Effect of a guideline-based multicomponent intervention on use of physical restraints in nursing homes: a randomized controlled trial

Abstract

Context: Despite unambiguous legal regulation and evidence for lack of effectiveness and safety, physical restraints are still frequently administered in nursing homes.

Objective: To reduce physical restraint prevalence in nursing homes using a guideline- and theory-based multicomponent intervention.

Design, setting, and participants: Cluster randomized controlled trial of 6 months' duration conducted in 2 German cities between February 2009 and April 2010. Nursing homes were eligible if they had 20% or more residents with physical restraints. Using external concealed randomization, 18 nursing home clusters were included in the intervention group (2283 residents) and 18 in the control group (2166 residents).

Intervention: The intervention was based on a specifically developed evidence-based guideline and applied the theory of planned behavior. Components were group sessions for all nursing staff; additional training for nominated key nurses; and supportive material for nurses, residents, relatives, and legal guardians. Control group clusters received standard information.

Main outcomes measures: Primary outcome was percentage of residents with physical restraints (bilateral bed rails, belts, fixed tables, and other measures limiting free body movement) at 6 months, assessed through direct unannounced observation by blinded investigators on 3 occasions during 1 day. Secondary outcomes included restraint use at 3 months, falls, fall-related fractures, and psychotropic medication prescriptions.

Results: All nursing homes completed the study and all residents were included in the analysis. At baseline, 30.6% of control group residents had physical restraints vs 31.5% of intervention group residents. At 6 months, rates were 29.1% vs 22.6%, respectively, a difference of 6.5% (95% CI, 0.6% to 12.4%; cluster-adjusted odds ratio, 0.71; 95% CI, 0.52 to 0.97; P = .03). All physical restraint measures were used less frequently in the intervention group. Rates were stable from 3 to 6 months. There were no statistically significant differences in falls, fall-related fractures, and psychotropic medication prescriptions.

Conclusion: A guideline- and theory-based multicomponent intervention compared with standard information reduced physical restraint use in nursing homes.

Trial registration: isrctn.org Identifier: ISRCTN34974819.