Differences in predictors of 5-year survival over a 10-year period in two cohorts of elderly nursing home residents in Sweden

Aim: The aim was to compare 5-year survival in two included cohorts (from year 2000 and year 2007) of 249 nursing home residents (NHR) in this retrospective, comparative study. Methods: The cohorts were compared regarding chronic diseases, medication, physical/cognitive/nutritional status, body mass index, body temperature and 5-year mortality. Factors correlated with 5-year survival were determined using Cox regression analysis. Results: In average, cohort 2007 survived 31 +/- 16 months and cohort 2000, 38 +/- 13 months, p < 0.001. Dementia, ageing and circulatory failure were more common as cause of death 2007, while stroke, chronic obstructive pulmonary disease (COPD) and pneumonia were less common, compared with 2000. NHR belonging to cohort 2007 were significantly older when admitted to nursing homes (NH), more dependent in activities of daily living (ADL), had dementia, stroke, autoimmune disease and treatment with antidepressants, while malnutrition and treatment with paracetamol were more common 2000. In 2000, medication with antidepressants, the presence of stroke and diabetes, irrespective of gender, and in women cardiovascular disease, two to threefold significantly increased survival, while autoimmune disease, influenza vaccination and dependency in ADL decreased survival. In 2007, maintaining BMI, irrespective of gender, and autoimmune disease and COPD in women significantly increased survival, while malnutrition, influenza vaccination, dependency in ADL and medication with sedatives/tranquillisers or paracetamol severely reduced survival. Conclusions: The present results indicate a trend that individuals are older and frailer when admitted to NH and that survival time after admission has been shortened. Hence, the need of daily support and care has increased, irrespective of housing. Also, predictors of survival, possible to influence, have changed.