Abstract P169: Incidence and Predictors of Outpatient Heart Failure in a Large Insured Population

Abstract

Background: While the determinants of inpatient heart failure (HF) have been described, less is known about the incidence and predictors of HF diagnosed in outpatient settings. Methods: We analyzed data from 359,948 adults (≥18 yrs old) without HF at baseline and who were enrolled in Kaiser Permanente Georgia any time during calendar years 2000-2005. Diagnosis of incident HF required one inpatient or two outpatient visits with an ICD-9 code for HF. We developed separate models for incident HF diagnosed in either the outpatient or inpatient setting using multivariable logistic regression with repeated measures to assess the association of incident HF with the following antecedent risk factors: coronary artery disease (CAD), hypertension (HTN), diabetes mellitus (DM), atrial fibrillation (AF), and valvular heart disease. Results: During 1,015,794 person-yrs follow-up, there were 4,001 new cases of HF (50% in women). Two-thirds of all cases were diagnosed in the outpatient setting (n=2,639; 66%). The prevalence for each risk factor was lower in outpatients compared with inpatients: HTN (74% vs. 87%), DM (27% vs. 39%), CAD (19% vs. 30%), AF (6% vs. 15%), and valvular heart disease (6% vs. 15%). The adjusted odds ratios for most factors were higher for HF diagnosed in the inpatient compared with the outpatient setting (Table), but all factors together yielded good discrimination (C index > 0.85) in both inpatient and outpatient settings. Conclusion: In our population, twice as many incident HF cases were diagnosed in the outpatient compared with the inpatient setting. Further research on HF in outpatient settings is essential to reduce the growing burden of HF on society. Table: Multivariable predictors of incident heart failure diagnosed in the outpatient compared with the inpatient setting. * Adjusted odds ratio (95% confidence interval) † Risk Factor Model for Outpatient HF ‡ Model for Inpatient HF ‡ Age (per 10-yr increase) 1.99 (1.93 - 2.05) 1.85 (1.78 - 1.93) Female sex 0.81 (0.75 - 0.88) 0.95 (0.85 - 1.07) § Hypertension 3.23 (2.92 - 3.57) 5.58 (4.69 - 6.66) Diabetes mellitus 1.82 (1.66 - 1.99) 2.75 (2.45 - 3.08) Coronary artery disease 2.00 (1.80 - 2.23) 2.68 (2.35 - 3.05) Atrial fibrillation 1.76 (1.47 - 2.10) 3.78 (3.18 - 4.46) Valvular heart disease 1.35 (1.13 - 1.61) ¶ 2.63 (2.22 - 3.11) Model C-statistic 0.859 0.869 * Outpatient and inpatient incident HF cases were mutually exclusive. A patient who was initially diagnosed with outpatient HF but then subsequently had inpatient HF was counted as having incident HF as an outpatient only, and vice versa. † Each risk factor is adjusted for all the other listed risk factors. ‡ Corresponding P values for all odds ratios are < 0.0001, except where otherwise indicated. § P value > 0.40 ¶ P value = 0.0007