In the Interim...

In this episode of "In the Interim…", Dr. Scott Berry shares the framework he presented at a CSRC regulatory science hackathon on digital twins and synthetic controls in cardiovascular clinical trials. He separates three kinds of models that are often given the same name: population-level statistical models (a historical control rate), patient-level statistical models (propensity matching, Cox models, hot-deck imputation, and black-box machine learning predictions), and system-level digital twins of human biology. Drawing on ARPA-H's CIRCLE program for ICU immune modulation, the Framingham risk model, the EMA-qualified use of AI-generated prognostic scores as covariates, and Novartis's recent Lp(a) outcomes trial, Scott examines what each type of model can and cannot replace in a trial. Dr. Berry argues that virtual controls should provide a full predictive distribution for each patient rather than a single forecast, and that Bayesian hierarchical models with prospectively enrolled controls allow a trial to verify those predictions as it runs. He closes with how much smaller trials could become, and why clinical trial science must adapt as natural history models improve.

Key Highlights

• Reserving "digital twin" for system-level models of biology, not statistical predictions such as the Framingham risk model.
• ARPA-H's CIRCLE program: immune-system digital twins to guide ICU therapy, targeting a 25% reduction in ICU stay.
• Why a validated system-level digital twin would remove the need for a clinical trial.
• Prognostic scores as covariates versus virtual controls, and the counterfactual a control group is meant to supply.
• From a single historical control rate to patient-specific predictions built on patient-level data.
• Predictive distributions, not point forecasts, as the basis for comparing treated patients with virtual controls.
• Bayesian hierarchical models, MAP priors, and prospective controls to verify and calibrate virtual controls during the trial.

For more, visit us at https://www.berryconsultants.com

In this episode of "In the Interim…", Dr. Scott Berry recounts a presentation he gave to the Speak Foundation, a patient organization focused on limb-girdle muscular dystrophy and makes the case for master protocols directly to the stakeholder trials exist to serve. Opening with the deliberately absurd premise of a sports league that constructs a four-billion-dollar stadium for every single game and demolishes it at the final whistle, Scott reframes the conventional one-off trial as a two-year construction project dismantled the moment the contest ends. The episode distinguishes platform trials and basket trials as standing "arenas," explains the modular role of the master protocol described in Woodcock and LaVange's 2017 New England Journal of Medicine paper, and draws on worked examples including I-SPY 2, the HEALEY ALS Platform Trial, GBM AGILE, Precision Promise, and the ROAR basket trial. Scott takes each stakeholder in turn from the treatment arm sponsor, the site, the regulator, and the patient. Scott lays out how three-to-one randomization lets an arm fund 133 patients while reading out inferences with more than 200, how 667 patients in a single platform can do the work of 1,000 across five separate trials, and why far fewer patients are assigned to placebo. He closes on the genuine bottleneck: the first arena is the heavy lift, and in rare disease it is frequently patients themselves who make it happen.

Key Highlights

• The one-game stadium analogy of two years to build a trial, one contest, then demolition.
• Platform trials and basket trials as standing arenas for multiple arms and multiple subtypes.
• Master protocols as modular documents, with arms plugging in and out as appendices.
• Worked examples from I-SPY 2, HEALEY ALS, GBM AGILE, Precision Promise, and the ROAR basket trial.
• A stakeholder-by-stakeholder view: arm sponsor, site, regulator, and patient.
• 667 patients in one platform versus 1,000 across five separate trials, with far fewer on placebo.
• Bayesian borrowing across 39 limb-girdle subtypes, and the rare-disease shots never otherwise taken.

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry talks with Dr. Alun Bedding, a professional certified coach and statistician who spent decades in pharmaceutical statistics, including as global head of statistical methods, collaboration, and outreach at Roche, before founding his own coaching and consulting practice. The conversation traces Alun's path from statistician to coach and unpacks what coaching is: a forward-facing partnership built on exploring a client's own thinking, not on advice, mentoring, or therapy. Alun and Scott draw a direct line between good coaching and good statistical consulting as both begin by questioning the stated goal rather than accepting the first request at face value, whether that request is "I want to be a better manager" or "I need a power calculation." The episode covers why coaching must be uncomfortable to produce real change, why self-coaching runs into the limits of blind spots and self-bias and where AI coaching tools genuinely help and where they fall short.

Key Highlights

• Coaching is a partnership exploring a client's own thinking, not advice or mentoring.
• The parallel between coaching and statistical consulting: determining the real goal behind a request.
• Why effective coaching must feel uncomfortable to produce genuine progress.
• The limits of self-coaching, and why blind spots require an outside challenger.
• How sports and performance coaching shifted from instruction to inquiry.
• Where AI coaching helps and where human presence and challenge remain irreplaceable.
• The difference between coaching and therapy: forward-facing versus backward-facing.

For more, visit us at https://www.berryconsultants.com/

In this episode of “In the Interim…,” Dr. Scott Berry talks with Dr. Roger Lewis, Dr. Anna McGlothlin, and Dr. Nick Berry — all co-authors on the CHIPS trial results recently published in JAMA (Spinella et al., August 2026) — about the design, implementation, and results. The conversation covers why room-temperature platelets, which can be stored only five to seven days, leave rural hospitals, low-volume centers, and military and disaster settings without a reliable supply, and how a Bayesian adaptive design was used to find the maximum safe cold-storage duration rather than testing a single fixed duration. Roger, Anna, and Nick walk through the monotonic dose-response model that governed escalation, an unplanned mid-trial complication when the FDA independently authorized fourteen-day cold storage, and how the Data Safety Monitoring Board reviewed results within days of each interim. The trial ultimately demonstrated non-inferiority of cold-stored platelets out to twenty-one days with a Bayesian probability greater than 99.9%, offering a path to expanding platelet access in settings where it was previously very challenging.

Key Highlights

• Cold-stored platelets tested as a potential answer to platelet shortages in rural, low-volume, and military/disaster settings.
• Bayesian adaptive design used to find the maximum safe cold-storage duration, not just test a single fixed duration.
• A monotonic dose-response model constrained escalation to a pre-specified, safety-first ladder across four interim analyses.
• Mid-trial complication: an independent FDA decision allowing 14-day cold storage, absorbed into the design without unblinding.
• Non-inferiority demonstrated out to 21 days of cold storage, with a Bayesian probability greater than 99.9%.
• Interim data turned around by the unblinded implementation team in five business days, versus the six weeks often assumed for adaptive trials.
• Implications for platelet access in disaster, military, and low-volume hospital settings, and for how shelf-life-dependent products are tested going forward.

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry examines three recent health-research headlines through a statistician's lens: an Adventist Health Study-2 analysis claiming eggs reduce Alzheimer's risk, an Emory University trial on high-dose vitamin D and cognition (Zhao et al.), and a British Medical Journal study led by Dr. Daniel Danishvar (Harvard, Boston University) reporting that 25 to 97 percent of deceased NFL players showed evidence of chronic traumatic encephalopathy (CTE). After flagging multiplicity and small-sample issues in the first two studies, Scott spends most of the episode on the CTE study's central flaw: because CTE can only be diagnosed after death, the published prevalence is calculated from a sample of deceased players rather than the full population of NFL players — a distinction he illustrates using a thought experiment on sudden infant death syndrome (SIDS) and a breakdown of the study's own age-stratified death data. He traces the resulting bias, a form of differential mortality, to a single caveat buried deep in the study's limitations section.

Key Highlights

• Adventist Health Study-2's "27% fewer Alzheimer's diagnoses in egg-eaters" finding, and why it's likely multiplicity-driven and observational, not causal.
• The Emory University vitamin D study (Zhao et al.): a 13% MoCA improvement drawn from roughly eight patients, presented as a headline finding.
• The British Medical Journal NFL CTE study (Danishvar et al., Harvard / Boston University): a headline prevalence of "25% to 97%" of NFL players.
• Why 215 of 235 donated brains (97.7%) is a hugely biased numerator — CTE is only diagnosed posthumously, and families of symptomatic players are most likely to donate.
• The corrected denominator: 878 NFL players who died between 2016 and 2021, yielding roughly 24.5% — still biased, because CTE itself accelerates death.
• A SIDS thought experiment showing how building a "population" from those who have already died systematically overstates prevalence through a form of differential mortality bias.
• The key limitation, buried three-quarters of the way through the paper's limitations section, quietly mentions the selection bias behind the headline number.

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim...", Dr. Scott Berry dissects prevailing concepts of “clinically meaningful difference” in clinical trials focusing on progressive diseases. Through detailed examples from pancreatic cancer (Ben Sasse, Revolution Medicines), emphysema (Elevair), Alzheimer’s disease (lecanemab), and IVF, Scott challenges the adequacy of the population-mean of a continuous outcome in reflecting true patient benefit. The episode discusses inconsistent usage and interpretation of acronyms such as MCID, CSD, and Target Product Profile (TPP). Dr. Berry advises trialists to resist interpreting the mean difference using patient-level minimal effects, and adopt responder analyses and cumulative probability approaches to enhance patient-level relevance. Guidance is offered for analyzing the effect of time-saved instead of a mean differences in a clinical endpoint at a single time for progressive diseases – measuring “sweet time.”

Key Highlights

• Focus on added time not the change from baseline as the most meaningful outcome for a progressive disease.
• In-depth evaluation of MCID, CSD, TPP, and risk of misinterpretation.
• Critique of trying to interpret mean-based endpoints for clinical meaningfulness such as six-minute walk distance and CDR sum of boxes.
• FDA advisory panel guidance on MCID for IVF live birth endpoints and dichotomous versus continuous endpoints.
• Advocacy for responder analyses and cumulative probability of achieving thresholds in reporting the clinical effect of a treatment.

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Don Berry provides a detailed account of co-chairing the 1997 NIH consensus development conference on mammography for women in their 40s. His conversation with Dr. Scott Berry covers the statistical and clinical complexities of breast cancer screening, addressing lead time and length bias, trial design limitations. Don discusses the panel’s finding, based on randomized trials and meta-analysis, that the average benefit of screening women in their 40s is modest (an estimated 1.4-day average life extension and 18% hazard reduction). The panel recommended individualized decision-making rather than universal screening. The episode follows the reaction: heated debate with radiologists, scrutiny from journalists and policymakers, Senate testimony, and personal threats. Don explains how these findings became distilled into soundbites, evidenced by coverage in The New York Times, Chicago Tribune, and a reference in James Patterson’s Murder Games. The conversation addresses overdiagnosis, false positives, and the consequence some called the “Berry effect”—an observed drop in mammogram rates after the panel’s recommendations.

Key Highlights

• Don Berry’s NIH consensus conference leadership and approach to breast cancer screening recommendations.
• Statistical bias and trial limitations inherent in mammography evidence.
• Meta-analysis findings: 18% hazard reduction; 1.4-day average life extension.
• Policy guidance supporting individualized patient decision-making over universal screening.
• Strong backlash from advocacy, media, radiology, and government—including Senate hearings and personal threats.
• Enduring debates, public misunderstandings, and continued citation of these events in scientific and popular sources.

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry speaks with Dr. Srinivas Murthy, Dr. Thomas Hills, and Dr. Lindsay Berry about the REMAP-CAP trial results on oseltamivir in critically ill influenza patients. The trial used a Bayesian covariate-adjusted platform design and found oseltamivir was not effective at reducing 90-day mortality with a “98% and 99% probability of harm in 90-day mortality” compared to control. Covariate adjustment addressed baseline and site variation. Subgroup analyses showed greater harm in patients with higher illness severity. Sensitivity analyses using alternative neutral, optimistic, and pessimistic priors produced important scientific exploration of the results. No evidence was found for benefit over control in any subgroup.

The discussion highlights the first randomized, controlled evidence in this patient group, contrasting prior observational studies and clinical guidelines. A mechanism of harm remains unclear. REMAP-CAP is continuing enrollment in moderate severity and pediatric cohorts to further examine population-specific effects. The episode also addresses the broader challenges of trial design and interpretation in acute care research, the limitations of nonrandomized evidence, and the importance of ongoing Bayesian analyses and transparent reporting.

Key Highlights

• Pre-print is available: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7172531
• REMAP-CAP platform trial, Bayesian logistic regression, covariate adjustment
• Oseltamivir arms: statistical trigger for inferiority, “98 and 99% probability of harm”
• Greater harm in sicker subgroups, consistent results across sensitivity analyses
• Ongoing arms: moderate severity and pediatric cohorts, mechanistic questions unresolved
• Context: limitations of previous historical data studies, clinical practice impact, future research directions

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry leads a comprehensive discussion of the ICECAP trial results with four Principal Investigators: Dr. Will Meurer (Professor, Emergency Medicine and Neurology, University of Michigan; consultant to Berry Consultants), Dr. Robert Silbergleit (Professor, Emergency Medicine, University of Michigan Medical School), Dr. Romer Geocadin (Professor, Neurology, Neurosurgery, and Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine), and Dr. Sharon Yeatts (Professor of Biostatistics, Public Health Sciences, Medical University of South Carolina). The panel dissects the ICECAP trial’s multi-arm Bayesian adaptive design, response-adaptive randomization, and population-level approach to cooling duration after out-of-hospital cardiac arrest. Emphasis is placed on methodological transparency, direct operational experience, absence of evidence for incremental benefit beyond six hours of cooling, and future direction for neurocritical care trials.

Key Highlights

• Detailed review of adaptive design methodology, Bayesian interim analyses, and stopping criteria for futility based on posterior probabilities
• Analysis of flat duration-response curve: no clinical benefit seen for extended hypothermia, trial triggered to stop per prespecified rule
• Cohort discussion: representation of U.S. epidemiology, inclusion of heterogeneous etiologies (notably respiratory and overdose) and bystander CPR rates
• Operational challenges: running frequent interim analyses, maintaining trial integrity during COVID-19, site-level differences, statistical reporting timelines
• Panel consensus on the need for continued equipoise in temperature management, caution against misinterpretation, and priority for precision subgroups in future research
• Directions: implementation lessons, pediatric ICECAP, PRECISE-CAP phenotyping study, ongoing subgroup analyses

For more, visit us at https://www.berryconsultants.com/

In this episode of "In the Interim…", Dr. Scott Berry and Dr. Elizabeth Lorenzi systematically examine the analytic pitfalls in recent acute ischemic stroke trials, especially the implications of violating the proportional odds assumption on the modified Rankin Scale. The discussion draws on the DISCOUNT, INSTANT, ESCAPE-MeVo, and ORIENTAL-MeVo trials, spotlighting frequent reactive shifts to proportional odds violations. Scott and Liz detail how such approaches obscure clinically relevant heterogeneity and react by creating analysis methods that obscure the clinical relevance of a violation of proportional odds. The episode underscores the necessity for trial designs that explicitly address heterogeneity of treatment effect. Listeners gain an unvarnished critique of prevailing reporting practices and an actionable vision for future stroke trial designs.

Key Highlights

• DISCOUNT trial’s interim analysis and futility stopping.
• Issues with endpoint dichotomization after proportional odds violations.
• Comparison across multiple recent stroke trials with inconsistent endpoint definitions.
• Obscuring of proportional odds violations, which may be the most important result of the trial.
• Adaptive strategies in STEP platform.

For more, visit us at https://www.berryconsultants.com/