Lung ultrasound (LUS) as a predictor of delirium in elderly patients with COVID-19 during hospitalization: A geriatric vulnerability-stress model

Introduction

Delirium, an acute confusional state, is common in elderly hospitalized COVID-19 patients and is linked to poor outcomes. Despite this, underdiagnosis persists due to symptom overlap with dementia. This study investigates whether lung ultrasound (LUS) and cognitive screening can predict delirium risk in elderly COVID-19 patients.

Methods

A prospective study at San Luigi Hospital enrolled 64 COVID-19 patients (mean age 82.6). Pulmonary involvement was assessed using the LUS Extension Score (LUSext), while cognition was evaluated via MMSE and 4AT tests. Clinical data, including comorbidities and inflammatory markers, were analyzed. ROC curves and multivariate regression identified delirium predictors.

Results

Delirium occurred in 61.4 % (35/57). Univariate analysis linked delirium to pre-existing dementia (p = 0.03), higher LUS scores (p = 0.004), and lower MMSE scores (p = 0.0002). A LUS score >3 (sensitivity 74 %, specificity 82 %) best predicted delirium. Multivariate analysis confirmed LUS >3 (OR=4.22, p = 0.023) and lower MMSE (OR=0.87, p = 0.007) as independent risk factors.

Discussion

LUS and cognitive impairment are strong predictors of delirium in elderly COVID-19 patients. The LUSext score provides an objective, rapid bedside measure of lung pathology severity, with a score >3 indicating significantly elevated delirium risk. The link between low MMSE and delirium underscores cognitive vulnerability.

Conclusion

While delirium is multifactorial, integrating LUS into routine assessment may improve early detection and management in resource-limited settings, effectively identifying high-risk patients using an accessible bedside investigation. A LUS score >3 and low MMSE should prompt preventive measures. Further studies should validate these findings and explore pathophysiology.