Prognostic value of estimated glomerular filtration rate in hospitalized elderly patients

A multicenter observational study, REPOSI (REgistro POliterapie Societa Italiana di Medicina Interna), was conducted to assess the prognostic value of glomerular filtration rate (eGFR) on in-hospital mortality, hospital readmission and death within 3 months, in a sample of elderly patients (n = 1,363) admitted to 66 internal medicine and geriatric wards. Based on eGFR, calculated by the new Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula, subjects at hospital admission were classified into three groups: group 1 with normal eGFR (>= 60 ml/min/1.73 m(2), reference group), group 2 with moderately reduced eGFR (30-59 ml/min/1.73 m(2)) and group 3 with severely reduced eGFR (<30 ml/min/1.73 m(2)). Patients with the lowest eGFR (group 3) on admission were more likely to be older, to have a greater cognitive and functional impairment and a high rate of comorbidities. Multivariable logistic regression analysis showed that severely reduced eGFR at the time of admission was associated with in-hospital mortality (OR 3.00; 95 % CI 1.20-7.39, p = 0.0230), but not with re-hospitalization (OR 0.97; 95 % CI 0.54-1.76, p = 0.9156) or mortality at 3 months after discharge (OR 1.93; 95 % CI 0.92-4.04, p = 0.1582). On the contrary, an increased risk (OR 2.60; 95 % CI 1.13-5.98, p = 0.0813) to die within 3 months after discharge was associated with decreased eGFR measured at the time of discharge. Our study demonstrates that severely reduced eGFRs in elderly patients admitted to hospital are strong predictors of the risk of dying during hospitalization, and that this measurement at the time of discharge helps to predict early death after hospitalization.