iiinova — where your digital self keeps evolving

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

Potential solution? The ASBI (Alcohol Screening and Brief Intervention). Takes about five minutes at a routine visit, roughly doubles the odds of reported abstinence, and is almost never used. Full piece on Medscape: buff.ly/MqpzstA

020

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

All of this is self-reported over the phone, including the answer to "are you pregnant." Social desirability bias should push the alcohol numbers down, not up. Unless people simply got more honest with telephone surveyors over 14 years. I can't think of a reason they would.

120

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

The women who drank were older, more likely to be college graduates and employed, less likely to be married or partnered. Tobacco use was the strongest correlate: 26.3% of pregnant smokers drank vs 11.3% of nonsmokers. Frequent mental distress, 22.8% vs 11.1%.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

Binge and heavy drinking rose faster in relative terms. Binge drinking (4 or more drinks on an occasion) 1.77x. Heavy drinking (8 or more per week) 2.66x, from 0.8% to 2.0%. The trend models are significant for all three.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

Among pregnant women, current alcohol use went from 9.1% in 2011-2012 to a peak of 15.2% in 2021-2022, and it has not come back down. 14.5% in 2023-2024. Prevalence ratio 1.59.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

Among nonpregnant women of reproductive age, rates over those same 14 years are basically flat (maybe a hint of a decline recently). So this isn't a story about American drinking in general.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

The data come from BRFSS, the CDC's telephone survey of US adults. 33,729 pregnant women aged 18 to 49, surveyed from 2011 through 2024.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

The Surgeon General issued a formal advisory in 1981, and in the Book of Judges an angel tells Samson's mother to drink no wine while pregnant. So you'd expect the rates to be falling. A research letter out this week in JAMA finds the opposite: buff.ly/BXAltCj

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 27/07/2026

The word is definitely out that alcohol should be avoided in pregnancy. 90% of women now know drinking in pregnancy isn't safe, up from 82% last year. But reported rates of drinking while pregnant are, broadly, increasing. 🧵

131

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

Full piece on Medscape: buff.ly/G3KZQ59

020

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

So it's the rare medical conclusion where you get to tell patients to keep doing the thing they enjoy. Moderate coffee (roughly 3-5 cups) isn't just safe, it may be cardioprotective. Take the W. Maybe run it through a paper filter. And skip the Red Bull.

141

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

But the panel draws a firm line at synthetic, high-dose caffeine. Energy drinks, pre-workout powders, and pills don't inherit coffee's benefits, and there are case reports of young, healthy people suffering cardiac arrest after drinking them: buff.ly/Cbrhdak

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

Heart failure risk bottoms out near 4 cups a day: doi.org/10.1161/CIRCHEARTFAILURE.112.967299. And coronary disease shows no added risk out to six cups: doi.org/10.1161/CIRCULATIONAHA.113.005925

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

On hard outcomes the curve is J-shaped and friendly. Lowest stroke risk lands around 3-4 cups a day, a 21% reduction: buff.ly/v3wjOWP

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

Then the diabetes paradox. Coffee acutely impairs glucose tolerance in the lab, yet across 28 cohort studies habitual drinkers had lower type 2 diabetes risk (20% lower at 3.5 cups/day, 30% at 5): buff.ly/Oo1C3Sk. Probably the non-caffeine compounds.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

It isn't a clean win. The same data that clears coffee for the atria show it bumps up premature ventricular contractions: buff.ly/vrfsaTN. Soothes the top of the heart, irritates the bottom. No free coffee.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

The biggest surprise (to me, a non-cardiologist) is atrial fibrillation. For a generation the reflex was: palpitations, stop the coffee. Then a randomized trial had AFib patients drink a daily cup or abstain before cardioversion. The coffee group had 39% fewer recurrences: buff.ly/n6iK9Jj

210

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

Same theme with blood pressure. In habitual coffee drinkers, more coffee tracks with slightly lower BP: buff.ly/a99vBE2. But energy drinks (synthetic, high-dose) cause real acute spikes: buff.ly/HaRLPMV. Source matters more than caffeine.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

The cholesterol part is mechanical. Unfiltered coffee (French press, Greek, Turkish, boiled) has a diterpene called cafestol that raises LDL in randomized trials: buff.ly/K43TtT1. Run it through paper and the effect vanishes: buff.ly/GD0Ivok

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

The statement is in Circulation, and it pulls together decades of messy, mostly observational research. The through-line: concerns that coffee might exacerbate heart disease are largely unfounded, and the opposite may, in fact, be true. buff.ly/XT6NB8V

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 20/07/2026

Your morning coffee either raises your LDL cholesterol or does nothing to it, and the single variable that decides is whether you poured it through a paper filter. That's one of the odder findings in the new AHA scientific statement on caffeine and the heart. 🧵

131

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

The part I couldn't fit here: why polyethylene (yes, grocery-bag plastic) turned up in 97% of the positive samples, and why that matters less than you'd think. Full piece on Medscape: buff.ly/MpURib1

buff.ly

STEMI Patients Are Awash in Microplastics

Notwithstanding definitive proof of cause and effect, micro- and nanoplastics could have a major role in STEMI.

120

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

What I keep coming back to is the exposome (the sum of everything in our environment that ends up inside us). You can eat well, exercise, take your statin, and still absorb the world around you. Microplastics are really a public health risk factor.

130

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

The authors ran it both ways. Predicting who had microplastics: smoking won and everything else dropped out. Predicting who had obstructive coronary disease: microplastics held up and smoking didn't. With correlations this strong, you can really tell either story.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

Here's the rub. Smoking is such a strong risk factor for both microplastics and heart disease that a lot of this could just be confounding. Hard to untangle in 61 people.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

Where does it come from? Smokers had far more than nonsmokers. So did people breathing dirtier air. Put both together (a smoker in a high-pollution area) and detection hit 100%.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

And it wasn't just detection. Concentrations were markedly higher in the STEMI group. Within those patients, the plastic load was elevated everywhere they looked, not only at the blocked artery. These people were simply awash in more plastic.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

Microplastic studies have a contamination problem (plastic is everywhere, including the lab). These authors went hard: fume hoods, cotton coats, glass vials, blank saline runs, mass spec plus a laser method. The numbers look real.

120

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

61 patients at two Italian hospitals, all getting cardiac catheterization. 19 with a STEMI (a major heart attack), 20 with chronic coronary disease, 22 with completely clean arteries. Blood was pulled from the artery itself as well as a peripheral vein.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

We've known micro- and nanoplastics get into the body. Showing they actually do harm has been much harder. This new study in the European Heart Journal is one of the more provocative attempts. buff.ly/1sEeCVq

132

F. Perry Wilson, MD @fperrywilson.bsky.social · 15/07/2026

They sampled blood straight from the coronary arteries of people mid-heart-attack. 84% had detectable microplastics. In people with clean coronaries, it was 32%. 🧵

131

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

My full take (including whether any of this is actually legal) is on Medscape: buff.ly/h5apjHI

022

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

So either we enforce the rules and close the compounding loopholes, or we admit that for some of these drugs, if everyone can get a prescription, maybe no one should need one.

132

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

My major issue: this patient was an easy yes. I'd want to know what happens at a BMI of 26 with no real comorbidity. Do the sites wave him through too, or help him along? This study can't tell us.

122

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

Almost all of it was compounded, not branded. Compounding a copy of an available drug generally isn't allowed, so 60% of the prescriptions added a supplement (usually B12 or glycine). That's one way to argue you're not simply copying Ozempic.

122

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

One of those two initially refused to fill the duplicate. Then the shopper asked for a refund, and the site decided to ship it anyway.

122

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

Other signs the supervision was thin: the same provider writing scripts across multiple sites, prescriptions issued despite missing required photos, and a couple written in under five minutes. Only two sites noticed he'd ordered the same drug somewhere else.

122

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

Turns out, not much. Every site had a questionnaire. Only 55% asked about eating disorders. 13 of the 49 required a video visit and 3 required a phone call. The rest was forms.

122

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

The persona was built to be an easy yes: a 26-year-old man with a BMI of 35, hyperlipidemia, hypertension, prediabetes, sleep apnea, and no contraindications. So the question was never whether he'd get a prescription. It was how little anyone would ask first.

122

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

We put some drugs behind a prescription because the risks are real enough that you want an expert in the room first. This is a clean test of whether that gate still exists online. New research letter in JAMA: buff.ly/Rv0KtIT @yale_crrit

132

F. Perry Wilson, MD @fperrywilson.bsky.social · 06/07/2026

A team @Yale invented a fake patient and had him try to buy GLP-1 drugs on 49 different websites. 45 of them wrote him a prescription. A couple in under five minutes. 🧵

274

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

We're not there yet. But I think we get there, and there's a future where NOT having an AI reading over your shoulder counts as malpractice. Interesting times. Full piece on Medscape: buff.ly/OKKtvli

020

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

FWIW, these were early-2026 models (GPT-5, Gemini 3 Pro, Opus 4.6), and in this field, if you don't like a model's performance you just wait a couple weeks. They also ran them at temperature zero, which can quietly cost accuracy.

120

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

False positives are poison for trust in medical AI. Cry wolf enough and clinicians dismiss the alert the moment it fires, including the time it's right. Better to miss a few than to nag constantly. My gut says a tool with these numbers doesn't get adopted yet.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

Run 100 discharges through the best model. About 48 trigger a pop-up. 9 are real catches. 39 are false alarms (you'll start tuning them out fast). And 2 patients bounce back that the AI never flagged at all.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

But sensitivity is a double-edged sword. The model that catches the most real misses also flags the most patients who are perfectly fine. Most sensitive, least specific, and vice versa. Every model landed somewhere on that tradeoff.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

Then they handed the same charts to the models and asked, simply, is something being missed here? Sensitivity varied wildly. Claude Sonnet 4 caught the most real misses (about 86% in the discharge cohort). GPT-5 mini caught the fewest (43%).

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

Not every bounce-back is a miss. Back pain that later turns out to be an epidural abscess (no fever, normal labs at first) isn't one. A DKA hiding behind an unaddressed anion gap is. Of 288 reviewed cases, 39 (13.5%) were genuine misses.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

How do you even know a diagnosis was missed? They took two ED cohorts (patients discharged then readmitted within 72h, and floor patients bumped to the ICU within 24h) and had two physicians review each chart to decide if something was genuinely missable.

110

F. Perry Wilson, MD @fperrywilson.bsky.social · 30/06/2026

The catch is alert fatigue. Beep at a doctor enough and they tune the beep out, even when it's right. An AI safety net has to be accurate without firing on every patient. Six LLMs put to the test in JAMA Network Open: buff.ly/fWd94A9

110