In the Interim...

In the 60th episode of “In the Interim…”, Dr. Scott Berry, Dr. Nick Berry, and Dr. Joe Marion discuss how Berry Consultants uses AI in clinical trial design and software development. The conversation addresses current applications, limitations, implications for productivity, and the ongoing need for human expertise in clinical trial design. The team examines both promising use cases and the risks associated with security, compliance, and AI-generated statistical work.
Key Highlights
AI is used to develop user interfaces and code modules, notably expediting tasks like R Shiny app development and software prototyping.
Statistical coding for complex modeling and simulation—such as numerical integration and predictive probability calculations—remains unreliable when delegated to AI and still requires direct oversight and manual review.
Attention to security and confidentiality is central; Berry prohibits the use of client-sensitive or patient data within AI tools.
Generative AI assists with drafting and editing documents, but the output tends to be non-specific, generic, and sometimes imprecise, requiring expert editorial input before use.
While embracing AI to improve efficiency, the discussion is critical of current AI hype, especially around black-box modeling and pushes back against the perception that current AI can replace domain-specific statistical design or strategic judgment.
For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry and Dr. Nick Berry investigate how futility in clinical trials and stopping rules in sports illuminate very similar decision problems, albeit with very different consequences. Drawing from baseball’s 10-run rule, tournament cuts in golf, the discussion confronts traditional and Bayesian strategies for interim decisions. The episode explains why simulation, not historical trial review, provides the empirical backbone for futility boundaries in clinical trials, and details the mechanics and consequences of aggressive stopping criteria. Using the Biogen aducanumab Alzheimer’s trials, the conversation exposes how a futility rule based on 20% predictive probability halted trials even when meaningful probability of success remained. Scott and Nick address the influence of ethical considerations, cost, regulatory priorities, and statistical rigor, and contrast Bayesian predictive probability’s strengths over conditional power.
Key Highlights
Dissects sports futility rules (10-run rule, golf cuts, Bill James heuristic) and their application to clinical trial design
Argues for prospective simulation to define adaptive futility thresholds
Explains how Bayesian predictive probability provides a more robust framework than conditional probability for interim adaptive decisions
Details how aggressive futility criteria may prematurely stop trials and risk missing beneficial treatments, as in the aducanumab case
Explores the intersection of ethics, patient safety, operational efficiency, regulatory standards, and trial cost

In this episode of "In the Interim…", host Dr. Scott Berry undertakes a detailed, methodical critique of ICH-E20 draft guidance language as applied to adaptive clinical trial design. Focusing on an innocuous but corruptible paragraph in Section 3.1, Scott scrutinizes the logic behind regulatory reluctance to appreciate multiple or complex adaptations in confirmatory trials. Drawing on extensive experience, he highlights how such restrictive interpretations do not reflect practical development realities, instead setting up “false choices” where alternative designs desired by regulators are infeasible. Through operational scenarios—including the SEPSIS-ACT trial, an enrichment design, and sample size re-estimation examples—Scott illustrates the empirical benefits of seamless and multi-adaptive trials for sponsors, patients, and regulators. Technical discussion addresses misconceptions about complexity and bias and stresses the value of presenting realistic alternatives when engaging with regulatory authorities. The episode ultimately encourages a more nuanced dialogue to advance efficient and scientifically robust clinical trials.
Key Highlights
Discussion of ICH-E20 section 3.1 guidance and its operational impact on adaptive designs.
Dissection of “false choice” dilemmas in regulatory interactions, referencing real adaptive trial submissions.
Case-based examples: SEPSIS-ACT, enrichment, and sample size adaptation trials.
Highlighting myths regarding bias and operational burden from multiple interim analyses.
Emphasis on practical strategies for more effective regulatory communication about adaptive trials and realistic alternatives.
For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…" Dr. Scott Berry is joined by Professors Victoria Cornelius, Danny McAuley, and Anthony Gordon, for a technical review of the PANTHER trial—an international, Phase 2 adaptive platform evaluating pharmacologic interventions for ARDS. The trial is open-label and does not employ blinding, as discussed in the episode. The primary endpoint is 28-day organ support-free days (death as -1, survivors 0–28 days), analyzed with a Bayesian proportional odds model. PANTHER uses  stratification by hyper- and hypoinflammatory subphenotypes, with fixed, equal randomization within each stratum. Analyses for treatments are separated by stratum, reflecting the potential of differential treatment effects. Quarterly interim analyses allow early stopping by stratum for efficacy or futility. Content includes explicit discussion of infrastructure: rapid device deployment, centralized data for trial and future biological discovery, and governance challenges in multinational collaboration. Funding is provided by NIHR (UK), US Department of Defense, CIHR (Canada), NHMRC and MRFF (Australia), HRB (Ireland), and additional support from Germany and Japan. PANTHER is positioned to streamline Phase 2 critical care drug testing and facilitate graduation to larger platforms such as REMAP-CAP, with potential to expedite pharmaceutical evaluation and accelerate ARDS therapeutic development.
Key Highlights
Real-time phenotyping (Randox device) to stratify ARDS patients.
Separate Bayesian analyses by phenotype stratum.
Open-label, fixed randomization within stratum.
28-day organ support-free days as a composite endpoint.
Quarterly interim analyses enable early dropping or graduation of arms by strata.
Central data resource and biosample collection for future research.
Operational, funding, and device logistics for global trial deployment.
Transition of Phase 2 results to established Phase 3 platforms (e.g., REMAP-CAP).
For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry connects with Dr. Byron Gajewski, professor of biostatistics and data science at the University of Kansas Medical Center (KUMC), for a detailed discussion on the design, simulation, and operational realities of Bayesian adaptive clinical trials in academic environments. Gajewski discusses his academic background, training at Texas A&M, and progressive adoption of Bayesian principles based on direct experiential advantages in complex data settings. The conversation highlights KUMC’s Fixed and Adaptive Clinical Trial Simulator Working Group, which utilizes FACTS for faculty, staff, and student collaboration, enabling practical simulation, trial protocol development, and in-house applied statistical training. The PAIN-CONTRoLS Trial serves as a practical example of multi-arm Bayesian adaptive design, using response-adaptive randomization for comparative effectiveness in neuropathy research. The NIH-funded HOBIT trial is examined in detail: multi-arm structure, adaptive allocation among investigational arms, fixed control randomization, group-sequential interim analyses, and sliding dichotomy methodology for the Glasgow Outcome Scale Extended. The discussion stresses a shift to probabilistic, evidence-driven interpretation and reporting, shaping operational choices and academic culture for both investigators and trainees.
Key Highlights
Gajewski describes how practical challenges in real-world problems catalyzed his transition to Bayesian modeling.
KUMC’s working group integrates FACTS software in collaborative simulation and operational trial planning.
The PAIN-CONTRoLS Trial: multi-arm Bayesian adaptive design, response-adaptive randomization, real-time analysis, and endpoint-driven allocation.
HOBIT trial: Adaptive allocation, fixed control arm proportion, group-sequential interims, ordinal endpoint modeling.
Emphasis on probabilistic, quantitative reporting over binary outcomes in trial analysis and interpretation.
Cultural shift observed among academic collaborators and trainees embracing Bayesian adaptive strategies.