Evaluation of short training session for venous limited compression ultrasonography: prospective multicenter study

Background Venous limited compression ultrasonography (VLCU) is recommended in case of suspicion of deep venous thrombosis (DVT). Current training pathways are rather long and include experiential phase. This aim of this study was to investigate the efficacy of a short training session for VLCU without experiential phase. The training session was applied in residents without previous ultrasound skills. Program included operation of ultrasound device and interpretation of venous images. Included patients were older than 18 years and had a suspicion of DVT. After realization of VLCU using usual technique, residents reported the dynamic compressibility of the femoral and popliteal veins, the presence or not of a visible thrombus, self-reported difficulty and duration. Patients then underwent a whole leg ultrasonography (WLCU) in the local laboratory which was blinded to VLCU results. The main criterion was the negative-predictive value (NPV) of VLCU for the absence of proximal DVT diagnosed with WLCU. Secondary criteria were VLCU diagnostic performances, rate of inability to conclude, difficulty and duration. For a NPV of 95 +/- 6%, the needed number of patients was 96. This study was approved by the ethical committee. Results 102 patients were analyzed. 46 residents were trained. A DVT was diagnosed by WLCU in 18 patients (prevalence of 17.6% [95% CI 11-26%]). VLCU detected 15 DVT (NPV of 96% [95% CI 89-99%]). The positive likelihood ratio was 9.9, the negative likelihood ratio 0.19 and Cohen's Kappa 0.62 [95% CI 0.52-0.71]. The sensitivity was 83% [CI 95% 60-94%] and specificity 88% [CI 95% 79-93%]. The mean number of VLCU by residents was 2.3 +/- 2.1, median 2 (minimum 1, maximum 8). Mean duration was 3.4 min, difficulty was 3.7 +/- 2. Conclusion The principal objective, NPV 96% [95% CI 89-99%], was achieved. However, this short training session was inadequate to allow ruling-out a DVT with sufficient security. Thus, the experiential phase seems to be essential.