Recent changes in stroke history, mobility status and life expectancy at admission among nursing home residents in Japan

Abstract

We investigated changes in the condition of elderly people at admission to nursing homes, their prognosis after admission and recent changes of mortality risk factors among the residents. We then examine the possible relationship between the changes identified and discuss the possible effects of changes in social policy. A retrospective cohort study using survival analysis was conducted among 407 residents of three nursing homes in Toyama, Japan who were admitted between 1976 and 1990. By reviewing medical and care notes, we abstracted the admission date, gender, age at admission, stroke history, physical activity and date of discharge or death for each subject. Positive stroke history was defined as medical certification of thromboembolic or haemorrhagic stroke before admission. The level of physical activity at admission was classified as either mobile or immobile, according to whether it was possible to get out of bed without assistance from another person. Data were compared for subjects admitted during the two periods of 1976–82 and 1983–90. The latter period includes the year, 1985, when the Medical Service Law in Japan was amended for the purpose of effective allocation of medical resources. In comparison with that for the first admission period, the mean age at admission for the second period was significantly higher in both males and females. The age- and gender-adjusted odds ratio of stroke history to immobile status decreased from 4.7 for the first admission period to 2.2 for the second. The 5-year survival rate decreased, from 51 to 30% (P≤ 0.0001). In multivariate analysis, the relative hazard (RH) of mortality of the subjects in the later admission period to those in the earlier period was 2.0 [95% confidence interval (CI): 1.4–2.8], while the RH was not significant for either gender or age. Immobile status without stroke history showed the highest RH. The RH of the interaction of mobile status with stroke history and late admission period was significantly lower than 1.0. The study indicates that: the relative importance of stroke history in immobile status has decreased among the nursing home population, stroke history among the mobile subjects has become less predictive of poor prognosis after admission, the life expectancy after admission has declined, and the increased mortality is related to factors other than physical activity status and may be attributable to the recent reorganization of medical resources.