Predictive Value of Red Flags for Communication at 6 and at 12 Months for M-CHAT Results at 18 Months In Children with Autism Spectrum Disorders

Background:  Early identification of autism spectrum disorders (ASD) is associated with improved outcomes.  Few studies have examined the predictive utility of an autism screening tool at 6 and 12 months, especially in community-based populations.

Objectives:  To assess the ability of Red Flags for Communication (RFC), a new autism screening tool administered at 6 and 12 months, to predict M-CHAT results (i.e., pass or fail) at 18 months.

Methods:  Educators from 26 Parents as Teachers Programs, a community-based, in-home parent support service sponsored by public school districts, screened more than 3,000 children ages 5 to 27 months for indicators of autism and other developmental delays. These children were assessed every 6 months using the Ireton Child Development Chart (ICDC); RFC was used for children ages 6 and 12 months; M-CHAT was used for children ages 18 and 24 months.  Using our database of screening tool responses, we determined patterns of passing and failing 3 consecutive observations using RFC and M-CHAT to estimate the association between RFC results (at 6 and 12 months) and M-CHAT at 18 months.  We employed the Fisher exact test to assess statistically significant differences in passing or failing M-CHAT between two groups: (1) consecutive passing RFC at 6 and 12 months vs. (2) consecutive failing RFC at 6 and 12.  With the same statistical method, we also examined the association between RFC results at 6 months alone and M-CHAT results.

Results:  Of the 260 subjects with complete screening data at 6, 12 and 18 months, 8 (or 3.1%) failed RFC at 6 and 12 months.  Three of these consecutive RFC failing subjects (or 37.5%) also failed M-CHAT (at 18 months).  Two of the three subjects with this screening result pattern were diagnosed with autism; the diagnosis of the other one is unknown (i.e., loss to follow-up).  By comparison, 3 of 252 subjects or (1.2%) with consecutive RFC passing also failed M-CHAT (p = 0.0004).  There were 324 subjects with complete screening data at 6 and 18 months.  Among these subjects, 3 of 13 (or 23.1%) who failed RFC (at 6 months) also failed M-CHAT; while 5 of 311 (or 1.6%) who passed RFC (at 6 months) failed M-CHAT (p = 0.0025).

Conclusions:  A child who fails the RFC at 6 months has a 14.1 times higher risk than a community-based population to fail the M-CHAT at 18 months.  The risk for failing the MCHAT at 18 months further increases if the child fails consecutive RFC at both 6 and 12 months.  By contrast, children who passed the RFC at 6 and 12 months had a high probability (96.7%) of passing the M-CHAT at 18 months.  Use of the RFC screening tool at 6 months of age may improve the early identification of autism spectrum disorders in children.