The American Academy of Pediatrics has recommended autism screening for all children at 18 and 24 months, but empirical data are lacking to determine the added value of screening at two time points. Furthermore, diagnostic stability has been shown to be good past the second birthday, but limited data bears on the stability of diagnosis made during the second year of life.
To investigate the incremental validity of a second screen at 24 months, and to establish the stability of diagnosis made at 18 vs. 24 months of age.
Seven thousand and sixteen 18-month olds were screened using the MCHAT (mean age = 18 months 15 days.). Of those, 122 failed both screener and follow-up interview and agreed to a developmental-diagnostic evaluation. Three thousand eighty four 2-year-olds were screened (mean age = 26 months 14 days); of these, 243 failed both screener and follow-up interview and agreed to an evaluation. Further, a subset (n = 612) of the 18-month olds who passed the screen at 18 months received a second screen with the MCHAT at 24 months; of these, 26 failed the screener, but only 2 failed the follow up interview, both of whom received a developmental and diagnostic evaluation. Forty-six children who were evaluated at 18 months were re-evaluated at 4 years old while 141 children evaluated at 24 months received a re-evaluation.
65.6% (n = 80) of 18-month olds who were evaluated were diagnosed with an ASD and 34.4% (n = 42) were diagnosed as non-ASD. 70% (n = 170) of 24-month olds were diagnosed with an ASD and 30% (n = 73) were diagnosed as non-ASD. Only 1 of the 2 children who were evaluated after passing the screen at 18 months but failing at 24 months received an ASD diagnosis. Thirty-two (69.6%) of the ASD and non-ASD diagnoses made at 18 months remained stable at the 4-year old re-evaluation. Sixty seven (47.5%) of the diagnoses made at 24 months remained stable at the 4-year old re-evaluation.
69.6% of 18-month olds maintained the same ASD vs. non-ASD diagnosis at a second evaluation while 47.5% of the 24-month olds maintained their diagnoses (p<.05). Thus, diagnoses made during the second year appear to be as stable, or more so, as those made at 24 months. In this small data set, a second screening at age 24 months picked up very few additional cases.