Screening for fetal alcohol spectrum disorder in infants and young children

Abstract

Introduction: With an estimated prevalence of up to five percent in the general population, fetal alcohol spectrum disorders (FASD) are the most common neurodevelopmental disorder and more prevalent than autism. Early identification and subsequent early intervention have the potential to improve developmental trajectory of children with FASD. In addition, new research suggests supplementation with choline may ameliorate the developmental impairments associated with prenatal alcohol exposure. Availability of a screening tool with acceptable epidemiologic performance criteria may be clinical useful in identification of young children at increased risk for FASD. In this paper we describe the Early Fetal Alcohol Spectrum Disorder Screening Test (E-FAST) to identify young children at increased risk for an FASD. Methods: We developed the E-FAST dataset from previously published studies, comprised of 281 children under 5 years of age, 180 (64.1%) were diagnosed with FASD and 101 (35.9%) were non-FASD. Analysis: The analysis identified seven useful variables (prenatal alcohol exposure, ADHD (Attention Deficit Hyperactivity Disorder), foster care or adopted, small OFC (occipital frontal circumference), communication impairments, impaired social skills, and cognitive deficits. All variables were categorized as yes/no for ease of use in a screening tool. Risk ratios for each of the seven indicators were estimated using two-way table analyses. Weights for each variable were estimated based on the relative strength of their odds ratios. Results: The average age was 2.7 years of age (S.D. 1.29) and ranged from infant (6.4%) to 4 years old (35.9%). Maternal alcohol use alone had a sensitivity of 0.97, specificity 0.65, and accuracy 0.86. For the combined seven variables, sensitivity was 0.94, specificity 0.74, and accuracy 0.87. Thus, the seven-item E-FAST screen had acceptable epidemiologic screening characteristics. Discussion: In the United States, up to 547 infants with FASD are born each day which far exceeds the capacity of multidisciplinary diagnostic clinics. During routine clinical management of infants and young children the use of an evidence-based screening tool provides a time efficient means to exclude large numbers of young children from further follow-up for FASD. Conversely, a positive screen identifies a smaller number of children at increased risk for FASD requiring more intensive evaluation and follow-up.

Keywords: fetal alcohol spectrum disorder; infancy; neurodevelopmental disorder; prenatal alcohol exposure; screening.

FIGURE 1

The number of young children with fetal alcohol spectrum disorder (FASD) and non-FASD comparison children who had or did not have one of the seven E-FAST risk factors included in the analysis. PAE, prenatal alcohol exposure; ADHD, attention deficit hyperactivity disorder; OOH, foster care, adopted, living out of parents home; HC, head circumference; Talking, verbal communication impairments; Social, social skill deficits; Cognitive, cognitive impairments.

FIGURE 2

The two receiver operating characteristics (ROC) curves and the area under the curve for each of the two models. (A) Included six of the seven E-FAST Screening variables (prenatal alcohol use was not included in this model). (B) Models all seven E-FAST screening variables.

FIGURE 3

The early fetal alcohol spectrum disorder screening test for infants and young children (E-FAST).