Matthew A Levin, MD

Dr. Levin is a practicing board-certified Cardiac Anesthesiologist, Associate Vice Chair of Research and Director of Research Informatics for the Department of Anesthesiology, Perioperative & Pain Medicine, and Medical Director of the Clinical Data Science Team for the Mount Sinai Health System. In addition, he holds joint appointments in the Department of Genetics and Genomics Sciences and Department of AI and Human Health. After graduating from M.I.T., Dr. Levin began his career by working in the software industry during the dot-com boom. Throughout his career in medicine he has continued to maintain an interest and expertise in technology, obtaining board certification in Clinical Informatics and serving as a board member for the Society of Technology in Anesthesia. As Medical Director of the Clinical Data Science Team, he has successfully overseen the implementation of a real-time multi-modal streaming data science pipeline that is used to deploy operational machine learning models at the point of care. The Clinical Data Science team recently won the prestigious Hearst Health Prize for its machine learning application called NutriScan AI that facilitates faster identification and treatment of malnutrition in hospitalized patients.

Dr. Levin’s primary research interests are clinical informatics, intraoperative physiology, and perioperative genomics. His research expertise lies in designing and executing large database driven retrospective studies using high granularity EHR data. He has published multiple large cohort studies examining the relationship between intra-operative physiology and postoperative outcomes. Several of these studies have been widely cited within the anesthesiology literature, for example the finding that low intraoperative tidal volume without PEEP is associated with higher mortality. In the area of clinical informatics he has been PI or co-PI on two large pragmatic decision support trials examining the utility of machine learning generated clinical alerts: the "Double-Low" trial testing where alerts for concomitant low blood pressure and low brain activity can improve outcomes after surgery and ReSCUE-ME, a trial to prevent in-hospital clinical deterioration. The results of the latter trial were recently published in Critical Care Medicine, demonstrating that machine learning alerts for clinical deterioration can lead to faster escalation of care and are associated with decreased mortality. The accompanying editorial emphasized that this is one of the few studies to truly examine what happens when a prediction model is moved from "byte to bedside". 

During the COVID-19 crisis Dr. Levin led a multi-disciplinary group to drive innovation in ventilators (the Mount Sinai HELPS Innovate Lab) by working with both the NASA Jet Propulsion Laboratory, and industry partners to develop a novel flow control valve that allowed two patients to be ventilated with a single ventilator. 

More recently, Dr. Levin has focused on differences in treatment and outcomes between racial/ethnic groups. He has investigated the relationship between self-reported ancestry and intra-operative blood pressure response to phenylephrine, a medication commonly used to treat low blood pressure, finding that there is a significant difference in response between patients of Hispanic, African American, and European ancestry. In the area of oximetry, he has demonstrated that cerebral oximetry monitoring devices appear to have no racial/ethnic bias, whereas pulse oximeters may.

Dr. Levin sits on the Mount Sinai Health System Data Use Committee, where he helps ensure the protection of patient data.

LinkedIn