Volume 29, Issue 5, 2020
DOI: 10.24205/03276716.2020.1071
Application of Comprehensive Geriatric Assessment Combined with Tilburg Scale-Based Frailty Model in Health Evaluation of Community-Dwelling Elderly Individuals
Abstract
Objective: To apply comprehensive geriatric assessment (CGA) combined with Tilburg scale-based frailty model in the health evaluation of community-dwelling elderly individuals.
Methods: The community-dwelling elderly individuals in our city were enrolled as subjects through cluster sampling and assessed using the Tilburg frailty indicator (TFI), CGA phenotype of frailty (PF) and Ability Assessment for Older Adults. Then receiver operating characteristic (ROC) curves of TFI and PF alone or combination in predicting the disability status of the elderly were plotted by the nonparametric method, and corresponding area under curve (AUC) and 95% confidence interval (95% CI) were calculated. The prediction function of TFI and PF alone or combination for the disability status of the elderly was analyzed.
Results: In this study, 230 elderly individuals were surveyed, meeting the sample size estimation requirements. Among them, there were 110 males (47.8%) and 120 females (52.2%) aged 60-103 years old with (81.4 ± 8.6) years old on average. As to marital status, 47 elderly individuals (20.4%) were married, and 183 elderly individuals (79.6%) were unmarried. In terms of education level, there were 78 cases of illiteracy (33.9%), 45 cases of primary school (19.6%), 107 cases of middle school and above (46.5%). In addition, 110 of them (47.8%) suffered from two or more chronic diseases. The TFI score was 0-13-point(s), with a mean of (3.7 ± 2.7) points, and TFI ≥5 points (frailty) was found in 75 elderly individuals (32.6%, 95% CI: 28.5-37.0%). The PF score was 0-5-point(s), with an average of (2.1 ± 0.71) points, and PF ≥3 points (frailty) was detected in 74 elderly individuals (32.2%, 95% CI: 26.0-35.2%). The incidence rate of frailty screened by TFI and PF showed no statistically significant differences (χ2=0.184, P=0.652). In the ability assessment for older adults, the activity of daily living disability, mental status disability, sensory and communication disability, social involvement disability and comprehensive disability were found in 8 (3.5%), 24 (10.4%), 26 (11.3%), 20 (8.7%) and 21 (9.1%) elderly individuals, respectively. There was a statistically significant difference in AUC of TFI and PF alone or combination in predicting the disability status of the elderly (P<0.01). AUC of TFI in predicting the activity of daily living disability, mental status disability, sensory and communication disability, social involvement disability and comprehensive disability was higher than that of PF, and AUC of TFI in combination with PF in predicting the above items was the highest. The discriminant analysis was conducted with whether there was the activity of daily living disability, mental status disability, sensory and communication disability, social involvement disability and comprehensive disability as dependent variable and TFI and PF scores as independent variables. The cross-validation accuracy of TFI in predicting the activity of daily living disability, mental status disability, sensory and communication disability, social involvement disability and comprehensive disability was higher than that of PF, and the cross-validation accuracy of TFI in combination with PF was the highest
Conclusion: Both TFI and PF can be adopted for screening the frailty of the community-dwelling elderly individuals, but combination has higher value in predicting the disability status.
Keywords
Tilburg scale; comprehensive geriatric assessment; frailty model; disability; community