Abstract 17195: Heart Failure Case Identification Criteria and One Year Fatality Following Hospitalization - Does a Criterion Matter?
Background/Aim: The prevalence and rate of heart failure (HF) vary by criteria used to identify cases and the criteria are often discordant. We compared one year case fatality of HF confirmed by different extant criteria among hospitalized patients.
Methods/Results: We identified potential HF hospitalizations (2005-2009) from 18 hospitals in four US communities using a combination of ICD codes and presence of at-least one sign/symptom of decompensation in the ARIC HF surveillance study. Charts were abstracted for data-elements by trained abstractors using standardized protocols. The potential cases were finally categorized as HF vs. not-HF by applying the following criteria using computerized classification: ICD code 428 in the primary position, NHANES, Modified Boston, Gothenburg, Framingham, and the ARIC panel (standardized panel of physicians examining acute decompensation). One year case fatality was obtained after linking with the National Death Index.
We included 6382 hospitalized patients with potential. The proportion classified as positive varied by criteria (60% to 89%). Those classified as HF by any criteria had a 30-32% one year case fatality as compared to 24% in those classified not-HF by ARIC panel (Figure 1, p<0.01 for any given comparison with no HF).
There were no appreciable differences in the one year case fatality among sub-groups classified as HF by different criteria .
Conclusions: One year case fatality following a hospitalization is >30% among confirmed HF patients. Surprisingly, the case fatality among those suspected but unconfirmed HF patients was high as well, though lower than confirmed HF cases. The case fatality rate did not vary much across different criteria for identifying HF cases.
- © 2013 by American Heart Association, Inc.