Week-in-Review | October 8, 2025
Dear DGIM,
Congratulations to Bob Burke, Rachel Werner, Rebecca Brown (Geriatrics), and the VA SAGE team. Our QUERI Program Center was refunded for another 5 years. In this iteration of the Center we will focus on bringing evidence-based age friendly health system practices to the inpatient setting. With this focus on the inpatient setting the program has rebranded to iSAGE (Inpatient Safer Aging through Geriatrics-Informed Evidence-Based Practices). I will let you know how the work proceeds.
Congratulations to Nichole Scaramella. She was named the new PAH APD for Curriculum .
Bob Burke and Rachel Werner have a new article in JAMA Network Open. This paper is the second evaluating the SNF Value Based Payment (VBP) program. In the first paper (as I wrote about previously) they found the program was ineffective in its primary aim - reducing readmissions from skilled nursing facilities. This paper seeks to understand why, and looks at the size of the financial incentives used in the program. They calculated each SNF’s total Medicare fee-for-service revenue, and figured out how large their bonus or penalty was for each of the first three years of the program, finding that the median incentive is just over $10K and highly variable year to year, making it hard for SNFs to plan on having the money. Bob says this may go a long way to explaining why the program didn’t work - 10K is simply not enough to hire a nurse, invest in lab equipment, etc. Bob has presented the results to a CMS representative and learned that for larger incentives to be tested, it would need to be authorized by Congress (i.e. it is unlikely to happen any time soon). The paper has received some attention from within the world of long term care.
Medical student John Lin, Jaya Aysola, and collaborators have a paper in JAMA Network Open looking at long COVID and food insecurity. Using data from the National Health Interview Survey from 2022 to 2023 they find a strong association between long COVID and food insecurity (15% among those with long COVID as compared to 7% among those without it). This association was modified by SNAP receipt and being unemployed. It is not immediately clear why unemployment modifies this relationship between long COVID and food insecurity, but Atheen Venkataramani's research showed that receipt of unemployment insurance was associated with decreased food insecurity during the COVID-19 pandemic, and many food insecure individuals may have been forced back into work. This is yet another important study showing the benefits of food assistance programs.
Matt Klebanoff, Alex Li, Jalpa Doshi and I (Judith Long) have a new research letter in JAMA looking at Medicare Part D plan coverage of GLP-1 receptor agonists from 2020 to 2025, when the Inflation Reduction Act (IRA) was implemented. The law introduced an out-of-pocket (OOP) maximum for Part D beneficiaries, which was financed, in part, by increasing Part D plans’ financial liability for high drug costs. Between 2020 and 2025, coverage for GLP-1s was widespread and over 85% of the plans covered them. During this time prior authorization for these medications went from nearly 0 to 100%. In 2025, there was also a significant jump in OOP costs for these medications, especially in stand-alone plans compared to Medicare Advantage plans. The increases in costs resulted from rising use of co-insurance (a percentage of drugs’ list price) rather than fixed copayments. These changes suggest that Part D plans are responding to their new financial liabilities under the IRA by increasing beneficiary cost-sharing for high-cost drugs, including GLP-1s. This paper is also getting quite a bit of media attention.
Kaustav Shah (Gen Med Fellow) and Kevin Johnson (CHOP) have an invited commentary in JAMA Network Open discussing the AI scribe “revolution.” The invited commentary relates to an article finding AI scribe use was associated with a 30-day reduction in burnout symptoms, cognitive task load, and time documenting (using a variety of different AI scribes). Kaustav feels the findings are exciting, but worries that the 30-day follow-up is not adequate and might be just the bloom of a new rose. Given that time savings, while evident, are not extensive it is unlikley that this is the underlying cause for reduced burnout. Kaustav and Kevin ask: 1. Is the expense worth it (understanding worth it may be less about ROI in terms of patient revenue; 2. What do patients think about the risk; and 3. What are the unintended consequences of using these tools. I had an interesting conversation where David Aizenberg asked when is it the right time to introduce these tools to learners and do they undermine their critical reasoning skills? These are all good questions. As someone who loves my ambient AI I am on the side of those who see it as the first tool I have dealt with that has reduced and not expanded the work load.
LauraEllen Ashcraft (DBEI) and many from the SAGE team including Bob Burke, Rachel Werner, Rebecca Brown (Geriatrics), and I (Judith Long) have an article in BMJ Open Quality. LauraEllen is an implementation science methodologist and this paper describes a novel approach to implementation evaluation that she developed during the first 5 years of the SAGE Program (see above). The SAGE Program was a large hybrid type III implementation-evaluation trial implementing four different evidence-based practices at nine medical centers. Given the diversity of the practices and geography, tracking implementation was complex. The process she developed had seven steps: “1) select a template and identify point person; 2) complete initial development; 3) obtain targeted asynchronous feedback; 4) identify and analyze gaps; 5) conduct targeted virtual synchronous discussion; 6) finalize working document; and 7) apply the plan and solicit ongoing feedback.” This was not a simple process and it is hard to describe succinctly. I suggest you read the article if you want to learn more. However, it is exciting to see new methods develop in real time that address the need to assess complex implementation.
In another interesting methods focused paper Stephanie Teeple (MD-PhD candidate) and Jaya Aysola have a paper in HealthAffairs Scholar evaluating how well synthetic data from a popular open-source EHR data generator identifies racial, ethnic, and sex disparities. The source they evaluate is agent-based (e.g., does not use of any real patient records to generate the data) and what they find is higher rates of completing the care than the existing literature and either no disparities in care or disparities that are much lower than would be expected. They then used the Dartmouth Atlas of Health Care basically to create weights relating to acute MI care and applied them to the synthetic data, finding both rates and disparities more consistent with existing literature. The paper draws attention to the limitations of synthetic data and a clever way to deal with the limitation. I wonder if weights like the ones Stephanie and Jaya developed could be built into open-source synthetic data – making it much more useful for real-life questions.
Faculty Spotlight:
Nichole Smith is an Assistant Professor of Clinical Medicine, splitting her time between the Cooper Clinic and the Philadelphia City Health Center #3. She was born in Chicago, IL, grew up in Atlanta, and attended Vanderbilt University for college before returning to The University of Chicago for medical school. She completed her internal medicine primary care residency at Penn in 2024, and stayed on as the ambulatory chief resident for the 2024-2025 academic year. Her academic interests include medical education, navigating high risk transitions of care, and sustainable community engagement. She is excited to help build an Urban Health Equity Track for internal medicine residents starting summer 2026. Outside of work, she enjoys reading fiction and non-fiction, listening to all kinds of music, and traveling with her husband, Isaiah.
Keep the news coming.
Judith
Judith A. Long, MD
Sol Katz Professor of Medicine
Chief, Division of General Internal Medicine
University of Pennsylvania Perelman School of Medicine
Core Faculty, VA Center for Health Equity Research and Promotion
Corporal Michael J. Crescenz, VA Medical Center