Abstract Introduction Insomnia symptoms are highly prevalent during systemic cancer therapy and may exacerbate fatigue, cognitive dysfunction, and treatment intolerance. Despite recognition as a distressing symptom, the impact of insomnia on acute health-care utilization remains poorly quantified. Sleep disturbance represents an emerging, potentially modifiable toxicity of cancer treatment. This study evaluated whether patient-reported insomnia predicted unplanned emergency department visits or hospitalizations during active chemotherapy or immunotherapy to identify sleep as a target for supportive-care intervention. Methods We performed a multicenter retrospective cohort study of adults receiving systemic therapy for solid tumors or hematologic malignancies between 2019 and 2024. Insomnia was assessed at baseline and each treatment cycle using a standardized 0–10 numeric scale; clinically significant insomnia was defined as ≥4. The primary outcome was any unplanned emergency-department visit or hospitalization within 30 days of a treatment cycle. Secondary outcomes included treatment delay, dose reduction, and corticosteroid escalation. Multivariable mixed-effects logistic models (patient random intercept) adjusted for age, sex, cancer type and stage, performance status, pain score, anxiety and depression screening, corticosteroid use, and treatment line. Propensity-score weighting and exclusion of planned admissions were performed as sensitivity analyses. Results Among 2486 patients across seven centers (mean age 61 ± 11 years; 54 % female; 34 % hematologic malignancy), clinically significant insomnia occurred in 46 % of treatment cycles. Unplanned acute-care use occurred in 22 %. Insomnia independently predicted higher odds of unplanned care (adjusted odds ratio 1.48, 95 % confidence interval 1.24–1.76; p 0.001). A dose–response relationship was observed per one-point increase in insomnia score (adjusted odds ratio 1.07 per point, 95 % confidence interval 1.04–1.10; p 0.001). Associations remained consistent across cancer types and after propensity weighting. Insomnia also predicted treatment delay (adjusted odds ratio 1.22, 95 % confidence interval 1.03–1.45; p = 0.02) and corticosteroid escalation (adjusted odds ratio 1.31, 95 % confidence interval 1.10–1.56; p = 0.003). Conclusion Patient-reported insomnia during active cancer therapy independently predicted increased unplanned acute-care utilization and treatment disruption, supporting recognition of sleep disturbance as a clinically meaningful toxicity. Routine insomnia screening and rapid, non-pharmacologic management may reduce avoidable hospital visits and enhance treatment continuity. Support (if any)
Elzein et al. (Fri,) studied this question.