The USD FDA's four medical product centers — the Center for Drug Evaluation and Research (CDER), Center for Biologics Evaluation and Research (CBER), Center for Devices and Radiological Health (CDRH), and Oncology Center of Excellence (OCE) — jointly published a cross-center paper in August 2026 establishing key considerations for the development and use of digitally derived measures (DDMs) in clinical investigations. The publication synthesizes existing FDA guidance into a unified framework, signaling coordinated agency-wide expectations for sponsors seeking to incorporate wearable sensors, AI-enabled platforms, and other digital health technologies (DHTs) as sources of clinical trial endpoints. The paper is not a binding guidance document but reflects current agency thinking.
What it covers
The paper defines DDMs as measures derived from data collected via DHTs — including wearables, sensors, and AI-enabled platforms — and addresses their potential use as clinical outcome assessments (COAs), biomarkers, or components of multimodal endpoints. The framework establishes a three-stage evidentiary pathway: verification (confirming that the DHT accurately measures the underlying physical or chemical parameter), analytical validation (demonstrating that the DHT appropriately captures the clinical event or characteristic in the target population), and clinical validation (confirming that the DDM reflects a meaningful aspect of health and tracks clinically relevant change).
The paper introduces a tiered evidence standard calibrated to regulatory context — DDMs serving as primary endpoints in pivotal studies are expected to have prospective validation with pre-specified performance thresholds, while exploratory endpoints may rely on retrospective or bridging evidence. For AI-driven DDMs, the paper explicitly links credibility assessment requirements to FDA's draft guidance on AI in regulatory decision-making for drugs and biologics, requiring risk-proportionate evaluation of model performance. The framework also addresses usability studies, error characterization, and the management of software and algorithm updates mid-trial.
Why it matters
The paper reduces regulatory ambiguity for drug and device sponsors around what evidence standard a digital endpoint must meet before FDA will accept it in a pivotal study. By anchoring DDM requirements to the existing Patient-Focused Drug Development (PFDD) guidance series and the Biomarkers, EndpointS, and other Tools (BEST) biomarker framework while layering in DHT-specific considerations, the agency has provided a structured evidentiary pathway. The explicit tiering of validation requirements by endpoint role gives sponsors a practical basis for phasing DDM development investments across their pipeline.
The paper's emphasis on patient and caregiver involvement in defining "meaningful aspects of health" reinforces that digital endpoints cannot be justified on technical grounds alone — conceptual grounding in patient experience is framed as a prerequisite. For decentralized trial sponsors, the paper's guidance on remote-versus-clinic measurement comparability and user burden assessment addresses a persistent regulatory gap. The cross-center authorship signals that FDA intends DDM standards to apply consistently across drug, biologic, and device review pathways, reducing the risk of divergent expectations across review divisions. This publication arrives alongside a related FDA request for public input on regulatory frameworks for generative AI-enabled medical devices, indicating a broader agency effort to establish coherent standards across AI-dependent clinical and regulatory tools.
What to watch
- Early engagement is now effectively required. The paper explicitly instructs sponsors to engage with the relevant FDA review division before incorporating a DDM into a clinical investigation for medical product development. Sponsors should factor pre-submission timelines into DDM development plans accordingly.
- AI algorithm credibility assessments will be scrutinized. The cross-reference to FDA's draft guidance on AI in regulatory decision-making signals that AI-derived DDMs will face a formal, risk-stratified credibility review. Sponsors using machine learning models to generate endpoints should monitor finalization of that draft guidance, as it will directly govern the evidentiary bar for AI-enabled DDMs.
- Mid-trial technology updates carry compliance risk. The paper requires sponsors to validate that operating system and algorithm updates do not affect DDM outputs — an immediate consideration for trials using consumer devices subject to automatic software updates. Sponsors should audit DHT vendor agreements and trial protocols for update-management provisions.
- Qualification pathways for novel DDMs remain underdeveloped. The paper acknowledges that some novel DDMs will have no appropriate reference measure for direct comparison and defers to division-level discussion in those cases. The absence of a formal DDM qualification pathway analogous to FDA's biomarker qualification program remains an open structural question.
Source attribution
FDA — Key Considerations for the Development and Use of Digitally Derived Measures for Clinical Investigations, August 2026.
https://www.fda.gov/medical-devices/digital-health-center-excellence/key-considerations-development-and-use-digitally-derived-measures-clinical-investigations-fda-paper
Access the AllSci platform to explore the science behind the news.
Spot something wrong? Report an issue with this article