Evaluating a two-level screening strategy for the early detecton of Autistic Spectrum Disorders using routine paediatric surveillance

Background: General agreement exists about the need to identify Autistic Spectrum Disorders (ASDs) early.

Objectives: Evidence supports a two-level screening strategy, with the use of primary care facilities at the first level, followed by specialist screening, and clinical evaluation at the second level. The effectiveness of this strategy in identifying ASD in children younger than 3 years was investigated.

Methods: Physicians of well-baby clinics identified children at risk of ASD during routine developmental surveillance. In total 41,445 children living in a geographically defined area (the province of Utrecht, the Netherlands) were examined. Children considered at risk were screened with the 14-item Early Screening of Autistic Traits (ESAT). Those who tested positive were then clinically evaluated at the UMC Utrecht. Diagnoses were confirmed at follow-up, when the children were about 42 months.

Results: Physicians identified 109 children as being at risk of ASD, and these children were tested with the ESAT when they were 27±6 months old (mean±SD). Of these children, 73 tested positively on the ESAT and 66 (90%) were clinically evaluated. Thirty-nine children (53%) were diagnosed with ASD. The positive predictive value of the ESAT was 59% (39 of 66 children who tested positive); 81% of the children who tested false positive with the ESAT had other developmental disorders according to DSM-IV. False negatives included three children that were diagnosed with ASD, after a positive result at follow-up screening.

Conclusions: The two-level screening strategy appears to be effective, especially in lowering the age of diagnosis; however, the rate of detection of ASD is moderate. Recommendations include providing physicians red flags for ASD.