Pneumothorax CTA 32 year old woman arrives in your emergency department after being in a motor vehicle collision where she was the seat-belted driver. She undergoes chest CT imaging despite a negative chest x-ray because of her ongoing anterior chest wall diffuse tenderness. You discover a small 10% pneumothorax (PTX), but no other associated thoracic injuries. Should you place a tube thoracostomy (chest tube)? Should this patient be admitted to the hospital? A 2019 Annals of Emergency Medicine paper by the NEXUS Chest research group tackles these questions.1

Featured Paper

Pneumothorax and Hemothorax in the Era of Frequent Chest Computed Tomography for the Evaluation of Adult Patients With Blunt Trauma

Editorial Note on SOCTO Terminology

We are going to refer to PTXs and HTXs that are seen only on CT but not on CXR as “SOCTO” (Seen On CT Only – think a toe in a sock) injuries and not “occult” as it once was traditionally termed. This aligns with the NEXUS Chest research team’s terminology. “Occult” seems to imply a magical/mystical thing or something that is not detected. These injuries are neither magical or invisible.

Study Design

This multicenter study was a planned subanalyses from their 2 large prospective observational studies (NEXUS Chest and NEXUS Chest CT).2,3

  • Sites: 10 U.S. Level 1 trauma centers
  • Inclusion criteria: Blunt trauma victim, >14 years, and received chest imaging (CXR and/or chest CT)

Primary Outcome Measures

  1. Incidence of PTX and hemothorax (HTX) which are SOCTO as compared to PTX and HTX seen on both CT and CXR
  2. Incidence of “isolated” PTX and HTX (no other thoracic injuries)

Secondary outcome measures

There were a several secondary outcomes, focused primarily on the hospital course of these patients with PTX and HTX:

  1. Hospital admission rate
  2. Hospital length of stay
  3. In-hospital mortality rate
  4. Frequency of chest tube placement

Key Results

The research team enrolled an impressive 21,382 subjects over the 2009-2014 period. The overall incidence for PTX and HTX was 5% and 1.8%, respectively.

For the 8,661 patients, who received both a chest CT and CXR:

  • Overall incidence: PTX (10.5%), HTX (3.7%)
  • SOCTO incidence: PTX (67%), HTX (80%)
  • Bottom line: Most cases of PTX and HTX are SOCTO.

For the 1,117 patients with a PTX, HTX, or both:

  • Incidence of isolated PTX or HTX (no other thoracic injuries): 10%
  • Bottom line: Most PTXs and HTXs have some other thoracic injury and uncommonly occur in isolation. Be mindful to carefully check for at least the top 4 causes of associated injuries (Table).

Table: Most common thoracic injuries associated with a PTX or HTX as found on CT

Associated Thoracic InjuryIncidence
Rib fracture60%
Pulmonary contusion40%
Thoracic spine fracture3%
Sternal fracture1%


The secondary outcomes (hospital course) demonstrated that:

  • Patients with SOCTO PTX, when compared to patients with a PTX seen on both CT and CXR, had:
    1. About half the rate of chest tube placement (30% vs 65%)
    2. A slightly lower admission rate, but still high (94% vs 99%)
    3. A slightly shorter median hospital length of stay (5 days vs 6 days)
    4. Similar in-hospital mortality (5.3% vs 6%) despite fewer chest tubes
  • Patients with SOCTO HTX, when compared to patients with a HTX seen on both CT and CXR, had:
    1. Fewer chest tubes placed (49% vs 68%)
    2. Similar admission rates (97.6% vs 98.5%)
    3. The same median hospital length of stay (7 days)
    4. Similar in-hospital mortality (7.9% vs 7.7%) despite fewer chest tubes
  • Patients with an isolated PTX or HTX, when compared to patients with a PTX or HTX who also sustained a concurrent thoracic injury, had:
    1. About half the rate of chest tube placement (20% vs 43%)
    2. Less than half the admission rate (44% vs 97%)
    3. A shorter median hospital length of stay (4 days vs 5 days)
    4. Less than half the mortality rate (2.8% vs 6.2%) despite fewer chest tubes

Clinical Relevance

On initial read of this paper, it would seem that more patients should receive chest CTs because many are missed on CXR alone. However the true message is that a SOCTO PTX and/or HTX are of lesser clinical significance than a PTX or HTX seen on both CT and CXR. Furthermore, a PTX and HTX rarely occur in isolation.

Given that patients with an isolated SOCTO PTX and/or HTX had admission rates and mortality that was comparable to those without PTX or HTX, it is likely that morbidity and mortality are driven by the other associated injuries. Practitioners are now commonly sending these patients with isolated PTX/HTX home (about more than half were discharged home in this study).

The authors should be lauded for tackling a growing problem in the age of increasing CT use for blunt trauma evaluation, in which we are detecting many more minor injuries and “incidentalomas”. This study helps us to address the SOCTO PTX and HTX issue using a data-driven approach.

Final Thoughts

Is this a practice-changing publication? Not quite because it is an observational, non-interventional, non-randomized trial. But it is potentially a practice-nudging publication, because of its large sample size of over 8,000 patients who underwent both CT and CXR imaging.

The data suggest that the management of patients (decision to place a chest tube and admit the patient to the hospital) with PTX and HTX SOCTO should be driven more by the patient’s overall condition and associated injuries.

Author Insights: Dr. Robert Rodriguez
Professor of Emergency Medicine
Associate Chair for Clinical Research
University of California, San Francisco
This is the 5th in a series of papers that uses the largest database of prospectively enrolled adult blunt trauma patients (NEXUS Chest) to describe thoracic injuries in the age of frequent head-to-pelvis CT (pan-scan) imaging. Similar to our findings with rib fracture, sternal fracture, pulmonary contusion, and scapular fracture, we have demonstrated that most PTX and HTX are seen on CT only and that these SOCTO injuries have less clinical significance than injuries seen on CXR. So the message is NOT that we should be performing more chest CTs. We should use clinical decision rules to guide selective CT. NEXUS Chest CT detects major clinical injuries with extremely high sensitivity (>99%), allowing for a safe reduction of chest CT in 25-38% of patients.3



Rodriguez R, Canseco K, Baumann B, et al. Pneumothorax and Hemothorax in the Era of Frequent Chest Computed Tomography for the Evaluation of Adult Patients With Blunt Trauma. Ann Emerg Med. 2019;73(1):58-65. [PubMed]
Raja A, Lanning J, Gower A, et al. Prevalence of Chest Injury With the Presence of NEXUS Chest Criteria: Data to Inform Shared Decisionmaking About Imaging Use. Ann Emerg Med. 2016;68(2):222-226. [PubMed]
Rodriguez R, Langdorf M, Nishijima D, et al. Derivation and validation of two decision instruments for selective chest CT in blunt trauma: a multicenter prospective observational study (NEXUS Chest CT). PLoS Med. 2015;12(10):e1001883. [PubMed]
Michelle Lin, MD
ALiEM Founder and CEO
Professor and Digital Innovation Lab Director
Department of Emergency Medicine
University of California, San Francisco
Michelle Lin, MD


Professor of Emerg Med at UCSF-Zuckerberg SF General. ALiEM Founder @aliemteam #PostitPearls at Bio:
Michelle Lin, MD