Identifying profiles and symptoms of Long Covid patients: a data mining infodemiology study based on French social media (Preprint)
Authors:
- Amélia DéguilhemADAmélia Déguilhem
- Joelle MalaabJMJoelle Malaab
- Manissa TalmatkadiMTManissa Talmatkadi
- Simon RennerSRSimon Renner
- Pierre FoulquiéPFPierre Foulquié
- Guy FagherraziGFGuy Fagherrazi
- Paul LoussikianPLPaul Loussikian
- Tom MartyTMTom Marty
- Adel MebarkiAMAdel Mebarki
- Nathalie TexierNTNathalie Texier
- Stéphane SchückSSStéphane Schück
Preprint
- 25 May 2022
- preprint
- Published by JMIR Publications Inc. in JMIR Preprints
Abstract
BACKGROUND Long Covid – a condition with persistent symptoms post Covid-19 infection - is the first illness arising from Social Media (SM). Faced with a lack of recognition from medical and official entities, patients formed communities on SM and described their symptoms as long-lasting, fluctuating and multi-systemic. While many studies on Long Covid relied on traditional research methods that include lengthy processes, SM offers a foundation for large-scale studies with a fast-flowing outburst of data. OBJECTIVE To identify and analyze Long Haulers’ main reported symptoms, symptom co-occurrences, topics of discussion, difficulties encountered, and patient profiles. METHODS Data extraction was performed based on a list of pertinent keywords. Reported symptoms were identified via the MedDRA dictionary, displayed according to the volume of posts mentioning them, then aggregated at the user level. Associations were assessed by computing co-occurrences in users’ messages, as pairs of Preferred Terms (PTs). Discussion topics were analyzed using the Biterm Topic Modeling; difficulties and unmet needs were explored manually. To identify patient profiles in relation to their symptoms, the total of each PT was used to create hierarchal clusters at the user level. RESULTS A total of 15 364 messages were identified as originating from 6 494 patients of Long Covid or their caregivers. Our analyses revealed three major symptom co-occurrences: asthenia|dyspnea (35,3%), asthenia|anxiety (22.5%), and asthenia|headaches (17.3%). The main reported difficulties were “the management of symptoms” (35.4% of messages), “psychological impact” (15.1%), “significant pain” (12,0%), “deterioration in the general well-being (12,3%), and “impact on daily and professional life” (9,4% and 8,0% of messages, respectively). We identified three profiles of patients in relation to their symptoms: Profile A (n= 406 patients) reported exclusively a symptom of Asthenia (100% of patients), Profile B (n= 129 patients) expressed Anxiety (100% of patients), Asthenia (22%), Dyspnea (12%), Ageusia (2%); and Profile C (n=141 patients) described Dyspnea (100% of patients) and Asthenia (32%). About 50% of users continued expressing symptoms after more than 6 months post-infection, and 16.1% after 1 year. CONCLUSIONS Long Covid is a lingering condition that affects people across the globe, physically and psychologically. It impacts Long Haulers’ quality of life, everyday tasks, and professional activities. Social Media played an undeniable role in raising and delivering Long Haulers’ voices. It also has the potential to rapidly provide large volumes of valuable patient-reported information. Considering the fact that Long Covid was a self-titled condition by patients themselves via SM, it is imperative to continuously include their perspectives in related research. The data presented in this study can help design patient-centric instruments to be further used in clinical practice to better capture meaningful dimensions of Long Covid.Keywords
This publication has 42 references indexed in Scilit:
- Twitter as a Tool for Health Research: A Systematic ReviewAmerican Journal of Public Health, 2017
- Evaluating Google, Twitter, and Wikipedia as Tools for Influenza Surveillance Using Bayesian Change Point Analysis: A Comparative AnalysisJMIR Public Health and Surveillance, 2016
- Social media use in healthcare: A systematic review of effects on patients and on their relationship with healthcare professionalsBMC Health Services Research, 2016
- Identifying Adverse Effects of HIV Drug Treatment and Associated Sentiments Using TwitterJMIR Public Health and Surveillance, 2015
- Disease Detection or Public Opinion Reflection? Content Analysis of Tweets, Other Social Media, and Online Newspapers During the Measles Outbreak in the Netherlands in 2013Journal of Medical Internet Research, 2015
- Cumulative Query Method for Influenza Surveillance Using Search Engine DataJournal of Medical Internet Research, 2014
- Analysing Twitter and web queries for flu trend predictionTheoretical Biology and Medical Modelling, 2014
- Postmarket Drug Surveillance Without Trial Costs: Discovery of Adverse Drug Reactions Through Large-Scale Analysis of Web Search QueriesJournal of Medical Internet Research, 2013
- A biterm topic model for short texts2013
- Prise en charge d'un patient adulte présentant une hémorragie méningée standard suite à une rupture d'anévrisme2011