Objectives & Background International guidance has recently recommended serial proximal compression ultrasound (CUS) as first line imaging for suspected deep vein thrombosis (DVT). Limitations with this strategy include attrition, lack of a clear diagnosis, and increased costs associated with serial resource use / clinical review. Single whole-leg CUS is a routine alternative diagnostic strategy that can reduce repeat attendance and identify alternative pathology. We sought to assess the performance characteristics of an established emergency department ambulatory protocol incorporating whole-leg CUS by non-physicians for exclusion of DVT.

Methods A prospective observational cohort study, conducted between July 2011 and April 2012. Consecutive, ambulatory, adult patients with suspected DVT and negative or inconclusive whole-leg CUS had anticoagulation initially withheld and were followed up after three months. The primary outcome was a predefined clinically relevant adverse event rate: a subsequent diagnosis of symptomatic venous thromboembolism (VTE) or VTE related death during three month follow up. Secondary outcomes included alternative diagnoses, technical failure rate and characteristics associated with failure.

Results 212 patients agreed to participate and were followed for three months. One patient was subsequently diagnosed with an isolated distal DVT. The adverse event rate was thus 1/212, 0.47% (95% confidence interval 0.08 to 2.62%). 150/212 patients were provided with a clear documented alternative diagnosis. CUS directly contributed to or confirmed the alternate diagnosis in 55/150 patients. Technical imaging failure occurred in 11.3% of suspected cases (95% CI 7.7 to 16.3). Several potential predictors of an inconclusive result were identified on multivariate analysis, including obesity, active infection, immobilisation and active cancer.

Conclusion Patients who have anticoagulation withheld following a negative or inconclusive whole leg CUS for suspected DVT have a low rate of adverse events at 3 months. Including the calf in ultrasound examination aided and clarified diagnosis in approximately one third of patients. Technical failure remains an issue: several factors were significantly associated with inconclusive results in our cohort and may warrant an alternative diagnostic approach⇓