journal articles
COMORBIDITY MEASURES AND MORTALITY IN INSTITUTIONALISED ELDERLY
H.Y. Neo, M. Chan, T.L. Tan
J Aging Res Clin Practice 2013;2(1):110-116
Background: Objectives: The utility of comorbidity indices for the assessment of frail institutionalised elderly have not been fully explored. Such information may prove useful for individualized advanced care planning and healthcare resource allocation.We aimed to compare the predictive properties of four indices (CIRS-G total score, CIRS index, Charlson Score and Charlson age-adjusted) in the setting of a multi-racial Asian long term care facility. Design and Setting: We conducted a cross-sectional study with prospective collection of mortality data for 158 patients (mean age 76.6±12.3 years) at a nursing home in Singapore. Measurements: A multi-disciplinary team evaluated baseline demographics, disease number, medication burden, Mini Nutritional Assessment (MNA) score and modified Barthel Index (MBI). Correlations with baseline measures, univariate and multivariate regression analyses were performed to determine the impact of comorbidity indices on 2-year mortality. Results: Baseline correlations were significant but modest between the 4 indices and medication burden, MBI and MNA (Pearson's R range: 0.23-0.31, all p<0.05).Two year all-cause mortality was 25.8% (n=41). Upon univariate analyses, mortality was significantly associated with MBI (OR 0.99, P=0.016), MNA (OR 0.87, P=0.006), number of diseases (OR 1.21; P=0.048), CIRS-G total score (OR 1.14; P=0.010) and age-adjusted Charlson Score (OR 1.25; P=0.032). After accounting for age, gender, race, MBI and MNA, only CIRS-G total score significantly predicts mortality in the multivariate analysis (OR 1.14; P=0.02). Conclusion: Beyond its association with baseline demographics, nutritional and functional measures, the CIRS-G total score remained a significant predictor of mortality compared to indices derived from the Charlson Score. Inclusion of said comorbidity variable will be useful when interpreting mortality data of institutionalized elderly and planning of residential care resource allocation.