Predictive Formula for COVID-19 Consolidation: Clinical-HRCT Relationship

Abstract

Background: COVID-19 pneumonia kills thousands of patients daily, requiring early detection and early intervention. Healthcare systems struggle with the overwhelming number of patients, necessitating an automated method for lung disease measurement in the early stages.
Objective: To find the relationship between pulmonary HRCT findings and different clinical signs and symptoms of COVID-19 patients.
Methodology: The study examined 113 COVID-19 patients in four months using a retrospective and cross-sectional approach. Three radiologists independently reviewed HRCT. Data was collected from both genders and age groups. Statistical analysis, such as cross tabulation, logistic regression, Chi-square, and Fisher’s exact test, was conducted using SPSS V23 software.
Results: The study found 70% of patients were over 45 years old, with males being more susceptible to COVID-19. The study examined the relationship between fever, cough, fatigue, myalgia, anosmia, and ageusia with GGOs, consolidation, lung nodules, air bronchogram, crazy paving sign, and pleural effusion using crosstabulation and logistic regression. Results showed significant correlations between these symptoms and consolidation, with 83.2 % accuracy predicted.
Conclusion: In conclusion, the current study revealed that ground glass opacities and consolidation are typical findings in COVID-19. Significant relationships were found between the primary clinical signs & symptoms and pulmonary HRCT findings. For the prediction of consolidation, the binary logistic regression model is exceptionally good from a clinical aspect.