Ultrasound-guided lung lesion index as a predictor of treatment outcomes: cohort study of 388 patients with coronavirus infection

Abstract

INTRODUCTION. Currently there are a lot of articles of lung ultrasound (LUS) in COVID-19 both in the diagnosis and in the prognosis of the disease. OBJECTIVE. Evaluation of the relationship between the ultrasound-guided lung lesion index (UIL) with the volume of lung involvement determined by computed tomography (CT) and disease outcomes in patients with COVID-19. MATERIALS AND METHODS. A prospective observational cohort clinical study included 388 patients aged 18–75 years; diagnosed with COVID-19 pneumonia or suspected COVID-19. Lung ultrasound was performed according to the 16-zone “Russian Protocol” within 24 hours after CT scan of the chest organs. RESULTS. The median lung lesion volume on CT was 55 (35; 74) % and UIL was 46 (28; 60) points. UIL had a strong direct correlation of 0.873 (95% CI 0.842; 0.897, p < 0.01) with the change in the volume of lung involvement determined by CT and the inverse with the SpO2/FiO2 index –0.850 (95% CI 0.827; 0.871, p < 0.01). Mortality was 56 patients (14.4 %) (p = 0.018). The optimal cut-off point for ROC analysis in predicting mortality was 55 points and had a sensitivity of 97.6 % and a specificity of 73.9 % with an area under the curve of 0.896 (95% CI 0.861; 0.931). Kaplan–Meier analysis on the entire data set (n = 388) demonstrated a survival rate of 97.6 % in the group with a UIL score less than 55 points and 62 % in the group with a UIL score more than 55 points. Differences between groups were statistically significant (Log Rank test p < 0.001; Breslow test p < 0.001). As a result, multivariate Cox regression analysis, using the stepwise exclusion method, only UIL remained a significant predictor of adverse outcome (p < 0.01). CONCLUSIONS. UIL determined by 16-zone “Russian protocol” correlated with severity of respiratory failure and volume of lung injury and was a predictor of adverse prognosis of disease outcome.