Accuracy of Self-Reported Heart Failure. The Atherosclerosis Risk in Communities (ARIC) Study

Area of Science:

• Cardiology
• Public Health
• Epidemiology

Background:

• Heart failure (HF) is a significant public health concern.
• Accurate estimation of HF prevalence is essential for resource allocation and patient management.
• Discrepancies between patient self-reporting and clinical diagnosis can impact prevalence estimates.

Purpose of the Study:

• To assess the agreement between self-reported heart failure (HF) and physician-diagnosed HF.
• To compare the prevalence of HF based on different ascertainment methods.

Main Methods:

• Analysis of the ARIC cohort (ages 60-83) with annual self-report surveys on HF.
• Physician confirmation of self-reported HF cases.
• Inclusion of hospitalized HF surveillance and administrative claims data (hospitalized and outpatient) for physician-diagnosed HF.
• Calculation of sensitivity, specificity, kappa, and prevalence, including bias-adjusted kappa (PABAK).

Main Results:

• Self-reported HF demonstrated low sensitivity (28%-38%) but high specificity (96%-97%) compared to physician diagnoses.
• Initial agreement was poor (kappa: 0.32-0.39), improving significantly when adjusted for prevalence and bias (PABAK: 0.73-0.83).
• Prevalence estimates were similar for self-report (9.0%), ARIC hospitalizations (11.2%), and hospitalization claims (12.7%), but increased to 18.6% when outpatient claims were included.

Conclusions:

• Self-reported HF requires confirmation through diagnostic tests or medical records for accurate burden estimation.
• Improved patient awareness and understanding of HF diagnosis are necessary for effective condition management.
• Ascertainment methods significantly influence HF prevalence estimates, highlighting the need for robust diagnostic validation.