Reliability of the Japanese SOS-PD.

Background
Detecting pediatric delirium in critically ill children is important. The Sophia Observation withdrawal Symptoms and Delirium scale (SOS-PD) is a tool for assessing both pediatric delirium and iatrogenic withdrawal symptoms and contains a question for parents to assist in detecting pediatric delirium.
Objectives
The aim was to translate the Japanese SOS-PD and to perform a cross-culture validation of the J-SOS-PD pediatric delirium dimension while confirming the accuracy of family assessments of pediatric delirium.
Methods
The translation was undertaken with the internationally established forward- backward translation method. Pediatric delirium was simultaneously evaluated and compared between psychiatric diagnoses based on assessment by a pediatric intensivist and the Japanese version of the SOS-PD as evaluated by PICU researchers. We evaluated the criterion validity (sensitivity and specificity), cut-off point using a receiver operating characteristic (ROC) curve, and reliability with Cohen’s κ coefficient and intraclass correlation coefficients (ICC).
Results
A total of 125 independent assessments were performed in 67 children with a median age of 15 (IQR 5, 54) months and with a pediatric delirium-positive rate of 30% based on psychiatric evaluation. Based on the ROC curve analysis, the cut-off point of 4 was the most appropriate within the original scale and the Japanese SOS-PD version showed high sensitivity (0.92, 95% CI 0.84–1.00) and specificity (0.97, 95% CI 0.94–1.00) at a cut-off point of 4 and high reliability within the researcher assessments (κ = 0.95). We also verified family assessments of pediatric delirium as showing high sensitivity (0.90) and specificity (0.89) over 36 assessments.
Conclusions
The Japanese version of the SOS-PD shows a high accuracy similar to the original. Moreover, we revealed high accuracy in family perception of pediatric delirium that could promote family presence in PICU settings.