Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
Introduction: Approximately 20-80% of head and neck cancer (HNC) survivors experience sleep disturbance and the corollary adverse health outcomes associated with this problem. However, the prevalence, risk factors, and nature of sleep disturbance in this population are poorly characterized.
Methods: This study applies a cross-sectional, mixed methods design. The population includes HNC survivors within a single surgeon’s practice from 2/23-6/24 at an academic medical center. Survivors were evaluated at a single timepoint within their survivorship journey, ranging from the pretreatment phase to any time after 3-months posttreatment. The study was framed by the socio-ecological model of sleep health by Grandner et al.
In the quantitative analysis, the primary outcome was moderate-severe sleep disturbance and/or sleep impairment, as identified by the corollary 4-item PROMIS measures. The co-primary exposures include clinically-significant pain, anxiety, and mood, as identified via the University of Washington Quality of Life v4 measure. The hypothesis was that in adjusted analyses, pain, anxiety, and depressed mood will correlate significantly with moderate-severe sleep disturbance and/or impairment. The following potential sleep-related clinical confounders were accounted for: patient-, tumor-, and treatment-related variables, as well as sleep disorders and medications. Descriptive statistics and multivariable logistic regression were performed using STATA 15. Statistical significance was defined as p<0.05 and p-values were reported as two-sided.
In the qualitative analysis, at the time of survey completion, the surgeon’s clinical visit note was thematically analyzed using a mixed deductive-inductive approach. We aimed to ascertain the nature of the sleep disturbance or impairments, along with the potential determinants of these problems. The first 20% of notes were independently coded, discrepancies were resolved through discussion, and the codebook was iterated. Concordance was 0.81, the codebook was finalized, and the remaining transcripts were single coded using NVivo 12.0. Thematic findings were identified and then triangulated with quantitative findings by the research team.
Results: Among 145 eligible, consecutive HNC survivors, 126 completed relevant questionnaires and were included in the study. A representative population was sampled (female: 23.8%; median age: 65 years; non-Hispanic white: 69.8%; stage III-IV: 73.9%; p16+ disease: 29.4%; primary radiation: 36%; pre-cancer treatment phase: 25%; >3-months posttreatment phase: 75%). The primary exposures and outcome were prevalent (pain: n=17, 14%; anxiety: n=21, 17%; depressed mood: n=4, 3%; moderate-severe sleep disturbance and/or impairment: n=25, 20%). On adjusted analyses, moderate-severe sleep disturbance and/or impairment significantly associated with anxiety (odds ratio [OR] 6.76, 95% confidence interval [CI]: 2.21-21.78), pain (8.37, 95% CI: 2.15-32.52), and moderate-severe comorbidity (OR 8.02, 95% CI: 2.51-25.64). In thematic analysis, among patients with moderate-severe sleep disturbance or impairment, several patients reported < 4 hours of sleep per night (n=6, 24%) and patients or providers often attributed sleep disturbance to pain (n=10; 40%) and cancer-related worry (n=5; 20%).
Conclusion: In our sample, 25% of HNC survivors experienced significant sleep disturbance or impairment, which often correlated with pain, anxiety or cancer-related worry, and moderate-severe comorbidities. Longitudinal studies following HNC survivors throughout the survivorship journey are needed. These data may inform the design of interventions to improve sleep health in this population.