Abstract 15023: Cardiovascular Magnetic Resonance Imaging-Based Multivariable Risk Score for the Prediction of Heart Failure Admission or Death

IntroductionEstablished heart failure (HF) risk scores lack patient-specific markers available from contemporary cardiovascular magnetic resonance (CMR). We developed a novel risk score, inclusive of these markers for the prediction of heart failure admission or death.MethodsA total of 1393 consecut...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Circulation (New York, N.Y.) N.Y.), 2019-11, Vol.140 (Suppl_1 Suppl 1), p.A15023-A15023
Hauptverfasser: Cornhill, Aidan K, Dykstra, Steven, Mikami, Yoko, Flewitt, Jacqueline, Seib, Michelle, Faris, Peter, James, Matthew, Lydell, Carmen, Howarth, Andrew, Heydari, Bobby, Fine, Nowell, Exner, Derek V, Lee, Joon, White, James
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:IntroductionEstablished heart failure (HF) risk scores lack patient-specific markers available from contemporary cardiovascular magnetic resonance (CMR). We developed a novel risk score, inclusive of these markers for the prediction of heart failure admission or death.MethodsA total of 1393 consecutive patients with HF referred to CMR underwent standardized consent, health questionnaire completion and diagnostic test reporting using a commercial data management platform (Acuity, Cohesic Inc). Eligible patients were required to have LVEF≤ 50%, no myocardial infarction or acute myocarditis within 3-months, and no congenital heart disease. All routinely captured patient variables were considered for risk prediction modelling, including patient-reported health, laboratory, and CMR-based markers. Patients were followed for the composite outcome of heart failure hospitalization or death. A multivariable cox proportional hazards model was developed then used for variable selection and development of a risk score model.ResultsThe cohort (23% female, median age 61 years) was followed for a median of 878 days with 199 patients (14%) experiencing the composite outcome. A stepwise multivariable model was developed using all eligible variables identified by univariable analysis. The final model provided a concordance index of 0.762. A clinically translatable risk score for estimating two-year risk of the composite outcome was developed using coefficients from the multivariable model (Figure 1). Risk score estimates of risk showed good agreement with multivariable cox model risk (Cohen’s weighted kappa = 0.75 (0.72-0.78)). Cumulative two-year risk increased with number of points; ≤52.8%, 6-1010%, 11-1526%, >1540%. The risk score model provided an AUC of 0.742.ConclusionsThe described multivariable risk score inclusive of CMR-based phenotypic markers provided robust prediction of heart failure admission or death in patients with HF.
ISSN:0009-7322
1524-4539
DOI:10.1161/circ.140.suppl_1.15023