A Discriminant Analysis of the Factors Affecting Abnormalities in Chest Computed Tomographies of 608-Group Patients with COVID-19 Pneumonia

Abstract

Objective: To study factor correlation and classification affecting abnormalities in chest computed tomographies (CTs) of 608-group patients with coronavirus disease 2019 (COVID-19) pneumonia. Material and Methods: We retrospectively collected data of 608-group patients with COVID-19 pneumonia from medical records combined with data from chest CTs which were interpreted by a radiologist for CT abnormalities. The findings were analyzed by descriptive statistics, Fisher’s Exact Test and multiple discriminant analysis (MDA) by a stepwise method. Results: The majority of the 161 patients were female (55.9%), with an average age of 62.90 years (S.D. 16.68) and average weight of 63.07 kg (S.D. 16.18), non-smoking and non-alcohol drinking (71.4% and 61.5%, respectively) and with underlying respiratory diseases (28.6%). The important symptoms brought to a doctor were main symptoms including fever, chills, cough, nasal congestion, sore throat, difficult breathing, shortness of breath (74.5%). The average duration from onset of the symptoms to perform chest CTs was 11.18 days (S.D. 5.42). The abnormalities of CTs chest such as characteristics and locations were periphery (54.7%) with ground-glass opacity (44.7%). The CT severity score was level 2 (24.8%) from 5 levels. MDA revealed there were 5 factors affecting the abnormalities in the chest CTs of 608-group patients with COVID-19 pneumonia. CT severity score, peripheral location, body weight, age and location in the lower lungs. These factors accurately predicted abnormalities in chest CTs (60.2%). Conclusion: Abnormalities in chest CTs, and factor correlation and classification that affect abnormalities in chest CTs of 608-group patients with COVID-19 pneumonia will benefit the medical and multidisciplinary team in helping to determine treatment method, accurately prognosing severity and reducing mortality.