Written by Megan Hilbert
Spoon Feed
Proximal compression ultrasound (PUL) in the ED and risk stratification (via Well’s score) can result in appropriate evaluation and management decisions for patients presenting with concern for DVT.
While my title may be tortured, this algorithm is not
This was an observational cohort study with convenience sampling at a community hospital in Scandinavia. The goal was to evaluate the efficacy of a management algorithm for evaluation of DVT in the ED (see Figure 1 from the paper inserted below).
This had a negative predictive value of 99.5% (95%CI 98-99.9%) and sensitivity of 97.8% (95%CI 92.4-99.7%) for ruling out DVT.
[Please note: this study used operators with varying levels of ultrasound experience who underwent structured training and followed the same scanning protocol. This writeup is not intended to further clarify the best scanning protocol or training for providers with regard to this imaging modality.]
How does this change my practice?
I follow a 3-point compression protocol and risk stratification in my own practice, and this affirms my practice. Anecdotally, this has greatly improved throughput in my department for patients with concern for DVT.
- If you practice in the U.S., the ACEP Policy Statement regarding Ultrasound Guidelines, published in 2016, supports multilevel compression US on proximal veins, with a pooled sensitivity of 95% and specificity of 96%, with noted added advantage of decreased ED length of stay.
- A more recent Clinical Policy Statement from ACEP, published in Annals of Emergency Medicine in 2003 regarding work-up for patients with suspect lower-extremity DVT, concluded Level B recommendations regarding ruling out DVT with negative ultrasound as long as the patient is low risk.
- Moral of the story: no matter where you practice, if this is not your current approach to DVT, then maybe it should be.
Source
Proximal venous ultrasound with risk stratification safely excludes deep venous thrombosis in emergency department routine care: an observational study. Scand J Trauma Resusc Emerg Med. 2025 May 14;33(1):85. doi: 10.1186/s13049-025-01382-7. PMID: 40369661
