Validation of the SIRENA score for assessing the risk of inhospital mortality in patients with acute pulmonary embolism in an independent sample

Abstract

Aim. To validate the SIRENA score in assessing the risk of inhospital mortality in patients with pulmonary embolism (PE) in an independent sample.Material and methods. This retrospective, single-center study was based on the Samara Regional Cardiology Center. The risk of inhospital mortality was assessed using the SIRENA score, which includes such parameters as left ventricular ejection fraction <40%, immobilization in prior 12 months, creatinine clearance <50 ml/min, syncope, cyanosis on admission. For each positive sign, 1 point is assigned. Low risk is set at score of 0-1, high — ≥2.Results. The study included 452 patients with PE hospitalized from 2004 to 2019, of which 221 (48,9%) were men (mean age, 60,0 years (50,5-70,0)). With SIRENA score of 0, 1, 2, 3, and 4, inhospital mortality was 4,1%, 10,8%, 18,8%, 40,0%, and 100%, respectively. Mortality at SIRENA low risk (<2) was 7,1%, and at high risk (≥2) — 20,5% (odds ratio (OR), 3,34; 95% confidence interval (CI), 1,74-6,43; p<0,001). The predictive sensitivity and specificity for inhospital mortality for the SIRENA score were 70,5% and 60,8%, respectively. Area under the ROC-curve for the SIRENA score was 0,71 (95% CI, 0,63-0,79), while for Simplified Pulmonary Embolism Severity Index (sPESI) — 0,69 (95% CI, 0,60-0,77). With high risk on both scales (sPESI and SIRENA), inhospital mortality was 24,2% (OR, 4,09, 95% CI, 2,07-8,09; p<0,001).Conclusion. On an independent sample, the SIRENA score showed a high predictive ability in predicting adverse outcomes in patients with PE with a sensitivity of 70,5% and a specificity of 60,8% (AUC=0,71, 95% CI, 0,63-0,79), comparable with the sPESI.