Comparison of Four Views to Single View Ultrasound Protocols to Identify Clinically Significant Pneumothorax
Academic Emergency Medicine October 2016 - Pubmed Link
Background
Ultrasound to rule out pneumothorax is commonly used. However, there is not much standardization- some do single view, some protocols advocate 4 views of each hemithorax. Sure, you can imagine that if you do more views you might find more pneumothoracies, BUT are these important? Will you do anything for them? Maybe, maybe not. This study tried to look at only "clinically significant" pneumothorax to see if there is a difference when you select out the pathology you think might change management. The sensitivity and specificity of the two options are calculated for patients presenting to the emergency department with trauma.
Questions
Is one view as good as 4 views for identifying clinically significant pneumothorax?
Population
Single center, urban academic emergency department, volume 130,000, level 1 trauma center
Inclusion:
-18 yo or older
-Met trauma criteria + underwent CT chest as part of work up
Exclusion:
-too unstable
-required care that prevented chest wall US
-chest tube already
-pregnant
-prisoners
Who did the scans?
Emergency medicine attendings and residents - all credentialed (≥25 FAST exam and ≥25 chest wall exams). Residents credentialed in both 1 view and 4 view protocols prior to trauma rotation.
OR
Trauma attendings (training not specified)
Design
Randomized, prospective study, convenience sample
Once they determined they would receive CT chest, they were randomly selected into single view vs four views, performed before CT
Blinded to xray while performing US
Looking for significant pneumothorax: insignificant = if on fewer than 5 contiguous slices or up to 1 cm thick or less
Power calculation for noninferiority (beta 0.8 used), determined needed total of 236 patients (to determine difference of less than 5%)
All ultrasounds reviewed by author for presence or absence of pneumothorax (blinded to all information and other imaging)
Intervention
Everyone received an ultrasound and then CT chest.
Most patients also received a 1 view supine chest xray.
Scan
Linear probe, digital clips
Single view: longitudinal, midclavicular line in third intercostal space
Four views: first image as above, then moved inferiorly and laterally for additional three images
Interpreted as absence or presence of dynamic lung sliding indicating pneumothorax (does not note if using m-mode, power doppler, or blines to assess)
Learn how to find pneumothorax on ultrasound - 5 Minute Sono
Results
298 patients, 38 excluded so 260 patients over 2 years
139 single view, 121 four view
Patient characteristics: pretty similar
-more operative cases in single view (36% vs 26%)
-more pneumothorax in single view (22% vs 16%)
-similar operator experience (PG2/PG3 or attending)
Pneumothorax incidence was 19% (49/260)
29/49 (59%) were clinically significant - all admitted, 86% required chest tube (test of their definition of “significant”)
20/49 (41%) were insignificant - 100% admitted for observation, none had chest tube
Primary Findings
For Clinically Significant Pneumothorax
1 view: 93% sens, 99.2% spec
4 view: 93.3% sens, 98% spec
CXR: 48% sens, 100% spec
For any PTX
1 view: 54% sens, 99.1% spec
4 view: 66.7% sens, 97.6% spec
CXR: 32.6 sens, 100% spec
All of the clinically significant pneumothoraces except 1 that were diagnosed with four view, were seen on first view (so would have been seen on single view).
Other Findings
All of the clinically significant ptx except 1 that were diagnosed with four view, were seen on first view
100% agreement with interpretation during review.
US False Positives: 3 total, two right mainstem intubations, 1 soft tissue defect thought to be pleural line
US False negatives: 19 total, 18 of these were “insignificant” - did not require chest tube. 1 had tiny pneumo but also large hemo → got Chest tube
Single patient had significant pneumothorax missed by US - needle decompresssed by paramedics, lung sliding intact at that site but large pneumo elsewhere in thorax
Limitations
Single center - important that their trauma protocol resulted in patients supine with spinal immobilization prior to arrival
Excluded unstable patients
Convenience sample
Wide confidence intervals
Could have missed some pneumothoraces that did not get a chest CT
Take Home Points
1. With regard to ultrasound for clinically significant pneumothorax in trauma patients, 1 view has comparable sensitivity to 4 views.
2. Don’t mistake mainstem intubation for pneumothorax.
Our score
