Dataset Information

Monitoring Depression Trends on Twitter During the COVID-19 Pandemic: Observational Study.

ABSTRACT:

Background

The COVID-19 pandemic has affected people's daily lives and has caused economic loss worldwide. Anecdotal evidence suggests that the pandemic has increased depression levels among the population. However, systematic studies of depression detection and monitoring during the pandemic are lacking.

Objective

This study aims to develop a method to create a large-scale depression user data set in an automatic fashion so that the method is scalable and can be adapted to future events; verify the effectiveness of transformer-based deep learning language models in identifying depression users from their everyday language; examine psychological text features' importance when used in depression classification; and, finally, use the model for monitoring the fluctuation of depression levels of different groups as the disease propagates.

Methods

To study this subject, we designed an effective regular expression-based search method and created the largest English Twitter depression data set containing 2575 distinct identified users with depression and their past tweets. To examine the effect of depression on people's Twitter language, we trained three transformer-based depression classification models on the data set, evaluated their performance with progressively increased training sizes, and compared the model's tweet chunk-level and user-level performances. Furthermore, inspired by psychological studies, we created a fusion classifier that combines deep learning model scores with psychological text features and users' demographic information, and investigated these features' relations to depression signals. Finally, we demonstrated our model's capability of monitoring both group-level and population-level depression trends by presenting two of its applications during the COVID-19 pandemic.

Results

Our fusion model demonstrated an accuracy of 78.9% on a test set containing 446 people, half of which were identified as having depression. Conscientiousness, neuroticism, appearance of first person pronouns, talking about biological processes such as eat and sleep, talking about power, and exhibiting sadness were shown to be important features in depression classification. Further, when used for monitoring the depression trend, our model showed that depressive users, in general, responded to the pandemic later than the control group based on their tweets (n=500). It was also shown that three US states-New York, California, and Florida-shared a similar depression trend as the whole US population (n=9050). When compared to New York and California, people in Florida demonstrated a substantially lower level of depression.

Conclusions

This study proposes an efficient method that can be used to analyze the depression level of different groups of people on Twitter. We hope this study can raise awareness among researchers and the public of COVID-19's impact on people's mental health. The noninvasive monitoring system can also be readily adapted to other big events besides COVID-19 and can be useful during future outbreaks.

SUBMITTER: Zhang Y

PROVIDER: S-EPMC8330892 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundWith the spread of COVID-19, significant concerns have been raised about the potential increased risk for electronic cigarette (e-cigarette) users for COVID-19 infection and related syndromes. Social media is an increasingly popular source for health information dissemination and discussion, and can affect health outcomes.ObjectiveThis study aims to identify the topics in the public vaping discussion in COVID-19-related Twitter posts in order to get insight into public vaping-related perceptions, attitudes and concerns, and to discern possible misinformation and misconceptions around vaping in the COVID-19 pandemic.MethodsUsing the tweets ID database maintained by Georgia State University's Panacea Lab, we downloaded the tweets related to COVID-19 from March 11, 2020, when the World Health Organization declared COVID-19 a pandemic, to February 12, 2021. We used R to analyze the tweets that contained a list of 79 keywords related to vaping. After removing duplicates and tweets created by faked accounts or bots, the final data set consisted of 11,337 unique tweets from 7,710 different users. We performed the latent Dirichlet allocation (LDA) algorithm for topic modeling and carried out a sentiment analysis.ResultsDespite fluctuations, the number of daily tweets was relatively stable (average number of daily tweets = 33.4) with a sole conspicuous spike happening on a few days after August 11, 2020 when a research team published findings that teenagers and young adults who vape face a much higher risk of COVID-19 infection than their peers who do not vape. Topic modeling generated 8 topics: linkage between vaping and risk of COVID-19 infection, vaping pneumonia and the origin of COVID-19, vaping and spread of COVID-19, vaping regulation, calling for quitting vaping, protecting youth, similarity between e-cigarette or vaping-associated lung injury (EVALI) and COVID-19, and sales information. Daily sentiment scores showed that the public sentiment was predominantly negative, but became slightly more positive over the course of the study time period.ConclusionsWhile some content in the public discourse on vaping before the COVID-19 pandemic continued in Twitter posts during the COVID-19 time period, new topics emerged. We found a substantial amount of anti-vaping discussion and dominantly negative sentiment around vaping during COVID-19, a sharp contrast to the predominantly pro-vaping voice on social media in the pre-COVID-19 period. Continued monitoring of social media conversations around vaping is needed, and the public health community may consider using social media platforms to actively convey scientific information around vaping and vaping cessation.

Project description:BackgroundSince the beginning of the coronavirus disease 2019 (COVID-19) outbreak, the Critical Care Outreach Team (CCOT) remained operational to provide critical care support to acutely ill and deteriorating patients on the wards. We aimed to evaluate the demand and efficacy of the critical care outreach service during the COVID-19 pandemic.MethodWe prospectively evaluated all patients referred to critical care outreach enrolled during a twelve-month period. We reported the cumulative number of activities and interventions and baseline characteristics, acuity level and patients' clinical outcome. The rate of ICU admissions, activity plan, patients' acuity and mortality are compared to historical data pre-pandemic.ResultsAmongst 4849 patients referred, 3913 had a clinical review and of those 895 were COVID-19 positive. Non-invasive ventilation was mostly delivered to COVID-19 patients (COVID-19 +VE: 853/895, 95% vs. COVID-19 -VE: 119/3018, 4%) alongside awake prone positioning (COVID-19 +VE: 232/895, 26% vs. COVID-19 -VE: 0/3018, 0%). Compared to pre-pandemic, the cumulative number of patients assessed increased (observed: 3913 vs. historical: 3615; p = 0.204), patients meeting Level 2 acuity were higher (observed: 51% vs. historical: 21%; p = 0.003), but ICU admission rate did not increase significantly (observed: 12% vs. historical: 9%; p = 0.065), and greater mortality rate (observed: 14% vs. historical: 8%; p = 0.046) was observed.ConclusionCritical care outreach bridges the gap between the intensive care unit and general wards and supports the concept of 'critical care without walls' acting as a valuable resource in optimizing and triaging acutely unwell patients and potentially averting critical care admissions.Relevance to clinical practiceThe COVID-19 pandemic has generated an unprecedented surge of deteriorating and critically ill patients with has caused severe and sustained pressures on intensive care units (ICUs) and general wards. Acutely ill patients can deteriorate quickly, and early recognition is vital to commence critical intervention on the wards or transfer timely to ICU. The Critical Care Outreach Team can help staff and optimize acutely ill and deteriorating patients by providing timely critical care interventions at the patient bedside.