Severe pneumonia requires rapid and accurate diagnosis for targeted treatment, but single lung CT has limitations in identifying pathogens and distinguishing infectious/non-infectious etiologies. This is a retrospective self-controlled study enrolling patients diagnosed with severe pneumonia at the institution between 2024 and 2025 (recruitment will be extended 6-12 months if fewer than 400 patients are enrolled), all of whom underwent both single lung CT and bronchoscopy-combined CT examinations.

Clinical data will be collected retrospectively, including demographic information, bronchoscopic mucosal findings (e.g., congestion, exudation), lung CT lesion characteristics (e.g., consolidation, ground-glass opacity), and gold standard diagnostic results (pathogenic detection or clinical comprehensive diagnosis). The core objective is to compare the diagnostic precision between single lung CT and bronchoscopy-combined CT, focusing on accuracy, sensitivity, and specificity across three etiological subtypes (bacterial/fungal, viral, non-infectious).

Bronchoscopy complements CT by directly visualizing airway mucosal changes, while CT provides panoramic views of pulmonary lesions. Their combination is hypothesized to improve diagnostic accuracy. The findings aim to optimize diagnostic strategies for severe pneumonia, guiding clinicians to select more effective imaging approaches.

Inclusion Criteria:

• Patients admitted to the study institution between January 2024 and December 2025, with a clinical diagnosis of severe pneumonia (consistent with the diagnostic criteria of the Chinese Guidelines for the Diagnosis and Treatment of Community-Acquired Pneumonia.
• Patients who underwent both standalone lung CT examination and bronchoscopic imaging examination during hospitalization; complete imaging reports and bronchoscopy operation records are available.
• Patients who completed at least one type of microbial standard test (as defined in Primary Outcome Measure: e.g., pathogen culture, nucleic acid detection, GM test); complete test reports are available.
• Medical Records: Complete electronic medical records are available, including demographic information (age, gender), clinical symptoms, treatment regimens, and prognosis data (length of hospital stay, in-hospital mortality).

Exclusion Criteria:

• Imaging/Microbial Data Deficiency: Patients with incomplete imaging data (e.g., missing lung CT images, unrecorded bronchoscopic mucosal changes) or unavailable microbial gold standard test results.
• Bronchoscopy Contraindications: Patients who underwent bronchoscopic imaging for non-diagnostic purposes (e.g., foreign body removal, hemostasis) or had bronchoscopy-related complications (e.g., severe hemorrhage, pneumothorax) that affected examination completion.
• Duplicate Enrollment: Patients who were admitted multiple times for severe pneumonia during the study period; only the first admission is included to avoid duplicate data.

The study population includes adult patients (≥18 years old) admitted to the study institution between January 2024 and December 2025, with a confirmed diagnosis of severe pneumonia (per Chinese clinical guidelines for community-acquired or hospital-acquired pneumonia). All included patients must have undergone both standalone lung CT and bronchoscopic imaging (CT-guided) during hospitalization, with complete imaging records. They also need to have completed at least one microbial gold standard test (e.g., pathogen culture, nucleic acid detection) with available reports, and intact electronic medical records (demographics, treatment, prognosis).

Patients are excluded if they have incomplete imaging/microbial data, bronchoscopy for non-diagnostic purposes, duplicate admissions, or are under 18 years old.

We plan to share de-identified individual participant data (IPD) that includes:

Demographic information (e.g., age, gender); Baseline clinical characteristics (e.g., disease stage, results of relevant laboratory tests); Intervention-associated data (e.g., diagnostic findings from computed tomography (CT) alone and combined with bronchoscopy); Outcome measures (the accuracy of diagnosis for severe pneumonia); Safety data (e.g., documentation of adverse events). To safeguard participant privacy, all shared data will be fully de-identified by removing direct identifiers (e.g., full name, contact details, medical record number).