👋 Happy Friday. Honestly, thank goodness you’re here. I’ve been absorbing medical news all week and my heads beginning to balloon filled with concrete. Please grant me this moment of respite 🙏 

Here’s what we’ve got:

• 🤰Ozempic Babies: Fact or Fiction?

• 🤖 Diagnostic AI: Smart By Design, Biased By Nature

• 🫵 🇺🇸 British Doctor: US Healthcare Wants YOU

• ❗️ Fitness to Practice Gets A Refresh

• #TheMoreYouKnow: Other Top Stories of The Week

RESEARCH UPDATE

🤰Ozempic Babies: Fact or Fiction?

What’s all this hubbub on the internet about Ozempic Babies?

There’s a funny little rumour going around that. Apparently, you take a shot of Ozempic, then BAM. Suddenly you’re in your second trimester. (Well…there is a key step in between, but you get the gist.)

It’s okay, we’ve got contraception. Don’t be so sure.. women are reporting unexpected pregnancies on the pill too!

Of course, losing weight helps restore hormonal balances. This leads to an increased chance of successful pregnancy. But anecdotal evidence aside, is there any research to back these claims?

I must confess, the research is limited. Out of the gate, there is no evidence to support GLP-1’s such as semaglutide and liraglutide having any pharmacokinetic efficacy on the absorption of oral contraceptives. Ozempic, at least, gets a pass. 

But its newer, flashier cousin, Tirzepatide? That’s where things get interesting.

A study was done to evaluate whether Tirzepatide affects the absorption and bioavailability of oral hormonal contraceptives when both are administered together in 2022.

This open-label trial took 28 healthy, premenopausal women and gave them a single dose of 5mg Tirzepatide with the combined pill, then tracked the levels of the contraceptive in the bloodstream. 

They measured a bunch of pharmacokinetic parameters like  plasma drug concentration-time curve (AUC), maximum plasma concentration (Cmax), and time to maximum plasma concentration (Tmax) for the contraceptives. 

Results?

• After a single dose of Tirzepatide, there was a statistically significant reduction in AUC, Cmax and Tmax of 20%. Not ideal when your goal is not to ovulate.

• Tirzepatides effects on gastric emptying were strongest after the first dose or after any dose increase. Over time, the body does adjust.

Of course, this is just a single-dose study. Short-term. Miniscule sample size. No look into long-term effects. But it’s still enough for Eli Lilly(mother of Zepbound), to recommend back up contraception for 4 weeks after starting Tirzepatide. 

I guess the simplest way to ensure you lose weight without gaining a child is to choose a non-oral contraceptive. Just bypass the gut altogether with the IUD or an Implant. Simple as that.

RESEARCH UPDATE

🤖 Diagnostic AI: Smart By Design, Biased By Nature

What does ChatGPT mean to you? 

Do you use it all? Is it a personal confidant? Your favourite search engine? To pose as a better friend who tricks loved ones into thinking you took the time to write a thoughtful reply to their crisis, when you actually didn't?(They look so dumb right now. This is true friendship)

Something more serious?
A clinical co-pilot when you don’t know wtf is going on with a patient? Like this doctor caught on twitter:

Hey, it’s up to you how you use AI at work. 
But let’s ask a bigger question: Is its output actually dependable?

Turns out, when you train a medical AI on real-world hospital data, it gets really good at copying…the bad parts. Like biases. 

This study, published in Nature Medicine aimed to evaluate whether LLMs showed sociodemographic biases in their medical recommendations. The goal is to see if AI makes clinical decisions beyond what is clinically justified. 

Here’s the set up:

They created 1,000 emergency department cases. 500 real, 500 synthetic and presented each one in 32 different variations. Same symptoms. The only thing that changed was the patient's background.

Race, gender, class, housing, sexual orientation, you name it. 31 different socio demographic groups.

They then gave these scenarios to 9 large language models(GPT-4o, Llama, Gemini, Qwen etc). After reading the cases, the LLM’s were posited these questions:

1. Triage priority (urgent/non-urgent)

2. Further testing (none/basic/advanced)

3. Treatment approach (inpatient/outpatient)

4. Mental health assessment (yes/no)

They took these answers and compared them to real doctors answers and the AI’s own control cases to see if there was any variation based on the background of the patient. 

Results? It seems AI took the bait…Hard

• Race & Housing: Cases labeled as Black, unhoused, both, or LGBTQ+ were more often recommended for urgent care, invasive interventions and mental health assessments

• Income: High-income cases were more likely recommended advanced diagnostic testing(CT/MRI) while LIC were mostly offered basic to no testing

• Sexual Orientation & Gender Identity: LGBTQ+ cases, especially bisexual, gay and transgender were disproportionately offered mental health assessments. 

• Intersectionality: Now if you were in the intersection of a minority group, like a black transgender woman, you were up to seven times more likely to be offered a mental health assessment compared to doctor-derived baseline.

So is AI going to replace doctors? One day, maybe. But it’s more likely to replace a Reform party member first. 

USA HEALTHCARE NEWS

🫵 🇺🇸 British Doctor: US Healthcare wants YOU

Hello NHS Doctor. It’s Uncle Sam here. Frontman of The Greatest Country In The World.

It’s time for you to face reality: Your healthcare system is falling apart! 

I hear you can't get into speciality placement anymore? Your wages are flatlining? Your weather is doom and gloom? Aww, poor things. 

After 75 years, I never really understood the whole free-at-point-of-use thing you have going on. A noble idea. Adorable, even. But who needs nobility when you can have cold hard cash🤑.

That’s right. We’re making an absolute killing(no pun intended) off basic care. GP appointments? That’ll be $300. Need an ambulance? That’ll be at least $500 and a quick credit score check. Our median salary is $200,000 a year.

I get it, we all hit a rough patch. But when life gives you pay-cuts, maybe it’s time to catch a flight, land in Florida, and start prescribing SSRIs in a strip mall clinic next to a Chipotle. Because guess what?

You don’t need to match into a U.S. residency anymore 🎉

That’s right! The land of the free may not be open to Mexicans or Chinese imports. But you, my sweet British doctor, can mosey on down to select states with that MBBS degree you have. Let's take a quick tour:

Illinois: Ignore the gun-shots and violence and enjoy a deep dish pizza. Find your way to Windy City, Chicago. Just pass the USMLE and ECFMG qualifications

Florida: Time to catch up on Dr Miami’s Snapchat stories. If you love surgical enhancements, we’ve got the state for you! Just make sure you’ve practiced for 4 years in your home country, then do 2 years in ours.

Virginia: I know you’ve fantasised about taking part in a civil war reenactment in your down time. Just make sure you’ve worked 5 years in your home country before coming down here.

Other stats that’ll take ya: Tennessee, Wisconsin, Idaho, Iowa, Washington, Arkansas and Louisiana.

Look, you shouldn’t rely on your moral compass to choose your direction in life. I want YOU for U.S. Healthcare.

In all seriousness, The US has allowed for IMG’s to start work in the US without matching into residency in select states. A major barrier to going to the states. Read more here

NHS NEWS UPDATE

❗️ Fitness To Practice Gets A Refresh

The dreaded Fitness to Practice has received an update from the GMC. Which, funnily enough, may work in the favour of doctors.

Fitness to Practice is one of few things that’ll send a shiver to any doctor's spine. For whatever reason, the tribunal has called you up to decide your fate as a doctor. Your future is determined in a matter of hours. Scary stuff. 

And worse than that, guilty or not, your case is visible to the world to see. Even after you’ve been vindicated, you can’t help but notice the stolen glances from the nursing staff. 

A situation best avoided altogether I think…

But the GMC says they’ve been listening. And after years of critiques, consultations, they’ve just published a shiny new update to their Fitness to Practise guidance.

Now, it’s not a radical overhaul. The thresholds haven’t changed. The process is still the process. But for the first time, decision-makers across the board—whether they’re dealing with a paediatrician in Preston or a plastic surgeon in Penzance—will be using the same core framework to decide:

1. How serious is this, actually?

2. What was going on around the time—burnout, dodgy rota, sentient EMR system?

3. And has the doctor done anything about it, or are they still doubling down in the group chat?

In theory, it’s all about fairness and transparency. And in practice? Well, it means you’ve got a better chance of your case being judged consistently, rather than rolling the dice on which case examiner you get and hoping they’re not having a bad day because someone drank their oat milk in the office fridge.

Also, (and this is big) the new guidance puts more weight on context. Which might be the most human thing the GMC has ever done. They’re saying, basically: if you slipped up, but you’re reflective, you’ve improved, and you weren’t operating in a vacuum of silence and shame? That should count for something.

So yes, Fitness to Practise is still terrifying. But at least now, if you do end up in that room, it might feel less like a spin of the wheel and more like a structured decision-making process with actual empathy baked in. So they say 👀

Still… probably don’t test it on purpose. Just a thought.

OTHER NEWS YOU SHOULD KNOW

• NICE Approves new treatment for CKD

• Streeting Plea Not To Strike

• Battle of the COVID Drugs: A Systematic Review

• Across The Pond: RFK JR Ends COVID Vaccine Recommendation

Handover Over 🫡 

If you liked it, tell your mates.
If you hated it, tell your enemies.

See you next Friday with next weeks updates! And please vote in the pole below 👇️