Frailty degree and illness trajectories in older people towards the end-of-life: a prospective observational study.

OBJECTIVES

To assess the degree of frailty in older people with different advanced diseases and its relationship with end-of-life illness trajectories and survival.

METHODS

Prospective, observational study, including all patients admitted to the Acute Geriatric Unit of the University Hospital of Vic (Spain) during 12 consecutive months (2014-2015), followed for up to 2 years. Participants were identified as end-of-life people (EOLp) using the NECPAL (, palliative care needs) tool and were classified according to their dominant illness trajectory. The Frail-VIG index (, Comprehensive Geriatric Assessment) was used to quantify frailty degree, to calculate the relationship between frailty and mortality (Receiver Operating Characteristic (ROC) curves), and to assess the combined effect of frailty degree and illness trajectories on survival (Cox proportional hazards model). Survival curves were plotted using the Kaplan-Meier estimator with participants classified into four groups (ie, no frailty, mild frailty, moderate frailty and advanced frailty) and were compared using the log-rank test.

RESULTS

Of the 590 persons with a mean (SD) age of 86.4 (5.6) years recruited, 260 (44.1%) were identified as EOLp, distributed into cancer (n=31, 11.9%), organ failure (n=79, 30.4%), dementia (n=86, 33.1%) and multimorbidity (n=64, 24.6%) trajectories. All 260 EOLp had some degree of frailty, mostly advanced frailty (n=184, 70.8%), regardless of the illness trajectory, and 220 (84.6%) died within 2 years. The area under the ROC curve (95% CI) after 2 years of follow-up for EOLp was 0.87 (0.84 to 0.92) with different patterns of survival decline in the different end-of-life trajectories (p<0.0001). Cox regression analyses showed that each additional deficit of the Frail-VIG index increased the risk of death by 61.5%, 30.1%, 29.6% and 12.9% in people with dementia, organ failure, multimorbidity and cancer, respectively (p<0.01 for all the coefficients).

CONCLUSIONS

All older people towards the end-of-life in this study were frail, mostly with advanced frailty. The degree of frailty is related to survival across the different illness trajectories despite the differing survival patterns among trajectories. Frailty indexes may be useful to assess end-of-life older people, regardless of their trajectory.