Purpose: To identify symptom clusters in advanced cancer patients attending a palliative radiotherapy clinic using the Edmonton Symptom Assessment System (ESAS).
Methods: Principal component analysis (PCA), exploratory factor analysis (EFA), and hierarchical cluster analysis (HCA) were used to identify symptom clusters among the nine ESAS items using scores from each patient's first visit.
Results: ESAS scores from 182 patients were analyzed. The PCA identified three symptom clusters (cluster 1: depression-anxiety-well-being, cluster 2: pain-tiredness-drowsiness, cluster 3: nausea-dyspnea-loss of appetite). The EFA identified two clusters (cluster 1: tiredness-drowsiness-loss of appetite-well-being-pain-nausea-dyspnea, cluster 2: depression-anxiety). The HCA identified three clusters similar to the PCA with an exception of the loss of appetite item being classified under cluster 1 rather than 3. Two to three symptom clusters were identified using three analytical methods, with similar patterns reported in the literature. Particular groups of items co-occurred consistently across all three analyses: depression and anxiety; nausea and dyspnea; as well as pain, tiredness, and drowsiness.
Conclusion: Three similar symptom clusters were identified in our patient population using the PCA and HCA; whereas, the EFA produced two clusters: one physical and one psychological cluster. Given the implications of symptom clusters in the management of quality of life, clinicians should be aware of these clusters to aid in the palliative treatment of patients.
Keywords: Cancer; Edmonton Symptom Assessment System; Quality of life; Symptom clusters.