Lung US vs CXR for Traumatic Pneumothorax

Diagnostic accuracy of bedside chest ultrasonography vs supine chest radiography in ED trauma patients
Bedside lung ultrasound was nearly twice as sensitive as supine CXR for traumatic pneumothorax, with similarly excellent specificity

Study design

  • Cochrane systematic review and diagnostic accuracy meta-analysis
  • 13 prospective paired comparative studies included
  • 9 studies contributed patient-level data to the primary analysis
  • 4 additional studies used lung fields as the unit of analysis
  • CUS and supine CXR were performed in the same trauma patients
  • Reference standard: chest CT or tube thoracostomy findings

Population

Primary analysis
  • 1,271 trauma patients
  • 410 traumatic pneumothoraces
  • Median pneumothorax prevalence 30%
  • Emergency department setting
9 studies
Included studies
  • Prospective design
  • Paired CUS + supine CXR
  • CUS performed by frontline non-radiologist physicians
  • Trauma patients with suspected pneumothorax
  • CT or tube thoracostomy reference standard

Diagnostic tests

Chest ultrasonography
  • Bedside lung ultrasound
  • Performed by frontline non-radiologist physicians
  • Multiple probe types and operator backgrounds represented
CUS
Supine chest radiography
  • Portable supine CXR
  • Standard initial trauma imaging comparator
CXR

Primary outcome

Sensitivity for traumatic pneumothorax
91% with chest ultrasound vs 47% with supine CXR
Absolute sensitivity difference 44%
95% CI 27–61%, P < 0.001

Sensitivity for pneumothorax

Primary patient-level meta-analysis

100% 80% 60% 40% 20% 0%
91%
Lung US
95% CI 85–94%
47%
Supine CXR
95% CI 31–63%
Sensitivity 91% vs 47%  |  Difference +44%  |  P < 0.001

Lung ultrasound detected substantially more traumatic pneumothoraces than supine CXR

Diagnostic accuracy

CUS sensitivity

91%
95% CI 85–94%

CXR sensitivity

47%
95% CI 31–63%

CUS specificity

99%
95% CI 97–100%

CXR specificity

100%
95% CI 97–100%

Specificity difference

−0.7%
P = 0.35

Sensitivity range across studies

CUS 82–98%
CXR 9–75%

What this means clinically

Hypothetical 100 trauma patients with 30 pneumothoraces
Lung US would miss about 3 pneumothoraces vs 16 with supine CXR
CUS false negatives ≈3
CXR false negatives ≈16

Clinical interpretation

  • Lung ultrasound is substantially more sensitive than supine CXR for traumatic pneumothorax
  • Both tests have excellent specificity
  • A negative supine CXR does not reliably exclude pneumothorax in trauma
  • Bedside ultrasound can identify pneumothorax rapidly without moving an unstable patient
  • Diagnostic performance remained favorable across trauma type, operator type, and ultrasound probe type
  • CT remains the definitive reference when clinically necessary and the diagnosis remains uncertain

Limitations

  • All included studies had high or unclear risk of bias in at least one domain
  • 11 of 13 studies had high or unclear risk of bias related to patient selection
  • Substantial heterogeneity in CXR sensitivity
  • Operator experience and ultrasound technique varied between studies
  • Results apply specifically to trauma patients undergoing supine chest radiography
  • Four included studies used lung fields rather than patients as the unit of analysis and were excluded from the primary analysis
Chan KK, Joo DA, McRae AD, Takwoingi Y, Premji ZA, Lang E, Wakai A. Chest ultrasonography versus supine chest radiography for diagnosis of pneumothorax in trauma patients in the emergency department. Cochrane Database Syst Rev. 2020;7(7):CD013031. doi:10.1002/14651858.CD013031.pub2. Cochrane