Abstract 14632: Poor Health Status is Associated with Increased Mortality and Direct Medical Costs in Community Heart Failure Patients

Abstract

Background: Health status assessed with the Kansas City Cardiomyopathy Questionnaire (KCCQ) has been shown to predict healthcare resource use in trial participants with reduced ejection fraction (EF) following myocardial infarction. However; these associations have not been studied in community heart failure (HF) populations or in those with preserved EF.

Methods: We evaluated 12-month mortality and direct medical costs among HF patients enrolled in a community study from October 2007 to December 2011. Costs were obtained using population-based administrative data and inflated to 2011 U.S. dollars using the General Consumer Price Index. Health status was assessed by the KCCQ and patients were categorized according to their score with scores of 0 to <25 reflecting the worst health status, 25 to <50 poor, 50 to <75 fair, and 75 to 100 good health status. Preserved EF was defined as ≥50%. Cox proportional hazard regression models were used to examine mortality and generalized linear models to predict costs.

Results: Of 332 HF patients (mean age 73 years, 57% male, 42% preserved EF), 46% had good health status, 30% fair, 21% poor, and 3% the worst. After 1 year, 30 (9%) patients had died, and worse health status was associated with increasing mortality, with 1-year mortality rates of 4%, 8%, 14%, and 67% for those with good, fair, poor and the worst health status respectively (p for trend <0.001). The distribution of health status and its association with mortality was similar in patients with preserved and reduced EF. One year direct medical costs increased with worse health status with mean (median) costs of $38339 ($22630), $38765 ($25716), $54487 ($29774), and $77777 ($37671) for those with good, fair, poor, and the worst health status, respectively (p for trend <0.001 adjusting for age, sex). Differences were mainly due to higher inpatient costs, which more than doubled from a mean of $27279 (median $14811) for patients with good health status to $64363 (median $29933) for those with the worst health status.

Conclusion: Health status predicts death and direct medical costs in community HF patients with both preserved and reduced EF. Health status may be useful in risk-stratifying patients so that more intensive outpatient care may be applied to those with the worst status.