Three Part Question
In [patients with suspected acute DVT] can [Emergency Physician performed 2-point compression ultrasound] [accurately diagnose DVT]?
A 62-year-old man presents with an acutely swollen left calf. You suspect that he may have a deep vein thrombosis (DVT). Can a 2-point bedside compression ultrasound, performed at the bedside by an Emergency Physician confirm or refute this diagnosis accurately?
Medline 1946 to 2013 using OVID interface.
Embase 1988 to 2013 using OVID interface.
Other Evidence Based Medicine (EBM) resources including Google Scholar, Best Bets, Up to Date, Cochrane, CINAHL, TRIP, ACP Journal Club and grey literature.
[exp Ultrasonography OR bedside adj ultrasound.mp] AND [exp Emergency Medicine OR emergency adj doctor.mp OR emergency adj physician.mp] AND [exp Venous Thrombosis OR deep adj2 thrombosis.mp OR DVT.mp].
Medline search found 17 papers. Review of abstracts and titles identified 7 to be relevant to topic, 10 were excluded. Applying the above criteria to the full text excluded a further 5 papers.
Embase search revealed 85 papers of which 82 were excluded, 6 being duplicates of papers found through Medline.
Other EBM resources were searched using keywords. Best Bets identified two unpublished relevant Best Bets. Up To Date, Google Scholar, CINAHL, TRIP and ACP Journal Club did not highlight any additional papers. Search of the grey literature found four abstracts from the Society for Academic Emergency Medicine 2008 annual conference. EBM search for these abstracts revealed no full text articles and so they were excluded from this review. Hand search of the references of each of the included papers and Best Bets articles found an additional two papers included for review.
One good quality systematic review and meta-analysis was found which incorporated all of the found papers predating it. Two papers published subsequently to the systematic review were included. 3 papers were found which answered the question
|Author, date and country
||Study type (level of evidence)
|Torres-Macho et al|
|74 patients (76 limbs) with suspected DVT presenting to a single ED from June 2010 to March 2011. Five Emergency Physicians with no prior ultrasound experience underwent 10 h of training on 4 different indications; suspected cholecystitis, hydronephrosis, DVT and varying cardiovascular problems. ||Prospective observational study.
||Bedside ultrasound accuracy compared with reference standard radiology departmental ultrasound||Sensitivity 92% (95% CI 82% to %100), specificity 98% (95% CI 94% to 100%), positive predictive value 96% (95% CI 88% to 100%), negative predictive value 96% (95% CI 90% to 100%), 96% global inter-relator reliability. Prevalence of DVT 35.1%.||Study focused on Emergency Physicians performing bedside ultrasounds for 4 different indications. No stated inclusion or exclusion criteria. No follow-up of patients performed. Reference standard ultrasound not well described. |
|Pomero et al. |
|16 studies included, 12 full text and 4 abstracts. 2379 patients. Subgroup analysis of 13 studies, 1806 patients focused on proximal BUS. Studies assessed with QUADAS2. Searched without language restrictions. Only included studies with reference standard. Grey literature searched with abstracts from conferences included. ||Systematic review and meta-analysis of emergency physician-performed ultrasonography in diagnosis of DVT
||Accuracy of proximal EP performed BUS compared with reference standard||Kappa 0.83 between 2 independent investigators. Of proximal BUS weighted mean sensitivity 96.5% (CI 90.1% to 98.8%), weighted mean specificity 96.8% (95% CI 94.7% to 98.0%). Pooled DVT rate 23%||Long-term follow-up in only 2 studies. No specification of exact technique used, i.e., 2-point or 3-point. No studies met all of 8 QUADAS2 criteria. 11 defined with high quality, meeting only 3 of QUADAS2. Training highly variable throughout studies. One of the included papers studies patients with suspected PTE as opposed to DVT. Index test varied; 14 proximal BUS, 1 whole-leg, 2 colour-flow Doppler. Reference standard varied; 12 colour-flow duplex, 3 angiography, 1 either of aforementioned. |
|Crowhurst et al,|
|178 patients (189 limbs) who were referred for Radiology Department scans for suspected DVT. 15 Emergency Physicians performed the scans after 2 h training session. The 1st three scans for each physician were supervised and subsequent were independent.||Prospective diagnostic study||Emergency Physician ultrasound compared with reference standard of Radiology Department scan||134 scans performed independently after withdrawals. Sensitivity 66.7% (95% CI 35.4% to 87.9%), specificity 86.1% (95% CI 78.8% to 91.1%). Prevalence of DVT 13.4%. ||Convenience sample. Small number of patients considering number of clinicians involved.|
The systematic review and meta-analysis by Pomero et al (2013) suggest that a high sensitivity is possible for ultrasound examinations performed by Emergency Physicians in patients with suspected DVT. However, there was considerable variation in the diagnostic performance of the included studies with the sensitivity ranging from 88.9% to 100%. Specificities ranged from 75.9% to 100%, and excluding one study, i.e., Frazee et al (2001), the only study to include an indeterminate option for venous compression, this increased to 96.8–100%. The studies published subsequent to this review article have less impressive results with Crowhurst et al (2013) reporting a disappointing sensitivity of 66.7%.
The prevalence of DVTs varied widely throughout the included studies from 7% to 47.3% suggesting significant heterogeneity between the populations. This may be partly accounted for by the definition of a positive reference standard result. Most of the studies used Radiology Department ultrasound scans as the reference standard; although this is not a perfect gold standard, it is appropriate as this is the standard used in clinical practice.
The advantage for patients, assuming equity of Emergency Physicians versus Radiology Department scans, is the significant reduction in time to scan and therefore to diagnosis. This is particularly pertinent for patients who are scanned on an outpatient basis and are required to commence anticoagulant treatment on the basis of the assumed diagnosis while awaiting their scan. While unfractionated heparin carries a low risk of adverse events, it is not a no-risk treatment. The main disadvantage of using ED scans, aside from the variability in quality, is the time that it takes to perform the scan for the Emergency Physician.
Although the scans can be performed in some ED settings and had very high sensitivity, the variation reported in the studies suggests the presence of a significant learning curve associated with this skill. And, as a complex skill it is likely to require frequent practice to maintain competency.
BUS, bedside ultrasound; DVT, deep vein thrombosis; EP, Emergency Physicians; PTE, pulmonary thromboembolism.
Clinical Bottom Line
In conclusion, the results from the literature show studies of limited quality with poor follow-up and sample size calculations. Although many of the studies demonstrate high sensitivity and specificity, there is considerable variation that may reflect differences in study populations or degree of training. These results suggest that it may be possible for Emergency Physicians to attain a level of competence equivalent to that of radiologists or ultrasonographers but at a cost of substantial training and practice to achieve and maintain this performance.
- Torres-Macho J, Anton-Santos JM, Garcia-Gutierrez I et al. Initial accuracy of bedside ultrasound performed by emergency physicians for multiple indications after a short training period. Am J Emerg Med 2012;30: 1943–9.
- Pomero F, Dentali F, Borretta V et al. Accuracy of emergency physician-perfomred ultrasonography in the diagnosis of deep-vein thrombosis: a systematise review and meta-anaysis. Thromb Haemost 2013;109: 137–45.
- Crowhurst T, Dunn R. Sensitivity and specificity of three-point compression ultrasonography performed by emergency physicians for proximal lower extremity deep vein thrombosis. Emerg Med Aust 2013;25:588–96.
- Frazee BW, Snoey ER, Levitt A. Emergency Department compression ultrasound to diagnose proximal deep vein thrombosis. J Emerg Med 2001;20:107–12