Retrospective analysis of the diagnostic accuracy of lung ultrasound for pulmonary embolism in patients with and without pleuritic chest pain

Abstract

Abstract Background Lung ultrasound (LUS) has a role in the diagnosis of pulmonary embolism (PE) mainly based on the visualization of pulmonary infarctions. However, examining the whole chest to detect small peripheral infarctions by LUS may be challenging. Pleuritic pain, a frequent presenting symptom in patients with PE, is usually localized in a restricted chest area identified by the patient itself. Our hypothesis is that sensitivity of LUS for PE in patients with pleuritic chest pain may be higher due to the possibility of focusing the examination in the painful area. We combined data from three prospective studies on LUS in patients suspected of PE and extracted data regarding patients with and without pleuritic pain at presentation to compare the performances of LUS. Results Out of 872 patients suspected of PE, 217 (24.9%) presented with pleuritic pain and 279 patients (32%) were diagnosed with PE. Pooled sensitivity of LUS for PE in patients with and without pleuritic chest pain was 81.5% (95% CI 70–90.1%) and 49.5% (95% CI 42.7–56.4%) (p < 0.001), respectively. Specificity of LUS was similar in the two groups, respectively 95.4% (95% CI 90.7–98.1%) and 94.8% (95% CI 92.3–97.7%) (p = 0.86). In patients with pleuritic pain, a diagnostic strategy combining Wells score with LUS performed better both in terms of sensitivity (93%, 95% CI 80.9–98.5% vs 90.7%, 95% CI 77.9–97.4%) and negative predictive value (96.2%, 95% CI 89.6–98.7% vs 93.3%, 95% CI 84.4–97.3%). Efficiency of Wells score + LUS outperformed the conventional strategy based on Wells score + d-dimer (56.7%, 95% CI 48.5–65% vs 42.5%, 95% CI 34.3–51.2%, p = 0.02). Conclusions In a population of patients suspected of PE, LUS showed better sensitivity for the diagnosis of PE when applied to the subgroup with pleuritic chest pain. In these patients, a diagnostic strategy based on Wells score and LUS performed better to exclude PE than the conventional strategy combining Wells score and d-dimer.