Letters Section Editor: Jody W. Zylke, MD, Senior Editor.
Author Affiliations: Department of Epidemiology Research, Statens Serum Institut, Copenhagen, Denmark (email@example.com).
In Reply: Dr Lund and Ms Benson are concerned that those with candesartan use had more severe heart failure, leading to confounding by indication.
First, Lund and Benson question the general validity of ICD codes for the identification of heart failure, and reference a systematic review1 showing significant variation across different data sources. However, the specific data source used in our study for the identification of heart failure patients has been validated with positive predictive values ranging from 81% to 100%.2,3 This should be well within what is considered an adequate predictive value. Additionally, the study patient characteristics (Table 1 in article), such as the high baseline prevalence of cardiovascular disorders and risk factors (eg, 49% with ischemic heart disease and 18% with diabetes), together with the 37% mortality (during a median 1.9 years of follow-up), reflect a study population with severe morbidity, typical for patients with heart failure.
Svanström H, Pasternak B, Hviid A. Losartan vs Candesartan for Heart Failure—Reply. JAMA. 2012;308(4):336-337. doi:10.1001/jama.2012.7644