Preferences for treatment outcomes in patients with heart failure: Symptoms versus survival

Background: Patient preferences for congestive heart failure therapy outcom es may vary depending on the goals of improving symptoms versus survival, b ut this has not been extensively investigated. Our objective was to analyze patient preferences for congestive heart failure therapy outcomes based on the goals of symptom versus survival improvement. Methods and Results: This was a prospective, full-profile conjoint analysis study of individual preferences for congestive heart failure treatment out comes. Conjoint analysis was based on ratings of 16 treatment-outcome profi les, each consisting of 4 attributes (tiredness, shortness of breath, depre ssion, and survival) varied across 4 severity levels. Part-worths (utilitie s) and importance weights were calculated for each attribute to determine t heir relative contribution to the full-profile rating decision using standa rd full-profile conjoint analysis techniques. Fifty-one patients with conge stive heart failure from our medical center (University of Pennsylvania Med ical Center, Philadelphia, PA) and 47 age-, gender-, and race-matched contr ol subjects were studied. Part-worths and importance weights were significa ntly different for shortness of breath and depression between patients and control subjects. Symptom-sensitive (n = 33) and survival-sensitive (n = 17 ) treatment outcome preference segments were identified within the patient group. Importance weights for symptom-sensitive versus survival-sensitive p atients were as follows: tiredness 0.30 +/- 0.10 versus 0.16 +/- 0.09 (P < .01); shortness of breath 0.26 +/- 0.08 versus 0.21 +/- 0.08 (P =.07); depr ession 0.26 ir 0.09 versus 0.19 +/- 0.09 (P = .01); and survival 0.18 +/- 0 .07 versus 0.43 +/- 0.11 (P < .01), There were no significant predictors of which treatment outcome preference segment a patient belonged. Control sub jects did not display similar preference segmentation. Conclusions: Symptomatic congestive heart-failure patients were clustered i nto symptom-sensitive and survival-sensitive segments in a manner suggestin g that treatment outcomes of improved symptoms were of greater importance t o the majority than longer survival. A full understanding of these individu al preferences may have important implications for the design of therapy fo r heart-failure patients.