Early Identification of Autism Using Social Attention and Communication Surveillance in Tianjin, China (SACS-C)

Background: Advances in understanding the early autism phenotype means that earlier identification and diagnosis of Autism Spectrum Disorders (ASD) is now possible within the second year of life. However, this knowledge is largely based on studies undertaken in the Western world and published in English, with findings rarely impacting practices in non-English speaking nations. This situation is exacerbated in developing nations and low resource communities. The Tianjin Women and Children's Health Center sought a reliable and valid tool suitable for use by doctors in community hospitals, with high sensitivity and specificity, to detect ASD at younger ages within its community-based three level childcare network.

Objectives: Our objective was to translate evidence-based knowledge in developmental surveillance for ASD, by training early childhood medical professionals on Social Attention and Communication Surveillance - Chinese (SACS-C) in a pilot program in Tianjin, China.

Methods: Two-hundred and fifty two (252) medical professionals were trained to monitor early signs of ASD at 12-, 18- and 24-months of age as part of their routine developmental checks of babies. Sixty-two community hospitals in six urban districts in Tianjin participated in the study with 10,516 children monitored between 12 – 24 months of age: 3,178 at 12-months; 3,758 at 18-months; and 3,580 at 24- months (between May 2013 and October 2014). Children identified at risk of an ASD were referred to the Tianjin Women’s and Children’s Health Centre for further assessments and diagnosis.

Results: Eighty-nine (89) children were referred, and the referral rate was 0.85%; however, only 56 children (63% of referred families) attended the follow-up assessment session. Of these, 24 met the criteria for ASD (43% positive predictive value), 24 had a developmental and/or language delay (43%), and 8 were typically developing (14%). 50% (n=12) of those who had developmental and/or language delay were diagnosed with ASD at 3 years of age, resulting in a final Positive Predictive Value of 64%. Combining the number of children assessed who had a classification of an ASD, with 64% of those who were referred as "at risk" and not assessed, results in an estimated rate of 1:200 children with ASD in the sample monitored using SACS-C.

Conclusions: The findings indicate that it is possible to translate evidence-based practice from one cultural context to another. Developmental surveillance of social and communication behaviors, which differ according to the age at which the child is monitored, enables the accurate identification of children at risk for ASDs between 12- to 24-months. Clearly greater education about the early signs of ASD for all primary health-care professionals is needed to increase awareness of these conditions more widely and to highlight the importance of earlier diagnosis and intervention.