcare system (e.g. individual clinical unit, hospital system). Below, we discuss core medical and technological concepts; key stakeholders; emerging trends; and representative clinical examples. # Medical Resilience and Failure Medical physiology is the study of the physical processes of human life in health and disease. Physiological systems are able to maintain function and stability in the face of internal dysfunction and external insult. The capacity to do so is called resilience. Failure is a loss of resilience and represents a threat to the continued integrity and viability of a living system. Failure can be caused by discrete subsystem failure (e.g. heart attack) or by systemic dysregulation (e.g. septic shock). Here, we give a “clinical-first view” for resilience and failure. We take a broad scope to include both acute disease as well as chronic disease and aging-associated disease. From an engineering perspective, we have thematic emphasis on the emergence of resilience from nonlinearity and interconnected systems. Likewise, we discuss the flip-side of nonlinearity and interconnection: the transition from resilience to failure can be abrupt and difficult to predict. Finally, given the importance of resilience and failure in several key challenges in clinical care, we make connections to data and medical decision-making. %% Data and computation are reshaping medicine and clinical decision-making. Examples include acute states of physiological failure such as shock and sepsis as well as failure modes associated with aging (e.g., delirium/acute brain failure, falls). This seminar course provides (1) a modern, clinically facing view of physiological failure and (2) a survey of how data and computation are reshaping clinical concepts and practice, including decision-making. Key topics and concepts include medical data types (e.g., imaging, lab values, oximetry); nonlinearity and complexity in physiological resilience and failure; clinically relevant AI/ML methods; data-driven definitions of medical disease; predictive modeling as a distinct field in AI/ML; and clinical decision-making using modern data and computational tools. %% ___ *Author: Michael A. Choma, MD, PhD (Yale University)*