Reducing the use of diagnostic imaging in patients with suspected pulmonary embolism: Validation of a risk assessment strategy

Abstract

The aims of this study were to measure the: (i) effects of implementation of a new risk assessment strategy for patients with suspected pulmonary embolism (PE) on the use of imaging and D-dimer assay; (ii) negative predictive value for PE of a combination of low risk and negative D-dimer assay; and (iii) compliance of ED clinicians with the strategy. A non-randomized clinical trial was conducted in the ED of a 720-bed teaching hospital between November 2002 and August 2003. Study subjects with suspected PE were compared with 191 randomly selected historical controls. The risk assessment strategy of Kline et al. was disseminated and implemented. The negative predictive value for PE was 99% (95% confidence interval [CI] = 97-100%) in 114 patients with low risk and negative D-dimer. There was a 21% absolute reduction in the rate of imaging following the implementation of the risk assessment strategy (56% vs 77%, P < 0.001). Low risk combined with a negative D-dimer result may allow exclusion of PE without imaging.