Screening for autism in extremely preterm infants: problems in interpretation


  • This article is commented on by Lipkin on page 485 of this issue.

Dr Samantha Johnson, Research Fellow, at Department of Health Sciences, University of Leicester, 22–28 Princess Road West, Leicester LE1 6TP, UK. E-mail:


Aim  The aim of this article was to report the prevalence of, and risk factors for, positive autism screens using the Modified Checklist for Autism in Toddlers (M-CHAT) in children born extremely preterm in England.

Method  All children born at not more than 26 weeks' gestational age in England during 2006 were recruited to the EPICure-2 study. At 2 years of age, postal questionnaires incorporating the M-CHAT and additional developmental questions were sent to the parents of each survivor (n=1031; 499 male, 532 female), of which 523 (266 male, 257 female; 51%) were returned completed.

Results  The prevalence of positive M-CHAT screens in this extremely preterm population was 41% (216/523; 130 male; 86 female). Severe bronchopulmonary dysplasia, administration of postnatal steroids, late-onset bacteraemia, and being male were statistically significantly associated with a positive screen. Coexisting disabilities were present in 320 (62%) children. Of 200 children without disability, 16.5% screened positive. In contrast, 63 (95.5%) of those with severe motor impairment (odds ratio 42; 95% confidence interval [CI] 12.9–135) and 175 (55.9%) of those with cognitive impairment (odds ratio 5.3; CI 3.5–8) screened positive. All children with a significant vision or hearing impairment screened positive.

Interpretation  The prevalence of positive M-CHAT screens in extremely preterm children is high, especially in children with neurodevelopmental impairment. Positive screens should be interpreted in the light of other neurodevelopmental sequelae in clinical practice to avoid false-positive referrals.