OBJECTIVES. Various individual characteristics increase a nursing home
resident's likelihood of being mechanically restrained. However, ther
e is not good information on the role played by facility characteristi
cs and geographic area in determining the likelihood of physical restr
aint use. This study investigates the importance of individual, facili
ty, and area indicators in determining the likelihood of restraint use
. METHODS. This research uses data on 2,000 nursing home residents rec
eiving care in more than 250 nursing homes in 10 different areas. Resi
dent-level data were collected in 1990 by research nurses using the Mi
nimum Data Set for Nursing Home Resident Assessment and Care Screening
. Facility-level data were gathered by telephone interviews with facil
ity staff. The relations among the variables were investigated using c
ontingency tables and multivariate logistic regression models. RESULTS
. Controlling for residents' physical and cognitive function, facility
characteristics and location had a significant impact on restraint pr
actices. Those facilities with low nurse staffing were more likely to
restrain residents. Furthermore, controlling for these factors, restra
int practices varied significantly across the geographic areas in the
study. CONCLUSIONS. Like hospitals, nursing homes exhibit significant
interfacility and regional variation in their clinical practices. Futu
re analyses of nursing home care might benefit from considering both f
actors in much finer detail.