Autism Wandering and Elopement: Building a Safety Plan That Holds

Why this is treated as a safety issue, not a behaviour problem

In a study of 648 autistic children aged 30 to 68 months in the Study to Explore Early Development, 59% wandered sometimes or often and 27% had had a serious injury — one resulting in an emergency-department visit or hospitalisation — at some point since birth.

That study also found the relationship is not uniform. Among children with a significant early learning delay, those who wandered had a higher adjusted prevalence of serious injury than those who did not (30% against 18%). Among children without that delay, wandering was not significantly associated with serious injury. The authors describe a complex interplay rather than a single rule, which is a reason to plan around your own child's pattern instead of a general risk figure.

Record the pattern before you buy anything

NICE guidance on supporting autistic children and young people recommends assessing the factors that raise the risk of behaviour that challenges as part of routine assessment and care planning, including the physical environment such as lighting and noise, and the social environment across home, school and leisure. It also recommends developing a risk-management plan rather than responding case by case.

In practice that means writing down what happened immediately before, not only what happened. A fortnight of short notes usually shows something a memory will not: that leaving clusters around a particular transition, a particular room, or the end of a noisy activity.

• Time of day, and what was happening in the ten minutes before.
• Where the person was heading, if that is knowable — water, a road, a specific place they like.
• Who was present, and what changed just before.
• Sensory conditions: noise, crowding, lighting, temperature.
• What ended it, and what helped afterwards.

What caregivers rated as worth the effort

The Interactive Autism Network surveyed 526 caregivers of autistic children about the measures they had tried for elopement. Most families had tried several. Overall they rated their measures as effective but burdensome, and the balance between those two things differed sharply by measure. These are caregiver ratings from one survey rather than trial evidence, and the authors concluded that more study is needed to find measures that are effective, affordable and easy to put in place.

• Fencing and window locks were among the measures with the most favourable effectiveness-to-burden profile.
• Behavioural specialists were commonly used and rated effective; in that survey they were most often funded through insurance.
• Tracking devices were used infrequently and were rated low in effectiveness by the caregivers who had tried them.
• Medications were rated low in effectiveness for elopement specifically, whether taken for that purpose or another one.

Put the school on the same plan, in writing

A plan that lives only at home stops at the school gate. Send the school the same short document you use yourself, and ask for it to be attached to the student record so it survives a change of teacher or a supply day.

Keep it to one page. A plan long enough to need reading is a plan nobody reads during the two minutes that matter.

• The specific triggers and times your own records show, not general autism information.
• What the person is drawn towards, which is where to look first.
• What helps them re-regulate, and what makes it worse — including whether being approached quickly escalates things.
• How they communicate under stress, and whether they respond to their name.
• Who to phone, in what order, and who is allowed to collect them.

What someone who finds your child needs within seconds

The person who finds a child alone is usually not a professional, and rarely has more than a moment. Anything you want them to know has to be immediately visible and short enough to take in at a glance.

Autism Resource Hub can generate a shareable safety card carrying these details, and keep it current across the people who need it, so that the version at school is not last year's copy.

• Name, and how the person prefers to be addressed.
• That they may not answer to their name or make eye contact, and may not speak when distressed.
• How they do communicate, including any device or cards they use.
• Immediate dangers to check first, such as nearby water or roads.
• Two contact numbers, and anything that reliably calms rather than escalates.

What this page does not decide

This page summarises published research and sets out how to organise a plan. It is not clinical, legal or safeguarding advice, it cannot say which measure suits a particular child, and it does not evaluate specific products or services.

If wandering is putting someone at immediate risk, treat it as urgent and raise it directly with your clinical team, your school and your local emergency services rather than working through it alone.