A secure children's home (SCH) mentioned in the title of this post typically refers to a place where care and accommodation is given to "children and young people who have been detained or sentenced by the Youth Justice Board (YJB) and those who have been remanded to secure Local Authority (LA) accommodation." The paper published by P.J. Kennedy and colleagues  provides some interesting details about the prevalence of autism or "symptom complexes compatible with ASD [autism spectrum disorder]" among young people residing in two SCHs here in Blighty, alongside some initial demographic data around those detected.
Over 110 adolescents housed in SCHs were included for study, where their support workers completed the Social Communication Questionnaire (SCQ) on their behalf. The SCQ is one of the more important autism screening questionnaires. It's not however considered 'diagnostic' and is not without it's issues (see here), hence the use of the words "Symptom Complexes Compatible with Autistic Spectrum Disorder" by Kennedy et al. The results suggested that approaching 15% of their cohort were judged to present with symptoms/traits compatible with a diagnosis of ASD. Authors also mentioned how certain 'aspects' of positive screeners might also be important; for example, "differences in gender, legal status and a history of Child Sexual Exploitation (CSE)." There's also mention of an important word - 'vulnerability' - in the Kennedy paper, which I've often said is still very much under-used when it comes to the autism spectrum (see here and see here for examples).
Whilst further work is required on autism (whether in diagnosis or 'symptom complexes') in the context of SCHs, the suggestion that autism or autistic traits may be over-represented among SCH residents is an important one. I should add that whilst SCHs house those who have been 'detained or sentenced by the Youth Justice Board' they do also cater for quite a wide range of issues/difficulties outside of those presenting with offending behaviour. This includes those who may be at risk to themselves and others alongside those with mental health difficulties. This is an important detail in the context that I don't want to promulgate the idea that towards 15% of 'offenders' have an undiagnosed autism spectrum disorder (ASD). We don't yet know this, and need to keep in mind other important issues such as the growing realisation that autism rarely exists in a diagnostic vacuum (see here) and what that might mean for those on the autism spectrum who do offend (see here).
 Kennedy PJ. et al. Brief Report: Using the Social Communication Questionnaire to Identify Young People Residing in Secure Children's Homes with Symptom Complexes Compatible with Autistic Spectrum Disorder. J Autism Dev Disord. 2018 Jul 17.