The minimum data set prevalence of restraint quality indicator: Does it reflect differences in care?

Purpose: This study investigated whether the use of restraining devices and related measures of care quality are different in nursing homes that score in the upper and lower quartiles on the Minimum Data Set (MDS) prevalence of restraint quality indicator, which assesses daily use of restraining devices when residents are out of bed. Design and Methods: The study was a cross-sectional study, with 413 residents in 14 nursing facilities. Eight homes scored in the lower quartile (25th percentile; low prevalence, 0-5%) on the MDS restraint prevalence quality indicator, and six homes scored in the upper quartile (75th percentile; high prevalence, 28-48%). Eight care processes related to the management of restraints and gait and balance problems were defined and operationalized into clinical indicators. Research staff conducted direct observations during three 12-hr days (7 a.m. -7 p.m.) to determine the prevalence of restraining devices and identify resident and staff behaviors that may be affected by restraint use. Results: Residents in high-restraint homes were in bed during the day on more observations than residents in low-restraint homes (44% vs. 33%; p < .001), were more frequently observed with bed rails in use (74% of residents vs. 64% of residents; p < .03), and received less feeding assistance during meals (2.7 min vs. 4.1 min; p < .001). There were no differences between homes in the use of out-of-bed restraints, nor were there any differences on any care process measure related to the management of restraints, gait and balance problems, or measures of physical or social activity. Implications: A home's score on the MDS-generated prevalence of restraint quality indicator was not associated with differences in the use of restraints, physical activity, or any care process measure when residents were out of bed. However, there were differences in the use of in-bed restraining devices, and residents in high-restraint homes were in bed more often during the day. These differences were associated with poor feeding assistance and reflect important differences in quality of care between homes, even though these differences are not what the restraint prevalence quality indicator purports to measure. Methods to monitor and improve the quality of care related to exercise, in-bed times, and resident freedom of movement are discussed.