E-values acquire the missing pieces — but not the validation
Four preprints close the main gaps in anytime-valid inference for sequential trials. None have touched a clinical dataset, and one mechanism won't survive a DMC charter review.
Biostatistics Desk
Three preprints worth filing: ReLU regression, safer multi-task borrowing, and a simplex trap
Three stat.ME preprints actually worth a trial statistician's time this cycle — and one of them says your microbiome robust-stats pipeline is solving the wrong problem.
Holm, Hochberg, and Rank-Based Co-Primary Methods Each Gain Formal Extensions in One Cycle
Multiple FWER-controlling procedures have landed simultaneously — filling genuine gaps in the co-primary toolkit while raising a shared question regulators haven't yet answered.
Project Optimus Has a Menu Problem: Too Many Designs, Not Enough Regulatory Signal
The simulation-evidence burden has shifted to sponsors — and a crowded design landscape makes the justification conversation harder, not easier.
Five Methods, One Problem: The Crossover-Adjustment Landscape Just Got Crowded
A single publication cycle has delivered peer-reviewed backing for five distinct OS-under-switching frameworks — more validated tools, but a heavier burden of justification every time you open an SAP.
SBR Stages a Structured Debate on Whether Pharma SMGs Still Earn Their Keep
AI is automating the routine methodology work SMGs once owned while regulators demand more sophisticated innovation. The journal exchange asks whether the model still fits.
Win Ratio's Stress Test: Four Papers, Three SAP Fixes
A concentrated burst of methodology work shows the win ratio can contradict its own component results, is noncollapsible, and needs RMST as a co-report.