Risk stratification of elderly patients with acute pulmonary embolism
Roland Klingenberg,Oliver Schlager,Andreas Limacher,Marie Méan,Nicolas Vuilleumier,Juerg H. Beer,Daniel Staub,Beat Frauchiger,Markus Aschwanden,Bernhard Lämmle,Marc Righini,Michael Egloff,Joseph Osterwalder,Anne Angelillo‐Scherrer,Nils Kucher,Martin Banyai,Nicolas Rodondi,Arnold Eckardstein,Drahomir Aujesky,Marc Husmann,Roland Klingenberg,Oliver Schlager,Andreas Limacher,Marie Méan,Nicolas Vuilleumier,Juerg H. Beer,Daniel Staub,Beat Frauchiger,Markus Aschwanden,Bernhard Lämmle,Marc Righini,Michael Egloff,Joseph Osterwalder,Anne Angelillo‐Scherrer,Nils Kucher,Martin Banyai,Nicolas Rodondi,Arnold Eckardstein,Drahomir Aujesky,Marc Husmann,Christian M. Matter,Roland Klingenberg,Oliver Schlager,Andreas Limacher,Marie Méan,Nicolas Vuilleumier,Juerg H. Beer,Daniel Staub,Beat Frauchiger,Markus Aschwanden,Bernhard Lämmle,Marc Righini,Michael Egloff,Joseph Osterwalder,Anne Angelillo‐Scherrer,Nils Kucher,Martin Banyai,Nicolas Rodondi,Arnold Eckardstein,Drahomir Aujesky,Marc Husmann,Christian M. Matter
DOI: https://doi.org/10.1111/eci.13154
2019-07-16
European Journal of Clinical Investigation
Abstract:BackgroundCombining high‐sensitivity cardiac Troponin T (hs‐cTnT), NT‐pro‐B‐type natriuretic peptide (NT‐proBNP) and high‐sensitivity C‐reactive protein (hs‐CRP) may improve risk stratification of patients with pulmonary embolism (PE) beyond the PESI risk score. MethodsIn the prospective multicenter SWITCO65+ study, we analyzed 214 patients ≥65 years with a new submassive PE. Biomarkers and clinical information for the PESI risk score were ascertained within 1 day after diagnosis. Associations of hs‐TnT, NT‐proBNP, hs‐CRP and the PESI risk score with the primary endpoint defined as 6‐month mortality were assessed. The discriminative power of the PESI risk score and its combination with hs‐cTnT, NT‐proBNP and hs‐CRP for 6‐month mortality was compared using integrated discrimination improvement (IDI) index and net reclassification improvement (NRI). ResultsCompared with the lowest quartile, patients in the highest quartile had a higher risk of death during the first 6 months for hs‐cTnT (adjusted HR 10.22; 95% CI 1.79‐58.34; p=0.009), and a trend for NT‐proBNP (adjusted HR 4.3; 95% CI 0.9‐20.41; p=0.067) unlike hs‐CRP (adjusted HR 1.97; 95% CI 0.48‐8.05; p=0.344). The PESI risk score (c‐statistic 0.77 (95% CI 0.69‐0.84) had the highest prognostic accuracy for 6‐month mortality, outperforming hs‐cTnT, NT‐proBNP and hs‐CRP (c‐statistics of 0.72, 0.72, and 0.54), respectively. Combining all three biomarkers had no clinically relevant impact on risk stratification when added to the PESI risk score (IDI = 0.067; 95% CI 0.012‐0.123; p=0.018; NRI = 0.101 95% CI ‐0.099‐0.302; p=0.321). ConclusionsIn elderly patients with PE, 6‐month mortality can adequately be predicted by the PESI risk score alone.This article is protected by copyright. All rights reserved.
medicine, general & internal, research & experimental