Usefulness of Clinical Risk Index for Babies based on birth weight in predicting hospital death and severe intraventicular hemorrhage in the SEN 1500 Spanish neonatal network

• Published in: Anales de pediatría (Barcelona, Spain : 2003) Vol. 71; no. 2; pp. 117 - 127
• Main Authors: Guzmán Cabañas, J M, Párraga Quiles, M J, del Prado, N, Ruíz, M D, García Del Río, M, Benito Zaballo, F, Fernández, C, Figueras Aloy, J
• Format: Journal Article
• Language: Spanish
• Published: Spain 01-08-2009
• Subjects: Birth Weight; Cerebral Hemorrhage - epidemiology; Cerebral Hemorrhage - mortality; Female; Hospital Mortality; Humans; Infant, Newborn; Male; Prognosis; Prospective Studies; Risk Assessment; Risk Factors; Severity of Illness Index
• ISSN: 1695-4033
• DOI: 10.1016/j.anpedi.2009.04.007

To evaluate the usefulness of the Clinical Risk Index for Babies (CRIB) in predicting hospital mortality and severe intraventricular hemorrhage (IVH) in very low birth weight infants stratified by weight groups, in the Spanish neonatal network SEN 1500.
A prospective cohort study was made. Morbidity-mortality data and CRIB were collected in newborns weighing below 1500 g and admitted to 68 neonatal intensive care units between January 2002 and December 2006. Data were analyzed globally and stratified by weight groups (< 501 g, 500-750 g, 751-1000 g, 1001-1250 g, 1251-1500 g). Multivariate models were generated and ROC curves were plotted for estimating predictive values.
A total of 10,608 patients were analyzed. The mean weight was 1116 g (SD 267), and gestational age 29.5 weeks (SD 2.9). Low birth weight for gestational age was 34.3% and the multiple birth rate 36%. Prenatal corticoids were given in 78.2%. Severe intraventricular hemorrhage was diagnosed in 8.5%. Gender, prenatal corticoids, birth weight, gestational age and CRIB proved significant for the outcomes. CRIB showed the highest predictive accuracy in all strata (P < 0.001) except in the 501-750 g group, where it was similar to gestational age. Body weight showed the lowest AUC in all groups, except in the 1251-1500 g group, where it was no different to gestational age. Gestational age and CRIB yielded greater AUC values than weight (P < 0.001) in all groups. No significant differences were found between CRIB and gestational age, except in the 751-1000 g group, where gestational age was greater (P = 0.029).
The CRIB is the best predictor among newborns below 1500 g, except in the 501-750 g group, where CRIB is similar to gestational age. Body weight is the worst predictor, except in the group 1251-1500 g, where it is similar to gestational age. The accuracies of CRIB and gestational age in the prediction of IVH are similar, and both superior to body weight. This similarity persists in all the groups, except in the 751-1000 g interval, where gestational age is a better predictor.