Sleep and Executive Functioning in Pediatric Traumatic Brain Injury Survivors after Critical Care

Abstract

Over 50,000 children are hospitalized annually for traumatic brain injury (TBI) and face long-term cognitive morbidity. Over 50% develop sleep/wake disturbances (SWDs) that can affect brain development and healing. We hypothesized SWDs would portend worse executive function outcomes in children aged 3–18 years with TBI 1–3 months after hospital discharge. SWDs were defined using the Sleep Disturbances Scale for Children (t-scores ≥ 60). Outcomes included the Global Executive Composite (GEC, t-score) from the Behavior Rating Inventory of Executive Function, Second and Preschool Editions, and multiple objective executive function assessments combined through Principal Components Analysis into a Neurocognitive Index (NCI, z-score). Multiple linear regression evaluated associations between SWDs and executive function outcomes, controlling for covariates. Among 131 children, 68% had clinically significant SWDs, which were associated with significantly worse median scores on the GEC (56 vs. 45) and NCI (−0.02 vs. 0.42; both p < 0.05). When controlling for baseline characteristics and injury severity in multivariable analyses, SWDs were associated with worse GEC (β-coefficient = 7.8; 95% Confidence Interval = 2.5, 13.1), and worse NCI (β-coefficient = −0.4; 95% Confidence Interval = −0.8, −0.04). SWDs in children with TBI are associated with worse executive function outcomes after hospital discharge, and may serve as modifiable targets to improve outcomes.

Exposure to Sedation and Analgesia Medications: Short-term Cognitive Outcomes in Pediatric Critical Care Survivors With Acquired Brain Injury

“…19,70,71,77 Given the extent to which scores on the utilized assessments correlate with one another, and consistent with previous research, principal component analysis (PCA) was conducted to create a single, cumulative neurocognitive index (NCI) to serve as a proxy estimate of performance-based cognition as evaluated by the above tests. 19,70,78 Rather than including each neurocognitive measure individually within our model, the PCA allowed for creation of an outcome measurement that is a single z-score approximating overall ability (the NCI), thereby reducing the possibility of a Type I error in our statistical models. Coefficients less than 0.4 were suppressed from the PCA, and no rotation was applied.…”

“…80,81 As expected, prior analyses show WRAT Word Reading scores account for a significant amount of variance in regression models using NCI as the outcome. 78,82 To account for this strong association between NCI and WRAT Word Reading, we calculated standardized residuals using simple linear regression of NCI predicted by WRAT Word Reading. With this method, the variability in residuals can be considered independent of baseline estimates of cognitive function estimated using WRAT Word Reading.…”

“…We also acknowledge that, ideally, in lieu of a PCA, an exploratory factor analysis (EFA) would be evaluated via a maximum likelihood factor analysis to explore the possible presence of a latent construct related to executive functioning within the well-established PCCNRP assessment battery. 19,71,77,78,122 An EFA approach would provide understanding of dimensionality of the items within the NCI and more robustly assess the amount of variance explained by the single factor. In light of the abundance of research literature suggesting that a factor structure should be stable with about 300 participants for a "scale" such as the NCI, [123][124][125] we utilized a PCA approach over an EFA given our limited sample size.…”

Variability Across Caregiver and Performance-Based Measures of Executive Functioning in an Acute Pediatric Neurocritical Care Population

“… 49–51 ISS scores were assigned by trained trauma program staff in accordance with National Trauma Data Bank standards. 52 , 53 ISS and GCS scores were utilized in principal analyses. Notably, GCS scores given at admission are known to have multiple confounders, limiting its utility in predicting trajectory and outcomes in pediatric TBI, although research has also highlighted that GCS produced the strongest associations with more severe injuries.…”