Digital Endpoints Are Transforming CNS Clinical Trials
Digital endpoints in CNS trials have moved past the feasibility question. The EMA has already qualified a wearable-derived measure as a primary endpoint, and ICH E6(R3) now explicitly addresses digital health technologies within the global GCP framework. The question for sponsors is no longer whether digital measurement works, but how to design, validate, and operationalize it in a way regulators will accept.
This white paper from Koneksa’s Head of Biomarker Exploration and SVP of Biometrics and Data Science draws on peer-reviewed research, practical program experience, and current regulatory guidance to lay out the scientific and operational foundations of digital measurement in CNS research.
Why CNS Trials Need Digital Endpoints
CNS conditions like Parkinson’s, Alzheimer’s, and MS share a reliance on subjective, episodic, clinician-rated scales that require substantial change to detect progression and misalign with how symptoms actually fluctuate. Sparse clinic visits infer linear trajectories from nonlinear disease dynamics, conflating biological signal with noise. The statistical cost is real: in one simulation cited in the paper, moving from conventional modeling to individual-specific thresholds derived from continuous monitoring cut per-arm sample sizes from roughly 490 to 26.
What’s Inside
The paper works through the biometric decisions that determine whether digital endpoints in CNS trials produce decision-grade evidence: statistical modeling for dense longitudinal data, slope-based versus snapshot endpoints, covariance structure and missing data strategy, machine learning within regulatory constraints, and how raw sensor signals get translated into clinically meaningful constructs, illustrated through a Parkinson’s disease case study.
Download the Full White Paper
Get the complete framework for implementing digital endpoints in CNS clinical trials.
Complete the form to access the full resource.
What You'll Learn
About the Authors
Robert Ellis, Ph.D.
Head of Biomarker Exploration
Robert Ellis, PhD, leads the development of model-driven approaches to solving read more…
Michael Mendoza
Senior Vice President, Biometrics and Data Science
Michael Mendoza leads the company’s biometrics strategy and oversees the design read more…
View the Transcript
00:09
CNS diseases represent one of medicine’s most urgent challenges with the impact of Alzheimer’s disease alone projected to approach seventeen trillion dollars globally by 2050.
00:20
Yet we measure them with infrequent, episodic clinic visits…assuming a simple story in diseases that are anything but.
00:27
Sparse assessments suggest a linear progression. (pause) But that’s not how diseases behave.
00:32
Continuous measurement reveals something very different: non-linear change, overlapping signals, and true disease dynamics.
00:40
Digital endpoints promise a better way. But most approaches miss a critical step.
00:46
One that determines whether data becomes insight… or just noise.
00:51
More data alone isn’t enough. Without the right design, variability can overwhelm the signal, reducing statistical sensitivity instead of improving it.
01:00
Digital measures now capture key domains of CNS function, from gait and tremor to speech, sleep, and cognition, in patients’ daily lives.
01:09
But translating raw digital signals into clinically meaningful endpoints is not straightforward, and correlation alone isn’t enough.
01:17
Because success depends less on the technology (pause) and more on how measurement is designed.
01:23
Download this Koneksa white paper to see what most digital endpoint strategies are missing in CNS clinical trials.
Related Resources
Human-in-the-Loop: Why Experts Matter More in the Age of AI
Meet Lodestar™: Your North Star for Regulatory Submissions
