The application of disability data from epidemiological surveys to the development of indicators of service needs for dependent elderly people

Age and Ageing,  Jan, 1995  by Didier Leibovici,  Sarah Curtis,  Karen Ritchie

• E-mail
• Print
• Link

Introduction

The significant decrease in mortality rates at higher ages which has been observed in European countries over the last decade has led to concern over the increase in morbidity which such demographic changes might engender. Calculations of healthy life expectancy in Europe, Canada and the United States suggest that a significant proportion of the years of life expectancy which have been gained by older people from improvements in social conditions and advances in medical technology are spent in a state of disability and dependency. It has been estimated for example that whereas life expectancy at age 65 in France and the United Kingdom is around 14 years for men, and 19 for women, life expectancy without disability is only 8 years for men and 10 for women [1]. The Royal College of Physicians estimated in 1986 [2] that 60% of disabled persons in western countries were over the age of 60. The provision of services for dependent elderly people is thus clearly emerging as a principal target of public health policy. For both health planners and policy makers reliable indicators of service need which take into account the changing health profile of elderly populations are essential to the efficient provision of aid to the relevant sector of the population.

In France, as in many other European countries, concern has been expressed regarding present resource allocation systems which are often based on inadequate criteria such as service consumption or environmental circumstances rather than empirical observations of need. For example financial aid to adapt the physical environment of disabled persons in France is determined by reference to the physical state of the dwelling without reference to the health status of the individual [3]. Additionally in France, a movement towards decentralization of responsibility for service allocation to dependent elderly people to the French administrative counties or de'partements has led to increasing urgency for the development of methods for obtaining the necessary data to inform the planning process, and reliable indicators which summarize the social and medical services' needs of local populations [4].

On the other hand an abundance of epidemiological and health survey data relating to disability and dependency are presently available for most European countries, and a wide array of assessment instruments exists for the functional assessment of elderly people [5]. However, while such information might prove valuable in describing the health status of populations, it is generally too complex for use by health planning organizations. The development of strategies to summarize health survey data and thus produce simple indicators of need based on empirical observation would provide more relevant criteria for the distribution of formal support than criteria such as current service consumption. It would also help to avoid persistent 'social inequalities in the allocation of aid to elderly people due to inappropriateness of the geographical distribution of existing services [6, 7].

Studies of predictors of mortality in elderly populations in France have demonstrated that indicators of confinement to bed or to the home among elderly people are highly predictive of 3-year survival probabilities [7]. In view of these results, there was interest in whether these variables would also be good predictors of more general levels of disability and dependency. This paper addresses the question of whether level of confinement sufficiently reflects dependency status and living conditions to be useful for planning services for disabled elderly people. The study involves the use of epidemiological survey data relating to sociodemographic features, living conditions, social support and disability of elderly persons to test whether confinement level may constitute an efficient and economic system for classifying levels of population need which may be applied in the first instance to the French de'partements and perhaps more widely to communities outside France.

Subjects

The dependency and living conditions data for the present analysis were obtained from three independent surveys undertaken in Basse Normandie (1984), Languedoc-Rousillon (1986) and Ile de France (1982). The three studies are described in greater detail elsewhere [8]. The results from the three surveys have been combined, giving a total number of 5158 respondents. The survey samples were derived from stratified random samples of people over 65 years drawn from electoral registers. The response rates from the three surveys were 75% for Basse Normandie, 79% for Languedoc-Rousillon and 58% for Ile de France. In cases where the addresses drawn from the electoral register were out of date or the person had died, the individual was replaced from a supplementary sample drawn from the same source. In Languedoc-Rousillon and Basse-Normandie, respondents who refused were also replaced from the supplementary sample. The extent to which the samples reflect living conditions throughout France is unknown, however, the three groups are derived from three very different regions of France (the north-west, the south and the Paris region) and therefore cover a wide variety of health and living conditions.