The use of physical restraint in the treatment of self-injury and as positive reinforcement.

Hamilton et al. (1978) worked with three people with intellectual disability. Each person hurt themselves often.

The team tested if giving brief physical restraint right after self-injury would stop it. They also checked if restraint could make toy play go up.

Self-injury dropped to zero for every person as soon as restraint followed it. Toy play rose at the same time.

The effect flipped back when restraint stopped, then returned when restraint came back. Restraint worked like candy for these clients.

Rajaraman et al. (2022) now shows you can get the same zero-injury result without ever touching the child. Their kids simply chose to stay in treatment and earned prizes. The 2022 paper supersedes the 1978 idea by removing the need for restraint.

Lord et al. (1997) extends the 1978 finding. They gave preferred items and gentle prompts non-contingently during hygiene. Destructive behavior still fell, showing you can deliver comfort without waiting for problem behavior.

Allison et al. (1980) is a predecessor. They first proved escape can fuel aggression. Hamilton et al. (1978) then showed restraint itself can be the payoff, not just the break from work.

If your client leans into holds or hugs, probe whether restraint feels good. If it does, you can use brief, safe restraint as a reinforcer, but first check the 2022 choice model. Letting the learner pick treatment plus toys or attention may give you the same drop in self-harm with no physical management and far less risk.