Predicting mortality of residents at admission to nursing home: a longitudinal cohort study

Abstract

Background: An increasing numbers of deaths occur in nursing homes. Knowledge of the course of development over the years in death rates and predictors of mortality is important for officials responsible for organizing care to be able to ensure that staff is knowledgeable in the areas of care needed. The aim of this study was to investigate the time from residents' admission to Icelandic nursing homes to death and the predictive power of demographic variables, health status (health stability, pain, depression and cognitive performance) and functional profile (ADL and social engagement) for 3-year mortality in yearly cohorts from 1996-2006.

Methods: The samples consisted of residents (N = 2206) admitted to nursing homes in Iceland in 1996-2006, who were assessed once at baseline with a Minimum Data Set (MDS) within 90 days of their admittance to the nursing home. The follow-up time for survival of each cohort was 36 months from admission. Based on Kaplan-Meier analysis (log rank test) and non-parametric correlation analyses (Spearman's rho), variables associated with survival time with a p-value < 0.05 were entered into a multivariate Cox regression model.

Results: The median survival time was 31 months, and no significant difference was detected in the mortality rate between cohorts. Age, gender (HR 1.52), place admitted from (HR 1.27), ADL functioning (HR 1.33-1.80), health stability (HR 1.61-16.12) and ability to engage in social activities (HR 1.51-1.65) were significant predictors of mortality. A total of 28.8% of residents died within a year, 43.4% within two years and 53.1% of the residents died within 3 years.

Conclusion: It is noteworthy that despite financial constraints, the mortality rate did not change over the study period. Health stability was a strong predictor of mortality, in addition to ADL performance. Considering these variables is thus valuable when deciding on the type of service an elderly person needs. The mortality rate showed that more than 50% died within 3 years, and almost a third of the residents may have needed palliative care within a year of admission. Considering the short survival time from admission, it seems relevant that staff is trained in providing palliative care as much as restorative care.

Figure 1

Survival curves for the CHESS scale.

Figure 2

Survival curves for residents' ADL scores by quartiles.

Figure 3

Survival curves for the Index of Social Engagement scale.