| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2008;118:e8.)
© 2008 American Heart Association, Inc.
Correspondence |
Cardiovascular Division, Department of Medicine, Brigham and Womens Hospital, Boston, Mass
London School of Hygiene, London, UK
University of Glasgow, Glasgow, UK
Duke University Medical Center, Durham, NC
Hamilton Health Sciences and McMaster University, Hamilton, Ontario, Canada
Sahlgrenska University, Hospital/Östra, Göteborg, Sweden
Cleveland Clinic Foundation, Cleveland, Ohio
AstraZeneca LP, Wilmington, Del
We appreciate the insights provided by Drs Höglund and Nilsson regarding our recent publication on the risk associated with heart failure hospitalization.1 Although in principle their request that models should also adjust for time-updated covariates is of interest, in practice it is liable to be unrealistic in most follow-up studies of this kind. When a clinical event such as a hospitalization for heart failure occurs, the data collected are usually confined to the event itself, and there are no routinely available data on time-updated covariates up to that point. Studies may have planned visits where such covariates are collected, for example, New York Heart Association class. Although such regular covariate updates could be added into our models (with substantial added complexity for the reader to interpret), we doubt they would change the conclusions we derived. Indeed, because the clinically relevant message from these data is the risk associated with a hospitalization, including covariates that would be closely related to the risk for hospitalization in our model would attenuate that risk estimate and possibly give a lower—and unrealistic–estimate of the true risk related to a heart failure hospitalization.
Our findings as presented show a valid prediction of how a patients risk of death is changed after hospitalization for heart failure. Quite simply, that is a clinically evident and important event after which the patients risk of death is sharply elevated and then gradually declines again over time. We agree that there may well have been a cumulative deterioration in the patients condition leading to the admission, but that is best considered as water under the bridge once the clear event of hospital admission actually happens.
| Acknowledgments |
|---|
Drs Solomon, Dobson, Pocock, McMurray, Granger, Yusuf, Swedberg, Young, and Pfeffer have received research funding from AstraZeneca. Drs Pocock, McMurray, Granger, Yusuf, Swedberg, Young, and Pfeffer have consulted for or have received honoraria from AstraZeneca. Dr Michelson is an employee of AstraZeneca. Dr Skali reports no disclosures.
| Reference |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2008 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |