Population ageing in Scotland - implications for healthcare expenditure

POPULATION AGEING IN SCOTLAND - IMPLICATIONS FOR HEALTHCARE EXPENDITURE Population ageing is a major concern for developed countries in terms of public expenditure required to pay for health care (HC). The broad aim of this thesis is to contribute to and expand the debate on the independent effects that population ageing and the time immediately before death (TTD) have on HC expenditure in Scotland. This study analyses, for the first time in Scotland, how HC expenditure projections are influenced through the application of two approaches; the first only accounting for an increasing proportion of the elderly population, and the second also implementing a TTD component. Several issues that are under-researched or have not been addressed in TTD studies previously, are explored and alternative approaches are presented. Utilising two large linked datasets this thesis addresses important methodological issues. Alternative methods to cost inpatient hospital stays are examined as this has pivotal implications for any analysis undertaken to estimate the independent effect of TTD and age on HC expenditure. Explanatory variables that have previously not been considered, such as health risk and health status measures at baseline, are included in these analyses. The issue of sample selection, arising through the inclusion/exclusion of survivors in a TTD study is investigated and the impact of individuals’ socio-economic status on costs is examined. The analysis of alternative costing methods clearly showed that any inference that can be made from econometric modelling of costs, where the marginal effect of explanatory variables is assessed, is substantially influenced by the chosen costing method. The application of a Healthcare Resource Group (HRG) costing method was recommended. This study found that TTD, age and the interactions between these two factors were significant predictors for HC expenditure. The analysis further identified some of the health status and health risk measures to be important predictors of future HC expenditure. An examination of how sample selection impacts on estimated costs at the end of life showed that if survivors were excluded from the analysis, costs might be overestimated. Drawing on a representative sample of the Scottish population, the investigation of the association that the socio-economic status had with HC costs suggested that less is spent on individuals from more deprived areas. This might partly be explained through the decreased probability of accessing hospital services for individuals from more deprived areas. Furthermore, results showed that projected HC expenditure for acute inpatient care for the year 2028 was overestimated by ~7% when an approach that only accounts for the higher proportion of elderly people in a population in the future is being used as compared to an approach that also accounts for the effect that remaining TTD has on costs