Growing up in a large working-class family in West Belfast, Ireland, Paul Barr learned early on to take advantage of a good opportunity when it came up. So, after earning a BSc in psychology from Queen’s University in Belfast, he accepted a scholarship offer to pursue a master’s at the London School of Hygiene and Tropical Medicine in London.
The experience piqued an interest in public health he’d had since childhood. “My mom graduated with her nursing degree from university in her mid-40s, after raising six kids—I think her influence has always made me interested in caring for folks who are in need,” says Barr, who would go on to earn a PhD in health services research back at Queen’s University.
Today, as an associate professor of health policy and clinical practice, biomedical data science and community and family medicine at Geisel—Barr is one of the most innovative and productive researchers in his field, having published 14 papers last year alone, many in high-impact journals. In the following Q & A, he talks about his work and his efforts to develop tools that can improve communications between patients and their healthcare providers.
Q: What has been the primary focus of your research?
Barr: My focus has been on really amplifying the voice of patients and their families in healthcare through better patient-centered communication and using digital tools to support that. I’ve done that in a variety of ways but principally through the process of shared decision making as one of the major thrusts of my work.
In a nutshell, shared decision making is a collaborative process where healthcare providers and patients work together to make informed decisions about treatment options.
And then the bigger area for me now, the all-encompassing work, has been the Open Recordings research group that I launched in 2015. The Open Recordings team is an incredible group of people I have the good fortune to work with every day. We are a multidisciplinary group of researchers, patients, caregivers, and clinicians working together to understand how to improve the communication of clinic visit information. We also make great coffee for anyone who wants to visit us!
Q: Can you describe what the Open Recordings group does, and why its work is so important?
Barr: Broadly, we study patient-centered communication. Our team has a special interest in the clinical implications and applications of audio and video recording clinic visit interactions. We believe that reviewing recordings can be used to improve how patients and their care partners process information from their visits and use that information to better manage their health at home, where most of the care happens.
Traditionally, patients have received very short summaries of their visits. But they’re just not enough because so much is discussed. Studies have shown that patients forget 40 to 80 percent of the medical information they’ve been given within 15 minutes of leaving the doctor’s office, so having a recording of the visit can be immensely valuable. We really want to understand what’s happening in the visits, what’s important to the person, and how we can share this information in a comprehensive and helpful manner.
With the emergence of artificial intelligence (AI) and the various benefits it can offer, we’ve been exploring how to use new technologies like AI to analyze visit interactions in ways that can better support patients, their loved ones, and clinicians. For example, we have developed a platform, HealthPAL, where we use AI to highlight key visit information from clinic visit recordings for patients and provide links to trustworthy resources. We have also been exploring whether we can automate the assessment of patient-centered communication through the analysis of visit conversations.
Q: Are there some specific areas of care that your research has focused on?
Barr: My colleagues and I have worked in a number of different areas, such as diabetes, co-occurring disorders, older adults, and dementia. For example, we have two large clinical trials across the U.S., supported by the National Institute on Aging and the Patient-Centered Outcomes Research Institute (PCORI), in which we’re looking at the impact of patient visit recordings on the care of older patients.
While those trials will be key in helping us determine if recordings improve patient health outcomes, we’re very encouraged by the results of other research we’ve been conducting. We recently published a multisite study on the feasibility of sharing digital audio recordings of clinic visits online with older adults in primary care settings, where a lot of the focus of my research has been.
The main purpose of that project was to gauge older adults’ acceptance of visit recordings and to determine if they could be integrated into routine care. We found that among those who received audio, 85 percent logged into their recordings and the retention rate over a three-month period was 98 percent. This not only signified that there was a real interest from older adults in accessing their recordings, but also our team’s commitment to engaging them in the design of the trial from the outset. We also found that patients who received visit recordings had higher mental health scores after three months compared to those who received traditional care.
In collaboration with data scientists, our team plans to integrate more AI features into our work. The possibilities are endless, but we must be guided by theory and the patient communities we serve. For instance, in a current R01 project funded by the National Institute on Aging, VOiCES3, our team is building on the generative AI features available in HealthPAL to enhance patient-centered communication for individuals living with dementia, their loved ones, and their clinicians. We are working closely with patients and their care partners to identify and optimize aspects of visit communication that lead to better experiences and health outcomes.
Due to my work in visit recording, I was invited to join an international team of researchers, industry leaders, and advocacy groups to develop best practices for using AI in medicine, led by Professor Yuri Quintana from the Division of Clinical Informatics (DCI) Network at Harvard. We have held conferences and published papers on the responsible use of AI in medicine for clinical decision support and patients. I have learned a lot and am eager to learn more and contribute further in this rapidly evolving field.
Q: As you look ahead, what do you hope the impact of your research will be?
Barr: I’m hoping that we can show that it will help patients and families to manage their care better and by doing so improve their quality of life. My colleagues, including Dr. Soroush Vosoughi (Computer Science) and Prof. Bob Gramling (University of Vermont), and I recently wrote an editorial in New England Journal of Medicine AI (NEJM AI), highlighting the emergence of ambient visit recording across the country—we fully expect clinic visit recordings to become routine in the next five years.
We also want to understand the potential harms. As we discuss in our NEJM AI piece, visit recordings offer many wonderful opportunities. Still, with the emergence of AI, some very foreseeable risks exist. Through my research, I want to raise awareness of these risks to protect our patients and clinicians and realize the full potential of AI and recording in medicine.
