Soon Min Lee1, Myung Hee Lee2, Yun Sil Chang3. 1. Department of Pediatrics, Yonsei University College of Medicine, Seoul, Republic of Korea, smlee@yuhs.ac. 2. Statistics and Data Center, Samsung Medical Center, Seoul, Republic of Korea. 3. Department of Pediatrics, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea.
Abstract
BACKGROUND: A simple predictive indicator of mortality and morbidities is essential to assess neonatal illness severity and plan proper management. OBJECTIVE: This study aimed to test the time-dependent performance of the Clinical Risk Index for Babies (CRIB) II in predicting mortality and major short-term morbidities among very low birth weight infants (VLBWIs). METHODS: This population-based prospective study from 67 Korean Neonatal Network centers performed between 2013 and 2016 included 5,296 VLBWIs with CRIB II calculation and 6,398 infants with CRIB II calculation but without the base excess (CRIB II-BE). A regression model predicting time-dependent mortality and morbidities using the CRIB II score was designed. The discriminate ability of the CRIB II and CRIB II-BE scores in predicting mortality and morbidities was explored using receiver-operating characteristic analysis. RESULTS: CRIB II performed significantly better in predicting mortality than did gestational age or birth weight alone. The time-dependent performance of CRIB II was good in the first 30 days (area under the curve [AUC], 0.8435) and at 31-90 days (AUC, 0.8458). However, it was poor after 90 days (AUC, 0.6576). Specific CRIB II cutoffs were associated with severe intraventricular hemorrhage (AUC, 0.81), bronchopulmonary dysplasia (AUC, 0.77), and mortality or major morbidities (AUC, 0.80), respectively. The model using CRIB II-BE showed similar performance in predicting mortality and morbidities to that of the CRIB II model. CONCLUSION: Certain CRIB II cutoffs were significantly associated with time-dependent mortality, particularly within the first 90 days after birth as well as with short-term morbidities.
BACKGROUND: A simple predictive indicator of mortality and morbidities is essential to assess neonatal illness severity and plan proper management. OBJECTIVE: This study aimed to test the time-dependent performance of the Clinical Risk Index for Babies (CRIB) II in predicting mortality and major short-term morbidities among very low birth weight infants (VLBWIs). METHODS: This population-based prospective study from 67 Korean Neonatal Network centers performed between 2013 and 2016 included 5,296 VLBWIs with CRIB II calculation and 6,398 infants with CRIB II calculation but without the base excess (CRIB II-BE). A regression model predicting time-dependent mortality and morbidities using the CRIB II score was designed. The discriminate ability of the CRIB II and CRIB II-BE scores in predicting mortality and morbidities was explored using receiver-operating characteristic analysis. RESULTS: CRIB II performed significantly better in predicting mortality than did gestational age or birth weight alone. The time-dependent performance of CRIB II was good in the first 30 days (area under the curve [AUC], 0.8435) and at 31-90 days (AUC, 0.8458). However, it was poor after 90 days (AUC, 0.6576). Specific CRIB II cutoffs were associated with severe intraventricular hemorrhage (AUC, 0.81), bronchopulmonary dysplasia (AUC, 0.77), and mortality or major morbidities (AUC, 0.80), respectively. The model using CRIB II-BE showed similar performance in predicting mortality and morbidities to that of the CRIB II model. CONCLUSION: Certain CRIB II cutoffs were significantly associated with time-dependent mortality, particularly within the first 90 days after birth as well as with short-term morbidities.