Beyond the P Value Series #Ep 1 - Why RCTs Matter and Why They Mislead Us with Prof Paul Young

Randomised controlled trials shape almost everything we do in the ICU. So why do so many of them seem to show nothing? And when a trial says "no benefit," what does that actually mean for the patient in front of you?

This is the first episode of Beyond the P Value, a new eCritCare sub-series on interpreting RCTs in critical care with Professor Paul Young, intensivist and clinician researcher from New Zealand. Over the coming episodes, Dr Swapnil Pawar and Professor Young go past the abstract and the p-value to unpack how landmark ICU trials should be read, challenged and applied at the bedside.

In this conversation:

• Why the RCT is still the gold standard for causal inference, and the three things to look at first: population, outcome, and plausibility of the treatment effect
• The two big weaknesses: surrogate outcomes that leave the patient out of the centre, and the gap between average effects and the individual patient
• Why "negative" trials never really fail, and how a neutral average can hide big benefit for some patients and big harm for others
• Individualised treatment effects, machine learning on trial datasets, and the coming oxygen target trial built on them
• Reinforcement learning in the ICU: how the Instagram feed model could one day balance treatments to maximise survival
• Availability bias at the bedside: why a rising blood pressure after a fluid bolus tells you almost nothing about outcome
• Heterogeneous populations, disease trajectories, and why you still have to start somewhere
• The collapsed right upper lobe that fixed itself, and what it says about how clinicians learn
• Where the practice-changing evidence actually lives: three journals, fewer than 20 papers a year, and the Critical Care Reviews meeting

Next episode: how to interpret an RCT, and the anatomy of a trial.

Read this episode on Substack: https://critcareedu.substack.com/p/beyond-the-p-value-why-rcts-matter