C0454: Accuracy of Nurse-Performed Compression Ultrasonography in the Diagnosis of Proximal Symptomatic Deep Vein Thrombosis: A Prospective Cohort Study

Abstract

Summary. Background: Compression ultrasonography(CUS) has been recognized as the diagnostic procedure ofchoice for the investigation of patients with suspecteddeep vein thrombosis (DVT); the aim of this study was toassess the diagnostic accuracy of nurse-performed CUSfor symptomatic proximal DVT of the lower limb. Mate-rial and Methods: We prospectively evaluated all consecu-tive outpatients referred for suspected DVT from January2011 to December 2012. All patients underwent bilateralproximal lower limb CUS, first by trained nurses andthen by physicians expert in vascular ultrasonography,with every group blinded with respect to each other. Thistest was repeated after 5–7 days in all negative or unclearexaminations. Interobserver agreement and accuracy ofnurse-performed CUS were calculated, considering thephysician’s final diagnosis as the reference test. Results:Six hundred ninety-seven patients were included in thestudy. DVT was diagnosed in 122 patients by expertultrasound physicians with an overall prevalence of17.5% (95% confidence interval [CI] 15.8%–20.6%).Nurse agreement with the physician in DVT diagnosiswas excellent (Cohen’s j 0.82, 95% CI 0.79–0.85). Nurse-performed CUS had a sensitivity of 84.4% (95% CI81.7%–87.1%) and a specificity of 97.0% (95% CI95.8%–98.3%) with a diagnostic accuracy of 94.8% (95%CI 93.2%–96.5%). Conclusion: Our results suggest thatnurse-performed CUS may be a potential useful alterna-tive to physician performed CUS with a good accuracy.However, sensibility of nurse-performed CUS appearedsuboptimal and future studies should incorporate in theevaluation of this technique other pretest tools that mayincrease its accuracy.Keywords: diagnostic tests; Doppler ultrasonography;nurses; vascular diseases; venous thrombosis.IntroductionPrompt diagnosis of deep vein thrombosis (DVT) isessential to decrease the risk of potentially fatal pulmo-nary embolism and postthrombotic syndrome inuntreated patients [1,2]. Diagnostic strategies in DVTshould be accurate, simple to apply, and cost effective.Currently, compression ultrasonography (CUS) is consid-ered the method of choice for the diagnosis of DVT insymptomatic patients, and its use has been validated inseveral prospective studies [2–4]. The potential advantagesof this technique are the ease of performance by manyoperators and the availability in many hospitals with vir-tually all ultrasound scanners, regardless of generation,model, and even probe frequency [5].Although CUS has shown good accuracy for the diag-nosis of symptomatic proximal DVT, its specificity andsensitivity for calf vein thrombi are much lower [2,4]. Forthis reason, CUS may be limited to the proximal veinsand repeated within 1 week when negative (serial proxi-mal CUS) or extended to both proximal and distal veinsand performed on a single occasion with color-Dopplersupport (whole-leg ultrasonography) [2,5]; both strategiesare reliable diagnostic options for the management ofthese patients [2,5].