Challenges and Opportunities for Improving Patient Safety Through Data Science and Informatics
29th May 18:00-18:30
PhD, RN, FAAN, FACMI Program Director
Research Center for Patient Safety, Research, a
Patricia C. DYKES
In my career, I have focused on iteratively developing, testing and implementing technological solutions to improve patient care. To decrease falls in hospitals, I led the development of a fall prevention toolkit that captured the fall risk assessment data documented in the computer and reused it to dynamically generate tailored fall prevention plans. A randomized study demonstrated a 25% reduction in falls (JAMA, 2010) and 34% reduction in fall injuries (JAMA Network Open, 2020). These were the first published study in the United States that provides evidence for using an informatics intervention to reduce falls and fall related injuries in hospitals. I was the Partners HealthCare site PI for the STRIDE Fall prevention clinical trial, a national PCORI/NIA-funded study that seeks to integrate known fall prevention evidence into primary care practices. Currently, I am leading two federally-funded projects to develop patient-centered fall prevention clinical decision support for use in primary care settings (including mobile app for patients/family). In a CMS-funded project, I am also leading the development of electronic clinical quality measures (eCQMs) related to care of patients post total joint surgery including complications, opioid-related adverse events and opioid extended use. I am also the Partners HealthCare site PI for the CONCERN (Communicating narrative Concerns entered by RNs) project where we are using EHR nursing documentation machine learning methods to identify patients at risk for deterioration. Finally, I am leading an NIH/NIA funded study that uses machine learning methods to identify COVID + patients in long-term care settings who are at risk for deterioration.