💜Bieber, 🧠 Brain Trauma and🩸Bleeding?

I knew you’d be back.
You just love medical news. You always have. You’ll probably never beat the nerd allegations. That’s okay, this is a safe space. Anyway…

👋 Happy Friday. Here’s what we got:

If you want to read any previous editions of The Handover, you can on our website.

In medicine, we have our commandments of unholy combinations.

Somewhere at the top, it is written in the scripture:

And lo, it was decreed: Thou shalt not mix anticoagulants with antiplatelets in stroke patients without AF; for such practice bringeth forth bleeding and great harm. (NICE 12:6, NKJV)

The prevailing wisdom has always advised against it. The research concurs. 
Hell, even a certain unnervingly good newsletter wrote an article about the disaster-class of mixing these drugs almost exactly a year ago.

TLDR: No reduction in stroke recurrence. Significant increase in bleeding risk. Significant increase in major adverse cardiovascular events(MACEs). Bad, bad news

But as Justin Bieber reminded us all at Coachella this week, you should Never Say Never. And apparently, neither should clinical researchers…

Published in the NEJM, these researchers wanted to challenge the status quo in a way that hadn’t been done before: Looking at a new kind of anticoagulant. 

Step aside, apixaban.
Not today, dabigatran.
Warfarin, please don’t even look our way.

Today, it’s a factor XIa inhibitor called Asundexian to undergo the scrutiny of a randomised control trial. 

The logic is simple: research has found that genetically lower levels of Factor XI were linked to a lower ischaemic stroke risk. So… if that's the case, surely if we can reduce Factor XI levels, we can also reduce the risk of additional ischaemic events? 

The researchers got to work. 

They recruited 12,327 patients across 702 centres in 37 countries. To qualify, patients had to have had a noncardioembolic ischaemic stroke or high-risk TIA within the last 72 hours. They were then randomised 1:1 to either:

To be taken daily. Then they watched like hawks to see

After waiting patiently for 1.5 years… the marquee findings were this: 

The incidence of ischaemic stroke was significantly lower with asundexian than placebo AND the incidence of major bleeding was similar between groups

Was the trial perfect? They never ever are…

Limitations: The biggest concern is generalisability. The patient population was overwhelmingly mild stroke patients. And the population diversity that only Tommy Robinson would approve of (2% of participants were black).

Whilst not immediately practising changing and more studies to be done, there is hope yet for an anticoagulant and antiplatelet to live in perfect harmony.
Reducing stroke recurrence without increasing bleeding. 

So I hope it’s not sacrilegious to say, someday, we may all be beliebers in asundexian.

Congratulations on finishing your foundation school application 🥳

Unfortunately, the burden of manually dragging and dropping 268 jobs on Oriel is only the first step in the lead up to August…

Black Wednesday haunts every new doctor.
And you haven't clerked solo or written a drug chart since OSCEs 💀
Most FY1s get through it fine…but if the GMC ever comes knocking, what do you do?

Call Medical Protection.

Medical Protection has been in the business of having doctors' backs since 1892. Think of them as the workmate you call at 2 am when something's gone wrong. Backed by a team of actual lawyers, so you don't have to face it alone.

350,000+ clinicians trust them for:

No financial caps. No ceiling on what they'll do to defend you.

And better yet, it’s just £1 per year for foundation-year doctors.
One. Pound. Sterling. 

Join the world's largest MDO, Medical Protection, using the link below 👇

Membership is based on an application process. All the benefits of membership are discretionary, as set out in our Memorandum and Articles of Association. Among these benefits is the right to ask for assistance with a range of medicolegal issues, including indemnity for clinical negligence claims. We are owned by our members, so our starting point is always to see how we can help.

If you’ve recently been branded Unc* by a child born after 2010, this is a safe space.

*old, uncool, fossil

And if you’ve done the 6,7 hands or quoted Tim Cheese lore in the last month … take a seat.

As was always inevitable, Gen Z are becoming the internet dinosaurs.

And despite our best efforts to educate the youth on the cultural importance of One Direction or how to work a DVD player, it’s time to face the facts:

You can’t apply adult rules to the world of kids.

It’s all fun and games out-mogging Gen Alpha by dabbing in their face without a hint of shame …

But why are we pulling the same move in paediatric intensive care? 

When it comes to how to deal with TBIs (traumatic brain injuries), intracranial pressure(ICP) is king. 

Current guidelines recommend intervention at ICP ≥20 mmHg, but this is largely extrapolated from adult data and supported by limited paediatric-specific evidence.

We all know kids aren’t just small adults, even if paediatricians are.

So researchers, published in JAMA, challenged the status quo this week in their new STARSHIP study.

It was meant to fly, and to make sure it did, they analysed high-resolution ICP monitoring data from 135 children (≤16 years) across 10 UK paediatric ICUs between 2018 and 2023, with a 12-month follow-up.

And touch the sky, they did🚀

These associations held even after accounting for injury severity, treatment intensity and hospital level differences. (And get this, the intensity of treatment was not linked to outcomes as an independent factor)

What mattered was if ICP was actually controlled:

And yes, it was a small sample size, focusing mainly on older kids (median age 8 years) and with a wide range of causes for TBIs, it’s impossible to account for all variables.

But the takeaway is clear: The pressure’s not just on the brain anymore, it’s on paediatricians to think hard about what ICP threshold they choose. 

And it’s not just about how high ICP gets, but how long it’s allowed to remain up there. 

So when it comes to paediatric ICP, it’s not just a nuclear family of bears that should be worried about a goldilocks zone…

Not too high. Not too late.
14-15 mmHg? Jusssttt right🥣🤌

🤖 Robots Hallucinating Surgical References

Just like half the wellness influencers out there touting the benefits of unpasteurised milk, AI’s evidence backing up its medical claims essentially boils down to “Because I told you so.” 

If you’ve ever used AI as a shortcut to searching the literature, then you’ve likely experienced it before - that rush when it gives you just what you’ve been looking for. You excitedly search for the original, only to realise it’s fabricated, an entirely plausible-sounding paper. 

It doesn’t just feel like a betrayal; it feels unsettling. The URL seems legit. The title looks spot on.  And it’s by Zhang et al. It’s fooled me before, and it’s probably fooled you.

So how the hell do we know what to trust? 

A paper published by the Royal College of Surgeons put nine AI chatbots to the test by asking them all a set of standardised surgical questions. They found huge variability between bots - some cited evidence only when asked, others gave it consistently. 

To everyone’s great surprise, coming in ninth place was Grok, Elon Musk’s favourite child (sorry X Æ A-Xii), which hallucinated up to 34% of the time.

Luckily, surgery is a famously chill and low-stakes hobby, so as long as we don’t mind accidentally yoinking out a couple of gallbladders for each attempted appendix, we’re golden ☺️ 

And your average medical textbook can’t give you both surgical tips and 4k images of you starring in every Ghibli movie. Seems like a fair compromise to me…

🍃 Blinding in Psychadelic Trials: A POTential Problem

More and more research seems to be looking at the promising medicinal benefits of psychedelics for a host of illnesses, ranging from major depressive disorder to PTSD.

That’s great and all, but a systematic review published in JAMA Psychiatry pointed out a rather obvious problem … blinding becomes obsolete when half the participants are high as a kite. 

You see, most of the time, in a double blind trial, both participants and researchers have no idea if they’ve got the drug or the placebo. 

But if half your sample is baked on some laboratory-grade Hash, and the other half is just popping Tic Tacs, it becomes pretty obvious who’s in which group. 

And if people know they’ve got the drug that might make them better, they might start feeling better regardless - known as expectancy bias. 

But there’s no perfect solution.
So just be sure to take those trials with a pinch of coke if you’re a participant, or salt if you’re a Handover reader. 

🪶 Starmer’s Substack Musings

Our contemplative indie King, Kier Starmer, just dropped the most insightful Substack essay of 2026,  reflecting on the NHS’s role in our lives. 

A poet at heart, he waxed lyrical about how the NHS should “fit around your schedule”, inspiring thousands to pre-book their STEMIs for out-of-office hours. 

He was also sure to include a glowing endorsement of AI and its integration in the NHS, with one-third of CXRs already being AI-assisted. Luckily, we all now know AI is completely foolproof and reliable, so it’s really top-notch stuff from the PM.

In true artistic style, he signed his essay “Kier.” Moving stuff 🥹

We anxiously await news of Kier’s weekly substack newsletter (and Spotify playlist reveal) and will be sharing the emails of all Handover readers directly with government HQ to show our full support* 🤜🤛

You can read his musings here

*It’s a joke. We would never. We respect your privacy, and we fear GDPR almost as much as we fear microbiology

Handover Over 🫡

If you liked it, tell your mates
(forward along, please. It only takes 12 seconds, making this took an entire flight back from Japan)
If you hated it, tell your enemies.

Before you depart(if you haven’t voted already), please let us know your role in the poll below - It means the world when you do 🫶