Affiliation:
1. Department of Pediatric Medicine, Dr. B. C. Roy Post-graduate Institute of Paediatric Sciences, Kolkata, West Bengal, India
2. Department of Neonatology, Institute of Post-graduate Medical Education & Research and SSKM Hospital, Kolkata, West Bengal, India
Abstract
Premature births before 32 weeks of gestation pose a significant challenge in perinatal healthcare. The global incidence of premature births is on the rise. To ensure appropriate care and resource allocation, assessing the severity of neonatal illnesses is crucial. In this context, the study assessed CRIB-II and SNAPPE-II, two widely used scoring systems for newborn illness severity. These systems consider both perinatal and postnatal variables, providing comprehensive assessments. This prospective observational study conducted in an Eastern Indian tertiary care children’s hospital between February 2021 and August 2022 aimed to evaluate and compare the predictive capabilities of two scoring systems, CRIB-II and SNAPPE-II, for in-hospital mortality and long-term neurodevelopmental outcomes in preterm infants born before 32 weeks of gestational age. Results showed that both CRIB-II (AUC of 0.862, 95% CI: 0.745-0.939) and SNAPPE-II (AUC of 0.919 95%, CI: 0.816-0.975) demonstrated similar predictive abilities for neonatal mortality (difference: 0.057, 95% CI: -−0.044 to 0.16, P value .268). However, when it came to predicting long-term neurodevelopmental outcomes at 12 months of corrected gestational age, SNAPPE-II maintained significance, while CRIB-II did not. The discrepancy in their performance over time emphasizes the complexity of predicting long-term outcomes in very preterm infants. To conclude, this study suggests that CRIB-II and SNAPPE-II are comparable in predicting neonatal mortality. Given their similarity, CRIB-II may be preferred for its practicality in high-capacity NICUs. However, further research is needed to assess these systems’ utility in comparing later neurodevelopmental outcomes and to explore the impact of new factors on their predictive accuracy.