CAPD is an 8-item observational tool for paediatric delirium scored from 0 to 32. It was developed and validated in a tertiary 20-bed PICU by comparing nurse-completed CAPD screens with blinded child psychiatry diagnosis using DSM-IV criteria.

Study	Setting	Population	Cut-point	Sensitivity	Specificity
Traube et al. 2014	General PICU	111 patients, 0–21 years; 248 paired assessments	≥9	94%	79%
Hong et al. 2021	Postoperative surgical ward	170 children; median age 4 years	≥9	87%	98%
Rohmah et al. 2024	Pooled PICU studies	7 CAPD studies	Study-specific	95%	81%

In the original PICU validation, delirium prevalence was about 21%, CAPD took less than 2 minutes to complete, and inter-rater reliability and internal consistency were high.
CAPD remained highly sensitive across most age groups and in children with developmental delay, though specificity was lower in developmentally delayed children.
Multiple translations and validations (including Chinese, Italian, Turkish, Japanese and others) support use across languages and age groups, generally using a cut-point of 9 or higher, with some cohorts exploring 10–11 to increase specificity.
A systematic review and meta-analysis of paediatric ICU delirium tools reported pooled CAPD sensitivity around 0.95 and specificity around 0.81, with higher sensitivity but somewhat lower specificity than pCAM-ICU and psCAM-ICU.
2022 Society of Critical Care Medicine paediatric guidelines recommend routine delirium screening from admission to discharge in the PICU with a validated tool such as CAPD, pCAM-ICU or psCAM-ICU.
CAPD is usually completed by bedside nurses over the course of a shift and is suitable for most hospitalised infants, children and adolescents who are at least arousable to voice; performance may be less reliable in deeply sedated, comatose or severely neurologically impaired patients.

Cornell Assessment of Pediatric Delirium (CAPD)