Physician · Scientist · Methodologist
Statistical fragility and robustness in clinical trials.
Methodology for trustworthy evidence in clinical research.
Founder, Internet Medical Journal.
Background
Thomas F. Heston, MD, MSc is a physician-scientist and architect of the p–fr–nb framework — a unified system extending traditional statistical reporting in clinical trials by adding two orthogonal dimensions: fragility (classification stability) and robustness (distance from therapeutic neutrality). The framework has been validated across 129 clinical trials and 720,000 simulated trials.
His 2025 peer-reviewed validation study demonstrated that 50% of clinical trials show discordance between p-value classification and complete evidence assessment — a finding with direct implications for reproducibility in biomedical research. His 2026 Cureus paper formally proved that the fragility index is a partial function, establishing the mathematical foundation for the Global Fragility Index and the broader p–fr–nb framework.
He holds dual faculty appointments as Clinical Assistant Professor of Family Medicine at the University of Washington and Clinical Associate Professor at Washington State University’s Elson S. Floyd College of Medicine. He trained at Saint Louis University School of Medicine (MD, Distinction in Research), Duke University (internship), and Johns Hopkins University (fellowship, molecular imaging).
The Framework
The p–fr–nb framework provides complete statistical evidence assessment in clinical trials. It covers 100% of standard parametric study designs and is implemented in an open-source Python toolkit on Zenodo.
The traditional test of whether a result clears a pre-specified threshold. Necessary but insufficient as a standalone measure of evidence quality in clinical decision-making.
Classification stability: the minimum number of outcome events that must change before a significant result becomes non-significant, or vice versa. Quantifies how precarious the conclusion is.
Distance from therapeutic equivalence: a geometric framework quantifying how far a trial result sits from the point where treatment and control are indistinguishable. Measures robustness directly.
Selected Key Publications
Editorial
The Internet Medical Journal (ISSN 1093-7935) is an open-access, Google Scholar–indexed publication founded to provide rapid scientific commentary for clinical methodology, AI accountability, and evidence-based medicine.
Volume 1, Number 1 launched in 2026 with eight inaugural articles spanning fragility methods, AI accountability, and bioethics. The journal is hosted on Open Journal Systems (OJS) and serves as the primary platform for rapid dissemination of the p–fr–nb framework, commentary responses, and international mentee work.
IMJ accepts submissions in methodology commentary, rapid communications, letters, and brief original research. International collaborators and early-career researchers are especially welcome.
Visit the Internet Medical JournalAll content freely available. Indexed in Google Scholar with permanent DOIs via Zenodo.
Expedited review for methodology commentaries, letters to the editor, and brief communications.
Dedicated platform for early-career researchers in clinical methodology worldwide.
Collaborate
I welcome inquiries from researchers, clinicians, and methodologists interested in statistical fragility, robustness metrics, and clinical trial evidence quality.
Areas of particular interest: research collaboration, co-authorship on spoke keyword papers, mentorship for early-career researchers, editorial board inquiries, and written interview requests.
Written correspondence is strongly preferred. I respond to all substantive research inquiries.
Thank you — I will respond by written correspondence.