Initiatives to reduce the use of seclusion and restraints on people with developmental disabilities: a systematic review and quantitative synthesis.

Contrary to the expectations articulated in public policy, restrictive interventions are commonly used in support services for people with developmental disabilities. This systematic review and quantitative synthesis was undertaken to investigate whether the use of seclusion and restraints on people with developmental disabilities can be reduced. Searches of the Academic Search Complete, CINAHL, MEDLINE, and PsycINFO electronic databases returned 7226 records, of which 11 met the inclusion criteria for this review. A further 3 papers were obtained through scanning the reference lists of those articles included from the initial literature search. All 14 studies were single-subject designs focusing on initiatives to reduce physical or mechanical restraint. Between the baseline and intervention phases, there were mean reductions in the frequency and duration of restraint use of 79% (SD=21%, n=13 subjects from 7 studies) and 45% (SD=58%, n=10 subjects from 6 studies), respectively. For studies in which restraint use to manage agitation and aggression was targeted, there was a 79% (SD=21%, n=13 subjects from 7 studies) decrease in the frequency and a 28% (SD=67%, n=6 subjects from 3 studies) reduction in the duration of restraint. With respect to studies in which restraint use to prevent self-harm was targeted, there was a 71% (SD=34%, n=4 subjects from 3 studies) reduction in restraint use. Effect sizes were calculable, using non-overlap approaches, for 9 of the 14 studies. The magnitudes of the effect sizes suggest that, on average, the interventions were effective in reducing the use of restraints. The effects generated in studies where restraint use for self-harm was targeted were typically more pronounced than those in which restraint use for agitation and aggression was addressed. There were broad variations, however, in the percentage reductions in restraint use and in the magnitudes of the effect sizes. Although the findings of this review are encouraging, more research is needed, in which greater attention must be paid to rigorous research design, application, and analysis.