Functional status in older people is a dynamic situation, which makes it necessary to evaluate functional capacity at different times to determinate their prognostic value.
To examine the association between functional status (baseline and change after acute illness) and mortality and functional changes at 1 year.
Hospital-based prospective longitudinal cohort study of all patients over 65 years old, admitted for multidisciplinary treatment of functional impairment after acute illness in a medium-stay unit (post-acute geriatric unit) of a teaching hospital ascribed to the Spanish National Health Service from Spain during 15 consecutive months. Functional status (Barthel Index, BI) was assessed prior to the acute illness, at admission in a post-acute unit, at discharge and 1 year later. At admission, other variables were collected: sociodemographic, main diagnostic for hospitalization (stroke, orthopedic process, or deconditioning), serum albumin, comorbidity (Charlson Index), cognitive status (Pfeiffer s Short Portable Mental Status Questionnaire). In order to analyze mortality 1 year after discharge, a Cox regression analysis was performed.
Three hundred and sixty-nine patients constituted the study population, mean age was 80.74 years (SD 7.4), 66.6% were female and 1 year after discharge mortality was 20%. In the multivariate analysis, variables associated with a higher 1-year mortality were age (HR 1.06; 95% CI = 1.00-1.07) male gender (HR 2.11; 95% CI = 1.26-3.55), worse prior functional status (HR 0.98; 95% CI = 0.96-0.99), and higher functional loss in BI at admission (HR 1.02; 95% CI = 1.00-1.04). On the other hand, a greater functional gain in BI at discharge was associated with a lower 1-year mortality (HR 0.98; 95% CI = 0.96-0.99).
The main functional gain obtained after treatment in a multidisciplinary post-acute geriatric unit is independently associated with a reduction in long-term mortality. In addition to baseline functional status and after acute illness, the subsequent potential recovery is very important to predict poor long-term outcomes.