Measurement properties of the Clinical Risk Index for Babies--reliabilty, validity beyond the first 12 hours, and responsiveness over 7 days.

Objectives: Clinical Risk Index for Babies (CRIB) is a simple instrument used to measure clinical risk and illness severity in very low birth-weight infants. We assessed its reliability, validity beyond the first 12 hrs after birth, and responsiveness to individual change in condition after 7 days.
Design: Cohort study.
Setting: Three tertiary and three nontertiary UK hospitals.
Patients: Three hundred ninety-eight infants whose birth weight was <1501 g or who were born before a 31-wk gestation period.
Interventions: Inter- and intrarater reliability of data extraction were assessed by Pearson and intraclass correlation. To validate CRIB, we tested the correlation between clinical risk and illness severity with the risk of: a) death; b) prolonged treatment with supplemental oxygen; and c) disability at 2 yrs. Logistic regression models were fitted to assess validity and responsiveness.
Measurements and Main Results: Reliability coefficients ranged from 0.76 (95% confidence interval, 0.71 to 0.81) to 0.97 (0.94 to 1.00). Throughout the first week, CRIB correlated with the risk of death (p < .001), prolonged treatment with oxygen (p < .001), and disability (p < .001 to p = .033). Improved condition, represented by a reduction in CRIB within the first week, was independently associated with lower risks of each adverse outcome, p < .05.
Conclusions: During the first week, CRIB was reliable, valid, and responsive. These properties support the use of CRIB in the stratification of infants by risk and illness severity in cohort studies, and they also indicate that CRIB may have the potential to be used in other ways in the future.