INCEPT: The Intensive Care Platform Trial-Design and protocol.

Summary

INCEPT is a pragmatic, Bayesian adaptive platform RCT infrastructure for ICU patients, organized into domains that evaluate commonly used but uncertain interventions. It defines core outcomes, uses response-adaptive randomization and pre-specified stopping rules, and aims to deliver faster, high-certainty evidence across multiple questions within one platform.

Key Findings

• Investigator-initiated, pragmatic, randomized, embedded, multifactorial adaptive platform for adult ICU patients.
• Bayesian analyses with neutral/skeptical priors, adjustment for prognostic variables, and ITT effect estimates.
• Core outcomes include mortality, days alive without life support/out of hospital/free of delirium, HRQoL, cognition, and safety.
• Response-adaptive randomization and pre-specified stopping for superiority, inferiority, equivalence, or futility.

Clinical Implications

If implemented as designed, INCEPT can shorten time-to-evidence for ICU practices, allow early adoption or de-implementation of interventions, and standardize outcome measures across trials.

Limitations

• Protocol paper without outcome data to date
• Complex implementation requiring robust infrastructure and governance

Future Directions

Activate domains across priority ICU questions, share platform tools and code, and evaluate scalability, equity, and global generalizability.