COGNITIVE CHANGE AMONG NURSING HOME RESIDENTS; COGRISK-NH SCALE DEVELOPMENT TO PREDICT DECLINE

Abstract

Abstract Objectives Chronicle cognitive changes in nursing home residents; develop risk model identifying predictors of decline. Methods Using secondary analysis design with MDS data, cognitive status and change measures were calculated based on Cognitive Performance Scale (CPS). Baseline and quarterly follow-up analyses of US and Canadian interRAI data (n=1,257,832) were completed. Risk model from logistic regression analyses identified predictors of decline. Results Baseline 15% of residents were cognitively intact (CPS = 0); 11.2% borderline intact (CPS=1), 15% mild impairment (CPS = 2). 58.8% of residents fell into more severe CPS categories (3-6). Over time, increased proportion of residents declined – 17.1% at 6 months, 21.6% at 9 months, at 21 months, 34.0%. Over same time, more residents remained stable than declined. Baseline CPS score was strong predictor of decline. CPS categories 0-2 had 3-month decline rates in mid-teens, categories 3-5 had average decline rate of 9%. Two strata risk model construction was employed – one for CPS categories 0-2, second categories 3-5 and both were integrated into 6-category risk scale (CogRisk-NH). Mean decline rates at 3-month assessment ranged from 4.4% to 28.3%. Over time, distinction among risk categories continued – 6.9% to 38.4.% at 6 months, 16.2% to 61.4% at 21 months. Case distribution had 15.9% in category 1, 26.84% category 2, and 36.7% category 3. Three higher risk categories (4-6) represented 20.6% of residents. Conclusion CogRisk-NH scale differentiates among residents and likelihood of decline. Knowledge of risk for cognitive decline enables allocation of resources targeting amenable factors contributing to decline.