LIMITATIONS OF GENERIC QUALITY OF LIFE SCALES IN PATIENTS WITH SEVERE HEART FAILURE
S. Wright*, R.N. Doughty, H.J. Walsh, G.A. Gamble, N. Sharpe.
University of Auckland, Auckland, New Zealand.
Background: Quality of life (QOL) is often markedly impaired in patients with heart failure (HF) and is important to assess for clinical and research purposes. However, the choice of QOL instruments is difficult. One important potential limitation of all questionnaires is the floor effect: this occurs when more severely limited patients score a disproportionate number of minimum scores, limiting the scale's ability to discriminate between clinically relevant quality of life scores. Physical functioning is one aspect of QOL that is affected by HF. This study aims to compare the physical functioning (PF) dimensions of a common, popular and well-validated generic health questionnaire (the SF-36) and a disease-specific questionnaire (Minnesota Living with Heart Failure questionnaire, MLWHF).
Methods: The SF-36 and MLWHF questionnaires were administered by trained interviewers to patients hospitalised for an exacerbation of HF.
Results: 200 patients were included: mean age 73 (SD 10.5), 62% male, 71% NYHA class 4; 30% had 2 or more prior admissions for heart failure. Mean physical functioning scores: SF-36 36.9 (SD 23.6), MLWHF 22.1 (SD 10.8). Overall Spearman correlation between the two scores was 0.43. Mean percent of the cohort scoring minimum scores (% floor effect) was 45.8% (95% CI 13.2, 78.4) for the SF-36 PF score, and 18.5% (95% CI 0, 45.4) for the MLWHF PF dimension. Floor effects of up to 82% were observed for some items in the SF-36 scale.
Conclusion: the physical functioning scores from the SF-36 and MLWHF scales were poorly correlated because of important floor effects. The marked floor effects in the SF-36, suggest poor validity of the SF-36 in patients with moderate to severe HF and support the use of a disease-specific questionnaire in these patients.
