A New Quantitative Triage System for Hospitalized Neonates to Assist with Decisions of Hospital Evacuation Priorities

Hospitalized neonates are vulnerable to natural and man-made disasters because of their persistent requirement for medical resources and may need to be evacuated to safe locations when electricity and medical gas supply become unreliable. In Japan, a triage system for hospitalized neonates, or the Simple Triage and Rapid Treatment for Neonates, Revised (START-Neo-R), has been used to determine whether neonates are in suitable conditions for transportation. However, this scale is not useful to determine the evacuation order of neonates because a considerable number of evacuees are classified into the same categories. To solve this problem, a novel triage system, Neonatal Extrication Triage (NEXT) was developed. This study tested the validity and reproducibility of both triages and compared them with a standardized prognostic scoring system for hospitalized neonates, the Neonatal Therapeutic Intervention Scoring System (NTISS). In this retrospective observational study, physicians and nurses independently assessed each neonate hospitalized at a tertiary neonatal intensive care unit (NICU) twice weekly using NEXT and START-Neo-R. The NEXT system comprises six questionnaires regarding medical resources required during transition and transportation, providing composite scores on a 12-point scale. The START-Neo-R classified neonates into five levels based on the severity of disease and dependence on medical care. Inter-rater reliability of both systems was assessed using Cohen's kappa coefficient, whereas the criterion validity with NTISS was assessed using Spearman's correlation coefficient. Overall, 162 neonates were assessed for 49 days, resulting in triage data for 1,079 accumulated patients. Both NEXT scores and START-Neo-R ranks were well-dispersed across different levels without excessive accumulation in specific categories. Inter-rater reliability of NEXT (kappa coefficient, 0.973; 95% confidence interval, 0.969-0.976) and START-Neo-R (kappa coefficient, 0.952; 95% confidence interval, 0.946-0.957) between physicians and nurses was sufficiently high. The correlation coefficient of NEXT and START-Neo-R scores with NTISS scores were 0.889 (P