Technical Reference
Technical Documentation
Zetyra's documentation is designed as a technical whitepaper for statisticians and regulatory professionals. Each page covers the mathematical foundations, explicit assumptions, and references required for Statistical Analysis Plans (SAPs).
Built for Regulatory Rigor
Statisticians don't just want to know how to use a tool—they want to see the math that proves it works. Each documentation page includes formal mathematical equivalences, explicit assumptions with testable criteria, regulatory guidance citations, and validation benchmarks against industry-standard software.
Frequentist Tools
CUPED & Covariate Adjustment
Theoretical foundation, statistical assumptions, and regulatory considerations for CUPED variance reduction.
Group Sequential Design
O’Brien-Fleming, Pocock, and spending functions. Interim analyses with rigorous Type I error control.
Survival Sample Size
Event-driven sample size for time-to-event endpoints. Logrank test, hazard ratios, accrual planning, and reverse calculation modes.
Adaptive Design
Adaptive Randomization (RAR)
DBCD, Thompson sampling, Neyman allocation for outcome-adaptive designs, plus Pocock-Simon covariate-adaptive minimization.
Blinded Sample Size Re-estimation
Nuisance parameter re-estimation from pooled data. Kieser–Friede methodology, Type I error preservation, and regulatory considerations.
Unblinded Sample Size Re-estimation
Promising zone approach (Mehta & Pocock 2011). Inverse-normal combination test, zone classification, and conditional power analysis.
Single-Arm SSR
Bayesian PPoS or CP promising-zone rules for Phase II ORR trials with decoupled gamma_efficacy / gamma_final thresholds and an in-product calibration helper.
Master Protocol
Master Protocol Guide
Comparison of basket, umbrella, and platform trial designs. Decision framework, shared control strategies, and regulatory considerations.
Basket Trial
Single treatment across multiple indications with independent, BHM, or EXNEX information borrowing. Binary endpoints.
Umbrella Trial
Biomarker-stratified sub-studies with shared control arm. Frequentist and Bayesian analysis with multiplicity correction.
Platform Trial
Multi-arm multi-stage (MAMS) design with staggered arm entry, non-concurrent control, and alpha-spending boundaries.
Bayesian Toolkit
Bayesian Toolkit Overview
5-step workflow for Bayesian trial design: prior elicitation, historical borrowing, sample size, two-arm design, and predictive power.
Prior Elicitation
Specify informative or vague priors from historical data. Supports Beta, Normal, and mixture priors.
Bayesian Borrowing
Incorporate historical control data using MAP, power, and commensurate priors with robust mixture components.
Bayesian Sample Size
Determine sample size for single-arm Bayesian trials with posterior probability decision rules.
Two-Arm Bayesian Design
Sample size determination for two-arm randomized Bayesian trials with allocation ratio optimization.
Bayesian Sequential Monitoring
Posterior probability stopping rules for two-arm trials. Efficacy and futility boundaries with operating characteristics.
Bayesian Predictive Power
Predictive probability of success (PPoS) at interim. Go/No-Go decision thresholds for clinical trials.
Guides
Start with the chooser, then the endpoint foundations, design choices, and method deep dives.
Which Calculator Should I Use?
Interactive decision guide to choosing the right statistical tool for your experiment.
Choosing the Right Calculator
Method-selection map across the Zetyra calculator suite. Decision table by design question, then methods organized by layer (sizing, precision, monitoring, adaptation, borrowing, master protocols, composition).
Sample Size for Continuous Endpoints
Guide to sample size calculation for continuous outcomes with worked examples.
Sample Size for Binary Endpoints
Guide to sample size calculation for proportions and binary outcomes.
Sample Size for Survival Endpoints
Event-driven sample size calculation for time-to-event studies using the logrank test.
Non-Inferiority Trials
Design considerations and margin selection for non-inferiority studies.
Single-Arm Trials
When single-arm is the right design, how to set the calculator's inputs (p0, p1, alpha, beta), and adaptive sample size re-estimation for Phase II ORR trials.
Randomization Schemes
Choose between simple, blocked, stratified, minimization, and response-adaptive randomization before powering — and map each scheme to the right Zetyra calculator.
Cluster-Randomized Trials
Design effect, ICC estimation, and sample size for cluster-randomized designs.
Longitudinal Studies
Repeated measures, mixed models, and sample size for longitudinal designs.
CUPED Practical Guide
Worked examples and practical tips for applying CUPED variance reduction in A/B tests and clinical trials.
GSD Practical Guide
Worked examples for group sequential design with interim analysis planning and boundary selection.
SSR Practical Guide
Blinded vs. unblinded SSR decision framework, worked examples, SAP language templates, and R code.
Complete Guide to Bayesian PPoS
Conceptual overview of predictive probability methods, with worked examples and interpretation guidance.
Complete Guide to Bayesian Sequential
Posterior probability stopping rules, prior specification, boundary calibration, and worked examples for Bayesian sequential monitoring.
GSD vs. Bayesian Sequential
Head-to-head comparison of Group Sequential Design and Bayesian Sequential Monitoring, with decision framework and worked example.