Reposted by Ryan Radecki MD MSCiaran Martin @ciaranm.bsky.social · 29/09/2026I was wrong. It’s happening 222040242
Ryan Radecki MD MS @www.evidencetriage.com · 25/09/2026Perhaps not "snake oil", but a combination of "anxiety oil" and "false reassurance oil", with a dollop of unnecessary invasive procedures. #medsky www.evidencetriage.com/p/here-comes...evidencetriage.comHere Comes the Multi-Cancer PanelThe blood equivalent of full-body MRI. 020
Ryan Radecki MD MS @www.evidencetriage.com · 22/09/2026Fewer than three weeks to go until the Chicago Marathon – if the fight against ALS is personal for you, please consider donating to ALS United Chicago's research fund: secure.alsunitedchicago.org/site/TR/Endu...secure.alsunitedchicago.org 2026 Team ALS - Bank of America Chicago Marathon: Dr. Ryan Radecki - ALS United Greater Chicago 000
Ryan Radecki MD MS @www.evidencetriage.com · 16/09/2026Mildly entertaining study re: AI CT queuing – – if you're the clinician who frequently orders low-yield/low clinical likelihood scans, I wonder if you'd notice the increases in your ED LOS, whilst the most effective clinicians would have deceased ED LOS .... pubs.rsna.org/doi/10.1148/...pubs.rsna.orgSimulation of AI-driven CT Queue Prioritization in the Emergency Department | Radiology: Artificial IntelligenceIn discrete-event simulations, artificial intelligence–driven emergency department CT queue prioritization (reordering by predicted probability of clinically actionable radiologic findings) reduced... 000
Reposted by Ryan Radecki MD MSDr Sarah Hendrica Bickerton 🏳️🌈 @sarahhbickerton.bsky.social · 07/09/2026Imagine if we had wind in Aotearoa New Zealand. *stares at the camera* 117823
Reposted by Ryan Radecki MD MS🇭🇲🐬🦘 @shatterwood.bsky.social · 07/09/2026In Australia, we get 3 hours of free electricity every day because of our solar 1334
Ryan Radecki MD MS @www.evidencetriage.com · 07/09/2026It's your chance to see the @annalsofem.bsky.social podcast team live! in person! And catch up on the last year of the top content from your @acepnation.bsky.social flagship journal. See you in Chicago! 000
Ryan Radecki MD MS @www.evidencetriage.com · 03/09/2026And we're at the "Biff Bucks" point of the timeline. 😑 110
Ryan Radecki MD MS @www.evidencetriage.com · 03/09/2026Sending low/intermediate-risk folks to CT coronary angiograms following an ED visit didn't do much for their CVD outcomes, but fewer people were dead within 3 years ... 🤔 #medsky www.evidencetriage.com/p/keep-ct-co...evidencetriage.comKeep CT Coronary Angiography Out of the Emergency DepartmentStill just a sales pitch without the outcomes to support. 021
Ryan Radecki MD MS @www.evidencetriage.com · 01/09/2026These 0/1-hour troponin algorithms are probably safe, but can we please stop trying to conflate the "ACS" question answered by the test with "30 day outcomes"? #medsky www.evidencetriage.com/p/are-tropon...evidencetriage.comAre Troponin Trials Asking The Right Question?Why are we measuring 30-day outcomes with such interest? 021
Reposted by Ryan Radecki MD MSsamuel mehr @mehr.nz · 01/09/2026This is absolutely insane. Don't pay for-profit publishers APCs. Don't let your collaborators do it either. Post preprints of your work, post your work on pubmed central, but for fucks sake stop sending hard-earned grant dollars to these publishing conglomerates 23113
Reposted by Ryan Radecki MD MSBrent Toderian @brenttoderian.bsky.social · 01/09/2026“When mayors in other cities ask me how #Copenhagen afforded to invest in its cycling networks, I ask them how on earth they have been able to afford highway projects. We invested in bike lanes because that was the cheapest option.” What the world can learn from Copenhagen’s cycling revolution.theprogressplaybook.comCopenhagen Cycling Revolution: Lessons for CitiesClose to two-thirds of the city’s residents commute by bike to school or work every day. 16801293
Ryan Radecki MD MS @www.evidencetriage.com · 30/08/2026Which AI drafts require the most revision? Which take the most time in revision? And which occur with greatest frequency? All these answers and more ... #medsky #MLsky www.evidencetriage.com/p/which-ai-d...evidencetriage.comWhich AI Drafts Ruin Our Inbox TimeThe culprits behind AI causing as much burden as relief. 110
Ryan Radecki MD MS @www.evidencetriage.com · 28/08/2026static.klipy.comStar Trek IV: The Voyage Home - Whales SceneALT: Star Trek IV: The Voyage Home - Whales Scene 182
Reposted by Ryan Radecki MD MSKarandeep(Singh) @kdpsingh.bsky.social · 28/08/2026When physicians start relying on AI-drafted content, how does a health system monitor that human-in-the-loop AI and fix what's broken? Our new @ai.nejm.org study from our team at @jchi-ucsd.bsky.social led by Dr. Lily Poursoltan found hundreds of physicians-hours spent editing AI content. 1104
Ryan Radecki MD MS @www.evidencetriage.com · 24/08/2026No cheating. Share the last photo of the sky in your camera roll 110
Ryan Radecki MD MS @www.evidencetriage.com · 21/08/2026A very interesting peek at a prospective evaluation of robust, integrated AI decision-support – and how the initial engagement rapidly evaporated from 70+% to lower than 25% within the first few weeks of deployment! #medsky #MLsky www.evidencetriage.com/p/when-cutti...evidencetriage.comWhen Cutting-Edge AI Decision-Support FailsClinical workflow acceptability beats clinical utility every day. 122
Ryan Radecki MD MS @www.evidencetriage.com · 21/08/2026In just under two months, I'll be running the Chicago Marathon for the ALS United charity! There are many worthy charities, and if you have the capacity to give, let this serve as a nudge to do so – and consider the fight against ALS as an option: secure.alsunitedchicago.org/site/TR/Endu...secure.alsunitedchicago.org 2026 Team ALS - Bank of America Chicago Marathon: Dr. Ryan Radecki - ALS United Greater Chicago 000
Ryan Radecki MD MS @www.evidencetriage.com · 14/08/2026Supposedly the new pulmonary embolism risk-stratification was dreamt up over drinks (née "expert consensus"), so, luckily, the categories probably do bin people into categories of increased risk. More prospective validation necessary, of course. #medsky www.medrxiv.org/content/10.6... 110
Ryan Radecki MD MS @www.evidencetriage.com · 13/08/2026Patients presenting with syncope sometimes have serious short-term outcomes! These data indicate selected patients might benefit from short-term observation to reduce missed opportunities for diagnosis and intervention, but that number remains small. #medsky www.evidencetriage.com/p/the-benefi...evidencetriage.comThe "Benefit" To Admitting SyncopeJust about none! But not none! 011
Ryan Radecki MD MS @www.evidencetriage.com · 10/08/2026my superpower is instantly forgetting people’s names 010
Ryan Radecki MD MS @www.evidencetriage.com · 10/08/2026Narrator: it did not stay in the box www.rnz.co.nz/news/world/9...rnz.co.nzAI assistant hacks gym website in first known Australian autonomous cyber attackAndrew asked his personal assistant to book him a spot in one of his gym's coveted morning classes. 010
Ryan Radecki MD MS @www.evidencetriage.com · 06/08/2026Three(!!!) jobs popping up here in Christchurch now advertised: 12-month Fellow parental cover: cdhb.careercentre.net.nz/Job/Emergenc... 0.5 FTE Permanent SMO position: cdhb.careercentre.net.nz/Job/Emergenc... 12-month fixed-term SMO position: cdhb.careercentre.net.nz/Job/Emergenc... 021
Ryan Radecki MD MS @www.evidencetriage.com · 05/08/2026In this fascinating pre-print, authors tried to address clinical safety in LLMs by just ... asking nicely for it to apply a cost function penalizing under-triage. And it kind of works! But probably not well enough to rely upon. #medsky #MLsky www.evidencetriage.com/p/llms-can-b...evidencetriage.comLLMs Can Be Commanded To Be Less RiskyIn their own strange, probabilistic way. 012
Ryan Radecki MD MS @www.evidencetriage.com · 04/08/2026It is not "the PERC rule". That is all.static.klipy.comI Have Spoken The MandalorianALT: I Have Spoken The Mandalorian 000
Ryan Radecki MD MS @www.evidencetriage.com · 04/08/2026Can I interest you in an other Epic/ML/RRT fantasy of "hundreds of lives saved" by an alert? #medsky #MLsky www.evidencetriage.com/p/ai-guided-...evidencetriage.comAI-Guided Rapid Response "Reduces Mortality"But what do these interventions actually *do*? 010
Ryan Radecki MD MS @www.evidencetriage.com · 03/08/2026Remember these Texas Medical Center dashboards where they documented how many extra ICUs they opened? 25020
Ryan Radecki MD MS @www.evidencetriage.com · 31/07/2026Which peptide fixes your back Asking for a friend 110
Ryan Radecki MD MS @www.evidencetriage.com · 25/07/2026New addition to the GRADE evidence framework: advertorial When a $180B company has to pay to publish a study from high-profile U.S.-based academic institutions as Open Access in The Lancet Regional Health Americas ... www.thelancet.com/journals/lan...thelancet.comWhole-blood point of care biomarker testing following traumatic brain injuryIn 2018, the U.S. Food and Drug Administration (FDA) first approved a blood test to aid clinical decision-making regarding CT imaging in patients following mild traumatic brain injury (TBI).1 This was... 000
Ryan Radecki MD MS @www.evidencetriage.com · 25/07/2026Let’s take two things that are almost free and make you pay money for them again. 020
Ryan Radecki MD MS @www.evidencetriage.com · 24/07/2026Another example of journals paying reviewers and reaping the benefits. It is a substantial investment on the part of the journal, however, and remains to be seen what sustainable models will ultimately emerge. #medsky www.evidencetriage.com/p/how-would-...evidencetriage.comHow Would You Like to Make £220?(You don't have to review for free!) 000
Ryan Radecki MD MS @www.evidencetriage.com · 22/07/2026It's all well and good to develop a predictive or prescriptive model, but what's the actual health system impact? The real world wants to know. #medsky #mlsky www.evidencetriage.com/p/the-actual...evidencetriage.comThe Actual Consequences of Decision-SupportAI edition! 010
Ryan Radecki MD MS @www.evidencetriage.com · 18/07/2026Stop scrolling and post 2 characters who bring you happiness.static.klipy.comAlex Horne & Greg Davies' HandshakeALT: Alex Horne & Greg Davies' Handshake 120
Reposted by Ryan Radecki MD MSPeter Tennant @pwgtennant.bsky.social · 18/07/2026"LLMs are structurally indifferent to truth" 193
Ryan Radecki MD MS @www.evidencetriage.com · 14/07/2026“Murderbot, I see your performance reliability is dropping. I believe I may have something for that.” cc: @marthawells.com 020
Ryan Radecki MD MS @www.evidencetriage.com · 12/07/2026Yep, the YEARS criteria are safe to use for exclusion of PE in patient with active cancer. #medsky www.evidencetriage.com/p/cancer-can...evidencetriage.comCancer Can't Stop YEARSKeep cutting down those CTPAs with "Hydra". 075
Ryan Radecki MD MS @www.evidencetriage.com · 03/07/2026Can we use LLMs to screen patient safety "eTriggers" for missed opportunities for diagnosis? Yes (probably) – but bland one-shot prompting of general purpose models isn't going to do the trick. #medsky #MLsky www.evidencetriage.com/p/the-llm-as...evidencetriage.comThe LLM As QA JudgeA good low-risk administrative application, but still a ways to go. 011
Reposted by Ryan Radecki MD MSBethany Black @bethanyblack.bsky.social · 27/06/2026🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️ One milkshake to rule the boys, one milkshake to find them, one milkshake to bring the boys and in my yard to bind them 🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️🕯️ 284232498
Ryan Radecki MD MS @www.evidencetriage.com · 26/06/2026Every city's got an ominous megastructure. 030
Ryan Radecki MD MS @www.evidencetriage.com · 24/06/2026Conceptually, this is a fun project – and I agree continuous feedback is an important component of practicing medicine. That said, tossing additional anecdotes into your decision-making workflow will contribute to recency and availability biases. 🙃 #medsky www.michiganmedicine.org/health-lab/h...michiganmedicine.orgHow AI is helping emergency physicians learn from their patientsHow the “Tell Me What Happens Next” initiative is being used by the Department of Emergency Medicine’s new Division of Clinical Informatics using artificial intelligence. 020
Reposted by Ryan Radecki MD MSScott McGrath @smcgrath.phd · 19/06/2026Nearly 3 out of 4 clinicians are concerned about widespread clinical deskilling from overreliance on automation. Also, as patients increasingly bring AI-generated research to medical appointments, doctors and nurses find themselves spending more time acting as validators of outside data. #MedSky 3/5 101