A few weeks ago, I was skimming this NEJM paper for UCSF's Division of Geriatrics Journal club on de-prescribing anti-hypertensive medications for older adults in nursing homes. Seemed to make a world of sense. The study found no difference between the deprescribing arm and the usual care arm in mortality, the primary study outcome. I thought, great! So we can deprescribe anti-hypertensives without changing mortality, that must be what the authors concluded.

I was shocked, therefore, to read in the first paragraph of the discussion that the deprescribing arm did not achieve the hypothesized 25% reduction in mortality. What?!? Why would deprescribing be associated with reduced mortality? That's not the main reason or even the first reason I think of for deprescribing. Reducing side effects that impair quality of life, sure. Less pill burden, of course. But prolonged life? Seemed a stretch.

Today we hear from the first author of this study, Athanase Benetos, an esteemed geriatrician-researcher from France. For context, we also interviewed Mike Steinman, co-chair of the Beers criteria and co-lead of the US Deprescribing Research Network.

We learned about:

• Why the hypothesis of reduced mortality in deprescribing was justified, based on natural decreases in blood pressure with aging, and the Partridge study, an observational study that found higher risks of mortality associated with using multiple anti-hypertensive and low blood pressure.

• Why mortality was chosen as the primary outcome.

• Is a negative superiority study the same as what they might have found in a non-inferiority study? (stay with us)

• Variation in outcome by frailty status

• How to place this study in context with other research, such as the Danton study mentioned on a recent podcast about deprescribing near the end of life. Dantos was a study of deprescribing for nursing home residents with dementia that was stopped early due to safety concerns. Other studies for context include Sprint, Optimize, and an observational study by Bocheng Jing (UCSF statistician in our group).

At the end, we ask our guests to put it together. With all that we know at this point, what's a clinician to do? To deprescribe or not to deprescribe?

And, zoot alors! I get to sing Hymne A L'amour in French! Athanase recounts the moving story of how Edith Piaf sang this song the night she learned the man she loved, Marcel Cerdan, died in a plane crash.

-Alex Smith

It's an era of breakthroughs in Alzheimer's research, yet for many clinicians, it's also a time of profound uncertainty. We are currently navigating competing definitions of the disease, multiple new biomarkers coming on market seemingly every week, and the clinical rollout of new amyloid antibodies.

How do we translate this rapid-fire science into daily practice? On this week's GeriPal podcast, we sit down with dementia experts Halima Amjad, Barak Gaster, and Heather Whitson. We dive deep into:

• The evolving definitions of Alzheimer's disease. Does someone have Alzheimer's disease if you have only an abnormal biomarker as defined by the Alzheimer's Association, or is amyloid pathology necessary but not sufficient to define Alzheimer's as per the International Working Group (IWG) recommendations?

• Where do blood-based biomarkers for Alzheimer's fit into the diagnostic workup, and should they be used at all in primary care? FYI - here is my take on that question in a recent JAMA IM article titled "The Limited Role of Alzheimer's Disease Blood-Based Biomarkers in Primary Care."

• What's the role of amyloid antibodies in the care of individuals with Alzheimer's disease, including who to use them on?

We covered a lot and discussed some of these resources that you can do a deeper dive on:

• Blood-based biomarker resources

• • JAMA article on Blood-Based Biomarkers for Alzheimer's Disease: Preventing Unintended Consequences

  • Alzheimer's Dementia article on Blood-based biomarkers for detecting Alzheimer's disease pathology in cognitively impaired individuals within specialized care settings: A systematic review and meta-analysis

  • JAMA IM article on The Limited Role of Alzheimer Disease Blood-Based Biomarkers in Primary Care

• Appropriate use recommendations for amyloid antibodies

• • Donanemab: Appropriate use recommendations

  • Lecanemab: Appropriate Use Recommendations

• Primary Care Resources

• • Cognition in Primary Care program

  • A JAGS article on "Large Health System Quality Improvement Intervention Providing Training and Tools to Improve Detection of Cognitive Impairment in Primary Care"

• Other resources

• • AGS's new online curriculum for Alzheimer's Disease

By Eric Widera

Today we're doing something different. Today, dear listeners, you get two podcasts for the price of one! (OK, our podcasts are both free, but you get the idea).

We're joined today by Chris Comeaux, host of TCN Talks, a podcast about leadership, strategy, innovation, and the future of serious illness care, and author of The Anatomy of Leadership. We are also joined by TCN Talks' frequent guest host Cordt Kassner, CEO of Hospice Analytics, which provides in depth data on hospice quality, utilization, and access, and publisher of Hospice and Palliative Care Today, a daily email about the hottest stories and news in the field.

This is an "ask us anything" style podcast in which we get to ask each other questions. Our discussions focus on concerning trends in hospice, Ira Byock's white paper, concerning trends in hospice, certificate of need, danger of losing a generation of junior researchers and hope in the form of ASCENT, various measures of hospice quality including Cordt's National Hospice Locator, which ranks all area hospice by quality, unlike CMS's Hospice Care Compare, which only has star ratings for about 30% of hospices.

Hospice and palliative care are going through a tough growth period, and sometimes being real with your friends and colleagues in your field means tough love. Love hurts. And no, I'm not attempting the Nazareth version!

-Alex Smith

Philippe Pinel remarked in 1800 that "It is an art of no little importance to administer medicines properly, but it is an art of much greater and more difficult acquisition to know when to suspend or altogether to omit them." This insight remains profoundly relevant today, especially in hospice care, where inappropriate prescribing is a common issue. Studies show that 20%–70% of hospice patients receive at least one unnecessary medication near the end of life, including drugs like antihypertensives, statins, and vitamins.

In this episode of the GeriPal Podcast, we tackle the pressing topic of deprescribing at the end of life with expert guests Jennifer Tjia, Jon Furuno, and Simon Mooijaart. The conversation focuses on identifying medications that should almost always be discontinued—such as statins, osteoporosis meds, finasteride, and vitamins, which offer minimal benefit for patients with limited life expectancy. We also delve into more nuanced cases, such as antithrombotics, which present complex decisions that challenge clinicians, particularly when prognosis spans the many weeks to months range.

Finally, we explore practical strategies for engaging patients and families in deprescribing conversations. Our guests highlight tools such as the FRAME mnemonic (Focus on the goals of care, Review current medications, Assess each medication's risk/benefit, Minimize the medication burden, and Evaluate regularly) and the Goal Concurrent Prescribing tool, which helps ensure medication decisions align with patients' values and end-of-life priorities.

By: Eric Widera

Other resources discussed in the podcast

• Prevalence and Factors Associated With Receiving a Prescription for Antithrombotic Therapy on Hospice Admission," JAGS. 2025

• Discontinuation of Anticoagulants and Occurrence of Bleeding and Thromboembolic Events in Vitamin K Antagonist Users with a Life-limiting Disease. 2025

• Effects of the discontinuation of antihypertensive treatment on neuropsychiatric symptoms and quality of life in nursing home residents with dementia (DANTON): a multicentre, open-label, blinded-outcome, randomised controlled trial. 2024

• Perspectives on deprescribing in palliative care. Expert Review of Clinical Pharmacology. 2023

• Developing a decision support tool for the continuation or deprescribing of antithrombotic therapy in patients receiving end-of-life care: Results of a European Delphi study. Thrombosis Research. 2025

• Human-Centered Design Development and Acceptability Testing of a Goal Concordant Prescribing Program in Hospice. JPM 2025

• Reduction of Antihypertensive Treatment in Nursing Home Residents. NEJM 2025

Do you think your hospital should allow unilateral DNR orders? Under what circumstances? Through what process? Do you think that when you obtain the assent of a family to not code their loved one, that assent DNR should be counted as a unilateral DNR order? Should we document unilateral DNR and the rationale? Why for DNR, when we don't document unilateral dialysis not offered, or unilateral no ECMO offered? Is the assent of a family member to a statement that we will not code their loved one a nudge, and is the assent approach ethical? Reasonable people will disagree, as we do on this podcast.

Our guests today are Gina Piscitello, Erin DeMartino, and Will Parker, authors of a terrific viewpoint in JAMA about the need to address inadequate documentation of unilateral DNR orders. You might recall Gina was a guest on our lively podcast about slow codes, and we pick up where that podcast left off.

We highlight the many clinical, practical, and ethical issues at stake, including Gina's finding that during Covid, 3% of critically ill patients receiving pressors had a unilateral DNR order. Black patients and those who spoke Spanish had higher rates of unilateral DNR. That variation should trouble those in favor of unilateral DNR orders. We talk about variation Gina found at the state and health system level, and what exactly is concerning, the variation itself, or the lack of thought and care that went into some of these policies.

Are you a heartbreaker? Dream Maker? Love taker? Don't you mess around with me. (song hint)

-Alex

The idea of embedding various forms of non-emergency care in the emergency department makes a WORLD of sense. If an older adult comes into the ED with a fall, the minimum the ED has to do is address the fall injury and send them out. But many emergency providers realize this is often a band aid. They see that patient again the next time they fall. And again. And again. The same could be said for the patient who is malnourished and dehydrated and admitted for "failure to thrive," again. And again.

Our two guests today, Liz Goldberg and Lauren Southerland, both emergency medicine physician-researchers, have had enough. On our podcast today they discuss how these sorts of experiences led them to argue that other services that can address the underlying causes that lead to ED visits. Liz Goldberg developed the GAPcare model to address falls, which includes a physical therapist and pharmacist seeing patients on the spot in the ED. Lauren Southerland got Columbus Ohio Office of Aging staff to re-locate from their desks to the emergency department, where they could sign patients up for home delivered meals, medical transportation, adult day services, home modification such as grab bars, and utility assistance for electricity, gas, and water bills.

With GAPcare, Liz saw a 66% drop in ED visits for fall over 6 months from her pilot (subsequent fall outcomes of the GAPcare II study will be linked here when published). Remarkable, particularly in the context of the primary care STRIDE intervention, which did not reduce injurious falls (e.g. the type that would result in an ED visit). Maybe the ED is just a better place to intervene? Patients are motivated to change. Get the physical therapist and pharmacist in there!

In a study published in JAGS, Lauren found 50% of participants were linked to a new Office of Aging service initiated during the ED visit, with no increase in ED length of stay or hospital admission rate. See also this terrific JAGS editorial on Lauren's paper by Liz. Putting on my JAGS editor hat - both the study and editorial have terrific color figures. A great way to increase your odds of review and acceptance at JAGS is to include one or more high-impact color figures that convey the main findings or points of your manuscript.

We talk about the potential downsides, real and perceived in embedding care in the ED. Should everything be embedded? We talk about how these interventions relate to geriatric ED certification. Lauren talks about a remarkable model in Australia that includes a geriatric RN embedded in the ED.

Most encouraging is that Liz and Lauren are finding other adopting these interventions. Word is spreading. Other emergency providers have had enough of the endless cycle. Enough.

And I got to belt out Gravity, by John Mayer!

-Alex

Today we interviewed Bob Wachter about his book, "A Giant Leap: How AI Is Transforming Healthcare and What That Means for Our Future." You may recall we interviewed Bob in April 2024 about AI, and at that time he was on the fence about AI - more promise or more peril for healthcare? As his book's title suggests, he's come down firmly on the promise side of the equation. On our podcast we discuss:

• Why Bob wrote this book, at this time, and concerns about writing a static book about AI and Healthcare, a field that is dynamic and shifting rapidly. He's right though - we've not had a "ChatGPT"-launch type moment recently.

• Top 5 or so ways in which Bob uses AI for work, from clinical care to book writing

• Concerns about job losses in healthcare, and will we still need doctors?

• AI for diagnosis, and the recent NEJM Clinical Case in which recent GeriPal guest and superstar clinician-educator Gurpreet Dhaliwal beats an AI.

• UpToDate vs OpenEvidence

• Trust issues - should we trust AI after being let down before? Clinicians felt burned by their experience with the hype and promise of EHRs - but they've been much less a game changer and much more a soul sucking chore designed to maximize billing rather than improve patient care. Yet early returns on AI have largely been positive. Time saved from writing notes, prior authorizations, and summarizing charts…all to the good!

Sadly, we didn't have Bob on piano singing the song for this one. He was in the office, not home. So I made do with ChatGPT's choice, Handle With Care, which has some surprisingly pertinent lyrics about AI in healthcare, including:

"Been beat up and battered around
Been sent up, and I've been shot down
You're the best thing that I've ever found
Handle me with care"

Enjoy!

-Alex Smith

Recent randomized controlled trials have shown that routine perioperative palliative care does not improve outcomes for patients undergoing curative-intent cancer surgery. No, that wasn't a typo. Regardless of how the data were analyzed, the findings remained consistent: perioperative palliative care DID NOT improve outcomes in the only two randomized controlled trials conducted in this area—the SCOPE and PERIOP-PC trials.

Null trials like these often receive less attention in academic and clinical settings, but they can be profoundly practice-changing. Consider the Shannon Carson study on palliative care for chronically critically ill patients. While some have argued it "wasn't a palliative care study," I've always regarded it as one of the most significant studies for understanding not what works—but what doesn't—for palliative care in specific patient populations.

The same holds true for the SCOPE and PERIOP-PC trials. Both were null, but their findings are deeply relevant to clinical practice. That's why we invited the lead authors, Rebecca Aslakson (PERIOP-PC) and Myrick "Ricky" Shinall (SCOPE), to share insights into what they did in their studies and why they think they got the results that they did.

One key takeaway for me from this discussion was the idea that patients undergoing curative-intent surgery might simply be too early in their cancer trajectory to derive meaningful benefits from palliative care, and maybe the focus should be more on geriatrics. I especially appreciated the closing discussion about the future of research in this area: if routine perioperative palliative care doesn't improve outcomes, what should the next generation of studies focus on?

Eric Widera

Studies we talk about during the podcast

• Aslakson et al. Effect of Perioperative Palliative Care on Health-Related Quality of Life Among Patients Undergoing Surgery for Cancer: A Randomized Clinical Trial. JAMA Netw Open. 2023

• Shinall et al. Effects of Specialist Palliative Care for Patients Undergoing Major Abdominal Surgery for Cancer: A Randomized Clinical Trial. JAMA Surg. 2023

• Carson et al. Effect of Palliative Care–Led Meetings for Families of Patients With Chronic Critical Illness: A Randomized Clinical Trial. JAMA. 2016

• Holdsworth et al. Patient Experiences of Specialty Palliative Care in the Perioperative Period for Cancer Surgery. JPSM. 2024

• Williams et al. Patient Perceptions of Specialist Palliative Care Intervention in Surgical Oncology Care. Am J Hosp Palliat Care. 2025

• Yefimova et al. Palliative Care and End-of-Life Outcomes Following High-risk Surgery. JAMA Surg. 2020