An authoritative algorithm most appropriate for the prediction of pulmonary embolism in patients with AECOPD

Abstract

Abstract Background Contemporarily authoritative algorithms for the prediction of acute pulmonary embolism (PE) comprise the Standard algorithm, the Age-adjusted algorithm, the YEARS algorithm, the PERC algorithm, and the PEGeD algorithm. To date, little is known with respect to which algorithm is most appropriate for the PE prediction in patients with acute exacerbation of chronic obstructive pulmonary disease (AECOPD). Methods The patients with AECOPD who underwent the confirmed chest imaging investigations of PE due to the likelihood of PE predicted by the Standard algorithm were retrospectively reviewed. The patients were reassessed by the other four algorithms to reveal which algorithm had the best diagnostic accuracy for the likelihood prediction of PE for patients with AECOPD. Results The results showed that the PEGeD algorithm(88.6, 80.7, 50.4, 97.0%, 4.591, 0.141, 0.693, 82.1%) performed better overall in the sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, negative likelihood ratio, Youden index, and diagnostic accuracy, in comparison with the Age-adjusted algorithm (78.6, 74.1, 40.1, 94.0%, 3.034, 0.289, 0.527, 74.9%), the YEARS algorithm (71.4, 76.6, 40.3, 92.4%, 3.051, 0.373, 0.480,75.6%), the PERC algorithm (98.6, 1.6, 18.2, 83.3%, 1.002, 0.875, 0.002, 19.2%). The difference of number of patients who were necessary to undergo chest imaging examinations and missed diagnoses resulted from each algorithm between the PEGeD algorithm and the Standard algorithm, the Age-adjusted algorithm, the YEARS algorithm, as well as the PERC algorithm were [− 789 (− 68.1%), N/A], [− 42 (− 3.6%),-21 (− 1.8%)], [− 3 (− 0.3%),-36 (− 3.1%)],[− 771 (− 66.6%), 21 (1.8%)], respectively. Conclusions To date, the PEGeD algorithm is the most appropriate strategy among the authoritative algorithms for the likelihood prediction of pulmonary embolism in patients with AECOPD.