Skip to main content

Table 2 Summary of design features for SHINE and the Goldilocks alternative

From: The Stroke Hyperglycemia Insulin Network Effort (SHINE) trial: an adaptive trial design case study

  SHINE Goldilocks
Hypothesis testing Two-sided One sided
Primary analysis Covariate adjusted Unadjusted
Randomization Response adaptive 1:1
Minimum sample size Approx 600 Fixed at 500
Maximum sample size Between 1,400 and 1,718 Fixed at 1,400
Planned interim analyses 4 9
Stopping boundaries Group sequential Predictive probabilities, incorporating early information
Complete follow-up of all patients in event of early termination Undefined critical value Explicitly defined critical value