Reliability, sensitivity and responsiveness of the Infant Behavioral Assessment in very preterm infants

Aim: The aim of this study is to investigate the reliability, sensitivity and responsiveness of the Infant Behavioral Assessment (IBA) to evaluate neurobehavioural organization in very preterm infants. Methods: Videotaped assessments of very preterm infants participating in a recent trial served to evaluate a standardized IBA observation. Inter-rater reliability was based on 40 videos scored by two independent observers, using percentage agreement and weighted Kappa's. Sensitivity was evaluated by comparing the IBA results of 169 infants at 35-38 weeks postmenstrual age, dichotomized according to two developmental risk factors. For responsiveness, the effect size (ES) was calculated between 0 and 6 months corrected age in all intervention and control infants and in subgroups of high-risk intervention and control infants with oxygen dependency 28 days. Results: Inter-rater agreement was 93% in the total assessment; Kappa agreement was moderate to good in the behavioural categories. Significant differences were found between groups with or without risk factors. Larger differences between ESs in the randomized groups with oxygen dependency 28 days than in the total randomized groups reflect the responsiveness of the IBA. Conclusion: In this study, we found satisfactory to good clinimetric characteristics of the IBA in very preterm born infants