[Predictive factors of hospital and 6-month morbidity and mortality in hospitalized elderly patients]

This article presents the results of a prospective multivariable study of elderly patients aged over 70 years, hospitalized in an Internal Medicine Department of a Central Lisbon Hospital. The study aimed to identify, at the beginning of hospital admission (HA), predictive factors of hospital mortality (HM) and mortality at 6 months, of duration of HA, of admission to a nursing home at the time of discharge and during a period of 6 months thereafter and of hospital readmission during the 6 months following discharge. The study included 158 patients with a mean hospital stay of 15 days and a hospital mortality of 12%. The main pathologies responsible for hospital admission were cerebrovascular accident (22%), heart failure (20%) and pneumonia (16%). Mortality at 6 months was 29% and hospital readmission in the 6 months thereafter was 24%. When the patient was cared for by the spouse there was a statistically significant correlation with a shorter duration of admission (p = 0.006). Mean hospital stay was not significantly associated with any other variable. A subjective medical evaluation (SME) at the start of HA (p = 0.001), a low Barthel score prior to and at the time of HA, low serum albumin (p = 0.001) and a high leucocyte count (p = 0.005) were correlated with a higher HM. Nursing home admission was only positively correlated with cerebrovascular pathology. Mortality at 6 months was significantly correlated with the SME (p = 0.001), a low Barthel score prior to admission (p < 0.008) and at the time of HA (p < 0.001), nursing home residency (p < 0.005) and a low mental test score (p < 0.01). Hospital readmission at 6 months was influenced by the SME (p < 0.04) and by the reduction in the Barthel score caused by the illness and HA (p = 0.004). These correlations enabled the development of mathematical models that predict HM and mortality at 6 months and admission to a nursing home at the time of discharge and during a period of 6 months thereafter. They could be important in identifying elderly patients' needs early in the hospital admission and in the improvement of the strategy necessary for a successful and dignified hospital discharge.