Everyday data for COVID-19 from mHealth devices: The PAIDUR framework

Brown, Colin Wylie

Abstract

This work examines the use of new data from mHealth devices in algorithmic Risk Predictor tools for conditions including the early diagnosis of COVID-19. The earliest signs of COVID-19 and other viral infection were from FitBit® devices, and data from these and other mHealth devices on three research platforms are analysed. MHealth devices are selected for everyday use for lifestyle. However users in healthcare may lack expertise, and staff can process only the least quantity and highest quality of mHealth data. On-device processing helps by deriving simple outputs such as Heart-Rate-over-Steps, but is not standardised. Clinicians have long processed simple data in the mind, as was prototyped for frailty with algorithms designed for mental arithmetic. Qualitative accuracy is analysed here as “Healthcare Veracity”, an issue familiar to clinicians and managed pragmatically, usually by time-serialisation. Such clinical quality assurance can also apply to mHealth devices, whose metrological precision and accuracy exceed that of older technologies. Accuracy further improves when each device is paired to its user. New tools could use machine data directly from local devices, or from remote devices if connected for interoperability. However, new devices require clinicians’ trust, and this has been evaluated in developing the PAIDUR framework:
Precision / Accuracy / Interoperation / Deployment / Use / Reuse
This structures how new mHealth systems can manage socio-technical issues to interoperate their new data across healthcare systems. Consumers’ inexpensive mHealth devices use the same technologies as branded and healthcare-certified products. Mass deployment is already here, and online platforms can already support multiple device types at scale. However, the costs of consumer platforms are also to privacy, which funds the business model, with implications for data quality. This suggests how privacy can be actively managed to keep proven healthcare benefits free at the point-of-care.

• Thesis Type: Thesis
• Deposit Date: Aug 21, 2023
• Publicly Available Date: Aug 21, 2023
• DOI: https://doi.org/10.17869/enu.2023.3175075
• Award Date: Jul 7, 2023