University of Tasmania
Whole-Lawrence-thesis.pdf (3.29 MB)

The trajectory of decline : a quantitative study to identify variations in the longitudinal functional profile of an Australian nursing home population

Download (3.29 MB)
posted on 2023-05-26, 01:21 authored by Lawrence, SJ
Population ageing is a public health success story. However, the perceived pressures that the arrival of the 'baby boomer' generation will demand of the health system, coupled with the financial implications of a shrinking tax base, are cause for concern. By 2056 it is estimated that a quarter of Australians will be 65 years and older, double the current proportion. Concurrently demand for residential aged care services will also increase. In Australia, aged care services are funded by the Australian Government via subsidies paid to service providers, based on the assessed care level required by each eligible client. The level of this subsidy is based on a validated measure of the resident's care needs that is assessed on admission to an aged care facility and then, at a minimum, annually thereafter. Residential aged care is the principal consumer of the aged care budget and it is predicted that costs will increase with the increasing proportion of aged in the population. Approximately 5% of Australians 65 years and older currently live in residential aged care, and the annual recurrent cost of their care is $6.6 billion. Review of resident care levels has shown that over the past decade the acuity of residents entering aged care facilities has increased while the length of stay of a newly admitted resident has decreased. The increase in resident separations via death will exert further pressure on care services already struggling with staff shortages and increased costs. Such changes highlight an imperative to develop models that will predict the cost basis of care for residents in the increasingly rapid resident turnover environment. Prediction of care needs and associated costs for an increasing number of elderly can be made based on population change estimates, but this is a coarse measure at best. Models that predict in advance the clinical and resource needs of the elderly described in the literature can potentially be used by policy makers and service providers to bring some certainty in planning future care and resource demands. One of these is the 'trajectory of decline' model which was articulated by Joanne Lynn in the 1990's. This model, empirically tested in an elderly USA community population in collaboration with June Lunney and colleagues in 2002, revealed five trajectories grouped by their cause of death. The model has been subsequently utilised in a number of policy documents in the UK, USA, EU and Australia and has also been used as a clinical tool for planning end-of-life care in the elderly population. The model predicts that as an elderly person progresses along their illness trajectory the resources required to support them will increase or decrease with the level of care needed. However, the trajectory of decline concept has not been tested in an elderly nursing home population, whose principal mode of separation is death. This thesis explores if the trajectory of decline, as proposed by Joanne Lynn and empirically tested by June Lunney and colleagues is existent in a nursing home population in Australia; and if there are comparable trajectories to those identified in their research. To replicate the existing empirical studies of this concept in a different population, two critical measures were required: (1) the cause of death to determine trajectory group membership; and (2) a measure of function in a nursing home population. In this study data were collected from the nursing home records of 247 deceased residents utilising the validated Resident Classification Scale (RCS) as the primary measure of function. This provided a total of 990 RCS scores available for analysis. Cause of death data were obtained from the Tasmanian Registrar of Births, Deaths and Marriages. The degree to which the elderly residents of nursing homes fit the trajectory profiles proposed by Lynn and empirically tested by Lunney and colleagues was examined with a novel statistical analysis using multivariate statistics which allowed the trajectory of each individual to contribute to the analysis. Using this statistic, a predictive model was developed to examine the effect of the variables on the measure of function, the RCS, over the whole admission period as well as the 12 months prior to death. These results inform the discussion examining the extent to which the trajectory of decline can prospectively predict the decline in functional ability of this elderly population. The principal findings of this research are that there are four trajectories of functional decline to death in this nursing home population when grouped by their cause of death. More than half the subjects comprised the frailty group with the remainder distributed in similar proportions as found by Lunney and colleagues in the cancer, heart and lung failure and 'other' group which is poorly specified. Using routinely collected data, the functional trajectory of each group can be identified from the point of admission to the nursing home to death and are generally comparable to the trajectories for each group described in the literature. However, in relation to the functional profile, the Frailty Group is statistically different to the other three groups, but the other three groups are not different to each other. A further finding is that there is poor agreement between what is written on the death certificate and the routine diagnosis made while the resident is still alive which reduces the utility of this concept for the purposes of prospective care planning. Repeating this study prospectively using a standardised medical admission will determine whether the four groups exist in this population or whether two groups, as suggested by this study, would be a better fit.


Publication status

  • Unpublished

Rights statement

Copyright 2013 the author

Repository Status

  • Open

Usage metrics

    Thesis collection


    No categories selected


    Ref. manager