Background/Objectives: Sleep disturbances have an impact on children’s physical and psychological development, yet little is known about the changes and factors influencing sleep after PICU discharge. To explore the trajectory of changes in sleep quality of critically ill children and to identify factors affecting sleep quality three months after Pediatric Intensive Care Unit (PICU) discharge. Methods: A longitudinal observation study was conducted between November 2022 and November 2023 at a tertiary children’s hospital. The Children’s Sleep Habits Questionnaire (CSHQ) was used at six time points: PICU-admission (T0), 1 week (T1), 2 weeks (T2), 1 month (T3), 2 months (T4), and 3 months (T5) after PICU discharge. The CSHQ is a 33-item parent-report outcome measure evaluating sleep problems. Total scores range between 33 and 99 points. A score of ≤41 indicates normal sleep; a score of >41 indicates sleep disturbance. Data were analyzed using the latent category growth model, univariate analysis, and multifactorial logistic regression. Results: Parents of 237 children completed all follow-up surveys. Prevalence of sleep disorders at T0-T5 of children with a score >41 were 46.5%, 83.5%, 69.7%, 54.3%, 50.2%, and 51.7%, respectively. General linear modeling revealed significant changes in CSHQ scores over time (F = 63.77, p < 0.05). The trajectories of identifying sleep changes were divided into three latent categories: High Sleep Disorder Group (n = 15, 6.33%), Moderate Sleep Disorder Group (n = 110, 45.2%), and No Sleep Disorder Group (n = 115, 48.52%). The trajectories were significantly different among preschool age, baseline sleep habit scores, surgery, and length-of-stay in pediatric wards (p < 0.05). The child’s age and surgical history were independent factors of sleep disturbance. Conclusions: The observed peak in sleep disturbances at 1-month post-PICU suggests that this period may be a critical window to develop and implement targeted interventions to improve sleep. The persistent sleep disorders highlight the need for long-term monitoring.
Wang et al. (Mon,) studied this question.