Objectives: To evaluate test validity of the Pediatric Confusion Assessment Method for the ICU, the Pediatric Anesthesia Emergence Delirium scale, and the newly developed severity scale for the Pediatric Confusion Assessment Method for the ICU; to prospectively assess covariates and their influence on test validity of the scores.
Design: Prospective observational cohort study.
Setting: PICU of a tertiary care medical center.
Patients: Critically ill patients 5 years old or older ventilated or nonventilated with an ICU length of stay of at least 24 hours.
Interventions: None.
Measurements and main results: Patients were scored with the Pediatric Confusion Assessment Method for the ICU and the Pediatric Anesthesia Emergence Delirium scale once daily for a maximum of 21 days. Validity was determined by comparing scoring results with the evaluations of the delirium experts who used the criteria of the Diagnostic and Statistical Manual, 4th Edition, Text Revision, for delirium diagnosis. Sixty-four patients were enrolled and 214 assessments were conducted and included in data analysis. The first assessments within each patient revealed sensitivities of 69.2% for the Pediatric Anesthesia Emergence Delirium scale, 76.9% for the Pediatric Confusion Assessment Method for the ICU, and 84.9% for the severity scale for the Pediatric Confusion Assessment Method for the ICU. Specificities were 98% for all scores. Considering repeated measurements, sensitivities decreased to 35.9% for the Pediatric Anesthesia Emergence Delirium scale and to 52.3% for the Pediatric Confusion Assessment Method for the ICU. The sensitivity of the severity scale for the Pediatric Confusion Assessment Method for the ICU dropped to 71.8%, which was significantly higher compared to the Pediatric Anesthesia Emergence Delirium scale (p = 0.0008). Receiver operator characteristic regression unveiled that sedation and mechanical ventilation had a significant negative effect on the validity of the Pediatric Anesthesia Emergence Delirium scale and the severity scale for the Pediatric Confusion Assessment Method for the ICU. Age and gender had a significant impact on the receiver operator characteristic curve of the severity scale for the Pediatric Confusion Assessment Method for the ICU.
Conclusions: The severity scale for the Pediatric Confusion Assessment Method for the ICU showed the best test validity when used in critically ill children of 5 years old or older. Nevertheless, validity of delirium screening itself depends on patient specific factors. These factors should be taken into consideration when choosing a delirium screening instrument.