Trauma showdown: pneumothorax. Chest x-ray or ultrasound?
The answer on the surface is pretty clear – unless you delve a little deeper into the precise question asked.
This meta-analysis of head-to-head ultrasound vs. CXR studies for the diagnosis of pneumothorax shows what we all essentially expect: the sensitivity of ultrasound is greatly superior, while specificity is statistically similar. Sensitivities for ultrasound were better for trauma, using the linear (high frequency) probe, and when performed by Emergency Physicians, and ranged from 73% to 85%. CXR sensitivities ranged from 32.6% to 49%, with heterogeneity based on study enrollment methods. Specificities for each were ~99%.
What the study does not address – are these pneumothoracies clinically meaningful? Ultrasound certainly finds more disease, but the newly identified disease will all be closer to the benign end of the spectrum. I guarantee there are patients out there with normal CXR in the setting trauma, who then receive an ultrasound positive for pneumothorax, and are then referred to CT scan and surgical evaluation – would have ultimately been fine. Before we move along to detecting more “disease”, we ought to examine the downstream consequences of missing or detecting these small pneumothoracies.
“Pleural ultrasonography versus chest radiography for the diagnosis of pneumothorax: review of the literature and meta-analysis”
ccforum.com/content/17/5/R208
What lung US finds are occult pneumothorax. The problem rather than US high sensitivity is that we don't really know how to approach occult PTx even when detected with the CT scan. Data available on this topic are pretty much contradictory. In my humble opinion they deserve observation, even in patients treated with positive pressure ventilation. Lung US should be the way to closely follow up their evolution. The finding itself shouldn't trigger a CT order. The vast majority of this small PTX have a clearly trackable lung point and it's going to shift over time if the pneumothorax extends. I think that even in mechanically ventilated patients with predictable short times of stay in the ICU observation is a reasonable strategy.
What lung US finds are occult pneumothorax. The problem rather than US high sensitivity is that we don't really know how to approach occult PTx even when detected with the CT scan. Data available on this topic are pretty much contradictory. In my humble opinion they deserve observation, even in patients treated with positive pressure ventilation. Lung US should be the way to closely follow up their evolution. The finding itself shouldn't trigger a CT order. The vast majority of this small PTX have a clearly trackable lung point and it's going to shift over time if the pneumothorax extends. I think that even in mechanically ventilated patients with predictable short times of stay in the ICU observation is a reasonable strategy.
It all comes down to, as we discussed on Twitter ad nauseum, the sophistication of the operator. Given the level of training in the U.S. and zero-miss culture we've got, I expect every suspicious looking ultrasound to result in a CT. We can scan you.
It all comes down to, as we discussed on Twitter ad nauseum, the sophistication of the operator. Given the level of training in the U.S. and zero-miss culture we've got, I expect every suspicious looking ultrasound to result in a CT. We can scan you.