Objective:
To investigate the potential of using smartphone data to infer changes in social behaviour in people with Parkinson’s and how this has been impacted by COVID19.

Background: Parkinson’s can lead to social withdrawal but little is known about how this changes with the progression of the condition.  Smartphones are now widely used and the hub of communications. Our initial aim was to explore the social withdrawal of people with Parkinson’s unobtrusively using smartphone data. However, social distancing because of COVID19 provides an opportunity to observe how the pandemic impacts people with Parkinson’s socially.

Method: A specially designed application was installed on six smartphones of people with mild to moderate Parkinson’s (age 63 -75, 2 females, Hoehn and Yahr stage one). It records all smartphone usage including calls, messages, application usage, etc before and during the pandemic (earliest start Oct. 2019 for a year). These usages are analysed by times, length and diversity. And smartphone sensors including Bluetooth, Wi-Fi, locations, microphones are utilized to infer social activities. [EP1] [HZ2] Semi-structured interviews were conducted with participants to confirm these observations.

Results: The smartphone data showed decreased overall social interactions and that participants spent almost all their time at home. Different purposes for leaving home and unique responses to social interaction changes were also revealed by smartphone data. However, face-to-face interactions are not fully replaced by other methods of communication, which may affect participants’ social wellbeing. Location data from smartphones can show which participants do essential shopping themselves. Usage data may suggest negative emotional states and extra attention or interventions could be offered to these individuals.

Conclusion: COVID19 has had an inevitable impact on the social lives of Parkinson’s patients. The smartphone data successfully illustrated these impact and provided individualised views for each participant, which shows its potential in detecting changes. The burden-free passive sensing gives detailed, unobtrusive, continuous descriptions of participants’ social lives. This could be applied to capture changes in behaviour over time produced by changed in their Parkinson’s symptoms, rather than by lockdown, to establish better care.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/using-smartphone-data-to-infer-social-behaviour-in-parkinsons-proof-of-principle-and-the-impact-of-covid19/