Main content area

A 10-min Targeted Geriatric Assessment Predicts Mortality in Fast-Paced Acute Care Settings: A Prospective Cohort Study

Aliberti, Márlon J. R., Covinsky, K. E., Apolinario, D., Lee, S. J., Fortes-Filho, S. Q., Melo, J. A., Viana, S. S. C., Suemoto, C. K., Jacob-Filho, W.
The journal of nutrition, health & aging 2019 v.23 no.3 pp. 286-290
cohort studies, comorbidity, confidence interval, death, elderly, females, graphs, health services, hospitals, income, intravenous injection, laboratory experimentation, models, mortality, nationalities and ethnic groups, patients, prediction, radiology, regression analysis, risk factors, screening, sociodemographic characteristics, therapeutics, Brazil
OBJECTIVES: To estimate whether a 10-minute Targeted Geriatric Assessment (10-TaGA) adds utility to sociodemographic characteristics and comorbidities in predicting one-year mortality in busy acute care settings. We have also compared the performance of 10-TaGA with the Identification of Seniors at Risk (ISAR) scale. DESIGN: Prospective cohort study. SETTING: Geriatric day hospital specializing in acute care in Brazil PARTICIPANTS: 751 older adults aged 79.4 ± 8.4 years (64% female), presenting non-surgical, medical illness requiring hospital-level care (e.g., intravenous therapy, laboratory test, radiology) for ≤ 12 hours. MEASUREMENTS: The 10-TaGA, an easy-to-administer screening tool based on the comprehensive geriatric assessment (CGA), provided a measure of cumulative deficits ranging from 0 (no deficits) to 1 (highest deficit) on admission. Standard risk factors, including sociodemographics (age, gender, ethnicity, income) and the Charlson comorbidity index, were evaluated. The ISAR, a well-validated screening tool, was used for comparison. RESULTS: During one year of follow-up, 130 (17%) participants died. Compared to the ISAR, 10-TaGA offered better accuracy in identifying older patients at risk of death (area under the receiver operating characteristic curve: [AUC] 0.70 vs 0.65; P = 0.03). In a Cox regression model adjusted for sociodemographics and comorbidities, each 0.1 increment in the 10-TaGA score (range 0–1) was associated with increased mortality (hazard ratio = 1.42, 95% confidence interval 1.27–1.59). The addition of 10-TaGA markedly improved the discrimination of the model, which already incorporated standard risk factors (AUC 0.76 vs 0.71; P = 0.005); adding ISAR (AUC 0.73 vs 0.71; P = 0.09) did not have this marked effect. CONCLUSION: The 10-TaGA is an independent predictor of one-year mortality in acute care patients. This multidimensional screening tool offers better accuracy than ISAR when differentiating between older people at low and high risk of death in healthcare settings where providers have limited time and resources.