Dr. Mafi is an assistant professor of medicine in the Division of General Internal Medicine and Health Services Research at the David Geffen School of Medicine at UCLA where he also practices and teaches primary care. He also serves as an affiliated natural scientist in Health Policy at RAND Corporation. Dr. Mafi completed his undergraduate studies at Northwestern University and then went on to complete medical school at Case Western Reserve University. He then finished his internal medicine residency training at Beth Israel Deaconess Medical Center in 2012, where he also served as chief medical resident in 2013-2014. Most recently, Dr. Mafi completed the Harvard Medical School Fellowship in General Internal Medicine and Primary Care at Beth Israel Deaconess Medical Center and earned his MPH at the Harvard T.H. Chan School of Public Health in 2015.
Dr. Mafi’s research focuses on quality measurement and improvement and how electronic health records can influence the value of care. He has led several national studies assessing the epidemiological trends and predictors of potentially harmful or “low-value care.” He has also studied the impact of electronic health record innovations such as OpenNotes, an initiative where doctors invite their patients to read their visit notes online. He is currently leading several national follow-up studies and working to leverage electronic health records to more effectively measure and improve the value of U.S. healthcare delivery.
Throughout my medical and research training, my interest in improving the quality and value of care has grown and developed into three separate but related trajectories: (1) describing the epidemiology and national trends of low quality or low-value care, (2) evaluating predictors of low quality or low-value care, and (3) assessing the impact of electronic health records on the quality and value of care.
Describing the epidemiology and national trends of low quality or low-value care
Because an estimated one third of health spending is considered wasteful, identifying examples of low-value care will be critical to ongoing efforts to safely reduce the costs of care. Working with my fellowship mentor Dr. Bruce Landon, I have spent the past 5 years publishing a series of national analyses that have important implications for U.S. health policy and the quality and value of healthcare. Most recently, I used an all-payer-claims database in Virginia to find that among 44 low-value health services, the majority of health spending is due to high-volume, low-cost services (e.g., medications, lab testing). These findings suggest that minor actions by all clinicians in the aggregate can make a sizable impact on reducing unnecessary healthcare spending. This work is currently in press at Health Affairs.
Evaluating the predictors of low quality or low-value care
I have expanded my research in the provision of low-value care by evaluating important structural predictors of low-value care, such as specialty, provider-type, and practice setting. For example, our recent national analysis revealed that hospital-based primary care practices provide more low-value care than office-based practices, raising concerns about the efficiency of hospital-based primary care. This manuscript was recently published in JAMA Internal Medicine.
Evaluating the impact of electronic health records to improve the quality and value of care
In my work on OpenNotes, I found that electronic patient reminders are critical for enhancing patient engagement. I discuss the implications of these findings in the Joint Commission Journal of Quality and Patient Safety. Working with my UCLA and RAND mentors, I also evaluated a multidisciplinary QI intervention (leveraging EHR data) to reduce low-value pre-op testing for older adults undergoing cataract surgery, and I presented these results in an oral presentation at the 2017 SGIM Annual Meeting, which won the SGIM 2017 Outstanding Quality/Patient Safety Oral Presentation Award. The manuscript for this work is in preparation.
UL1TR001881 (PI: Steve Dubinett) 07/01/2017 – 06/30/2020
Leveraging Electronic Health Records to Develop the Next Generation of Quality Measures: eMeasures of Low-Value Care
The goal of this project is (1) to develop eMeasures of low-value care at UCLA Health System. This includes (a) evaluating candidate eMeasure feasibility, clinical importance, and financial impact, and then selecting three to five eMeasures for development, (b) developing eMeasure specifications, and (c) testing eMeasure reliability and validity. The next aim will be (2) to identify important structural correlates of low-value care to identify opportunities where interventions could be targeted to reduce low-value care. This important bridge funding will help begin the important work outlined in this proposal until I receive a foundation or federal career development award, such as the AHRQ K08 Award for greater and longer-term support.
Innovations Grant (PI: Catherine Sarkisian) 07/01/2015 – 07/01/2018
American Board of Internal Medicine Foundation (ABIMF)
“Implementation and Evaluation of a Choosing Wisely Intervention in Los Angeles County to Reduce Unnecessary/Potentially Harmful Care”
We received a grant from the ABIMF to implement, evaluate and disseminate an intervention in order to reduce by 20% utilization of three treatments discouraged by Choosing Wisely: 1) imaging for nonspecific low back
pain; 3) pre-operative testing; 3) antibiotics for viral-based upper respiratory illness.
Innovations Grant: (PI: Michael Parchman) 11/01/2016-11/01/2018
Group Health MacColl Center and Robert Wood Johnson Foundation
“Taking Action on Overuse”
Dr. Mafi and his team will work with the MacColl Center for Health Care Innovation at Group Health Research Institute and the Robert Wood Johnson Foundation (RWJF) to build experiential evidence to support the usefulness of the Taking Action on Overuse Framework. They will also participate in ongoing evaluation efforts, share learnings to improve the Framework, and assist with the development and design of an accompanying “change package” and help MacColl Center build upon learnings to spread the Framework and its use as widely and rapidly as possible.