Distinguishing and predicting end-of-life trajectories in older people (WC2015-054)

Background

Starting date: 01/08/2015

In our ageing population, one of the most striking phenomena is the increasing likelihood for people to live until 90 or even 100 years of age. More and more, mortality has been concentrated in old age. In many countries, 7 in every 10 deaths occur at the age of 70 or over. Death and the last period of life are different for the ‘older-old’ than for the ‘younger-old’. But even within age groups there is heterogeneity in people’s development of functional decline and their need for care in the last years of life. This heterogeneity is probably due to a combination of biological, psychological and social factors. The complex dynamics between age and other factors in the way the last phase of life develops have, however, largely been neglected in old age policy. Yet, closeness to death is a strong determinant of the need for and costs of care. When getting close to death, advance care planning and palliative care have been proven beneficial for forgoing redundant care, saving costs, providing tailored care, and patients’ and relatives’ wellbeing in the last phase of life.

Archetypal trajectories of mortality related functional decline have been described: cancer patients experience rapid predicable decline; organ failure patients (e.g. heart disease, chronic obstructive pulmonary disease) experience a more gradual decline with serious intermittent episodes; and frail/dementia patients experience gradual prolonged decline. In principle, an understanding of trajectories of decline before death could be used to more accurately plan and initiate palliative care.  Our current understanding of end-of-life trajectories has, however, been criticized for not being sufficiently developed for such aims. The trajectories of decline before death were derived by grouping patients first by theoretical trajectory groups based on cause of death. A major problem with this approach is that there is wide variation in individual decline within groups. The appropriateness, therefore, of these archetypal trajectories for guiding patient care has been questioned. A better understanding is required of individual trajectories and the factors that influence them, such as socio-demographic characteristics and health factors like multi-morbidities.

Aims and objectives

This study, therefore aims to distinguish, understand and predict different end-of-life trajectories in older people. We will not predefine trajectory groups by diagnosis/cause of death, but rather attempt to identify individuals following similar trajectories. We will also look beyond physical functioning; not only will physical and cognitive trajectories be distinguished, but also decedents’ care use and associated costs will be mapped in order to identify main trajectories/pathways. Socio-demographic and health factors associated with trajectory groups will also be examined.

This study draws on data from two longitudinal cohort studies (LASA and InterRAI) and data from health insurance claims (Achmea). LASA and Achmea provide fairly representative samples of the Dutch population, whereas InterRAI provides a sample of long-term care residents.