Machine Learning Analysis of Post-Acute COVID Symptoms Identifies Distinct Clusters, Severity Groups, and
Medrxiv : the Preprint Server for Health Sciences
|December 3, 2025
Summary
Patient questionnaires reveal distinct long COVID symptom patterns and severity levels. Machine learning helps stratify patients for personalized treatment trials, improving understanding of disease heterogeneity.
Area of Science:
- Utilizes computational linguistics and machine learning for health data analysis.
- Focuses on post-acute sequelae of SARS-CoV-2 infection (PASC) or long COVID.
- Integrates multi-cohort patient-reported outcome data for comprehensive analysis.
Background:
- Electronic health records (EHRs) have often overshadowed patient-reported symptom data for COVID-19 analysis.
- Understanding the heterogeneity of long COVID symptoms is crucial for effective patient stratification.
- Patient questionnaires offer a valuable, low-cost data source for disease phenotyping.
Purpose of the Study:
- To leverage machine learning on questionnaire data to identify distinct long COVID symptom clusters and endotypes.
- To develop a framework for stratifying long COVID patients based on symptom profiles and severity.
- To investigate symptom trajectories, severity correlations, and recovery patterns in long COVID cohorts.
Main Methods:
- Applied topic modeling and unsupervised clustering to de-identified patient questionnaires from four cohorts.
- Mapped identified symptom clusters to organ systems to define endotypes.
- Conducted longitudinal analysis to identify symptom trajectories and severity correlations.
Main Results:
- Identified 9-12 endotypes per cohort, revealing significant heterogeneity in post-COVID-19 symptoms.
- Discovered three distinct symptom trajectories (resolving, persistent, progressive) and three severity levels (mild, moderate, severe).
- Found that individuals with non-mild acute COVID-19 symptoms have a 2.6x higher risk of moderate/severe long COVID.
Conclusions:
- Machine learning analysis of questionnaire data robustly identifies symptom clusters and endotypes.
- This approach provides a framework for stratifying long COVID patients for precision medicine and clinical trial design.
- Findings highlight the importance of patient-reported symptoms in understanding long COVID complexity and guiding treatment strategies.
Related Concept Videos
01:17Classification of Illness
8.5K
The meaning of illness is individualized to each person who experiences an alteration in health. In contrast, disease is a medical term indicating a pathological change in the structure and function of the body or mind. It is a condition that has specific symptoms and boundaries.
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
8.5K
01:18Steps in Outbreak Investigation
470
In the ever-evolving field of public health, statistical analysis serves as a cornerstone for understanding and managing disease outbreaks. By leveraging various statistical tools, health professionals can predict potential outbreaks, analyze ongoing situations, and devise effective responses to mitigate impact. For that to happen, there are a few possible stages of the analysis:
470