Identification of symptom clusters among patients with heart failure: An international observational study

摘要

Background: Virtually all patients with heart tailure experience multiple symptoms simultaneously, yet clinicians and researchers usually consider symptoms in isolation. Recognizing and responding early to escalating symptoms is essential to preventing hospitalizations in heart failure, yet patients have considerable difficulty recognizing symptoms. Identification of symptom clusters could improve symptom recognition, but cultural differences may be present that must be considered. Objectives: To identify and compare symptom clusters in heart failure patients from the United States, Europe and Asia. Design: Cross-sectional, observational study. Settings: In- and out-patient settings in three regions of the world: Asia (i.e., China and Taiwan); Europe (i.e., the Netherlands and Sweden); and the United States. Participants: A total of 720 patients with confirmed heart failure. Propensity scoring using New York Heart Association Classification was used to match participants from each of the three regions. Methods: Symptoms were identified using the Minnesota Living with Heart Failure Questionnaire. To identify symptom clusters we used cluster analysis with the hierarchical cluster agglomerative approach. We used the Euclidean distance to measure the similarity of variables. Proximity between groups of variables was measured using Ward's method. The resulting clusters were displayed with dendrograms, which show the proximity of variables to each other on the basis of semi-partial R-squared scores. Results: There was a core group of symptoms that formed two comparable clusters across the countries. Dyspnea, difficulty in walking or climbing, fatigue/increased need to rest, and fatigue/low energy were grouped into a cluster, which was labeled as a physical capacity symptom cluster. Worrying, feeling depressed, and cognitive problems were grouped into a cluster, which was labeled as an emotional/cognitive symptom cluster. The