LAUSR

e114 www.ccmjournal.org January 2017 • Volume 45 • Number 1 ratings by 41%. These findings suggested that sufficient education and training for delirium assessment could affect the quality of CAM-ICU scoring. Validation of sspCAM-ICU evaluated by trained or untrained bedside nurses is necessary before its practical use. Second, interrater reliability of sspCAM-ICU was not reported in this study. In sspCAM-ICU, scores for attention screening examination, altered level of consciousness, and disorganized brain were provided as continuous variables, whereas pCAM-ICU provides these variables as categoric variables. This characteristic of sspCAM-ICU could affect interrater reliability, especially in the repeated assessments of delirium. Provision of interrater reliability of sspCAM-ICU would be helpful for evaluating the reliability of this novel scoring method. Third, potential risk factors for developing delirium need to be simultaneously analyzed as possible confounders. Schieveld et al (4) demonstrated that underlying conditions including recent changes in sedative medication, neurologic disorders, infections, and respiratory disorders increased the risk of delirium in children. Silver et al (5) identified presence of developmental delay and need for mechanical ventilation as potential risk factors of delirium. These potential risk factors could have affected the diagnostic accuracy of sspCAM-ICU in delirium. Multivariate analysis (e.g., C-statistic) including various confounders would be helpful for better understanding of independent diagnostic accuracy of sspCAM-ICU. In conclusion, provision of additional data by the authors would help us better understand their conclusions regarding the validity and reliability of sspCAM-ICU as a novel delirium screening tool among pediatric ICU patients. Dr. Ohshimo’s institution received funding (supported, in part, by a JSPS KAKENHI grant number: JP 16K09541) from the Japan Society for the Promotion of Science. The remaining authors have disclosed that they do not have any potential conflicts of interest.