Objective: Demonstrate the impact of quality enhancing strategies in the Personalized Parkinson Project.

Background: Increasingly, the scientific community uses data from longitudinal observational patient cohorts to model disease expression, search for novel biomarkers, and understand disease progression. The data quality should be indisputable. As the primary objective of the Personalized Parkinson Project is to make the data available for researchers worldwide, high standards for quality assurance were implemented.

Method: The Personalized Parkinson Project is a single-center cohort study that started in 2017. It is enrolling 650 persons with Parkinson’s disease (diagnosed <5 years) for two years. During annual in-clinic visits, clinimetrics, biospecimens, and imaging data are collected. In addition, participants complete questionnaires and wear a wrist-worn device (the Verily Study Watch) for up to 23 hours per day for at least two years, with the possibility to extend to up to three years. The Study Watch continuously collects various types of physiological and environmental data. From 2020 on, the Study Watch also supports data collection during eight structured motor tasks.

Results: Out of 472 currently enrolled participants, 91% fulfilled a predefined stratification model (based on age, gender, and disease duration). Mean follow-up is 13 months (standard deviation = 6 months). The drop-out rate is 1% (n=5). This excellent retention is enabled by a series of measures: intense patient engagement in designing the study protocol; coupling each participant to their own personal study assessor; and rewarding participants with knowledge (monthly newsletters, regular vlogs from experts and annual participant events). Mean wear time of the Study Watch is 20.2 hours/day in participants who have been enrolled for at least 200 days (n=310). Participants mention the unobtrusive nature and attractive look of the Study Watch as motivational factors to wear the watch continuously. Inter- and intra-rater variability of clinical data are minimized by collecting the data in a single center, with a small team of eight assessors. Finally, a dedicated data storage and sharing infrastructure supports privacy of the study participants.

Conclusion: Data quality enhancing as well as patient engagement strategies are effective and should to be an integral part of longitudinal cohort studies.

« Back to MDS Virtual Congress 2020

MDS Abstracts - https://www.mdsabstracts.org/abstract/the-personalized-parkinson-project-data-quality-enhancing-strategies/