Get access

Sensitivity and specificity of three-point compression ultrasonography performed by emergency physicians for proximal lower extremity deep venous thrombosis

Authors


  • Thomas D Crowhurst, BMedSci (Hons), Medical Student; Robert J Dunn, MBBS, FACEM, Associate Professor.

Abstract

Objective

The present study aims to quantify the sensitivity and specificity of three-point compression ultrasonography for diagnosing proximal lower extremity deep venous thrombosis when performed by Australian consultant emergency physicians with limited specific training. Secondary aims included quantifying rapidity, technical adequacy, predictability of equivocal results and relationships between emergency physician experience and proficiency.

Methods

This prospective diagnostic study enrolled a convenience sample of adult patients presenting to a major ED with suspected lower extremity deep venous thrombosis. The index test was abbreviated compression ultrasonography examining three points: common femoral, proximal great saphenous and popliteal veins. Emergency physicians received specific training. The reference test was full-leg duplex ultrasonography in the Radiology Department.

Results

A total of 15 emergency physicians participated, enrolling 178 subjects. Sensitivity of the index test was 77.8% (95% confidence interval: 54.8–91.0%), specificity was 91.4% (95% confidence interval: 84.9%–95.3%) and accuracy was 89.6% (95% confidence interval: 83.1–94.2%). Median duration of the index test was 10 min 34 s (interquartile range: 6 min 31 s) and ED diagnosis occurred significantly before Radiology Department diagnosis. The only statistically significant relationship between emergency physician experience and proficiency related to rapidity, which increased from the 36th scan. Equivocal index tests occurred in 9.2% of examinations and emergency physicians predicted equivocal assessments with specificity of 86.1% (95% confidence interval: 78.8–91.1%).

Conclusions

Abbreviated ultrasonography performed by emergency physicians for proximal lower extremity deep venous thrombosis could be valuable. However, more precise estimates for sensitivity and greater understanding of relationships between emergency physician experience and proficiency are required.

Ancillary