Introduction Paediatric delirium (PD) in critically ill children has an estimated prevalence of 5 to 20%. The aim of this study was to determine the preliminary psychometric properties of the Sophia Observation withdrawal Symptoms-Paediatric Delirium Scale (SOS-PD).

Methods This prospective study included patients aged 3 months or older and admitted for more than 48 h. These patients were assessed three times a day with the SOS-PD. If the SOS-PD total score was 4 or higher the child psychiatrist was consulted to confirm the diagnosis of PD using the Diagnostic and Statistical Manual-IV criteria. The child psychiatrist was blinded for outcomes of the SOS-PD. The interrater reliability was independently tested in paired observations between the care-giving nurse and a researcher and calculated by using the Intra Class Correlation coefficient (ICC).

Results A total of 2088 SOS-PD assessments were obtained in 150 children (median age 54 months; IQR 14–146). Most frequent symptoms in PD patients (n = 13) were anxiety (13%), motor disturbance (13%), sleeplessness (16%) and less eye contact (19%) We compared 63-paired observations of the child psychiatrist versus SOS-PD assessments in 14 patients. The sensitivity was 90.1% (95% CI 75.6–97.9%) and positive predictive value 96.8% (95% CI 90.5–100). The specificity was 96.7% (95% CI 82.7–99.4) and the negative predictive value 90.6% (95% CI 80.5–100). The ICC of 16-paired observations was 0.90 (95% CI 82.7–99.4) for the SOS-PD scale.

Conclusions The SOS-PD scale shows promising validity for screening PD. Future research is aimed to establish the psychometric properties in a multicenter study.