Chest Computed Tomography Patterns Associated with Severity in Adults with Respiratory Syncytial Virus Infection: A...

There are limited data available on the relationship between chest computed tomography (CT) patterns and clinical severity in adult patients infected with respiratory syncytial virus (RSV). It was hypothesized that specific CT patterns would be associated with distinct clinical outcomes and would correlate with both immune status and viral load.

We conducted a retrospective, single-center study of adults aged ≥ 18 years with laboratory-confirmed RSV infection and chest CT performed within 7 days of diagnosis. A blinded radiologist reviewed the CT patterns and scored them using the Chest CT Score (CCTS). Clinical severity was defined as World Health Organization (WHO) clinical progression scale ≥ 5 (hospitalized patients requiring oxygen). Multivariable logistic regression was used to identify independent associations between CT features and disease severity.

A total of 113 patients (median age 72 years; 41.6% male) were included from January 2019 to March 2025, of whom 31.9% were immunocompromised. The median CCTS was 6 [25th-75th quartiles, 3-10], with consolidations (55.8%), micronodules (46.0%), mucoid impaction (73.5%), and bronchial wall thickening (65.5%) being the most common chest CT findings. Patients with severe infection (n = 66) had a higher CCTS (7 versus 4; p < 0.001) and were more likely to exhibit consolidations and emphysema. On multivariable analysis, CCTS > 6 (adjusted odds ratio (aOR) 3.66, 95% CI 1.63-8.62), consolidations (aOR 2.87, 95% CI 1.23-6.83), and emphysema (aOR 5.66, 95% CI 1.36-39.15) were independently associated with clinical severity. No significant differences in CT patterns were observed according to immune status, the presence of bacterial co-infection, or elevated versus reduced viral loads.

Higher CCTS, consolidations, and emphysema were found to be independently associated with clinical severity in adults with RSV infection. Standardized CT scoring could help with risk stratification, but this would need to be validated in a prospective study.