From the Editor
As psychotherapies have become increasingly more practical and relevant in recent years, we may ask: could a focused therapy help individuals who are suicidal?
In a new JAMA Psychiatry paper, Craig J. Bryan (of the Ohio State University) and his co-authors attempt to answer that question, reporting on a randomized clinical trial involving military personnel and veterans. 108 participants were offered brief cognitive therapy (BCBT) or another psychotherapy, present-centred psychotherapy (PCT), building on past work that has shown the potential of BCBT for those who are suicidal. “This randomized clinical trial found that BCBT reduced suicide attempts among US military personnel and veterans reporting recent suicidal ideation and/or suicidal behaviors compared with an active comparator.” We consider the paper and its implications.
In the second selection, from JAMA Internal Medicine, Dr. Daniel J. Morgan (of the University of Maryland) and his co-authors, ask what physicians can do to prepare for generative AI. They offer several useful suggestions. “All physicians will need to understand the basics of GenAI to practice medicine in the next decade. Those without this understanding may find themselves burdened by archaic workflows or responsible for errors that GenAI could have prevented.”
And in the third selection, Dr. Daniela J. Lamas (of Harvard University), an intensivist, looks at transplantation and those who have mental disorders. In a New York Times essay, she notes an historic bias against such individuals. Still, she wonders about the difficulties of the area. “Transplant is one of the most fraught decisions in medicine…”
DG
Selection 1: “Brief Cognitive Behavioral Therapy for Suicidal Military Personnel and Veterans: The Military Suicide Prevention Intervention Research (MSPIRE) Randomized Clinical Trial”
Craig J. Bryan, Lauren R. Khazem, Justin C. Baker, et al.
JAMA Psychiatry, 8 October 2025 Online First
Since 2001, suicide rates increased in the US by more than 30%. Among military personnel, suicide rates increased by nearly 50% during that same period. Half of military suicide decedents access mental health care in the months preceding their deaths. Mental health settings, therefore, present a critical opportunity to deliver suicide-focused interventions. Meta-analyses support the efficacy of outpatient cognitive behavioral therapies for reducing suicide attempts, with larger effects for therapies that target suicide risk directly, relative to treating psychological disorders in general. In the military, a previous randomized clinical trial (RCT) found that brief cognitive behavioral therapy (BCBT) for suicide prevention, a 10- to 12-session individual therapy protocol, significantly reduced the proportion of patients who attempted suicide vs treatment as usual (TAU) among active-duty US military personnel reporting recent suicidal ideation and/or suicide attempts…
Previous studies of BCBT and other suicide-focused treatments in military and veteran settings used TAU as the comparator, limiting the ability to draw conclusions about the treatment components that are essential for reducing suicide attempt risk.
So begins a paper by Bryan et al.
Here’s what they did:
- They conducted a two-arm, parallel, randomized clinical trial comparing BCBT, a psychotherapy that teaches emotion regulation skills, and PCT, a problem-solving psychotherapy, from 2020 to 2025.
- Three US-based outpatient psychiatric clinics were involved and “included US military personnel and veterans reporting suicidal ideation during the past week and/or suicidal behavior during the past month who were either self-referred or referred by their mental health clinicians.”
- Participants were randomly assigned to BCBT or PCT. The BCBT was “a 12-session psychotherapy that teaches patients how to use self-regulation strategies to manage emotional distress and change extreme negative thoughts and beliefs that sustain long-term vulnerability to suicide.” And the PCT was “a 12-session outpatient psychotherapy that provides psychoeducation about the typical symptoms and features of suicidal crises, normalization of symptoms, provision of emotional support and feedback about life problems and stressful situations, and positive interpersonal interactions.”
- The primary outcome: time to first suicide attempt, defined as “a deliberate, self-directed, and potentially injurious behavior enacted with the expectation or intent to die…”
Here’s what they found:
- Of 154 individuals assessed for eligibility, 108 were enrolled.
- Demographics. The mean age was 32.8 years with the majority being male (73.1%).
- Suicide attempts. Fewer patients receiving BCBT (5.6%) than PCT (27.9%) attempted suicide during follow-up. (!)
- Time. Mean time to first suicide attempt was 638.6 days in the PCT group vs. 755.9 days in the BCBT group.
- Hazard ratios. BCBT significantly reduced the risk of any suicide attempt (hazard ratio, 0.25) as well as the rate of follow-up suicide attempts (0.06 vs. 0.18 attempts per participant-year, risk ratio, 0.24).
- Suicidal ideation. It was significantly decreased in both groups.
A few thoughts:
1. This is a good study in a great journal.
2. The main finding in a sentence: “This RCT found that BCBT reduced the risk of suicide attempts among military personnel and veterans with recent suicidal ideation and/or suicidal behaviors and was more effective than an active psychotherapy that has been shown to reduce suicidal ideation.”
3. The study replicated an earlier RCT – showing a reduction in suicide attempts – but here the comparison was with an active control. (!)
4. Nice.
5. Psychotherapies continue to evolve and improve. Past work, for instance, has shown that BCBT is helpful for inpatients.
6. That said, the dropout rate was high, about 40%. Would a more intensive therapy (say, two or three times a week) have been better?
7. Like all studies, there are limitations. The authors note several, including “conclusions based on the present study may still be limited to military personnel and veterans.”
The full JAMA Psychiatry paper can be found here:
https://jamanetwork.com/journals/jamapsychiatry/article-abstract/2840019
Selection 2: “How Physicians Can Prepare for Generative AI”
Daniel J. Morgan, Adam Rodman, Katherine E. Goodman
JAMA Internal Medicine, 13 October 2025 Online First
Generative artificial intelligence (GenAI) large language models (LLMs) have rapidly integrated into everyday life, from internet search engines to chatbot customer support. Clinical medicine is no exception: early-adopter health care systems use LLMs to draft prior authorization letters and patient portal messages, and at many practices, artificial intelligence (AI) scribes have become routine.
Despite the visibility of these innovations, these first-generation GenAI tools represent the tip of the iceberg. By focusing primarily on documentation tasks that contribute to physician burnout, they differ fundamentally from the second-generation GenAI tools currently under development. These are expected to make day-to-day care better by assisting or outright performing cognitively complex clinical tasks, such as history taking and diagnosis. Some of these tools will eventually be integrated into physician workflows, such as clinical decision support with AI scribes. Others will be marketed directly to patients, circumventing physicians entirely.
So begins a paper by Morgan et al.
They make several suggestions for physicians; here, we focus on four.
Approach AI as a Partner, Not Merely as a Tool
“LLMs can be capable, fast, and articulate but still require clinician oversight in their current form.” In particular, the authors note that the questions asked are important. “Prompting is critical to current medical chatbots. Minor changes and different LLMs can alter responses substantially, which physicians can gain a feel for by rewording prompts and observing how outputs change. Some data suggest that chatbots provide higher-quality responses when given an identity or role, for example, ‘You are an emergency medicine doctor…’”
Clinicians Are Responsible for GenAI Output (for Now)
“The US Food and Drug Administration has not vetted the performance of documentation assistance tools such as AI scribes, and any erroneous information in an AI-generated note is the responsibility of the clinician who signed it. As the legal and regulatory landscape around clinical GenAI unfolds, physicians should be cautious and request evidence of tool accuracy and safety.”
Expect the Physician-Patient Relationship to Evolve
“GenAI has demonstrated remarkable technical skills for clinical medicine that exceed many physicians, including synthesizing the medical literature to inform clinical decision-making. These abilities will require clinicians to cede some authority while re-emphasizing particularly human aspects of care such as physical examinations, clinical judgment, and shared decision-making. Additionally, patients are already using GenAI to research symptoms and potential diagnoses and interpret clinical results. Physicians should anticipate that within the next few years, many patients will arrive with LLM-generated diagnostic suggestions or information. This will require humility as physicians see their role change from all-knowing medical expert to expert adjudicator.”
Develop Foundational Knowledge
“It is difficult to predict the exact knowledge necessary for GenAI. The foundational competencies include the basics of how LLMs work, facilitated hands-on experience, and training on ethical issues, limitations, and failures of GenAI. Institutions like NYU Langone and Harvard University now have this training. Likewise, professional societies like the American College of Physicians have developed short courses, which offer efficient training for those already in practice.”
A few thoughts:
1. This is a timely paper.
2. They make several strong points, including the implications of generative AI for the doctor-patient relationship.
3. AI has been considered in past Readings. In a recent selection, Dr. Kumara Raja Sundar (of Kaiser Permanente Washington) commented on patients who already have the answers. You can find that paper here:
The full JAMA Internal Medicine paper can be found here:
https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2840192
Selection 3: “Who Deserves an Organ? The Dilemma of Severe Mental Illness.”
Daniela J. Lamas
The New York Times, 12 October 2025
You walk into a room in the intensive care unit and find a man in his 30s, skin yellow and abdomen swollen with the fluid that accumulates in liver failure. He smiles, even though he is almost always uncomfortable, even though he has not left the hospital in more than a month. Family members sit anxiously at the bedside. The nurses stop by just to say hello.
Without a liver transplant, he will die. Imagine that you are on the committee charged with deciding who gets a transplant. Would you add him to the list of patients waiting to receive organs?
Now open the chart. This is what you’ll see: an unemployed man with a history of untreated anxiety and depression and a recent descent into alcohol abuse. Months of binge drinking have destroyed his liver.
So begins an essay by Dr. Lamas.
Dr. Lamas notes the complexity of the task since the transplant committee needs to determine if “the patient will realistically comply with the rigorous posttransplant regimen and remain healthy enough for a transplant to be worthwhile.” She writes: “My hospital declined to add this patient – my patient – to the waiting list for a transplant, but another hospital said it would consider doing so. Transplant committees regularly face complicated cases like these, and they also regularly come to different conclusions. These decisions can be especially thorny when the patient has a mental illness. Whom we choose to list for a limited resource, and whom we choose not to, reveals our biases about who is worthy of lifesaving care.”
She draws on a difficult case from residency. “I was on my I.C.U. rotation when we admitted an older teenager who was in liver failure because of a Tylenol overdose. I had never seen anyone that sick. Her blood would not clot. As the toxins that her liver could no longer filter accumulated, she became more and more confused… It was clear that without a new liver, she would die. But was the Tylenol overdose intentional? And, perhaps more important, was it a one-time event or part of a long pattern of psychiatric disease? If this act spoke to significant mental illness and risk of suicide, the transplant team would need to worry about whether the patient could reliably take the medications that would keep her body from rejecting a transplanted liver.”
Dr. Lamas notes the that medically she was a perfect candidate. “But the psychosocial questions gave them real pause. Ultimately, our transplant team decided to list her. She was young, she had a supportive mother, and she would die otherwise.”
Transplantation requires evaluation through two lenses: medical and psychosocial. “From the medical standpoint, the transplant program weighs whether the patient is sick enough that she could die without a transplant, but not so sick with other medical problems (like heart disease, cancer or severe obesity) that she won’t make it through the surgery and recovery.” She adds: “The psychosocial component is murkier, and this is where I have long worried that bias and judgment of social worth come into play. Here, the transplant committee – which often includes a psychiatrist and social worker – evaluates the patient’s mental health and ability to adhere to the daily medication required after transplant, while also looking into her so-called support system – the people who will care for her after the transplant.”
Progress has been made. “Once, a patient with severe mental illness would have automatically been denied a transplant. However, in recent years, the understanding of psychiatric disease and transplants has evolved. Patients with diagnoses like schizophrenia can be transplant candidates if their disease is controlled enough that doctors trust they will take their anti-rejection meds.” Still, there are inconsistencies. “A 2017 paper in BMC Medical Ethics noted that while attitudes in the field were changing, there remained significant variability among institutions.”
Alcohol problems are particularly challenging. “For years, most programs required patients be sober for six months before they would be considered transplant candidates. This served two purposes: It demonstrated whether a patient could stay abstinent after surgery, and it would give a patient’s liver enough time to potentially recover on its own. But for some patients with particularly severe alcohol-related liver damage, waiting six months meant death.”
What happened with the teenager? “She attended a few posttransplant appointments… But she was struggling with her mental health, asking why we had saved her life, expressing hopelessness about her life after transplant. Then she stopped showing up to appointments.”
Dr. Lamas has wondered if the transplant team was wrong. “In such decisions, there is no right or wrong. There is only a thoughtful balancing of risk, acknowledging the biases we bring to these decisions and then the commitment to do whatever can be done to give our patients the very best life possible, given their realities.”
A few thoughts:
1. This is a thoughtful and well-written essay.
2. It’s wrong to deny all patients with mental health problems the care they need (transplantation). But how much should the psychosocial weigh in on the decision making?
3. For another perspective, Mark Kingwell, a philosophy professor at the University of Toronto, wrote a long essay on his problems with alcohol and his resulting liver transplant. With regard to the six-month rule, he explained: “I’m almost certain I would have expired in that time, sober enough to reflect on my sins but not enough to survive them.” That essay can be found here:
https://www.theglobeandmail.com/opinion/article-and-then-every-day-reflections-on-a-life-of-drinking
The full NYT article can be found here:
https://www.nytimes.com/2025/10/12/opinion/organ-transplant-mental-illness.html
Reading of the Week. Every week I pick articles and papers from the world of Psychiatry.
Recent Comments