Background:
A foundation of current clinical trials is to categorize the severity of heart failure (HF) by New York Heart Association (NYHA) classification to ensure that enrolled patients have similar disease severity. Because the NYHA represents a clinician’s assessment of patients’ health status, it may vary from patients’ perspectives and can lead to more or less symptomatic patients being enrolled in clinical trials. We sought to directly compare the ranges of patient-reported health status, as assessed by the well-validated and reliable Kansas City Cardiomyopathy Questionnaire (KCCQ), with NYHA class in recent clinical studies.
Methods:
We used data from 2 contemporary HF clinical trials, HF-ACTION in patients with Heart Failure with Reduced Ejection Fraction (HFrEF) and TOPCAT in patients with Heart Failure with Preserved Ejection Fraction (HFpEF), and 1 prospective cohort study, the KCCQ Interpretability study (KCCQINT) in patients with HFrEF, where both NYHA and the KCCQ were contemporaneously collected. The distributions of KCCQ Overall Summary (KCCQ-os) scores by NYHA and the variation in assigned NYHA classes among patients with KCCQ scores ≥80 (congruent with NYHA Class I) were then described.
Results:
A total of 6,072 patients (mean age 64±12 years, 41% female) were included across the 3 studies. Figure 1 shows marked overlap in KCCQ scores across NYHA classes. In KCCQINT, 148 (27%) out of 545 patients reported a KCCQ-os score ≥80, of whom 39 (26%), 81 (55%) and 28 (19%) were coded as NYHA Class I, II and III. None were classified as NYHA Class IV. In HF-ACTION, 677 (32%) of 2129 patients reported a KCCQ-os score ≥80, of whom 548 (81%), 128 (19%) and 1 (<1%) were coded as NYHA Class II, III and IV, respectively. In TOPCAT, 484 (14%) out of 3398 patients reported a KCCQ-os score ≥80, of whom 410 (85%) and 74 (15%) were considered NYHA Class I-II and III-IV, respectively.
Conclusions:
Although the NYHA is used to identify similarly ill patients for enrollment in clinical trials, there is marked variability within and across studies in patients’ self-reported health status. Future trials should consider patient-reported outcome measures as the basis for defining patient eligibility to enroll a more homogenous cohort of disease severity.
Ventricular instability and sudden death in patients with heart failure: lessons from clinical trials
