Predictors and scoring system for health-related quality of life in an Indonesian community-dwelling elderly population.

AIM

to determine factors that independently predict health-related quality of life of Indonesian community-dwelling elderly and to obtain scoring system to predict their quality of life.

METHODS

this is a cross-sectional study covering a sample representative of the Indonesian community-dwelling elderly. The study was conducted among 487 elderly in 43 cities in Indonesia between November-December 2010. A structured questionnaire was applied to obtain independent variable such as age, sex, number of diseases, present activities, functional status (ADL score), cognitive status (AMT score), nutritional status (NRI score), and depression status. Quality of life as dependent variable was assessed by Euroqol-5D (Eq-5D). We defined good quality of life as 3 or more dimensions of Eq5D having score of 1. Predictors of quality of life were identified using multiple logistic regression analysis. We constructed a scoring system based on coefficients in multivariable analysis. Calibration performance of the score was evaluated by Hosmer-Lemeshow tests while discrimination performance was evaluated with the c-statistic.

RESULTS

most of the subjects were male (52.2%) and aged between 60-70 years old (75.8%). Prevalence of poor quality of life in this study was 14.9%. Bivariable analysis showed that all independent variables were significantly associated with quality of life. Multiple logistic regression analysis showed that factors predicting poor quality of life were female (OR 1.9; 95% CI 1.0-3.6), poor nutritional status (OR 2.7; 95% CI (1.2-0.8), having >2 chronic diseases (OR 3.6; 95% CI 1.9-6.8), low functional status (OR 4.2; 95% CI 2.2-8.2), and depression (OR 6.3; 95% CI 3.3-12.1). Performance of the risk score revealed Hosmer Lemeshow test p=0.86 and c-statistic 0.87 (95% CI 0.83-0.92).

CONCLUSION

the following factors were identified as predictive factors of HRQoL in Indonesian community-dwelling elderly: sex, nutritional status, number of chronic diseases, functional status, and depression status. The risk score developed based on those predictors showed good performance to accurately predict poor quality of life.