Body Temperature Trajectory Predictive of Delirium Duration in Critically Ill Children

Introduction: Delirium affects >30% of critically ill children and is associated with substantial morbidity and excess mortality. It is essential to identify children at highest risk for delirium soon after PICU admission to focus delirium prevention strategies. Studies in adults have utilized group-based trajectory modeling to identify four temperature trajectories (normothermic (N), hypothermic (H), hyperthermic fast-resolvers (HF), and hyperthermic slow-resolvers (HS)). These trajectories are associated with inflammatory markers and patient outcomes. We hypothesized that temperature trajectories would be independently associated with delirium burden in critically ill children.


Methods: In this prospective observational study, subjects were screened for delirium twice daily throughout their PICU stay using the Cornell Assessment for Pediatric Delirium. Children were classified by delirium status (ever/never) and delirium duration (number of days delirious). Temperature measurements within the first 72 hours of PICU admission were extracted from the electronic medical record. Using a validated model, children were classified into 4 temperature trajectories: N, H, HF, and HS. Linear and logistic regression generalized estimating equation models were used to assess the relationship between temperature trajectory group and delirium occurrence and duration, when controlling for relevant clinical covariates.


Results: Of the 1,482 subjects, 17.3% developed delirium (n=257), with a median duration of 2 days (IQR 1-5). 46% of the cohort were classified as N (n=688), 31% H (n=460), 11% HF (n=164) and 11% HS (n=170). Children in the hyperthermic trajectories were more likely to develop delirium (HF, 25%; HS, 32%) than the normothermic and hypothermic trajectories (N, 16%; H, 12%), p< 0.001. After adjusting for age, chronic conditions, developmental disability, and severity of illness, children classified as HS were estimated to have 5 more days with delirium when compared to children classified as HF (adjusted odds estimate -5.01, p=0.007).


Conclusions: Temperature trajectory is an independent and clinically relevant predictor of children at highest risk for delirium. This presents an opportunity for early identification of at-risk children to focus interventions to mitigate delirium burden and improve outcomes.