Assessing current sleep–cognition evidence and translation into oncology practice.

Abstract

e24129 Background: Despite strong evidence linking sleep disturbance with cognitive outcomes in cancer survivors, oncology practice lacks a standardized framework defining how sleep should be integrated into cognitive evaluation and management. The aim of this integrative review was to describe how sleep and cognition have been conceptualized and measured in existing studies and to examine the extent to which current approaches support or limit clinical interpretability and translation. Methods: We conducted an integrative review following Whittemore and Knafl methodology. A comprehensive search was performed in Embase, Web of Science, Ovid MEDLINE, PubMed, and CINAHL for studies published between 2016 and April 2025. Eligible studies included empirical data on both sleep and cognitive outcomes in cancer survivors. After systematic screening, 37 studies were included. Data extraction captured study design, population characteristics, cancer type and treatment exposures, sleep and cognitive measurement tools, analytic approaches, and reported associations. Sleep and cognitive outcomes were synthesized exactly as operationalized in the original studies, preserving their reported constructs and metrics, to evaluate the scope, consistency, and clinical interpretability of existing evidence. Results: Across 37 studies including 18,509 cancer survivors, sleep and cognition were conceptualized and measured using highly variable and largely non-standardized approaches. Nearly half of studies were cross-sectional (46%), limiting temporal or clinical inference. Sleep assessment relied predominantly on self-report instruments (89%), most often PSQI or ISI, with only 11% incorporating objective measures. In most studies, sleep was reported as a single global score rather than as distinct dimensions. Cognitive outcomes were similarly heterogeneous. More than two-thirds of studies used self-reported cognition, while only about half included performance-based testing, with inconsistent selection of cognitive domains and limited standardization of outcome definitions. Because sleep was rarely characterized by dimension and cognition was inconsistently assessed by domain, most studies reported broad sleep–cognition associations without providing clinically interpretable guidance on which sleep characteristics should be evaluated, which cognitive outcomes they relate to, or how findings could inform clinical decision making. Conclusions: Sleep–cognition evidence has not translated into oncology practice because of heterogeneous measurements, largely non-standardized approaches that do not specify which sleep dimensions relate to which cognitive domains. Without dimensional sleep assessment and domain-specific cognitive evaluation, findings remain descriptive rather than clinically actionable. Standardized frameworks are needed to enable meaningful clinical integration.