Send to:

Choose Destination
See comment in PubMed Commons below
J Heart Lung Transplant. 2014 Feb;33(2):163-9. doi: 10.1016/j.healun.2013.10.019. Epub 2013 Oct 21.

Impact of renal dysfunction on the Seattle Heart Failure Model.

Author information

  • 1Division of Cardiology, University of Minnesota, Minneapolis, Minnesota. Electronic address:
  • 2Division of Cardiology, University of Washington, Seattle, Washington.
  • 3Division of Cardiology, University of Minnesota, Minneapolis, Minnesota.
  • 4Division of Cardiology, Italian Association of Hospital Cardiologists Research Center, Florence, Italy.
  • 5Division of Cardiology, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts.



Renal dysfunction (RD) is a strong predictor of mortality in patients with heart failure (HF). However, its impact on the discrimination of the Seattle Heart Failure Model (SHFM) is poorly understood.


Serum creatinine (SCr) and creatinine clearance (CrCl) were reviewed for patients from four of the six cohorts originally used to derive and validate the SHFM. Patients were followed for death. The independent prediction of adding SCr or CrCl to the SHFM was assessed using multivariable Cox proportional hazards and the incremental value for prediction by changes in the ROC curves for 1- and 2-year event prediction.


Among 7,146 patients (mean age 63 ± 11 years), 1,511 deaths occurred during a mean follow-up of 2.04 years. SCr and CrCl had a modest positive correlation with SHFM (r = 0.30, p = 0.002). In combination with SHFM, SCr (hazard ratio [HR] per mg/dl 1.25, 95% CI 1.13 to 1.38, p < 0.0001) and CrCl (HR per 10 ml/min 0.95, 95% CI 0.93 to 0.97, p < 0.0001) were both multivariable predictors of events. When stratified by absolute risk based on the SHFM, SCr or CrCl provided more additional information in lower risk patients and less or no additional information in higher risk patients. The addition of SCr and the SHFM*SCr, or CrCl and the SHFM*CrCl interaction to the SHFM was associated with almost no change in the 1- and 2-year area under ROC curves for the SHFM score.


Compared with the SHFM alone, RD is independently predictive of mortality only in lower risk patients. Overall discrimination is only minimally improved with addition of SCr or CrCl to the SHFM.

© 2013 International Society for Heart and Lung Transplantation Published by International Society for the Heart and Lung Transplantation All rights reserved.


Seattle Heart Failure Model; chronic kidney disease; heart failure; risk prediction

[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Write to the Help Desk