A limitation of the present study is that the checklist of symptom assessment is a relatively short instrument, which assesses only 15 symptoms. It should be considered to use a more comprehensive tool to assess symptoms when exploring symptom clustering. In addition, the study design is cross-sectional, whereas patients' symptoms are dynamic over time along the disease trajectory. Studies on longitudinal follow-up should be designed to explore symptom clustering at different time points.
In conclusion, symptom clustering exists in advanced cancer patients. Survival, functional performance, bone metastasis, and fluid accumulation are associated with particular types of symptom clustering. Psychological distress is secondary to physical deterioration. To come to grips with the complex manifestations of symptoms in advanced cancer, symptom clusters should be considered in developing clinical guidelines for palliative care.