Risk factors for three-month mortality after discharge in a cohort of non-oncologic hospitalized elderly patients: Results from the REPOSI study.

BACKGROUND: Short-term prognosis, e.g. mortality at three months, has many important implications in planning the overall management of patients, particularly non-oncologic patients in order to avoid futile practices. The aims of this study were: i) to investigate the risk of three-month mortality after discharge from internal medicine and geriatric wards of non-oncologic patients with at least one of the following conditions: permanent bedridden status during the hospital stay; severely reduced kidney function; hypoalbuminemia; hospital admissions in the previous six months; severe dementia; ii) to establish the absolute risk difference of three-month mortality of bedridden compared to non-bedridden patients.

METHODS: This prospective cohort study was run in 102 Italian internal medicine and geriatric hospital wards. The sample included all patients with three-months follow-up data. Bedridden condition was defined as the inability to walk or stand upright during the whole hospital stay. The following parameters were also recorded: estimated GFR≤29mL/min/1.73m2 ; severe dementia; albuminemia ≪2.5g/dL; hospital admissions in the six months before the index admission.

RESULTS: Of 3915 patients eligible for the analysis, three-month follow-up were available for 2058, who were included in the study. Bedridden patients were 112 and the absolute risk difference of mortality at three months was 0.13 (CI 95% 0.08-0.19, p≪0.0001). Logistic regression analysis also adjusted for age, sex, number of drugs and comorbidity index found that bedridden condition (OR 2.10, CI 95% 1.12-3.94), severely reduced kidney function (OR 2.27, CI 95% 1.22-4.21), hospital admission in the previous six months (OR 1.96, CI 95% 1.22-3.14), severe dementia (with total or severe physical dependence) (OR 4.16, CI 95% 2.39-7.25) and hypoalbuminemia (OR 2.47, CI 95% 1.12-5.44) were significantly associated with higher risk of three-month mortality.

CONCLUSIONS: Bedridden status, severely reduced kidney function, recent hospital admissions, severe dementia and hypoalbuminemia were associated with higher risk of three-month mortality in non-oncologic patients after discharge from internal medicine and geriatric hospital wards.