Clinical reports for longitudinal monitoring of Parkinson's Disease

Abstract

Parkinson’s disease (PD) is a complex chronic neurodegenerative disease and achieving an optimal management is a rather difficult task, due to the fluctuating nature of the disease and the occasional appointments in-clinic. To ensure patients have the best quality of life, it is necessary to obtain a clear overview of full disease expression. Wearable devices are a promising solution for an accurate and continuous monitoring of symptoms, without compromising patient’s everyday activities and comfort. Currently, despite the oversupply of metrics to accurately and obtrusively monitor PD, a big portion of the data extracted from these systems is not being rightfully displayed and therefore, the utility of the technologies falls short. In this work, we seek to provide an overview of the sensor-based systems that generate clinical reports. Three databases (PubMed/MEDLINE, IEEE and ACM) were systematically searched for papers published until March 2021. Study eligibility criteria included: people with a PD diagnosis; the use of technology to monitor disease; data visualization reports. 38 were selected for data extraction. There were wide varieties of participant sample sizes (5 up to a hundred) and study duration (punctual tasks up to longitudinal study of 1 year). The metrics evaluated by technologies included motor symptoms such as, tremor, gait, tremor, activity, bradykinesia, dyskinesia and activities of daily living, as well as non-motor expressions, e.g., sleep, cognition or depression. Data visualization is presented in web or mobile applications. Involvement of the users in the design process improves user experience and utility of presented data. Secondly, we conducted a participatory study with 15 healthcare professionals from seven different areas of expertise, working at a tertiary specialised movement disorders center in Portugal. The sessions were divided in three parts: understanding current situation, ideating future solutions and creating ideal dashboard. We adopted thematic analysis with inductive and deductive coding. General demographic information and PD-related events were agreed to be of paramount importance and should be common in all reports; other metrics should be adapted per report due to the contrasting range of metrics each area evaluates. This work shows there is a paucity of clinical reports for longitudinal monitoring of PD and consequently, healthcare professionals lack the full picture of disease expression.