#NOharm
📣 We have results! In this week's PCT Grand Rounds, Andrea Cheville and Jon Tilburt of @mayoclinic.org will present “Boosting Peri-Operative Pain Care: Mixed Results (and Myriad Implications) From the NOHARM Pragmatic Trial.” Join us!

🔗 du.ke/pctnews20261...
November 3, 2026 at 5:06 PM
Thelminha é madrinha de startup que evita erros no SUS
Thelminha é madrinha de startup que usa IA para evitar erros no SUS: 'É tecnologia salvando vidas'
Ex-BBB incentiva doações para NoHarm, ONG de Porto Alegre que usa software para supervisionar tratamento de pacientes
f5.folha.uol.com.br
October 12, 2026 at 2:02 AM
World timekeepers vote on ending the leap second. Thousands of baby bath seats sold on Amazon are recalled due to drowning risks. Brazilian startup NoHarm uses AI to prevent medication errors in the public health system. Montana voters weigh a ballot measure to curb corporate political spending.
October 12, 2026 at 2:14 AM
“EHR-enabled strategies that emphasize patient-centered decision-making and education may reduce opioid exposure while maintaining patient-centered outcomes, reinforcing the ongoing need for patient education to address opioid-related harms,” said Dr. Andrea Cheville.

➡️ duke.is/pctnews20260...
October 9, 2026 at 5:47 PM
Dr. John Ferguson, MD objects to a proposal to license autonomous AI parrots as credentialled physician-like entities, noting software has no moral agency, accountability or ability to refuse unsafe action. NOHARM reported 22% of AI recommendations cause severe harm, despite competency on USMLE.
October 2, 2026 at 3:25 PM
In NOHARM, an NIH Collaboratory Trial newly published in @jamanetworkopen.com, a strategy to promote nondrug pain care after surgery reduced exposure to opioids without worsening pain interference, physical function, and other patient-reported outcomes.

Learn more ➡️ duke.is/pctnews20260...
September 28, 2026 at 6:06 PM
(9/9) Full episode with Marina Mogilko: https://www.youtube.com/watch?v=oOpF0ZXSXuk

State of Clinical AI Report: https://arise-ai.org/report

NOHARM preprint: https://arxiv.org/abs/2512.01241

Mailing list: https://arise-ai.org/contact
September 27, 2026 at 8:13 AM
(6/9) Hallucinations still happen but are much less of an issue than a few years ago.

Harms of omission are the major issue in our NOHARM study. An incidental lung nodule that doesn't make it into the summary could leave cancer undiagnosed.
September 27, 2026 at 8:13 AM
A new article from NOHARM, an NIH Collaboratory Trial, describes strategies and challenges for posttrial sustainment of pragmatic trial interventions. duke.is/pctnews20260...
August 25, 2026 at 2:08 PM
"MAST" presents a suite of clinical AI benchmarks testing safety, diagnostic/management reasoning, agentic EHR tasks and multimodal imaging performance across real medical scenarios. #HealthcareAI #MedicalBenchmarks #AIinMedicine www.arise-ai.org/mast/benchma...
Benchmarks | MAST - ARISE
Explore the MAST benchmark suite including NOHARM, SCT, MedAgentBench, and CPC-Bench for evaluating clinical AI systems.
www.arise-ai.org
August 14, 2026 at 8:36 AM
Du meinst die hier?

arxiv.org/abs/2512.01241
First, do NOHARM: towards clinically safe large language models
Large language models (LLMs) are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm Assessment for Risk in Medicine), a benchmark using 100 real primary care-to-specialist consultation cases to measure frequency and severity of harm from LLM-generated medical recommendations. NOHARM covers 10 specialties, with 12,747 expert annotations for 4,249 clinical management options. Across 31 LLMs, potential for severe harm from LLM recommendations occurs in up to 22.2% (95% CI 21.6-22.8%) of cases, with harm of omission accounting for 76.6% (95% CI 76.4-76.8%) of errors. Safety performance is only moderately correlated (r = 0.61-0.64) with existing AI and medical knowledge benchmarks. The best models outperform generalist physicians on safety (mean difference 9.7%, 95% CI 7.0-12.5%), and a diverse multi-agent approach improves safety compared to solo models (mean difference 8.0%, 95% CI 4.0-12.1%). Therefore, despite strong performance on existing evaluations, widely used AI models can produce severely harmful medical advice at nontrivial rates, underscoring clinical safety as a distinct performance dimension necessitating explicit measurement.
arxiv.org
April 9, 2026 at 4:07 PM
damn iPhone wouldn't sync. forcing a restore…. SUNUVABITCH!!!! #noharm meant since I don't know some of ya mommas name.
November 23, 2024 at 12:08 AM
As a former BDSM Domme, I can agree heartily. It's not necessary to restrict blood flow or breathing to achieve a beautiful bind. Shibari was in its infancy here in the US then, so I'm no expert. I did my best with a macrame book and what little I could find on the internet. #NoHarm #ConsentOnly
February 8, 2025 at 4:21 PM
Witches, if you have been acting for political purposes, I think it may be working. I'm 💙, so I'm celebrating the GOP infighting in Florida. Slow them down. Keep up the good fight. #witchsky #noharm
January 31, 2025 at 1:29 PM
The current system is collapsing — and staying in it means being part of the harm.

#zeroist #zero #noharm #zeroism
May 22, 2025 at 11:05 AM
Does your child need an excuse note for the #ClimateStrike on Friday? Physicians know there is a link between environmental pollutants (such as pesticides) and asthma -- the #1 reason kids miss school. So, they wrote a letter for your kid:

noharm-uscanada.org/documents/medi…
February 26, 2025 at 12:56 AM
FIRST DRAFT:NO HARM

Cat jaw
Furried flat against my paw
Bushy bum buried
Within the indent of my arm
Uncleft, closed clawed, no harm

#blueskypoet #blueskypoetry #blueskypoem #poem #poetry #poet #chapbook #firstdraft #noharm #poetry #author #selfpublished #bushybum #catpoetry #mainepoet #catpoem
February 2, 2025 at 8:20 PM
“First, do NOHARM: towards clinically safe large language models” 31 medical #LLMs using 100 real clinical consultation cases show risks of harmful recommendations, especially omission errors, while multi-agent #AI systems improved #PatientSafety.
First, do NOHARM: towards clinically safe large language models
Large language models (LLMs) are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm…
arxiv.org
May 24, 2026 at 4:50 PM
Population Health - #Ohio - Healthcare providers join advanced #energy companies to urge state lawmakers to support #cleanenergy:#renewable #renewableenergy #pollution

noharm-uscanada.org/articles/news/…
December 2, 2024 at 2:34 AM
I'm debating in my mind about the protestors on Wall Street and the different Metros...I want so badly for something good to happen. #noharm
February 21, 2025 at 12:29 AM