The Long COVID Symptoms and Severity Score: Development, Validation, and Application

Abstract

Objectives

The primary focus of this research is the proposition of a methodological framework for the clinical application of the long COVID symptoms and severity score (LC-SSS). This tool is not just a self-reported assessment instrument developed and validated but serves as a standardized, quantifiable means to monitor the diverse and persistent symptoms frequently observed in individuals with long COVID.

Methods

A 3-stage process was used to develop, validate, and establish scoring standards for the LC-SSS. Validation measures included correlations with other patient-reported measures, confirmatory factor analysis, Cronbach’s α for internal consistency, and test-retest reliability. Scoring standards were determined using K-means clustering, with comparative assessments made against hierarchical clustering and the Gaussian Mixture Model.

Results

The LC-SSS showed correlations with EuroQol 5-Dimension 5-Level (r_s = −0.55), EuroQol visual analog scale (r_s = −0.368), Patient Health Questionnaire-9 (r_s = 0.538), Beck Anxiety Inventory (r_s = 0.689), and Insomnia Severity Index (r_s = 0.516), confirming its construct validity. Structural validity was good with a comparative fit index of 0.969, with Cronbach’s α of 0.93 indicating excellent internal consistency. Test-retest reliability was also satisfactory (intraclass correlation coefficient 0.732). K-means clustering identified 3 distinct severity categories in individuals living with long COVID, providing a basis for personalized treatment strategies.

Conclusions

The LC-SSS provides a robust and valid tool for assessing long COVID. The severity categories established via K-means clustering demonstrate significant variation in symptom severity, informing personalized treatment and improving care quality for patients with long COVID.