Dr. Barnes is a cardiologist and health services researcher who is passionate about improving care for patients with thrombotic conditions using innovative tools within the electronic health record. He uses user-centered design principles to develop clinical decision support and population health tools that support clinicians in making evidence-based decisions. Dr. Barnes partners closely with data scientists to leverage machine learning methods to enhance the implementation and usability of various tools with the electronic health record.
For one project, his team developed a population management tool within the electronic health record to identify all patients using oral anticoagulant medication and to flag those whose prescription did not follow the FDA package label instructions. This tool, called the DOAC Dashboard, has now been implemented in five health systems nationally and has dramatically reduced off-label prescribing rates. Concurrently, he evaluated a similar tool within the Veterans Health Administration system and showed that use of their DOAC Dashboard was associated with a reduction in both off-label prescribing and a reduction in stroke and other thrombotic complications.
For another project, his team developed a series of medication alerts within the electronic health record that appear at the time a clinician prescribes an off-label oral anticoagulant. We developed a system for real-time randomization so that we could compare the two different alerts to see which was most effective at achieving a change in the prescription.
He started his research journey as a medical student, exploring the basics of anticoagulant prescribing. In medical school, he helped to write a grant application to Blue Cross Blue Shield of Michigan that would fund a multi-site quality collaborative focused on anticoagulation care. As a faculty member here at the University of Michigan, he partnered with Dr. Michael Lanham, an informaticist, to develop the DOAC Dashboard and implement it in five different health systems. he works closely with colleagues from the Medical School and School of Public Health to implement best practices in clinical care with a focus on innovative uses of the electronic health record to guide clinician behavior change.
In medicine, there is an overwhelming volume of data. No single person can keep up with the rapidly growing literature or the volume of data that is collected for individual patients. AI has the potential to rapidly summarize the literature and bring actionable data to the fingertips of clinicians in a way that guides them to make appropriate and safe decisions for their patients. Realizing this potential is what excites him most about data science and AI research in healthcare.