Ultrasound for Central Line Placement

March 9, 2020

Written by Clay Smith

Spoon Feed
Ultrasound was not sensitive for detecting IJ or subclavian central venous catheter (CVC) malposition but was comparable to CXR for detecting pneumothorax (PTX).

Why does this matter?
CXR is the study of choice after placing a CVC to check position and look for complications, such as a PTX. Would ultrasound do just as well?

Where is that cath tip…
This was a prospective study of post-op ICU patients which included 758 CVC placements. Malposition was determined by directly visualizing the tip, if possible, and also by using a bubble study looking for, “a laminar flow of microbubbles to appear in the right atrium within 2 seconds.” Sensitivity of ultrasound compared to CXR for malposition was 70% (95%CI 49% to 86%); specificity 99% (95%CI 98% to 100%). Ultrasound detected 5 pneumothoraces; CXR detected 11. Since CXR is a poor gold standard for PTX, they looked at agreement. U/S and CXR agreed on PTX 98.9% of the time, with a kappa of 0.5 (95%CI 0.19 to 0.80). That kappa isn’t great. They made the point that, “One clinically relevant malposition out of 758 placements (0.0013%) was missed by ultrasound. Ultrasound and chest x-ray film detected all clinically relevant pneumothoraces.” For example, although a CVC in the brachiocephalic was malpositioned, by study definition, that is still a safe location. And all PTXs that mattered clinically were found. Also, they did not do U/S before line placement. There was comment that some PTXs may have been present before line placement. My take is this: I don’t agree with the author’s conclusion that, “ultrasound is an accurate diagnostic modality to detect malposition and pneumothorax.” It simply was not. A test with a sensitivity of 70% is not a good screening tool. However, if you’re content with the line tip just outside the SVC (which I am), this approach looks pretty good.

Source
Ultrasound to Detect Central Venous Catheter Placement Associated Complications: A Multicenter Diagnostic Accuracy Study. Anesthesiology. 2020 Jan 21. doi: 10.1097/ALN.0000000000003126. [Epub ahead of print]

Open in Read by QxMD