Evaluation of an arm‐splint belt to reduce self‐injury
A behind-the-back belt plus arm splints can stop severe head-hitting in minutes and be faded without losing the gain, but only with full consent and a renewal watch plan.
01Research in Context
What this study did
One teen with autism hit his head 50-80 times per hour.
The team added a soft belt that pinned his arms behind his back.
They ran an ABAB design: baseline, belt on, belt off, belt on again.
Splints alone had failed; the belt was the new piece.
What they found
Self-hitting dropped to zero the moment the belt was added.
When the belt came off, the hits returned fast.
In the final phase the team loosened the belt a little each day.
The teen kept his hands safe even when the belt let him move more.
How this fits with other research
Fisher et al. (2023) warn against any painful or restrictive device. They say less-intrusive choices work just as well. The belt study shows one case where a mild restraint beat splints alone, but the warning still stands: try everything else first.
Falligant et al. (2024) remind us that automatic self-injury can come back when the setting changes. Morgan’s slow fade may lower that risk, yet you still need a plan to watch for renewal in new rooms or staff.
Wishart (1993) stresses consent. A belt is an aversive, so parents, the client, and the ethics board must all agree and review often.
Why it matters
If mild restraints like splints fail, a short-term belt fade can give immediate safety. Pair it with reinforcement for safe hands and a clear plan to remove the belt step-by-step. Document assent and review the plan weekly so you stay on the right side of ethics and renewal data.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Track baseline hits for one session, then add the belt only if splints alone failed and consent is signed—graph daily as you loosen one belt notch at a time.
02At a glance
03Original abstract
After a functional analysis yielded undifferentiated results, a subsequent assessment suggested self‐injury exhibited by a young boy with autism was sensitive to physical restraint. Canvas arm splints with metal stays were initially effective to reduce self‐injury. Although we successfully faded the number of stays in each sleeve to 3, self‐injury reemerged. We then used a withdrawal design to evaluate a behind‐the‐back belt connected to the arm splints. When the belt and splints were on, self‐injury did not occur. When the belt was removed, self‐injury increased, even though the splints remained intact. Finally, we faded the length of the belt to allow increased range of motion, and rates of self‐injury remained low.
Behavioral Interventions, 2017 · doi:10.1002/bin.1469