Restraint fading and the development of alternative behaviour in the treatment of self-restraint and self-injury.

Jarrold et al. (1994) worked with an adult who had severe intellectual disability. The person wore arm restraints almost all day to stop self-hitting. The team ran a functional analysis but got unclear results. They still moved ahead with a three-part plan: slowly shorten restraint time, reward compliance, and teach the client to ask for breaks with a micro-switch.

Restraint time dropped from 24 hours to under 2 hours a day. Self-injury stayed low. The client learned to press a switch to request short releases. Staff could remove the restraints completely during most activities without problem behavior returning.

Jones et al. (1992) showed that teaching a simple request keeps problem behavior down even with new teachers. C et al. built on that by adding gradual restraint removal and compliance rewards. O'Reilly (1996) later showed that when functional analysis is unclear, checking events outside sessions—like respite care—can explain rare self-injury. C et al. did not probe extra-session factors, so F extends their work by offering a next step when FA data are muddy. Greer et al. (2024) warns that big early cuts in reinforcement can cause resurgence; C et al. avoided this by thinning restraint time in small steps while keeping differential reinforcement strong.

You can start restraint fading even if the functional analysis is inconclusive. Pair each step down in restraint with immediate rewards for safe behavior and teach a simple mand. Move in tiny time blocks—minutes, not hours—to prevent resurgence. This gives you a road map for clients who enter services already physically restrained.