Multidimensional patient-reported outcomes in acutely decompensated heart failure: prognostic implications in a real-world cohort

Background

Patient-reported outcomes (PROs) are integral to patient-centered heart failure (HF) care. Although multiple PRO instruments exist to assess distinct domains of health status, how they correspond with clinician-assessed disease severity and whether each provides incremental prognostic information remain unclear.

Objectives

To clarify (1) how multidimensional PROs across disease-specific, generic, and mental-health domains correspond with physician-assessed disease severity as represented by established HF risk scores, (2) whether PROs provide incremental prognostic information beyond those risk scores.

Methods

In a cohort of 2,434 patients hospitalized with acutely decompensated HF, three PRO instruments were assessed at discharge: the Kansas City Cardiomyopathy Questionnaire overall summary score (KCCQ-OSS), EuroQol 5-Dimension level summary score (EQ-5D-LSS), and Patient Health Questionnaire-9 (PHQ-9). Disease severity was quantified using the Get With The Guidelines–Heart Failure (GWTG-HF) and Meta-Analysis Global Group (MAGGIC) risk scores.

Results

The mean age was 71.2 ± 14.4 years and 38.2% were women. Median discharge scores were 56.3 (IQR 39.3–73.0) for KCCQ-OSS, 8 (IQR 6–10) for EQ-5D-LSS, and 4 (IQR 2–9) for PHQ-9. 10.3%, 7.2%, and 7.4% of patients were categorized into the worst categories of the KCCQ-OSS (0–24), EQ-5D-LSS (16–25) and PHQ-9 (15–27), respectively. The three PROs demonstrated moderate intercorrelations (Spearman’s r -0.38 to 0.50), whereas correlations with either risk score were weak or absent (r −0.02 to 0.23). Over a median follow-up of 731 days, 386 deaths occurred. Across all PROs, worse categories were associated with progressively higher mortality risk after adjustment for either the GWTG-HF or MAGGIC risk score. These associations persisted even when PROs were analyzed as continuous variables (GWTG-HF-adjusted HR [95% CI] per 1-point change: 1.01 [1.004–1.01] for KCCQ-OSS, 1.07 [1.05–1.10] for EQ-5D-LSS, and 1.05 [1.03–1.07] for PHQ-9; all P<0.001).

When added to the GWTG-HF score, KCCQ-OSS improved risk reclassification for 2-year mortality, as assessed by net reclassification improvement (NRI) (0.14; P=0.02) and integrated discrimination improvement (IDI) (0.0063; P=0.004), without improving discrimination (C-statistic 0.71 vs 0.70; P=0.21). Adding EQ-5D-LSS and PHQ-9 improved discrimination (C-statistic 0.72 for each; P=0.02 and P=0.01 vs GWTG-HF alone) and reclassification (NRI 0.22 and 0.24; IDI 0.016 and 0.014; all P<0.001). Similarly, when added to the MAGGIC score, all PROs improved risk reclassification, as assessed by NRI (0.15, 0.23, and 0.21 for KCCQ-OSS, EQ-5D-LSS, and PHQ-9, respectively; all P≤0.01) and IDI (0.0063, 0.013, and 0.012; all P<0.01).

Conclusion

Discharge PROs provide independent and incremental prognostic information beyond established risk scores and may inform patient-centered post-discharge care in patients hospitalized for HF.Figure 1Figure 2