Association between self-reported sleep disturbance and other symptoms in patients with advanced cancer.

CONTEXT Sleep disturbance (SD) is a significant source of distress for patients with cancer. Studies of patients with advanced cancer receiving palliative care to identify symptoms associated with the severity of SD are limited. OBJECTIVES In this study, we sought to identify the symptoms measured by the Edmonton Symptom Assessment Scale (ESAS) that are associated with SD, as measured by the Pittsburgh Sleep Quality Index (PSQI). Secondary aims of the study were to determine the association between occurrences of SD with occurrences of other symptoms and screening performance of the ESAS-Sleep item against the PSQI. METHODS We reviewed the completed ESAS and PSQI assessments of 101 patients with advanced cancer who were receiving palliative care and had been admitted to prospective clinical trials previously initiated by us. Patients with a PSQI score of ≥ 5 were considered to have an SD. The frequency and severity of the ESAS symptoms items, their correlation with each other, the PSQI score, and the screening performance of the ESAS-Sleep item were calculated. RESULTS The median age of patients was 60 years. Most were white non-Hispanic (73%), had lung or breast cancer (41%), and were diagnosed with SD (85%). The PSQI score was correlated with the ESAS items of pain (r=0.27, P=0.006), dyspnea (r=0.25, P<0.001), well-being (r=0.35, P<0.0001), and sleep (r=0.44, P<0.0001). Compared with patients without SD, those with SD were more likely to report pain (P=0.0132), depression (P=0.019), anxiety (P=0.01), and a poorer sense of well-being (P=0.035). An ESAS-Sleep item cutoff score of ≥ 3 (of 10) resulted in a sensitivity of 74% and a specificity of 73%. CONCLUSION SD is associated with increased frequency of pain, depression, anxiety, and a worse sense of well-being. These four symptoms should be assessed in all patients with advanced cancer with a complaint of SD. The ideal cutoff point of the ESAS-Sleep item for screening for SD is a score of ≥ 3. More research is needed to better characterize this frequent and distressing syndrome.