1051 Level Of Agreement Between Objective And Subjective Sleep Measures In Lung Cancer Survivors With Insomnia

Abstract

Introduction: Lung cancer survivors (LCS) have the secondhighest incidence of impaired sleep among cancer population. Clinical studies use self-reported and objective measures to assess insomnia in LCS. However, negative beliefs of sleep patterns may impair sleep perception, impact psychological state, and influence insomnia treatment and survivorship outcomes. This study aims to assess the level of agreement between subjective and objective sleep measures in LCS. Methods: Forty-four non-small cell lung cancer survivors, stage I-III, at least 6 weeks after treatment completion, with Insomnia Severity Index >7, were recruited from two sites from 2014-2016. Individuals with sleep apnea, menopause, rotating shift work, uncontrolled substance abuse, or unstable medical/psychiatric illnesses were excluded from the study. Agreement between total sleep time (TST), sleep latency (SL) and sleep efficiency (SE) from Sleep Diary and wrist-actigraphy (i.e. CamNtech) were assessed with Intraclass Correlation (ICC) and Bland-Altman plots. Analysis of ICC, mean difference, standard deviation (SD) and 95% confidence interval (CI) were conducted using SPSS 24. Results: LCS were male (63.6%), 60-69 years old (52.3%), married or living with a partner (50%). Self-reported measurements underestimated TST (-68.2 minutes, ICC 0.45, CI 95%, SD 108.1) and SL (-25.9 minutes, ICC 95%, ICC -0.54, DS 17.5). Objective and self-reported measurements largely overlapped for SE, with a small actigraphy-based overestimation (9.2, ICC 0.2, CI95%, SD 7.3). Bland-Altman plots revealed that the agreement increased with lower SE values. Conclusion: LCS tended to underestimate TST and SL. An apparent agreement between objective and self-reported measures for SE may indicate sleep state misperception. Differences in various sleep parameters and potential covariates should be investigated for contribution to the unexplained considerable individual variability in behavioral treatment response among cancer survivors in larger sample sizes.