How FACTS is different

Learn how FACTS differs from EAST, rpact and ADDPLAN for adaptive clinical trial simulation.

Bayesian clinical trial design
Bayesian clinical trial software
adaptive clinical trial design
clinical trial simulator
FACTS
Berry Consultants
platform trials
dose response modelling

How FACTS Differs from EAST, rpact and ADDPLAN

FACTS (Fixed and Adaptive Clinical Trial Simulator), developed by Berry Consultants, is purpose-built for Bayesian clinical trial design and adaptive clinical trial simulation. While software such as Cytel’s EAST, rpact, and ADDPLAN (also supported by Berry Consultants) focus primarily on frequentist methods, sample size calculations, and analytical type I error control, FACTS takes a fundamentally different approach: it is a comprehensive clinical trial simulation platform designed to evaluate complex adaptive and Bayesian trial designs.

This makes FACTS especially useful for statisticians and trial designers exploring Bayesian adaptive trials, dose-response modelling, platform trials, and other innovative study designs that are difficult to evaluate using traditional analytical tools alone.

A Simulation-First Approach to Bayesian Clinical Trial Design

Most traditional clinical trial software packages focus on two-arm trials and frequentist hypothesis testing. Their workflows typically emphasize specific analytical methods for particular endpoints and standard sample size determination.

FACTS, by contrast, is built as a flexible clinical trial simulation toolbox. Rather than prescribing a narrow set of standard analyses, it allows users to define complex trial structures, simulate different scenarios (where each scenario is a different possible “truth”) thousands of times, and evaluate a design’s performance over this range of assumptions.

This simulation-driven approach is particularly valuable for Bayesian clinical trial software, where operating characteristics often need to be evaluated through simulation rather than closed-form calculations.

Support for Innovative Designs across many trial types

Phase 2 & 3 trials

  • Phase 2 PoC, and dose finding
  • Phase 3 confirmatory
  • Seamless Phase 2/3 trials
  • Sepatate Phase 2 followed by Phase 3
  • Trials can include one, two, or many treatment arms, optionally with control arms or active comparators.
  • Factorial trials
  • Multiple endpoints
  • Continuous, dichotomous, time-to-event, and ordinal endpoints
  • Adaptation at interims
  • Both Bayesian and frequentist analyses can be used.

Enrichment and Basket Trials

  • Population subgroup selection
  • Trials across related diseases or tumour types
  • Continuous, dichotomous or time-to-event endpoints

Platform Trials

  • Multiple treatments evaluated against a shared control
  • Treatments entering and leaving the platform over time
  • Continuous and dichotomous endpoints

Phase 1 Dose-Escalation Studies

FACTS includes simulation tools for modern Bayesian and model-based dose-finding methods, including:

  • i3+3, BOIN and mTPI
  • CRM using a Bayesian Logistic Regression Model (BLRM)
  • continuous enrolment, backfill, two populations, moderate and dose limiting toxicity,
  • Combination dose-finding with 2D-CRM

Advanced Bayesian Modelling Capabilities

FACTS was originally designed around Bayesian statistical methodology, and this remains one of its major strengths.

Key Bayesian features include:

  • Pairwise comparisons of treatment effects, including means, rates, log odds, hazard ratios, hazard rates, proportional odds, and utilities
  • Dose-response modelling using a range of model forms such as NDML (Normal Dynamic linear Model), Sigmoid, Hiearachical, Linear
  • Informative prior specification for model parameters
  • longitudinal data modelling for imputing the final score for patients who’ve not yet completed follow-up at an interim
  • Bayesian hierarchical models for borrowing external or historical control data
  • Calculation of Bayesian predictive probabilities for trial success
  • Probabilities that each treatment arm is optimal or minimally efficacious

These capabilities make FACTS particularly well suited for Bayesian dose-finding trials, adaptive dose selection, and model-based trial designs.

Adaptive Clinical Trial Simulation

FACTS enables simulation of a wide range of adaptive clinical trial designs, including:

  • Multiple interim analyses
  • Early stopping for futility or efficacy
  • Response-adaptive randomization
  • Dropping ineffective treatment arms

These adaptations can be evaluated across thousands of simulated trials to understand operating characteristics such as power, type I error, expected sample size, expected duration, *assurance, and probability of correct treatment selectio**.

These designs can also incorporate additional features such as efficacy endpoints, multiple populations, and flexible dose levels.

High-Performance Bayesian Computation

Unlike workflows that rely on external tools such as WinBUGS, JAGS, or Stan, FACTS performs Bayesian analysis directly within a compiled C++ library.

This allows simulations with Bayesian models to run significantly faster than approaches that call external Bayesian engines from environments such as R, making FACTS a scalable environment for large simulation studies.

Comprehensive Simulation Outputs

FACTS provides rich outputs for evaluating trial designs, including:

  • Standard graphical summaries and reports
  • Detailed operating characteristics
  • Complete simulation datasets, including every simulated subject and interim analysis, exported in CSV format

These outputs can be imported into R or other statistical software for additional analysis and visualization.

Summary: FACTS vs EAST, rpact and ADDPLAN

Feature FACTS EAST / rpact / ADDPLAN
Core focus Clinical trial simulation Analytical design tools
Support for Bayesian analysis Strong support for Bayesian designs Minimal
Support for Frequentist analysis Standard frequentist analysis included Strong support for frequentist designs
Trial complexity Complex adaptive and innovative designs Mostly standard designs
Dose-response modelling Extensive Bayesian modelling Limited
Platform and basket trials Supported Limited
Performance approach Optimized C++ Bayesian engine Typically analytical methods
Simulation Details Comprehensive simulation results Detailed simulation results only available from rpact

Note here we are comparing against the classic EAST software package (EAST 6.5) rather than the very different EAST Horizon.

In summary: FACTS is a specialized Bayesian clinical trial simulation platform designed to evaluate adaptive and innovative clinical trial designs, while EAST, rpact and ADDPLAN are primarily analytical tools aimed at traditional frequentist trial design and sample size calculation. FACTS is increasingly also useful for conventional non-Bayesian designs such as Group Sequential and Promising Zone.

Developed by Berry Consultants, FACTS helps teams design, simulate and evaluate Bayesian and adaptive clinical trials with speed, flexibility and depth.