Related Experiment Video
Updated: Jun 21, 2026
Assessment and Evaluation of the High Risk Neonate: The NICU Network Neurobehavioral Scale
Published on: August 25, 2014
Validation of CRIB II for prediction of mortality in premature babies
Pallav Kumar Rastogi1, V Sreenivas, Nirmal Kumar
1St Stephens Hospital, Department of Neonatology and Pediatrics, Tis Hazari, Delhi, India.
Objective:
Validation of Clinical Risk Index for Babies (CRIB II) score in predicting the neonatal mortality in preterm neonates < or = 32 weeks gestational age.
Design:
Prospective cohort study.
Setting:
Tertiary care neonatal unit.
Subjects:
86 consecutively born preterm neonates with gestational age < or = 32 weeks.
Methods:
The five variables related to CRIB II were recorded within the first hour of admission for data analysis. The receiver operating characteristics (ROC) curve was used to check the accuracy of the mortality prediction. HL Goodness of fit test was used to see the discrepancy between observed and expected outcomes.
Results:
A total of 86 neonates (males 59.6% mean birthweight: 1228 +/- 398 grams; mean gestational age: 28.3 +/- 2.4 weeks) were enrolled in the study, of which 17 (19.8%) left hospital against medical advice (LAMA) before reaching the study end point. Among 69 neonates completing the study, 24 (34.8%) had adverse outcome during hospital stay and 45 (65.2%) had favorable outcome. CRIB II correctly predicted adverse outcome in 90.3% (Hosmer Lemeshow goodness of fit test P=0.6). Area under curve (AUC) for CRIB II was 0.9032. In intention to treat analysis with LAMA cases included as survivors, the mortality prediction was 87%. If these were included as having died then mortality prediction was 83.1%.
Conclusion:
The CRIB II score was found to be a good predictive instrument for mortality in preterm infants < or = 32 weeks gestation.