Chest radiography (CXR) is not indicated for routine diagnosis of uncomplicated influenza. Its primary role is to evaluate for pulmonary complications — particularly influenza pneumonia and secondary bacterial pneumonia — in patients with lower respiratory tract signs (dyspnea, tachypnea, hypoxia, or clinical deterioration) or risk factors for severe disease.^[1][2]

Chest radiography should be obtained when clinical findings suggest pneumonia or when there is clinical deterioration. The 2019 CHEST guideline found insufficient evidence to recommend for or against routine CXR in adults with acute cough and no validated algorithm for determining which patients warrant imaging.^[3] The 2024 ACR Appropriateness Criteria support CXR in patients with acute respiratory illness who have abnormal vital signs, positive physical examination findings, or risk factors for poor outcomes.^[4]

Specific triggers for imaging in influenza include:

• Persistent or worsening fever beyond 3–5 days, or biphasic fever pattern (suggesting bacterial superinfection)^[1]
• Dyspnea, tachypnea, or hypoxia
• Auscultatory findings (rales, rhonchi, egophony)
• Elderly patients or those with comorbidities, in whom clinical features of pneumonia may be absent or atypical^[4]

Patients with the following risk factors are at increased risk for radiographic pneumonia and warrant a lower threshold for imaging:^[5][6]

• Age extremes (≥65 years or <2 years)
• Third-trimester pregnancy
• Chronic pulmonary disease (COPD, asthma)
• Cardiovascular disease (excluding hypertension alone)
• Diabetes mellitus
• Obesity (BMI ≥30 kg/m²)
• Immunocompromise (including HIV, malignancy, transplant recipients, chronic steroids)
• Functional dependency or ≥3 comorbidities

In uncomplicated outpatient influenza, most patients have normal CXRs. In patients requiring hospitalization or ICU admission, the frequency of abnormalities increases substantially.^[7][8]

Primary influenza pneumonia presents with diffuse bilateral infiltrates, with sputum cultures negative for bacteria.^[1] Characteristic CXR findings in severe cases include:

• Ground-glass opacities — most common finding on CT (69–89% of abnormal CTs)^[8][7]
• Consolidation — frequently coexists with GGOs (59–89% on CT)^[8][7]
• Bilateral distribution — present in 62–94% of severe cases, predominantly in mid and lower lung zones^[8][7]

A lag phenomenon may occur: infiltrates may be absent on initial imaging and appear 24–72 hours later, particularly in immunocompromised patients.^[1]

In a multicenter study of 3,423 ED patients with acute respiratory illness, CXR sensitivity for pneumonia compared to CT was 43.5% (95% CI 39.1–48.1%), with specificity 93.0% (92.0–93.9%), PPV 26.9%, and NPV 96.5%.^[4][9] A negative CXR has high negative predictive value, but a significant number of pneumonias are missed by plain film.

CT is more sensitive and should be considered when:

• There is discordance between clinical suspicion and a negative CXR
• Complications (empyema, abscess, necrotizing pneumonia, pulmonary embolism) are suspected
• The patient is immunocompromised^[10][4]

CT with IV contrast is indicated when there is suspicion for empyema, abscess, or necrotizing pneumonia.^[4] In immunocompromised patients, CT is preferred over CXR or lung ultrasound for evaluation of suspected pneumonia.^[10]

Common CT findings include mixed GGO and consolidation, peribronchovascular distribution, lower lobe predominance, and centrilobular nodules.^[11] Influenza CT features overlap substantially with COVID-19, SARS, and MERS, limiting specificity.^[4]

Invasive pulmonary aspergillosis is an increasingly recognized post-influenza complication in immunocompromised and critically ill patients. Distinctive imaging findings include nodules, cavitation, and halo sign — a high-mortality diagnosis that should be considered in appropriate clinical contexts.^[5]

Lung ultrasound (LUS) is an increasingly accepted alternative to CXR where expertise and equipment exist. Pooled LUS sensitivity for pneumonia is 92% (95% CI 88–95%) with specificity 89% (81–95%), compared to CXR specificity of 49% (40–58%).^[12] A 2025 meta-analysis of 26 studies (3,454 patients) found BLUE protocol sensitivity 0.88 and dynamic air bronchograms specificity 0.96 for pneumonia diagnosis.^[12] The 2025 ATS guidelines accept LUS as an alternative to CXR where expertise and equipment are available.^[12]

Radiographic pattern alone cannot reliably differentiate primary influenza pneumonia from bacterial coinfection; microbiologic testing remains essential.^[2] However, certain features may suggest bacterial superinfection:

• Predominant consolidation
• Pleural effusion (uncommon in primary influenza pneumonia)^[13]
• Centrilobular nodules
• Higher CT severity scores than primary influenza pneumonia alone^[14]

The classic clinical pattern of bacterial superinfection is a biphasic illness: initial improvement followed by recurrent fever, productive cough, and new or worsening consolidation 4–14 days later.^[1]

Radiographic extent at presentation correlates with clinical severity. In H1N1 patients, involvement of four or more lung zones and bilateral peripheral opacities are associated with adverse outcomes (mechanical ventilation or death).^[7] CT severity scores correlate with ICU days, mortality, peak LDH, and CRP levels.^[14]

There are genuine disagreements among guidelines:

• IDSA/ATS CAP guidelines (2019): Recommend radiographic confirmation for CAP diagnosis, with CT advised for uncertain or inconclusive cases^[15]
• UK NICE guidelines (2023): Restrict CXR to hospitalized patients with suspected pneumonia^[12]

• CAP overdiagnosis is significant: 10–30% of patients treated for CAP do not meet diagnostic criteria.^[9] Asymptomatic patients with incidental infiltrates should not be treated for CAP.
• Pediatric considerations: Mild influenza in children most commonly shows peribronchial thickening with hyperinflation; severe cases demonstrate bilateral multifocal consolidation with ground-glass opacities similar to adults.^[16]