Evaluation of the psychometrics properties of the alarm distress baby scale among 122 Brazilian children.

The Alarm Distress Baby Scale (ADBB) aims at assessing infant withdrawal behavior. A previous validation study revealed acceptable reliability and validity indices. The present study investigated the psychometric properties of the scale in a larger sample from a culturally different population. Pediatricians evaluated the behavior of 122 infants, 2 and 19 months old, using the ADBB during routine physical consultation. Four investigators (two pediatricians and two nurses not specialized in pediatric care) examined video recordings of the evaluations. Results showed good interrater reliability coefficients among pediatricians and poor correlation when all professionals were grouped together. Test-retest reliability revealed good intraexaminer agreement (r = 0.91). The pediatricians' evaluation using the ADBB was compared with a psychiatric examination to investigate the scale's criterion validity. The cutoff point of 5 provided the best clinical validity (sensitivity of 79%, specificity of 81%). Results from construct validity showed that the scale had three dimensions. Comparison of the factor solution with other construct validity studies of the same instrument revealed similarities and differences. Results suggest that the ADBB may be a useful screening instrument to detect signs of psychiatric alterations related to withdrawal behavior in primary care services, and it is likely to provide consistent information.