Objective: The 3DT initiative of the Critical Path for Parkinson’s Consortium (CPP) aims to advance the regulatory maturity of digital biomarkers, for clinical studies in Parkinson’s Disease (PD). This presentation highlights the progress of a Working Group established to identify sources of variability (SoVs) introduced by the implementation of digital measurements.

Background: Efforts to create robust and generalisable digital endpoints for clinical trials are still ongoing. Digital endpoints are affected by SoVs associated with the device used, environmental factors, and quality of information. Robust digital endpoints need to have these SoVs addressed before they can be confidently incorporated in clinical trial protocols. The CPP 3DT WG was established to propose approaches to mitigate SoVs effects.

Method: Drawing from the research literature, an inventory of SoVs was developed with reference to three specific areas of concern, i.e. device, environment, and quality of information. Further, using information from WATCH-PD, an observational cohort study, and supplementary data contributed by WG participants, a data-driven approach was adopted to conduct an in-depth investigation of the SoVs and suggest effective mitigation strategies.

Results: Four factors were explored in depth that related to the core sources of measurement SoVs namely, orientation and on-body sensor placement; number of sensors and sampling rate; environmental factors relating to the subject home setting; and, algorithm selection for classification employed for the prediction of clinical rating scales.

Conclusion: Clinical study protocols can effectively control SoVs affecting the reliability and reproducibility of digital biomarkers, through careful determination of data collection methods and requirements. In particular, quantitative analysis of the effect of key SoVs can determine their implications on recruitment requirements and statistical power calculations.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/identifying-and-controlling-variability-in-digital-health-measures-in-parkinsons-disease-clinical-trials/