Search | Search by Center | Search by Source | Keywords in Title
Sullivan JL, Shwartz M, Burgess JF. Examining the Relationship between a Minimum Data Set Composite Quality Indicator and the Artifacts of Culture Change Tool. Poster session presented at: Gerontological Society of America Annual Scientific Meeting; 2011 Nov 20; Boston, MA.
Objective: To examine the association between patient-centered care (PCC) practices and a composite measure of quality in Department of Veterans Affairs (VA) Community Living Centers (CLCs) which have undergone culture change activities to make care more person-centered. Study Design: We consider 107 VA CLCs that had at least 10 long-stay residents and in which at least one-third of the residents were long stay. We use a Bayesian hierarchical model to estimate a latent construct 'PCC' from the 6 sub-domains of the Artifacts of Culture Change Tool and, after controlling for case mix using RUG-III, to analyze the relationship of the latent construct to a composite measure developed from 24 adverse event quality indicators calculated from the Minimum Data Set (MDS). Analysis: We use a Bayesian hierarchical model to estimate the latent construct "culture change" as represented by the Artifacts of CC Tool and to analyze the relationship of the latent construct to the composite measure developed from the individual MDS QIs. We also adjust for case mix through RUGs scores. Principal Findings: There is a great deal of variability across facilities in the Artifacts Total Score and the MDS-based Quality Score: the high score is over 2.5 times the low score. In the Bayesian hierarchical model, the coefficient linking the Artifacts Latent Score to quality (which is also the correlation between the two variables) is -0.18 (95% credible interval: -0.38 to 0.01). Though the interval covers 0, there is a 97% chance the relationship is negative. The coefficient indicates that facilities with a one standard deviation higher Artifacts Latent Score have 13.2 fewer adverse events per 1,000 residents per year. Conclusions: Our results show that there is a relationship between PCC activities as measured by the Artifacts Tool and the MDS-based QIs. Though the strength of the relationship is relatively modest, the correlation was higher than most of the correlations among commonly-used hospital performance measures (Shwartz et al., 2011). From our cross-sectional analysis, there is no basis for drawing any conclusions about causality. Further research needs to test if variations in CLC care processes and policies may be driving the facilities doing better on the Artifacts Tool and the MDS measures compared to those that do poorly.