Typical symptoms for prediction of outcome and risk stratification in acute pulmonary embolism

Clinical presentation of pulmonary embolism (PE) comprises a wide spectrum from asymptomatic incidental finding to typical symptoms with chest pain, dyspnea, hemoptysis and syncope/collapse. We aimed to investigate typical symptoms of PE and increasing number of these symptoms to predict outcome in acute PE.

Data of PE patients were analysed retrospectively. According to the typical symptoms patients were subdivided in groups with 0, 1, 2, or ≥3 symptoms, which were compared with Kruskal-Wallis-Test. Logistic regression models were computed to investigate the association between the symptoms as well as the groups with the outcome parameters in-hospital death, myocardial necrosis, Shock-Index ≥1.0 and (right ventricular dysfunction (RVD). ROC curves were calculated to test the effectiveness of increasing number of symptoms to predict the outcome parameters.

One hundred eighty-two PE patients (61.5% female, mean age 68.5±15.3 years) were included in this study. 5 patients (2.7%) died in-hospital. Logistic regression models revealed associations between syncope/collapse and in-hospital death (OR 7.269, 95%CI 0.894-59.130, P=0.0636), myocardial necrosis (OR2.872, 0.904-9.130, P=0.0738), Shock-Index ≥1.0 (OR 4.906, 1.440-16.721, P=0.00110) and RVD (OR 5.265, 1.078-25.708, P=0.0401). Dyspnea and myocardial necrosis were also associated (OR 3.245, 1.127-9.348, P=0.0292). Increasing number of symptoms were not associated with in-hospital death, but absence of typical symptoms was associated with lower frequency of myocardial necrosis (OR 0.212, 0.046-0.976, P=0.0464). Effectiveness of increasing number of symptoms to predict myocardial necrosis was only moderate (AUC 0.608).

The symptom syncope/collapse is connected with poorer outcome in acute PE. An increasing number of symptoms failed to be useful for outcome prediction and risk stratification in acute PE.