Delirium and Developmental Disability: Improving Specificity of a Pediatric Delirium Screen*

Objectives: Children with developmental disabilities are at high risk for developing delirium when critically ill. However, existing pediatric delirium screening tools were designed for children with typical development. The objective of this study was to improve the specificity of the Cornell Assessment for Pediatric Delirium, to allow for accurate detection of delirium in developmentally delayed children admitted to the PICU. We hypothesized that the Cornell Assessment for Pediatric Delirium, when combined with fluctuation in level of awareness as measured by the Richmond Agitation-Sedation Scale, would be valid and reliable for the diagnosis of delirium in developmentally delayed children. Design: Prospective observational double-blind cohort study. Setting: Tertiary care academic PICU. Patients: Children with moderate to severe developmental delay. Interventions: Each child was evaluated by the bedside nurse with the Cornell Assessment for Pediatric Delirium once every 12 hours and the Richmond Agitation-Sedation Scale every 4 hours. Cornell Assessment for Pediatric Delirium (score >= 9) + Richmond Agitation-Sedation Scale fluctuation (change in Richmond Agitation-Sedation Scale score of at least 2 points during a 24-hr period) was compared with the criterion standard psychiatric evaluation for diagnosis of delirium. Measurements and Main Results: Forty children participated; 94 independent paired assessments were completed. The psychiatrists' diagnostic evaluations were compared with the detection of delirium by the Cornell Assessment for Pediatric Delirium and Richmond Agitation-Sedation Scale. Specificity of the Cornell Assessment for Pediatric Delirium + Richmond Agitation-Sedation Scale fluctuation was 97% (CI, 90-100%), positive predictive value of Cornell Assessment for Pediatric Delirium + Richmond Agitation-Sedation Scale fluctuation was 89% (CI, 65-99%); and negative predictive value remained acceptable at 87% (95% CI, 77-94%). In addition, to confirm interrater reliability of the criterion standard, 11 assessments were performed by two or more psychiatrists in a blinded fashion. There was perfect agreement (kappa = 1), indicating reliability in psychiatric diagnosis of delirium in developmentally delayed children. Conclusion: When used in conjunction with Richmond Agitation-Sedation Scale score fluctuation, the Cornell Assessment for Pediatric Delirium is a sensitive and specific tool for the detection of delirium in children with developmental delay. This allows for reliable delirium screening in this hard-to-assess population.