Giving a Voice to Patients With Smell Disorders Associated With COVID-19: Cross-Sectional Longitudinal Analysis Using Natural Language Processing of Self-Reports

Nick S Menger ¹, Arnaud Tognetti ^2,3, Michael C Farruggia ⁴

¹Institute of Medical Psychology and Behavioural Neurobiology, University of Tübingen, Tübingen, Germany.
²Department of Clinical Neuroscience, Division of Psychology, Karolinska Institutet, Stockholm, Sweden.
³Centre d'Economie de l'Environnement Montpellier, Centre National de la Recherche Scientifique, Institut National de Recherche pour l'Agriculture l'Alimentation et l'Environnement, Institut Agro, Université de Montpellier, Montpellier, France.
⁴Interdepartmental Neuroscience Program, Yale University, New Haven, CT, United States.
⁵Department of Molecular Medicine, University of Padova, Padua, Italy.
⁶School of Exercise and Nutritional Sciences, San Diego State University, San Diego, CA, United States.
⁷Department of Neurobiology and Behavior, University of California, Irvine, CA, United States.
⁸Department of Developmental and Educational Psychology, University of Basque Country, Leioa, Spain.
⁹Department of Anatomy, Howard University College of Medicine, Washington, DC, United States.
¹⁰Biological Sciences Department, Fisher College of Science and Mathematics, Towson University, Towson, MD, United States.
¹¹Goethe University of Frankfurt, Frankfurt am Main, Germany.
¹²Laboratoire Évolution et Santé Orale, Université Toulouse III, Toulouse, France.
¹³Department of Medicine, School of Medicine, Indiana University, Indianapolis, IN, United States.
¹⁴Language and Communication Department, Faculty of Arts, Radboud University, Nijmegen, Netherlands.
¹⁵Centre for Language Studies, Radboud University, Nijmegen, Netherlands.

⁰Institute of Medical Psychology and Behavioural Neurobiology, University of Tübingen, Tübingen, Germany.

Jmir Public Health and Surveillance +

|

May 10, 2024

View abstract on PubMed

Summary

COVID-19 survivors often experience prolonged smell disorders like parosmia and hyposmia. Analysis of patient reports reveals distinct keywords associated with persistent smell dysfunction in long-haulers.

Area Of Science

Medical research
Linguistics
Data science

Background

COVID-19 frequently causes smell disorders, including anosmia, hyposmia, parosmia, and phantosmia.
These smell dysfunctions can persist long after initial infection resolution.
Patient difficulty in describing smell symptoms complicates diagnosis and requires further investigation.

Purpose Of The Study

To explore patient concerns regarding smell disorders post-COVID-19.
To analyze self-reported data using natural language processing to understand smell dysfunction.
To extend knowledge on COVID-19-related smell disorders using a large dataset.

Main Methods

Utilized longitudinal survey data from 1560 participants collected by the Global Consortium for Chemosensory Research.
Compared self-reported smell disorder descriptions from long-haulers and non-long-haulers at two time points.
Employed machine learning to classify smell disorders, analyze sentiment, and identify salient keywords.

Main Results

Parosmia and hyposmia were significantly more prevalent in long-haulers.
A significant relationship was identified between long-hauler status and the sentiment of self-reported smell changes.
Specific keywords (e.g., fire, gas, wine, vinegar) were more characteristic of long-hauler reports.

Conclusions

Online self-reports provide valuable data for understanding smell disorders.
Findings align with previous research and offer new insights for future smell disorder investigations.
Further research into self-reported data can enhance healthcare understanding and patient support for smell dysfunctions.

Keywords:

COVID-19 anosmia hyperosmia hyposmia long-hauler parosmia phantosmia sentiment classification smell disorders web-based study