The largest portion of my research is on the factors influencing emergency cardiac care and its outcomes, and how it can be improved in actual practice. This work has included the development of clinical predictive instruments as decision aids for emergency physician clinical decision-making. Having been tested in large multicenter clinical effectiveness trials, these instruments are now incorporated into conventional computerized electrocardiographs. In related work, our group also evaluates and develops computer-based mathematical models for predicting other medical outcomes, and also uses these predictive models for retrospective evaluation of clinical care beyond the real-time clinical use. Our group also does research in community settings that test new approaches to the care of cardiac patients, including the use of metabolic support of the myocardium during heart attack.