Frailty Related Survival Risks at Short and Middle Term of Older Adults Admitted to Hospital

Abstract

In the context of healthy aging research, frailty is characterized by a progressive decline in the physiological functions of multiple body systems that lead to a more vulnerable condition for the development of various adverse events, such as falls, hospitalization and mortality. In this paper, we assess the impact of frailty on survival of older adults at three time horizonts: 1 month, 6 months, and 1 year, thus covering short and middle term survival. We consider frail and pre-frail patients recruited at hospital admission, according to the Fried phenotype. Cognitive status, comorbidities and pharmacology assessment was carried out at hospital admission and follow up survival was extracted from electronic health record (EHR). We compute Kaplan Meier estimates and log-rank test of survival probability functions at the selected censoring times for frail versus pre-frail patients. Additionally, we compute Cox regression for these censoring times over patient variables selected by independent sample t-test and logistic regression. We confirm that frailty predicts greater mortality at all censoring times. Significant hazard risks (HRs) identified in frail cohort were sex, age, weight, the occurrence of hospital readmission 30 days after discharge, sit and stand test scores, and the use of proton pump inhibitors, antiplatelet, quetiapine and paracetamol.