A prospective reassessment of the utility of the Wells score in identifying pulmonary embolism.

Abstract

OBJECTIVE: To assess whether a cumulative clinical risk score (Wells score) can optimise imaging strategies in patients with suspected pulmonary embolism (PE). DESIGN, SETTING AND PARTICIPANTS: Prospective, consecutive series of 633 studies on 595 patients referred to a major teaching hospital for ventilation/perfusion (V/Q) scanning for suspected acute PE between September 2004 and November 2005. Ventilation scintigraphy was performed using technetium-99m Technegas, and V/Q results were interpreted in conjunction with Wells scores. MAIN OUTCOME MEASURES: Likelihood of PE for each Wells score interval; overall prevalence of PE. RESULTS: The likelihood of PE for a given Wells score in our study was not significantly different from the likelihood in the original study by Wells et al. Scores of < 2 in our study were associated with a 4% risk of PE, scores between 2 and 6 with a 13% risk, and scores > 6 with a 67% risk. The overall prevalence of PE in our study was significantly less than that in the original study (9% v 16%; P < 0.01), attributable to a significantly larger proportion of our patients having scores of < 2 (66% v 40%; P < 0.0001). CONCLUSION: The Wells score remains a robust clinical tool for stratifying the likelihood of PE. Patients with Wells scores of > 2 warrant imaging assessment for PE, but for those with scores < 2, further imaging may be problematic.