Ratings of fetal alcohol syndrome facial features by medical providers and biomedical scientists

Abstract

Medical providers, (obstetricians, pediatricians) and biomedical research scientists engaged in fetal alcohol research for many years evaluated facial photographs of newborns previously diagnosed clinically as having or not having fetal alcohol syndrome (FAS). Medical providers and biomedical scientists did not differ significantly in their ratings. Children independently diagnosed as having FAS were distinguished from non-FAS children by both groups. Providing raters with additional information about children (e.g., birth weight) did not alter judgements significantly. Raters were highly consistent in the ratings they assigned to children (r = 0.96). Accuracy was assessed using signal detection measures (e.g., likelihood ratio, d'). Based on these measures, the two occupational groups did not differ significantly, and rater accuracy was highly significant. Mean ratings were highly correlated with positive maternal Michigan Alcohol Screening Test scores (r = 0.84). These results suggest that the facial features associated with fetal alcohol exposure are readily identifiable, and this, in turn, implies that recognition problems need not be a major contributor to ascertainment of FAS.