Cancer-symptom burden and health-related quality of life associated with sleep disturbance and hypnotic use

Objectives: To assess prevalence and correlates of cancer-related sleep disturbance (SD) and hypnotic use, and evaluate changes in cancer symptom burden and health-related quality of life (HRQoL) associated with sleep disturbance change and hypnotic use. Methods: Secondary analysis of the Symptom Outcomes and Practice Patterns (SOAPP) survey of 3,106 adult (aged ≥ 18 years) outpatients receiving treatment for breast, lung, prostate, or colorectal cancer between March 3, 2006 and May 19, 2008. At baseline and approximately four weeks later, patients scored severity of nineteen cancer symptoms from 0='Not present' to 10='As bad as you can imagine'. Both patients and clinicians scored symptom-burden interference for six HRQoL items (from 0='Did not interfere' to 10='interfered completely'). Correlates of SD and hypnotic use were identified using linear and logistic regression, respectively. Multivariate analyses tested whether hypnotic use (Hotelling's T²) or change in SD severity (multivariate linear regression) were associated with changes in symptom burden and HRQoL. Cancer-specific subgroups were evaluated when sample size permitted. Results: SD scores were available for 2,748 participants: 71% female, 86% white, mean age 61 years (range, 23-93 years). Primary cancer sites: breast (51%), colorectal (24%), lung (15%), and prostate (10%). The majority (62.8%) reported SD and 23.5% used hypnotics. The most important SD correlates were other cancer symptoms, regardless of their severity. Distress was the strongest and only universal correlate; cognitive difficulty, drowsiness, and fatigue were also common across cancer-specific subgroups. Hypnotic use was two to three times higher among whites, versus blacks. Several sedating medications correlated positively with hypnotic use, including opioid analgesics and promethazine. Hypnotic use correlated positively with clinician-identified distress, but not with clinician-identified sleep disturbance. Reduced sleep disturbance severity (SDS) correlated with improvement in nearly all symptoms (most notably: distress, dry mouth, and fatigue), and with improved HRQoL (possibly mediated by other symptoms). Hypnotic use correlated modestly with lower overall symptom burden, but not HRQoL change. Conclusions: Relationships observed between SD and many other cancer symptoms argue in favor of therapies that target multiple symptoms. Patterns of hypnotic use raise questions about whether cancer-related SD is being treated adequately, equitably, and effectively.