ADL CEO Jonathan Greenblatt reflects on Rosh Hashanah: after meeting Yunis Alkarnawi—a Bedouin father who risked his life on Oct 7 to save Jews & others—he writes that Yunis’ unimaginable courage will stay with him forever, a light reshaping how he sees this year. view.adlmail.org?qs=acfff68e6...

Today I will be chairing this webinar on the largest (less important) and successful (WOW!) suicide prevention trial from Australia. events.zoom.us/ev/AogXbbUoI...

Are they going to learn the first lesson of psychometrics?

And the only validated prescriptive model of alternative psychotherapies (i.e. based not only on prognostic factors but also on effect modifiers, and enabling differential therapeutics) was also based only on boring psychometric and demographic variables. www.nature.com/articles/s41...

Well, of course, it's not that strange in medicine. For heart failure, we have literally more than 100 prognostic indexes, none of which works very well ;)

Zach, why do people pursue this pipe dream when dozens of studies and millions of dollars have failed to find a replicable HTE subtypes based on imaging and/or genetics?

Yes, I think Dianne and Steve's scheme goes beyond the GRADEd systematic reviews, when they emphasize effectiveness research, heterogeneity in treatment effects and cost-effectiveness. But we set a higher bar for the very evidence base for "empirically supported therapies" today.

This 30-year-old scheme has served its purpose, but replication alone is insufficient today. We need systematic reviews with minimal publication bias, critically appraised by GRADE. Zach and Daniel, my respected BlueSky friends, shall we not teach this to our next generation?

An apparently fully automated generative AI analysis of our paper on precision psychotherapy (www.nature.com/articles/s41...) popped up in my mailbox. I think I am allowed to cite it here: bgpt.pro?q=Paper%20Re... (if not, pls let me know, BGPT!). It's amazingly accurate.

Very cool! @depressionlab.bsky.social does relevant work in this area 😊

#ClinicalPsychology, #Psychotherapy, #counselling

We believe this is the first-ever system to make evidence-based personalized psychotherapy recommendations about what to do for whom. Or does anyone know such a precedent? #Psychiatry #MentalHealthResearch #AcademicSky

Psychotherapists have claimed for >100 years that they provide/adjust interventions in view of the client's characteristics -- but all based on their personal experiences and opinions. Here we have the first-ever system to do so based on evidence.

Evidence-based personalized & optimized CBT skills for depression, based on a trial of 4469 adults. The Resilience Training App will augment the efficacy by 35% overall over the group average best, and by 64% for those whose best option wasn’t this group average best! www.nature.com/articles/s41...

@ryuheiso.bsky.social This large scale pragmatic cluster randomised controlled trial found no evidence to support the effectiveness of ultra-brief intervention in reducing alcohol consumption compared with simplified assessment only in primary care settings in Japan. bmj.com/cgi/content/...

« Weak campaign against predatory journals: structural issues » New commentary with Alain Braillon and Constant Vinatier. ijme.in/articles/wea...

Yep, but that’s a wrong way to compare treatments. The treatment effect is always a between-group matter.

How can this be? itnewsletter.itnewsletter.co.il/sending/webp...

2025 Youth Peace Camp of Bereaved Israeli and Palestinian teens mailchi.mp/theparentsci...

Really nice opinion in JAMA-Psychiatry about the challenges of #mediation modeling. The issues, such reverse causation and the need for longitudinal data, are clearly laid out. #AcademicSky pubmed.ncbi.nlm.nih.gov/40266608/

As if we could ever know that… Could someone please calculate the sample size necessary to answer these questions with confidence?

We published the MIC (Minimal Important Change. a.k.a. MID or MCID) of the Gambling Symptom Assessment Scale (G-SAS). A small study but we need more of these to make sense of the research literature. www.tandfonline.com/doi/full/10....

Excellent and important paper on the efficacy and safety of Esketamine 👇

O, no, no, NO!! You should have a model in your statistical analysis plan based on your prior knowledge before you analyze the data. This is statistics going against science!

One net result of which was: my eldest son believed in Santa Claus until the age of 13, when my youngest son knew that it was a fake story at the age of seven. Which is better?

Excited to share new work in @natmed.nature.com! 🌟 A global collaboration between Kyoto University and d3center director Inbal Billie Nahum-Shani explores smartphone-based CBT for subthreshold depression using an innovative factorial design. myumi.ch/z9ZzZ #BehavioralScience #MentalHealth #CBT

But what good does the knowledge of the "associational variables" do? Non-experimental cross-sectional data is rotten fish. Even the best sushi cook cannot make good sushi with rotten fish. The best you can do is to stay away from it.

It's good to know that there are sane people. itnewsletter.itnewsletter.co.il/sending/webp...

We finally managed to make sense of MASSIVE digital phenotype data gathered through wearables and smartphones. The ML model was able to predict the following week's worsening of depression recurrent depression on maintenance drug therapy with AUC=0.91 (0.89–0.93). doi.org/10.1016/j.ja...

Four water tankers arrived in Ras al-Ain today—a community that has not had access to water for two months. We are proud to support this work, stand with these activists, and be in solidarity with struggling Palestinian communities in the West Bank.

This week, NIF approved an emergency grant for Rabbis for Human Rights to bring water, food, medical equipment, and fuel to Palestinian communities in the West Bank that have been under siege since the start of the war with Iran.

Many Voices, One Prayer ⁠ Interfaith Prayer for Peace, Security and Human Rights www.youtube.com/watch?v=xSss...

I will be presenting the results of the RESiLIENT trial (www.nature.com/articles/s41... up to 26 wks; now extended to 50 wks) and the design of the BEATRICE platform trial tomorrow at the Int'l Consortium for Depression Prevention (www.groningendepressionprevention.nl/home).

So why do people think SpringerNature started this new series of predatory journals? Sick of having publishing fees sucked by MDPI? Wanted to blemish their hard-earned reputation?

More than >$775 million worth of data lost. Good that they know now. But I can totally understand any skepticism towards science and charities. Please fix a 95% system somehow.

Inspired to keep standing up for science and helping to ensure that new scientific knowledge continues to translate to collective wisdom, we, the #NEATLabs @PsychiatryUcsd co-directors, I and Dhakshin Ramanathan have just started the #NEATCast. Ep1 on #Youtube #Apple #Spotify 🧵

I am surprised (and honestly flattered) by this independent podcast by Drs. Mishra and Ramanathan from UCSD, focusing on our RESiLIENT trial that examined 5 CBT skills among >4000 participants. www.youtube.com/watch?v=_hGe...

We created a video byte for our smartphone CBT trial (Nature Medicine 2025). Hope you will enjoy! www.youtube.com/watch?v=m5jc...

And what was YOUR expectation before starting the trial?

Yes, small steps can go a long way. (And big talk may build dreams, but does not move you an inch.)

Here are some ways the world has gotten better.

Looks akin to contrast between YouTube and TikTok :)