5HT #130: the hottest protein, the bull case for AI prescribing, and 3 more

AI prescribing is here (in Utah, at least). I’m suuuper pro this, so here’s my bull case.

First, let’s zoom out and recap what’s been happening. 

Utah’s Office of Artificial Intelligence Policy runs a “regulatory sandbox,” where companies can test AI under state supervision, with temporary relief from specific rules (like, you know, requiring a clinician to authorize each prescription).

Since January, Doctronic has been renewing prescriptions for ~190 non-controlled medications for $4 a pop. In March, Legion Health got the green light for a similar pilot covering certain psychiatric medication refills. Earlier this month, Utah added a few more: August AI for routine chronic refills and Expect Fitness for AI-supported pelvic floor physical therapy.

But the big one added was Nolla Health—the first company in the U.S. authorized to let AI issue new prescriptions. Its app combines a face scan with medical-history questions to select from a shortlist of topical acne treatments. 🤳

To be clear, no one in Utah is prescribing without a human yet. Instead, these pilots start with actual clinicians reviewing every prescription. From there, it’s sort of designed like a ladder… Nolla’s doctors approve the first hundred patients, then the AI prescribes on its own, with reviews after the fact plus spot checks.

Okay, that sounds interesting and exciting and pretty practical for low-risk interventions. So what’s the issue? Well, I think the best argument I’ve heard against AI prescribing is that skipping a visit with a clinician has other downsides. For example, acne can signal PMOS, and ED can be an early manifestation of heart disease.

Fair. But that assumes the visit was catching it‚ and I think that’s unrealistic a lot of the time. One study estimated a PCP working alone would need 26.7 hours a day (!!) to deliver recommended care to a typical patient panel. And that’s even if you can find a PCP, where we’ve had a shortage for a long time. Those 15-minute slots just aren’t cutting it 🙅.

And, if anything, I think AI could catch more, and some research is pointing in that direction. For example, in a randomized, double-blind study published in Nature, Google’s AI outdiagnosed primary care doctors in text consultations with patient actors and scored better on 30 of 32 specialist-rated measures. Sure, those are simulated patients and, sure, it’s Google’s own study. But it’s still incredibly promising. 

Back to Utah. When Utah assessed letting an AI renew prescriptions, doctors agreed with 91% of the AI’s renewal recommendations 👍 (when they didn’t, it was often because the AI was too cautious/conservative).

In time, AI should be wayyy better at asking questions, gathering your health history, and assessing risk. That—and it can add safeguards to help prevent unauthorized prescriptions. Humans famously have sucked at this (I mean, remember when those telehealth execs were sent right to jail for prescribing ADHD meds every 30 seconds? Or, you know, the opioid epidemic?).

Looking to the future, I think this will go alllll the way. Renewals, then topicals, then the big stuff. My bet: By 2030, most routine prescriptions in America will be AI-written. I assume controlled substances, narrow-margin drugs where a small mistake is a big deal, and VERY expensive drugs won’t be included. I personally still want a real clinician when:

Point is, I don’t think your PCP should spend a whole visit deciding on if you need a tretinoin refill (true story). To me, AI prescribing removes an unnecessary bottleneck and can be used to apply consistent rules to get most people what they want while preventing fraud and human error. 

And bottom line for builders: if your business model is charging for the signature, start worrying. As for patients, the doctor may now see you less often, but hey—they’ll probably have more time to spend with you.

Reading has major health benefits—from stress relief to improved cognitive function—yet most of us don’t read enough. Truthfully, even I don’t read enough. The last book I read was Dungeon Crawler Carl Book 8, and that was about a month ago. 🫣

So when I see things like this 👇, it really makes me want to set a better example, both for myself and my two (soon-to-be three!) kiddos.

So consider this a PSA to read a book. (And by a PSA, I mean a pep talk I’m giving myself in writing. 😅)

Next up on my list is The Long Way To a Small, Angry Planet by Becky Chambers. What’s on your list? Hit reply and tell me what book you’re talking about at dinner parties.

Turns out America's leading protein source is… sandwiches?

In new research from the National Center for Health Statistics (NCHS), sammies took the number one ☝️ spot by a 5%+ margin. And, yes, they included a wide range of breaded combos, including burgers, breakfast sandwiches, and classic PB&Js.

According to the NCHS survey, the top five sources of protein intake were:

Maybe an interesting opportunity for healthyish food brands to create some viral sandwiches. (Buuuut maybe hold off until the nationwide meat recall is over—whhhhat 😵‍💫.)

Last week I was at the Assembly, one of my fave health conferences. What makes the conference so good is 1) it's put on by the amazing Jon Ward, who has a real knack for getting good, thoughtful people in the same room; 2) it’s < 100 people; and 3) it’s built for actual engagement.

Mornings have a few hours reserved for talks (all killer), afternoons are for group activities, and dinners encourage you to sit next to people you don’t know. And the connections started happening even before I got there! 

On the plane, I sat next to Nora Topholf, who runs wellness for Kohler Health and is killer. On the way back, I sat next to the lovely Dora Kamau, whose voice you’d probably recognize from Headspace.

A few highlights from the ground:

It was basically my dream room, annnnd I haven’t even gotten to the part where we handed out Fixie Dust. More on that next week 👀.

Oh—and you should join next year!

I have a looot of opinions about consumer health. (Shocking, I know.) 😆

My strongest opinion is that content marketing is hands down the most underrated, under-invested-in way for health companies to build trust, grow authority, and drive conversions. But only if you’re creating what I call “the best answer on the internet.”

Yes, it works (see here and here). And yes, most health co’s content is painfully average. But what does the best answer actually look like? Well, I’m bringing out the red pen for our first-ever webinar! 🤩

Introducing: The Healthyish Report Card: Grading the Best (+ Worst) Health Content on the Internet. 

On Thursday, Nov 19, from 12-1pm (ET), I’ll live-audit consumer health blogs using the ~official Healthyish Content scoring formula~ 😎. 

My promise: you’ll leave knowing what to fix and what to write next. (Ideally before spending another $ on an article no one asked for.) Expect tough love, constructive feedback, a bit of praise, and some Dad jokes. 

Annnd I’m keeping it spicyyy 🌶️ by taking audience submissions for rapid-fire audits. Submit your blog after registering… if you dare. 😏

3 other healthyish things I’m into right now: 

Most clicked last week: Ro’s ED pill commercial with Megan Fox. 💊

👋 Who are you again? I’m Derek Flanzraich, and I’m in the business of healthyish. I founded two venture-backed startups–Greatist (👍) and Ness (👎)—and helped dozens of brands you know grow including Ro, GoodRx, Midi, BetterHelp, and LMNT. I now run Healthyish Content, plus a few other healthyish things (like Fixie Dust, launching this fall).

Every Thursday, I share what’s hot, what’s not, and what’s next in consumer health. (Disclosure: I invest in and advise companies in this space—when I cover one, I'll say so.) I sometimes add affiliate links, but they influence my picks zero.

See something smart, strange, or seriously overhyped? Hit reply, I respond to every email.