eCritCare Podcast podcast

Beyond the P Value Ep #3: Adaptive Platforms, Bayesian Thinking and the Future of ICU Trials with Prof Paul Young

0:00
47:31
Manda indietro di 15 secondi
Manda avanti di 15 secondi

The large multicentre RCT has been the gold standard in critical care for decades. So why are trialists turning to cluster designs, adaptive platforms, Bayesian analysis and non-inferiority frameworks, and what do you give up when they do?

In the third episode of Beyond the P Value, the eCritCare sub-series on interpreting RCTs in critical care, Dr Swapnil Pawar and Professor Paul Young, intensivist and clinical trialist from New Zealand, work through the trade-offs behind modern ICU trial designs.

In this conversation:

  • Randomising patients or randomising ICUs, and how SPLIT enrolled about 2,500 patients in 28 weeks
  • Why low-risk patients in a broad trial add noise rather than signal
  • What Mega-ROX can teach us beyond its primary outcome, and why huge trials still matter
  • A first look at unpublished ICU-ROX data on haemoglobin and the response to conservative oxygen
  • Why "it looks fine at the bedside" can't tell you whether conservative oxygen is safe
  • REMAP-CAP: embedded, multifactorial, adaptive, and the problem of adjusting for time
  • Bayesian versus frequentist statistics, and why Paul thinks neither is better
  • Posterior probabilities for comparative effectiveness trials, and the "one minus the P value" shortcut
  • SuDDICU and the Bayesian meta-analysis of selective decontamination
  • Non-inferiority trials, the BALANCE trial, and choosing a margin that includes results you wouldn't like
  • The six possible results of a non-inferiority trial
  • Cluster, cluster crossover and stepped wedge designs, and the REVOLUTION trial of personalised oxygen targets

Previous episode: Beyond the P Value Ep 2, The Anatomy of an ICU Trial.

Read this episode on Substack: https://critcareedu.substack.com/p/beyond-the-p-value-icu-trial-design

References: Young P, et al. SPLIT. JAMA. 2015;314(16):1701-1710. Angus DC, et al. REMAP-CAP rationale and design. Ann Am Thorac Soc. 2020;17(7):879-891. Daneman N, et al. BALANCE. N Engl J Med. 2025;392(11):1065-1078.

Please send us your feedback @ [email protected]

Altri episodi di "eCritCare Podcast"