Patients with sepsis-induced acute respiratory distress syndrome (ARDS) have a high mortality rate. Early, targeted antibiotic therapy is crucial for patient survival. The clinical use of a next-generation sequencing (NGS)-based approach for pathogen identification may lead to improved diagnostic performance. Therefore, the objectives of this study were to examine changes in the pulmonary microbiome and resulting influences on patient outcome in sepsis-induced ARDS, and to compare NGS- and culture-based diagnostic methods for pathogen identification.

In total, 30 patients in two groups were enrolled in the study: 15 patients with sepsis-induced ARDS following major abdominal surgery and 15 patients undergoing oesophageal resection (serving as controls). In the ARDS group, blood samples were collected at ARDS onset (day 0), and 5 and 10 days later. Bronchoalveolar lavage (BAL) was performed at the same timepoints to collect epithelial lining fluid for culture- and NGS-based analyses, and to evaluate longitudinal changes in the pulmonary microbiome. In the control group, only one BAL and one blood sample were collected.

Patients with ARDS showed significantly lower α-diversity (P=0.007) and increased dominance (P=0.012) in their pulmonary microbiome compared with the controls. The α-diversity index correlated with the length of stay in the intensive care unit (P=0.015) and the need for mechanical ventilation (P=0.009). In 42.9% of patients with ARDS, culture-based results were negative while NGS findings indicated bacterial colonization.

Sepsis-induced ARDS is associated with significant dysbiosis of the patient's pulmonary microbiome, which is closely correlated with the clinical course of disease.