Skip to main content
Top
Published in: BMC Neurology 1/2021

Open Access 01-12-2021 | Parkinson's Disease | Research article

Social listening – revealing Parkinson’s disease over day and night

Authors: Hui Zhang, Fanwen Meng, Xingyu Li, Yali Ning, Meng Cai

Published in: BMC Neurology | Issue 1/2021

Login to get access

Abstract

Background

Nocturnal symptoms in Parkinson’s disease are often treated after management of daytime manifestations. In order to better understand the unmet needs of nocturnal symptoms management, we analyzed the characteristics and burden of nocturnal symptoms from patients’ perspectives and explored their changes over time. Overall symptoms (occurring at day or night) were collected to compare whether the unmet needs related to nocturnal symptoms and to overall symptoms are different.

Methods

We used a Social Listening big-data technique to analyze large amounts of Parkinson’s disease symptoms in dialogues available from social media platforms in 2016 to 2018. These symptoms were classified as either overall symptoms or nocturnal symptoms. We used share of voice (SOV) of symptoms as a proportion of total dialogues per year to reflect the characteristics of symptoms. Negative sentiment score of symptoms was analyzed to find out their related burden.

Results

We found the SOV for overall motor symptoms was 79% and had not increased between 2016 and 2018 (79%, p = 0.5). The SOV for non-motor symptoms was 69% and had grown by 7% in 2018 (p <  0.01). The SOV for motor complications was 9% and had increased by 6% in 2018 (p <  0.01). The SOV of motor symptoms was larger than non-motor symptoms and motor complications (p <  0.01). The SOV of non-motor symptoms was larger than motor complications (p <  0.01). For nocturnal symptoms, 45% of the analyzed PD population reported nocturnal symptoms in 2018, growing by 6% (p <  0.01). The SOV for nocturnal-occurring motor symptoms was higher than most non-motor symptoms. However, non-motor symptoms had the higher increases and evoked higher negative sentiment regardless of whether they occurred during the day or night. For symptoms that can occur at either day or night, each nocturnal symptom was rated with a higher negative sentiment score than the same symptom during the day.

Conclusions

The growing SOV and the greater negative sentiment of nocturnal symptoms suggest management of nocturnal symptoms is an unmet need of patients. A greater emphasis on detecting and treating nocturnal symptoms with 24-h care is encouraged.
Appendix
Available only for authorised users
Literature
5.
18.
go back to reference Lafferty J, McCallum A, Pereira FCN. Conditional random fields: probabilistic models for segmenting and labeling sequence data. In: Proceedings of the 18th international conference on machine learning 2001; 2001. p. 282–9. Lafferty J, McCallum A, Pereira FCN. Conditional random fields: probabilistic models for segmenting and labeling sequence data. In: Proceedings of the 18th international conference on machine learning 2001; 2001. p. 282–9.
Metadata
Title
Social listening – revealing Parkinson’s disease over day and night
Authors
Hui Zhang
Fanwen Meng
Xingyu Li
Yali Ning
Meng Cai
Publication date
01-12-2021
Publisher
BioMed Central
Published in
BMC Neurology / Issue 1/2021
Electronic ISSN: 1471-2377
DOI
https://doi.org/10.1186/s12883-020-02024-4

Other articles of this Issue 1/2021

BMC Neurology 1/2021 Go to the issue