11.8 Health of elderly people
Oxford Textbook of Public Health
Health of elderly people
Demographics and projections
Determinants of population ageing
Demographic and epidemiological transitions
Compression of morbidity
Ethnic minority elders
Older people as a special group
Icebergs of disease
Loss of adaptability
Costs of care
Health and social services for elderly people
Role of public health
Ageism and rationing
Community and social care
Health services for mentally ill people
Social support and informal carers
Health and social policy implications of population ageing
In this chapter, population ageing and its determinants are described. The case for considering elderly people as a special needs group is discussed, together with the health problems they face. The available evidence of efficacy of preventive interventions, primary care and hospital services, long-term care and social support are reviewed.
Elderly people are a heterogeneous group comprising those who are still actively pursuing careers in industry or politics, enjoying a healthy retirement, caring for frail and dependent relatives or friends, and others who are themselves very frail and dependent. Such diverse people are frequently grouped together as ‘the elderly’ as if they had similar needs and would respond in the same way to interventions. It is essential to recognize this heterogeneity in defining need, assessing the effects and relevance of intervention, and planning for the future. It is common to define the ‘young old’ as aged 60 to 69 years, the ‘old old’ as aged 70 to 79 years, and the ‘oldest old’ as 80 years and over. In the past, United Nations and other international agencies used 60+ years as a criterion of ‘the elderly’, but more recently there is a growing consensus that 65+ years should be used.
Demographics and projections
The world population is around 6 billion (that is, 6 × 109) people of whom 385 million (6.5 per cent) were aged 65 years and over in 1997. The largest numbers of elderly people are found in the most populous countries of the world—China, India, the United States, Russia, and Japan—rather than in the more affluent postindustrialized countries.
Focusing on a single age threshold hides important trends in the numbers of the oldest old who tend to present major challenges for health and social services. Figure 1 and Figure 2 show the numbers of people aged 65 years and over and 80 years and over, respectively, in the most aged countries in 1997 and projected to 2025. In developing countries the annual growth of people aged 65+ years is almost 3 per cent per year, whereas in industrialized countries it has fallen to below 2 per cent per year (Kinsella 1996). In industrialized countries the proportion of the oldest old is growing more rapidly than any other age group.
Fig. 1 Countries with the highest number of people aged 65+ years, 1997 and 2025. (Source: WHO 1998.)
Fig. 2 Countries with the highest number of people aged 80+ years, 1997 and 2025. (Source: WHO 1998.)
This growth in the numbers of elderly people is mirrored throughout most countries of the world but to different extents. The world can be divided into four main regional types. In type 1 countries of Europe, North America, Japan, Eastern Europe, and the former Soviet Union, the proportion of elderly people is high and will rise. Type 2 countries—China, Southeast Asia, Latin America, and the Caribbean—have a relatively small proportion of elderly people but rapid increases are predicted. Type 3 countries comprise India and the Middle East where the increase in the proportion of elderly people will be slower. The type 4 countries of sub-Saharan Africa are not expected to show any increase in the proportion of elderly people (Fig. 3).
Fig. 3 Oldest old (80+ years) as a percentage of all people aged 65+ in industrialized and developing countries. (Source: Kinsella 1996.)
Population ageing is an unprecedented phenomenon. It took over 100 years for Belgium to double the proportion of its 60+ population from 9 to 18 per cent. China will take 34 years and Singapore only 20 years to achieve the same population ageing. Projections suggest that the net world monthly gain in people aged 65 years and over by 2010 will be 1.1 million every month (Kinsella 1996), from a current level of about 800 000 every month (Fig. 4).
Fig. 4 The proportion of people aged over 60 years in similar types of countries estimated for the years 1990 and 2020. (Source: Dr M. Prince and Alzheimer’s Disease International.)
Projections of ageing of populations are important for planning services but are subject to error. Projections are particularly sensitive to assumptions about mortality rates at older ages and have always tended to underestimate the true numbers of older people. United Nations demographers give both high and low estimates for their projections.
Determinants of population ageing
The major determinants of population ageing are declines in fertility and mortality. It is tempting to attribute the declines to improvements in medical care. In industrialized countries, such as the United Kingdom, falls in infant and childhood mortality began in the nineteenth century, predating the era of modern effective medical care by many decades. However, it was not until the 1950s that declines in mortality among those aged 75 years and over were achieved (Charlton and Murphy 1997). During the 1970s the rate of decline in mortality increased at all ages. The trends over the nineteenth and early twentieth centuries in the United Kingdom were probably due to general improvements in social and economic conditions resulting in better nutrition, the public health reforms of sanitation, clean water, and control of common communicable diseases. The more recent mortality declines may be attributable to medical care (Bunker 1994), although the evidence is not strong. However, even in the recent past, mortality declines have only been experienced by the better off in the United Kingdom, demonstrating the continuing influence of social and economic factors (Drever and Whitehead 1997). In developing countries the successful infant and child health programmes (growth chart monitoring, oral rehydration for diarrhoea, breast feeding, and immunization for common infections—the GOBI programme) are probably responsible for the very rapid rates of mortality decline at younger ages. Studies of survival of animals in the wild and in captivity show a striking similarity to the survival curves of the last century and the present. Animals in the wild seldom achieve their maximum lifespan because of predation and other natural hazards. In captivity they are capable of reaching much older ages because the major risks of predation, starvation, and accidents have been largely eliminated (Kirkwood 1999). Much the same phenomenon may explain the so-called ‘rectangularization’ of survival in humans. Modern society is protective to its weakest members, has organized means of distributing food, and is arguably safer than in the past (Fig. 5).
Fig. 5 Declines in mortality at different ages for women, United Kingdom, 1841–1994. (Source: Charlton and Murphy 1997; Office for National Statistics 1998.)
Declines in fertility are attributed to urbanization, industrialization, and social mobility which have all had an influence on family structures. In addition the emancipation of women, their greater chances for education, later age at marriage, and access to contraception have all played a part. In developing countries, it seems likely that education of women and access to contraception explain the very rapid rates of population ageing in countries such as China and those of Southeast Asia. Japan and the United Kingdom are among 61 countries that are not having enough babies to replace their populations (Anonymous 1999). Replacement fertility rates are 2.1 babies per women; the figures for the United Kingdom and Japan are 1.7 and 1.4 respectively, and Spanish women have the lowest fertility in the world at 1.15. Such trends in fertility are not confined to industrialized countries. Women in India are having fewer babies than American women had in the 1950s and China, Cuba, and Thailand are already below the replacement rate. These very sharp declines in fertility worldwide have resulted in reductions in forecasts for population growth. United Nations estimates suggested a total population of 9 billion people by 2050 but it might be as low as 7.5 billion.
Demographic and epidemiological transitions
With a changing population structure came differing disease patterns, caused at least in part by the same factors that contributed to declines in mortality and fertility. In populations with high fertility, infections affecting those most susceptible, infants and children, were very common. As environmental conditions contributing to risk of infections improved, and as nutritional patterns changed from periods of starvation to a more plentiful food supply throughout the year, so were the seeds of chronic and degenerative diseases sown. These epidemiological transitions have been defined as the age of pestilence and famine, the age of receding pandemics, and the age of degenerative and man-made diseases. However, in the rapidly ageing developing countries the picture is not so clear-cut as in industrialized countries. Developing countries are experiencing the double burdens of both high levels of infectious diseases taking their toll on infants and children, and also emerging burdens of smoking- and diet-related diseases, in particular cardiovascular diseases, cancers, and diabetes mellitus.
William Farr produced the first English life table using mortality data for 1841. At this time, life tables were valued by friendly societies who used then to calculate the survival prospects of people at different ages to determine how much they should pay (Charlton and Murphy 1997). The attraction of the life table, then as now, is that it provides a simple and easily understood summary of the mortality experience of a population. In 1841, life expectancy at birth was 42 years for women and 40 years for men, reflecting the prevailing high rates of infant mortality. William Farr calculated average life expectancies in 1841 for Liverpool, Surrey, and London which were 26, 45, and 37. By 1992, while the rank order by place of residence was much the same, life expectancy above the age of 85 years showed even greater variation in 1992 than it did in 1841! Gains in life expectancy have been most pronounced at younger ages, but even at older ages increases have occurred due to falling death rates. Table 1 shows life expectancy at different ages for men and women in England and Wales.
Table 1 Life expectancy at different ages: England and Wales, 1841–1991
Information prior to the middle of the nineteenth century is difficult to find and interpret but supports a view that falls in infant and childhood mortality were the main driving forces in increased population growth during the period 1650 to 1850 (Gonner 1913), rather than increases in fertility (Wrigley and Schofield 1981). Comparisons between infant and child mortality among the British peerage and parishes (i.e. representative of the general population) in England show striking declines among both groups between 1650 and 1849; infant and childhood mortality showed a decline of about two-thirds among offspring of peers and about a third in parishes. Moreover, infant mortality among wealthy and poorer people showed little difference during the seventeenth and eighteenth centuries, suggesting that improvements in life expectancy were not driven by wealth but by reduced exposure to infectious diseases (Razzell 1998).
Life expectancy in England and Wales shows a marked social class gradient at birth and also at age 65 years. The difference in life expectancy at age 65 years between social classes I and II (professional and managerial) and IV and V (semi-skilled and unskilled manual) has increased from 1.4 years in 1972 to 2.6 years in 1987–1991 as shown in the Table 2. Similar, but less marked, gradients exist for women.
Table 2 Male life expectancy at birth and at age 65 years by social class, England and Wales, 1972–1991
Life expectancy in the former USSR and Eastern European countries show stark declines among men since the political reforms (Leon et al. 1997; Chenet et al. 1998). The explanation for such trends is not clear, but sudden deaths due to binge alcohol drinking, increased risks of violent death, and suicide all play a part (Chenet et al. 1996). Reunification of the former East Germany with West Germany shows a somewhat different picture. In the decade since reunification, starting with the collapse of the Berlin Wall in 1989, mortality rates in the east have fallen towards those in the west among women but not in men (Haussler et al. 1995). Differences in life expectancy at birth between East and West Germany, which favoured East Germany at the start of division, have fallen since reunification (Table 3). These natural experiments in social and economic engineering provide examples of how public health is influenced by political change. It also reinforces the importance of social and economic factors as major forces in determining life expectancy.
Table 3 Life expectancy in East and West Germany, 1949–1997
Life expectancy can be calculated using either period death rates—that is, the age-sex specific death rates for a particular year—or cohort death rates. The latter provide better estimates of true life expectancy as they adjust for the fact that people in successive birth cohorts are becoming healthier and—age for age—have a lower risk of mortality than their predecessors. Period death rates make an unwarranted assumption that current death rates will prevail unchanged. In 1990, cohort life expectancy at birth was 77 years and 83 years for men and women respectively in the United Kingdom, about a 5 per cent increase on period life expectancy estimates (Charlton and Murphy 1997). At 75 years, the relative difference is far greater at 27 and 22 per cent increases for men and women, respectively, reflecting the major improvements in survival prospects for successive cohorts of older people.
Life expectancy provides a crude monitoring tool for assessing the health status of a population. However, life expectancy does not tell the whole story. A major goal of health and social policy in old age is the increase in life expectancy free of disease or disability—healthy active life expectancy (Evans 1993a). The generic term, health expectancy, is best used to describe the growing number of indicators: disease-free life expectancy, dementia-free life expectancy, impairment-free life expectancy, disability-free life expectancy, handicap-free life expectancy, and activity restriction-free life expectancy. The precise form of health expectancy used has a profound effect on the duration of life estimated as the rates of different conditions vary markedly. For example, dementia is comparatively rare compared with all types of disability. Table 4 shows some estimates of different types of health expectancy for China.
Table 4 Health expectancy at age 65 years, China 1987 and 1992: effect of criterion used
Compression of morbidity
Human lifespan, unlike average life expectancy, is fixed and is about 120 years. Accounts of people living very much longer than this have not been corroborated. Life expectancy is increasing dramatically and will eventually approximate to our fixed maximum achievable lifespan. If the age at onset of chronic disability is increasing then, with a fixed lifespan, the period of serious ill health and disability will be compressed into a smaller proportion of the lifespan—the compression of morbidity hypothesis (Fries 1980). Fries’ original statement of the compression of morbidity hypothesis confused lifespan with life expectancy (the former is fixed and the latter is increasing) and he suggested that compression of morbidity would be apparent in populations where life expectancy was about 85 years (his ‘ideal average lifespan’). Revisions of the compression of morbidity hypothesis have emphasized that decline in the incidence of morbidity and disability must increase at a faster rate than the decline in mortality rates if compression of morbidity is to be achieved (Fries 1989) (Fig. 6).
Fig. 6 The observed mortality and hypothetical morbidity and disability survival curves for females in the United States in 1980: e0** and e60** are the number of years of autonomous life expected at birth and at age 60 years respectively. M50** is the age to which 50 per cent of females could expect to survive without loss of autonomy. (Source: WHO 1984.)
Fries et al. (1989) suggested that the compression of morbidity hypothesis was supported by evidence from randomized controlled trials of primary prevention that demonstrated no reduction in total mortality but reduction in the incidence of cardiovascular disease events. However, when all the evidence on the effects of primary prevention for cardiovascular disease is examined, no clear effects on morbidity or mortality are apparent (Ebrahim and Davey Smith 1997).
Alternative theories of expansion of morbidity and ‘steady state’ or equilibrium between morbidity and death have been put forward: the former on the assumption that increasing survival into old age will increase the numbers at risk of developing dementia syndromes (Olshansky et al. 1991). Manton’s steady state hypothesis suggests that there will be an equilibration of the amount of morbidity through parallel declines in both disease incidence and total mortality (Manton 1982).
Evidence to support or refute these hypotheses is difficult to find and hard to interpret. Successive cross-sectional studies of the prevalence of disability over a period of time may be used to calculate health expectancy using Sullivan’s method (Mathers and Robine 1997). This makes the assumptions that disability is irreversible and that duration of disability remains constant over the time period examined. Trends in health expectancy may then be compared with life expectancy to determine which is increasing faster. Such data are rarely available as it is not common for regular surveys using similar sampling methods and criteria for disability to be performed. Trends are particularly sensitive to the indicator of disability used (Table 5).
Table 5 Trends in disability-free life expectancy using different criteria of disability, England and Wales, 1976–1991
Cross-sectional data from the United States and from England and Wales suggest that health expectancy may be increasing more rapidly than life expectancy (Manton and Stallard 1991; Bone et al. 1995) if more severe disability is used as the criterion to define ‘health’. Figure 7(a) shows the trends in severe and milder levels of disability found in successive surveys in England and Wales, and Fig. 7(b) shows trends in limitation of activities among black and white people in the United States. In both England and Wales and the United States disability, however defined, is strongly age related. However, in England and Walse it appears that less severe disability—inability to get outdoors—shows no clear secular trend, whereas more severe disability—inability to manage basic self-care activities of daily living—show dramatic reductions in each age group over the same time period. In the United States, not only are disability rates higher among black than white people, but the trends between 1990 and 1996 show that white people have enjoyed a reduction in disability levels whereas black people have suffered an increase in levels (National Center for Health Statistics 1999). Unfortunately, such data are suspect owing to differences in the likelihood of selection bias in successive samples of older people and differences in how such questions on disability are interpreted by successive cohorts of older people. In the United States recent favourable trends in disability have been interpreted as demonstrating the effects of better health services (Manton et al. 1997). However, health services are unlikely to be the sole reason. Among university alumni followed up prospectively, those with high-risk health behaviours (i.e. smoking, high body mass index, lack of physical exercise) suffered twice the cumulative incidence of disability with the onset of disability postponed by an average of 5 years in those with low-risk health behaviours (Vita et al. 1998).
Fig. 7 (a) Trends in disability among older people in England and Wales, 1976–1991. (Source: Bone et al. 1995.) (b) Trends in disability among black and white people in the United States, 1990–1996 . (Source: national Centre for Health Statistics 1999.)
Trends in health expectancy are of major importance in monitoring health and social policy for older people. However, very few countries are capable of making reliable assessments over time, between geographic regions or socio-economic groups. Investment in research infrastructure is required to ensure that regular health and disability surveys using comparable methodology are established. The International Healthy Life Expectancy Network (REVES) has demonstrated the limitations of available data and have called for disability-free life expectancy to be promoted internationally as a health indicator (Robine et al. 1992). Application of earlier and later disability prevalence rates will make a major difference to projections of disability and health burden over the next two decades (Khaw 1999).
Population ageing is not uniform within most countries but tends to be geographically defined. For example, retirement communities have grown in areas of natural beauty and holiday resorts, and migration of younger people to cities has left older people behind in rural areas. The special circumstances of ageing in rural communities have not received much attention. However, the majority of the world’s elderly people live in rural areas. Living in rural areas is generally associated with economic hardship, traditional culture and organization of society, and limited availability and range of services (transport, shops, health and social support). Migration of young adults away from rural areas diminishes opportunities for family support and is so extreme in northern Italian hill villages that many are unlikely to be viable within the next decade (Golini 1997). Of particular concern from a public health viewpoint are poverty and depopulation, particularly in developing countries, which make traditional rural development programmes focusing only on mothers and children inappropriate. Environmental factors, including lack of safe drinking water and poor sanitation, inadequate housing, and absence of electricity, and poor roads and transport make life increasingly difficult for elderly people. Distance from major urban centres of population further increases the ease of neglect by government and health services. Traditional values of self-reliance are unlikely to operate sufficiently to reduce disadvantage given depopulation. Lack of an agreed definition of ‘rurality’ has limited information on elderly people in rural areas and has hindered development of specific policies that might improve their access to services.
Ethnic minority elders
Immigration associated with trade, the colonial past, wars, and famine have produced complex multicultural societies in many countries. Different countries will have experienced immigration at different times and contexts which will affect how ethnic minorities are viewed, their place in society, and their absolute and relative numbers. In the United Kingdom, immigration of ‘visible’ Caribbean, Somali, south Asian, and Chinese people occurred at different times and for different reasons (Ebrahim 1996a,b). The ‘push’ factors of poverty, political instability, and oppression and ‘pull’ factors of economic gain explain most migration patterns. Ethnic minorities tend to settle in distinct towns and regions of towns reflecting relative safety in numbers and work opportunities. Ethnic minority populations that migrated to the United Kingdom and other European countries are now ageing rapidly and may be at ‘triple jeopardy’ owing to their age, cultural and racial discrimination, and lack of access to health, housing, and social services (Norman 1985) (Fig. 8).
Fig. 8 Trends in world migration to the United Kingdom. (Source: Ebrahim 1996a.)
Census information shows that older people from the major ethnic minority groups in the United Kingdom, with the exception of Chinese, appear to have more long-standing disability than the white majority. The explanations for excess morbidity may be found in poverty, poor housing and diet, and limited physical activity, all of which are likely to contribute to higher rates of cardiovascular disease, diabetes, and other chronic health problems. The aetiology of ethnic minority health problems may be classified as influences of the ‘home’ country, influences of the new adopted country, selection of who migrates, and the processes of adaptation and adjustment to migration (Marmot et al. 1984). However, once people from ethnic minorities reach old age the selective ‘healthy migrant’ factors tend to have worn off and they become prey to the common causes of chronic ill health in the wider population.
Interestingly, black people in the United States have better life expectancy than white populations beyond the age of 80 years, which may reflect survival of the fittest (Manton 1980). Disability levels among American black and white people also show complex patterns as shown by data from New Haven and North Carolina. Black people were more likely than white people to become disabled, particularly in New Haven. This increased risk was explained by social and economic factors in predominantly rural North Carolina but not urban New Haven. It is possible that unmeasured social and economic factors or residual confounding explain this latter finding. Alternatively, other factors (e.g. weak family or community support networks) associated with black ethnicity in New Haven, but not North Carolina, result in an increased risk of disability even after socio-economic adjustment (Mendes de Leon et al. 1997).
Examining need and use of health and social services for ethnic minority elders is beset with problems. Denominators are not accurately defined as classification of ethnicity tends to be based on country of birth that fails to capture aspects of culture, language, religion, diet, beliefs, and behaviour. For example, in the United Kingdom, the classification used in the census developed by the Office for National Statistics is dominated by skin colour, reflecting British views of ethnicity (Table 6).
Table 6 Classification of ethnicity used in the United Kingdom Census 1991
Numerators for calculation of rates of morbidity are seldom complete owing to poor collection of information on ethnicity in routine health and social care settings. Ad hoc studies examining the use of health services have reported that use of primary care services is higher than in the white population (Balarajan et al. 1989), prompting concerns about ‘over-consultation’ and a desire to normalize rates to more acceptable levels. However, such work has not made appropriate allowance for higher levels of morbidity or more adverse social and economic factors which tend to have independent effects on primary care use. Inequality of access to primary health care may not occur but it is likely that more subtle barriers to referral, investigation, and treatment in secondary care exist (Ebrahim 1996a). For social services, it is far clearer that access is a major problem. Far fewer people from ethnic minorities receive social support services, yet from what is known of need, they should be represented in greater proportion. It is more likely that ‘the family will cope’ myth explains why fewer ethnic minority elders are deemed to need social services, together with culturally inappropriate and insensitive services (e.g. ‘meat and two veg’ meals on wheels, day care without appropriate religious observances). Much more work needs to be done to assess need in order to develop and evaluate more appropriate services for ethnic minority elders.
Older people as a special group
Like children, older people are often singled out for special status in public health plans and in prioritization exercises. While it is convenient for needs assessment to group older people by chronological age into young-old, old-old, and oldest-old, these groupings ignore very real differences in people and risk exacerbating the stereotyping of older people. People become more heterogeneous in their health status as they get older which would suggest that grouping elderly people together may be quite inappropriate. Indeed, much of the debate over geriatric versus general medical services for older people is due to the need to ensure that older people are treated decently and that they also obtain access to specialist services from which they are able to benefit. The physical and mental frailty of many older people, widespread ageism, and prejudicial views of their ability to benefit from health care increase the need for advocacy.
Icebergs of disease
Unrecognized health problems are more common among older than younger people and are caused by several distinct mechanisms: failure to report symptoms, denial of symptoms, underinvestigation, and poor diagnosis of disease by doctors. Under-recognition of health problems is particularly common in incontinence of urine, depression and dementia, visual and hearing impairment, and locomotor disability (Williamson et al. 1964). The implication of under-recognition is loss of potential health gain and attempts to improve under-recognition underpin much case-finding and screening for health problems in older people.
It is rare for any adult to have a single health problem. Typically most adults have a range of health problems, some of which are active, some inactive. In old age, it is more likely that such health problems will be active and will interact with each other, resulting in complicated patterns that tend to defy simple approaches to intervention. Multiple pathology is the rule rather than the exception and must be acknowledged in needs assessment and service delivery. At a population level, need is commonly assessed using mortality statistics. However, deaths are coded to a single underlying cause of death and routine mortality statistics do not use data from the second part of the death certificate on other diseases that may have contributed to the death. Indeed, such information is rarely complete, further limiting its value for assessing need. In developing countries it is common for a majority of deaths to be unattributed to International Classification of Disease groups. For example, in Thailand in 1996, 47 per cent of causes of death among people aged 60 years and over were recorded as ‘senility’ (R54), 6.7 per cent as ‘ill-defined and unknown causes of mortality’ (R95–R99), and 2.9 per cent were classed as ‘symptoms, sign and abnormal clinical and laboratory findings, not elsewhere classified’ (R00–R53; R55–R94).
From a therapeutic perspective, multiple pathology makes evaluation of interventions difficult as comorbidities may make treatment more hazardous or appear less effective. This may be a justification for excluding older people from randomized controlled trials but failure to include them perpetuates the current lack of evidence of benefits and harms of treatments for older patients.
Loss of adaptability
The physiological functional reserve capacity (e.g. in terms of muscle strength, cardiorespiratory fitness, skeletal integrity) of older people declines with increasing age (Young 1992). Older people have to cope with an increasing ‘fitness gap’ and it is this reduction in reserve capacity which places older people closer to thresholds that will limit their functional independence and will reduce their capacity to adapt to new challenges presented by disease and social and environmental factors (Gray et al. 1985; Evans 1990). Reduced adaptability makes elderly people vulnerable to increased risk of complications of disease and more likely to suffer a ‘cascade of disasters’ following an initial fairly trivial incident. Loss of adaptability also makes ‘atypical’ presentations of disease more common as thresholds for normal performance are so precarious that minor degrees of impairment, regardless of the organ system involved, result in more general physical and mental disturbances (Fig. 9).
Fig. 9 Relationship between age and physical capacity showing the effects of a threshold for independent living.
Loss of adaptability is of concern to public health in two main ways: firstly, maximizing functional reserve capacity through optimizing childhood and early adult life development will ensure that people cross thresholds that render them dependent at ever older ages. Secondly, public health should aim to identify and reduce social and environmental factors that present differential challenges to older people. For example, public transport in many countries is difficult to use without substantial reserves of lower-limb girdle muscle strength. Crossing a road within the time allowed on traffic light controlled crossings requires an average walking speed that is higher than that achievable by most 70-year-olds (Hoxie and Rubenstein 1994). Improving access to public transport and ensuring road crossings are safe for all should be part of wider public health concerns.
Costs of care
The potentially very high costs of care for older people make a compelling case for considering older people as a special group. However, the notion that health care costs increase inexorably with age is not supported by analyses examining the costs of care in relationship to proximity to death. Such analyses show that it is terminal illness that is expensive rather than age itself (Normand 1998). Very elderly people have lower costs than younger people with the same survival prospects (Himsworth and Goldacre 1999).
Nursing home costs in the United States have grown at a far faster rate since the 1960s than other sectors of health care spending (National Center for Health Statistics 1999), as shown in Fig. 10. British long-term care costs also dominate the overall projections over a period of decades, far outstripping primary care and hospital sector costs (Laing et al. 1991). Hotel ‘board and lodging’ costs make up a high proportion of long-term care costs and it is generally accepted that individual payment for these costs is acceptable (Royal Commission on Long Term Care 1999). Insurance schemes to meet long-term care costs have been introduced without much success. Commentators have noted that even with the most optimistic assumptions, long-term care insurance would meet only 10 per cent of total American government costs in this area (Weiner and Illson 1994).
Fig. 10 Trends in different sectors of health care costs, United States, 1960–1997. (Source: National Centre for Health Statistics 1999.)
Within the current health service system, both inequity of access to services and overuse of ineffective services occur. Older people who would benefit from specific interventions (e.g. coronary revascularization) do not receive them (Bowling 1999); simultaneously, acute hospital beds are inefficiently used by older patients waiting for transfer to a different care sector, either their own homes or institutional accommodation. Increasing the barriers to gaining entry into secondary care for older people is an obvious, but misplaced response, as it exacerbates inequity for some. Ensuring that both health and social care costs are considered in economic appraisals, rather than just the costs of one or other system should ensure that the balance of care is of net benefit to society and not simply to one payer rather than another.
Health and social services for elderly people
The purpose of health and social services for elderly people are to prevent disease, save lives, reduce suffering and disability, avoid complications of disease, and to ensure that resources are used efficiently. ‘Care in the community’ is a broad strategic goal which cuts across these other goals as it is expected to be less expensive than care in the hospital and to make better use of a wider range of community services. It is important to recognize both the breadth of health and social services for elderly people and their interrelationships. It is useful to have a comprehensive approach to services and this is best achieved by grouping services according to area of activity: health promotion, primary care, community services, social services, residential and long-term care, and acute hospital services. Typically, acute hospital services occupy a disproportionate amount of time and resources compared with other areas. The very wide range of services that may be offered to elderly people may present considerable problems of planning and communication if ‘seamless’ care is to be achieved (Box 1).
Ideally, health and social services for elderly people would be jointly planned and funded from a single budget. In practice, the historical development of services has resulted in separate planning and funding with consequent gaps in provision, and geographical differences in custom and practice. A ‘social bath’, for example, would constitute a bathing service that health services will not pay for but social services might! Health care systems in which social care is entirely provided by families also require some form of joint planning. This is to ensure that gaps in provision do not occur and that family support is feasible within the constraints of other aspects of development such as the emancipation of women, urbanization, and economic change.
Major health sector reforms—increased user charges, cost containment through standard care pathways, separation of health and social components of care—implemented worldwide, will tend to reduce access to both primary and secondary care. Their effects will be felt disproportionately by the most vulnerable sections of society—children and elderly people.
Role of public health
Public health is concerned with improving the health of populations through assessing need, defining priorities with local stakeholders (including patients), identifying and implementing cost-effective interventions, and monitoring and evaluating their impact. Too often in the United Kingdom, public health has become a means of commissioning health services from the lowest-cost provider. The new public health movement in the United Kingdom emphasizes improving population health by working with agencies other than health services, the need to tackle inequalities in health and in access to health care, and should provide a new impetus for making progress. However, too often public health policies fail to consider older people, and if they do, exclude older people as unable to benefit from primary prevention, for example (Department of Health 1998). Public health medicine practitioners are not well trained in the issues caused by population ageing and are inexperienced in working with elderly people, particularly in developing countries where difficult resource use choices will have to be made. New training initiatives on public health for ageing populations are urgently needed (Ebrahim 1999).
Ageism and rationing
In a society that holds ambivalent but essentially negative values towards older people, manifest as ageism and paternalism, rationing through denial of services or ‘knowing better’ will occur. How much should be spent on care for elderly people can only be determined by reference to ethical principles. If no value is placed on life once economic productivity ceases, then logically resources should not be wasted on health services for elderly people. National values are shaped over time but may change rapidly in response to practices perceived to be unacceptable. For example, the transfer of frail, demented elderly people to private sector nursing homes to enable closure of old outdated British National Health Service (NHS) property was common and accepted as it was assumed that the care on offer would be better and cheaper than the NHS alternative. However, a 1992 television programme that examined the plight of 40 demented elderly patients with no immediate next-of-kin transferred from an East London hospital to a Yorkshire nursing home offering abysmal standards of care caused public and government outrage. In this case, public health values were out of step with national values that appeared to find lowest cost care unacceptable.
How much one generation should spend on supporting earlier generations remains to be tested. The ‘fair innings’ argument—after so many years, old people should not expect any more—is used by health economists to argue for transfers from old to young rather than vice versa (Williams and Evans 1997). An obvious flaw in such arguments is that we only get sick and in need of health and social care when we get old; the ‘fair innings’ does not relate to health status or to ability to benefit from health care, only to years lived. Considerable effort is being made by Age Concern in the United Kingdom to persuade government to remove age discrimination in society and within the NHS (Age Concern 1999; Rivlin 1999).
Box 1 Health and social services for elderly people
lowering blood pressure
Case-finding and screening:
General practitioner services
General dental services
Community psychiatric nurse
Good neighbour schemes
Acute medical care
Acute psychiatric care
Early discharge schemes
Specialist units: stroke, orthogeriatrics
Day hospital: medical, psychogeriatric
Nursing home care
Family and informal carers
Nursing home care
Home care services
The ethics of state provision of any service is based on utilitarianism—the greatest good for the greatest number—and the British NHS was born out of the collectivist politics essential to surviving a world war. Health care policy should aim to target resources to where they will do most good rather than issue them in response to the distribution of disease. However, targeting requires relevant information on service inputs and health outcomes. In the face of inadequate information on costs and benefits—and the natural peacetime ethics of individualism (Evans 1997)—it seems inevitable that doing the best for the patient regardless of cost should gain hold over utilitarianism. With stronger evidence on the effects of pharmacological interventions, it is also inevitable that conflicts between groups holding different ethical precepts (and with different financial incentives) will emerge. Older people will stand to lose on two counts. Firstly, evidence is weakest and hardest to obtain for complex non-pharmacological interventions from which they may benefit. Secondly, the inherent ageism of society will ensure that elderly people will lose in an individual or collectivist fight for resources. In the United Kingdom, compared with European and North American countries, health care spending is now at such a low level that debates about age-based rationing are considered untenable until adequate resources for health services are found (Tonks 1999). However, without better and wider evidence of benefits of innovative patterns of care for elderly people, more resources will not necessarily mean better care.
Deciding on appropriate health outcomes for the evaluation of services for elderly people is not a simple matter. As different services have different objectives, general outcomes, such as survival or quality-adjusted life years, do not have wide currency, nor do they help make choices between different areas of activity. Several factors determine what should be considered a benefit of health or social care: the ethical stance taken; the type of intervention; the type of patient; and the viewpoints of other ‘stakeholders’—family, service providers, payers.
Interest in health-related quality-of-life indicators has increased as evaluation tools as their multidimensional approach and supposed focus on quality of life appear to tap into many of the objectives of health services for elderly people (Ebrahim 1995). Furthermore, health-related quality of life indicators are much more firmly rooted in patients’ experience of disease and should increase the chances of more patient-centred care. The goal of many health-related quality of life indicators, including the Euro-QoL and the 36-Item Short Form (SF-36) health questionnaires (Brazier et al. 1993; Jenkinson et al. 1997; Ware and Gandek 1998), is to provide an index or profile of the value or utility of different health states.
Valuations of different health states may be obtained by time trade-off methods, standard reference gambles, or by simpler visual linear analogue scales. But should patient, professional, or general public valuations be used? For dimensions of pain and mobility, there appears to be little difference in valuations provided by older and younger people or disabled people (Ebrahim et al. 1991). In a patient-centred service, it would be logical to use not only patient valuations of health states but also to permit the individual to define the health states of interest and concern to them. Such patient-generated health outcome indicators have been developed (Ruta et al. 1994), but require further development work before they could be widely used in health service evaluations.
When pooled to provide utilities, health-related quality of life indicators fall into a potential trap of giving a number that may be less informative than responses to a profile of different dimensions. The items making up health-related quality of life indicators include a mixture of symptoms of pain and depression, difficulties carrying out everyday activities, and impact on social and work life. It is highly unlikely that a specific intervention would affect each area equally, consequently there is a risk of making false-negative inferences if health-related quality of life indicators are used as trial outcomes.
Impairment, disability, and handicap
An alternative model for examining the effects of interventions for older people is the impairment, disability, and handicap model (Harwood et al. 1994b). Impairment is defined as the loss or abnormality of psychological, physiological, or anatomical structure or function. Disability is any restriction or lack of ability to perform an activity in a normal manner. Handicap is the disadvantage experienced by the individual as a result of impairments and disabilities that limits performance of a normal role, taking account of age, sex, and cultural background (WHO 1980). Impairment, disability, and handicap provide clearly defined areas that might be improved by specific interventions. The classification also enables a distinction to be made between the intrinsic consequences of disease (i.e. impairment and disability) and the external factors that mitigate or accentuate the degree of disadvantage suffered by the individual (i.e. handicap). In evaluation of health services, these distinctions are essential. For example, an intervention comprising walking and kitchen aids together with practice in their use might be expected to improve performance in locomotor and kitchen abilities, and hence these areas would be logical to measure as outcomes. An intervention comprising social services support and benefits would aim to reduce disadvantage and consequently should be assessed by measurement of handicap. If either of these interventions were assessed using health-related quality of life indicators, it is quite possible that no benefits would be found as the indicator would be insensitive to the interventions under investigation.
Disability scales abound and the most commonly used in hospital practice are the Barthel Index of basic self-care activities (Mahoney and Barthel 1965) and the extended or instrumental activities of daily living scales (Nouri and Lincoln 1987) (both originally developed for use with stroke patients). Extended activities of daily living scales are more suitable for community evaluations where people may be less disabled. There is little to choose between various scales as they all share similar items and are scored in similar ways. A caution that should be noted is the practice of examining changes over time in terms of Barthel Index points per week (Shah et al. 1990, 1991). Such an approach makes the assumptions that an activities of daily living scale is a continuous variable with interval scaling properties and that a comparable initial baseline point for each patient can be obtained. Thus, patients seen very early in the course of their illness may score zero and be completely unable to perform any tasks but within a day, may be able to feed themselves. If the former baseline is used, changes over time will be exaggerated compared with using the latter baseline.
Handicap measures have not been widely developed owing to the difficulty in operationalizing a concept that is intrinsically individualistic. One way forward has been to produce a multidimensional handicap scale using dimensions suggested by the original World Health Organization (WHO) report (Harwood et al. 1994a). Semantics of disability and handicap have changed recently and the new WHO classification uses the terms impairment, ability (i.e. disability), and participation (WHO 1999). While these latter two terms are more positive, the term participation does not capture the same essential characteristic of disadvantage implicit in the term handicap. Operationalization of these new terms to produce outcome measures is underway but it will take time before the performance and value of these new concepts can be assessed.
Health service evaluations usually consider outcomes at fixed time points which fails to take account of the pattern of recovery or natural history of the diseases suffered.
Patients will experience very variable ‘care careers’ despite receiving treatment from the same services. For example, patients with heart failure may suffer episodes of profound disability alternating with periods of reasonable ability whereas patients with Parkinson’s disease will tend to follow a progressive decline in ability. In these circumstances, it makes little sense to examine aggregated ‘outcomes’ of general services for elderly people. Heterogeneity between patients in their natural history and the goals of care make conventional evaluation methods much less robust and more likely to produce spurious data.
Routine information in most health care systems has failed to keep pace with demographic and epidemiological transitions. Summary mortality statistics are still published with an upper age band of 75 years and over, although in the last decade the British Office of National Statistics uses an upper age band of 95 years and over but the American Bureau of the Census uses an upper band of 85 years and over. International agencies should agree which thresholds to use in varying circumstances so that comparisons may be made more easily.
While routine reporting of common infectious diseases has been maintained despite the falling mortality attributable to this cause, no efforts have been made to make common chronic diseases notifiable to national health statistics centres. Consequently, we have little idea of the incidence of coronary heart disease or stroke and sources of variation within and between countries. By contrast, routine cancer notification has been established in many countries and can make a useful contribution to monitoring trends. Such information is essential for health service planning for an ageing population.
Surveys of disability have been performed in many countries but few countries have invested resources in maintaining regular repeated national surveys of disability. In the United Kingdom, the first national disability survey (Martin et al. 1987) conducted in 1985 was repeated in 1996 (Devore 1998) but differences in sampling and methods make comparisons between the two surveys difficult. In particular the 1996 survey did not include institutionalized elders. Ideally, a series of nationally representative cohort studies should be established using internationally agreed methods of sampling and measurement. This information resource would enable national and regional differences in health and social policy to be examined in ecological analyses. Projections of future disability levels would be aided by information on transitions in and out of disabled states, leading to more accurate estimation of health expectancies and future health and social care needs.
Increasingly, useful sources of up-to-date information on population ageing and the effects of health care interventions can be found on the World Wide Web (Jenkins 1999) (Box 2).
Health promotion is concerned with three interrelated activities: primary prevention of disease, health education, and health protection (i.e. legislation or fiscal methods) (Downie et al. 1990). It can be argued that effective health promotion can only occur when all three areas are working optimally. Increasingly, there is a recognition and desire to include elderly people in health promotion programmes that were formerly targeted to younger people. Disease-specific programmes for cancer and heart disease prevention have typically involved middle-aged adults and little work has been carried out focusing specifically on elderly people. Such programmes of multiple risk factor intervention, typically including dietary and smoking cessation advice, exercise, and sometimes treatment of hypertension and raised blood cholesterol, do not appear to be effective in reducing the incidence of disease when tested in large-scale randomized controlled trials (Ebrahim and Davey Smith 1997). Alternative study designs using quasi-experimental methods have also failed to establish a clear role for current health promotion methods for cardiovascular disease prevention in middle-aged adults (Ebrahim and Davey Smith 1996). It seems unlikely that the same methods applied to older adults will deliver markedly better results.
Box 2 World Wide Web sources of information on ageing
Effectiveness of interventions
By contrast similar multiple risk factor interventions applied to people who have already suffered a heart attack or angina are effective, reducing total mortality by about 20 per cent (Ebrahim and Davey Smith 1996). Secondary prevention is thus a major target for health promotion programmes and it is important that older people are allowed to take part in them. Reorientation of current primary prevention programmes is urgently required as resources are being wasted. It is also vitally important that these programmes are not sold to developing countries as the best means of coping with their increasing burdens of cardiovascular disease. Health protection appears to be more effective as shown by a recent account of the effects of a health promotion programme that achieved notable reductions in blood cholesterol only when the main source of cooking oil changed from palm oil to unsaturated oil (Dowse et al. 1995).
Smoking cessation advice amongst older people appears to be as effective as in younger adults (Vetter and Ford 1990; Vetter 1999), although in the United States brief interventions may not be as effective as in British primary care (Burton et al. 1995a). Contrary to popular belief, many older people do wish to stop smoking and do not view smoking as one of life’s little pleasures but as a positive harm that may have ill effects on their grandchildren (Vetter et al. 1988). The British Doctors’ Study highlights the years of life gained by avoiding smoking after the age of 35 years—over 7 extra years of life expectancy on average (Doll et al. 1994).
Exercise programmes are widely promoted among older people as a means of maintaining health. The supposed benefits of physical exercise are derived from observational epidemiological studies showing large differences in the rates of cardiovascular disease, depression, and osteoporosis in those taking moderate or greater levels of exercise (Fentem 1994; Wolman 1994; Tseng et al. 1995; NIH Consensus Development Panel on Physical Activity and Cardiovascular Health 1996; Buchner 1997; Rippe and Hess 1998; Fletcher 1999). However, attempts to test the effects of exercise in randomized controlled trials are less compelling (Hillsdon et al. 1995). Health benefits tend to be found in short-duration trials involving volunteers. Many older people do not wish to take part in exercise interventions, often because they feel too unwell or too busy to do so. Also, drop-out rates from trials are often large, emphasizing the difficulty of maintenance of increased levels of exercise, and have demonstrated that even brisk walking is not without adverse effects causing injury (Ebrahim et al. 1997). A systematic review has demonstrated that environmental interventions that encourage walking or cycling (e.g. safe cycle paths and walkways, local shops, and other amenities within walking distance) reach a greater proportion of the population than interventions aimed at increasing the use of exercise facilities (Hillsdon and Thorogood 1996).
Nutritional interventions for older people include the widespread ‘healthy diet’ approaches which have been modified over the years from low-fat low-cholesterol diets to fruit and vegetable diets. Evidence from randomized controlled trials of dietary advice provides limited support for the use of low-fat diets(Hooper et al. 2001) or antioxidant vitamins such as b-carotene (Egger et al. 1998). Among people with pre-existing heart disease, eating oily fish regularly has a significant and clinically important reduction in total and cardiovascular disease mortality (Hooper et al. 1999).
Age well programmes
More general health promotion programmes for elderly people have aimed to alleviate disengagement and loneliness and build on the perceived value of keeping active in old age. Examples abound and include programmes run by retirement associations and by voluntary sector charities such as Help Age International and Age Concern in the United Kingdom. Such programmes aim to provide opportunities for elderly people to participate in a range of activities that provide social engagement, role fulfilment, and physical and mental activity. In general, evaluation of these programmes has only been in terms of coverage and continued participation rates. Improvements in health status or in maintenance in the community have not been examined. Increasing the income of older people might be expected to increase health status from observational data showing social gradients in both disability and disease. Trials of increasing income have been conducted but appear not to have been targeted at older people (Connor et al. 1999). This is a major oversight and such trials are urgently required.
Cross-sectoral collaborative interventions aiming to improve environmental factors (e.g. transport, housing design) that contribute to disability have not been widely established or evaluated. Health improvement programmes are expected to involve local government departments and non-governmental agencies. These new alliances hold great promise for improving the health of older people. Evaluation of their costs and health benefits should be an integral component when establishing new programmes, but will be difficult (Judge 2000).
Elderly people are high users of primary care services, reflecting their greater levels of disease and disability. In the United Kingdom, elderly people are more likely to see their general practitioner at home than at the surgery, probably reflecting the increased frailty of very elderly patients (Fig 11 and Fig 12).
Fig. 11 Consultation rates in general practice by age and place of consultation. (McCormick et al. 1995, Table 2b.)
Fig. 12 Consultations in general practice by age and underlying diagnostic group. (McCormick et al. 1995, Table 19.)
The major causes of consultations for elderly people do not show marked differences from those at younger ages. However, there is a tendency for ill-defined conditions to contribute relatively more at the oldest ages.
In many developing countries, the role of primary care is paramount as secondary level services are inaccessible, and often inappropriate, for a majority of elderly people. However, training in health care for elderly people is not a high priority in undergraduate medical or nursing curricula of developing country medical schools. Despite the undeniable importance of maintaining strong maternal and child health training programmes in developing countries, it is essential that recognition of the growing numbers of elderly people should be reflected in increased opportunities to learn about the health care of elderly people. European and North American models of geriatric medicine are heavily reliant on hospital services and do not have an appropriate balance of care for the needs of developing countries. The first step is to examine the common health problems seen among older people in primary care and to match training to meeting these needs. Without appropriate training it is very likely that elderly people will simply receive a growing number of pills for each of their symptoms, running the risk of iatrogenesis, and their potentially diagnosable and treatable problems will be missed.
Immunization against influenza is strongly supported by evidence from cohort studies and randomized trials (Gross et al. 1995). It appears that elderly people who might benefit are frequently excluded or missed from immunization programmes (Centers for Disease Control and Prevention 1997; Watkins et al. 1999). In the United Kingdom, it has been suggested that immunization should be confined to high-risk and institutionalized elderly (Nicholson 1993). However, American cost-effectiveness studies demonstrated cost savings when universal, rather than selective policies of immunization were adopted (Mullooly et al. 1994; Nichol et al. 1994). It is likely that these findings would apply in most postindustrialized countries, and in 1999 British government medical advice was modified to bring it in line with many other countries, recommending the immunization of everyone aged 75 years and over (Watkins et al. 1999). It is probable that untargeted extension of immunization to those aged 65 years and over would also be cost-effective, given the problems associated with ensuring that high-risk people are identified. Additionally, there is trial evidence that health care workers in institutions for elderly people should be immunized against influenza, as this reduced the mortality risk among the residents (Potter et al. 1997).
Pneumococcal vaccination does not appear to give any protection against developing pneumonia or death when given to elderly people in the community (Fine et al. 1994). More recent trials confirm earlier findings (Koivula et al. 1997; Ortqvist et al. 1998) but suggest that those people at high risk of pneumonia (i.e. institutionalized, immobile, on immunosuppressive drugs, suffering with serious chronic diseases) would benefit. In cost-effectiveness analyses, a policy of widespread, untargeted vaccination of everyone aged 65 years and over was cost saving (Sisk et al. 1997). Further work is required to evaluate the cost-effectiveness of a combined strategy of universal vaccination of elderly people at different age thresholds with both influenza and pneumococcal vaccines compared with targeted vaccination to determine cost-effectiveness.
Case-finding and screening
Older people consult their general practitioner several times a year, which provides considerable opportunity for opportunistic case-finding. However, in areas such as high blood pressure, depression, and visual and hearing impairment, icebergs of unrecognized disease are common. Interestingly, non-consulters appear to be a fairly healthy group of elderly people and would not be a high priority for screening (Ebrahim et al. 1984). In the 1990 British general practitioner contract it was required that health checks be offered to people aged 75 years and over. These checks were expected to focus on mobility, mental state, hearing and vision, continence, general function, and medication review (Williams and Wallace 1993). While these areas are important and are the areas commonly causing problems to older people, the contractual obligation did not stretch as far as requiring any specific action to be taken in response to problems identified.
The potential benefit of health checks amongst the over-75-year-olds has been assessed: more dementia, depression, incontinence, and immobility would be identified (Iliffe et al. 1991a), but the costs would be considerable, amounting to about 150 h of primary care time per year for a typical general practice of 2000 patients of whom 130 are over 75 years old (Iliffe et al. 1991b). The evidence to support the effectiveness of health checks or case-finding carried out in primary care is very limited, although a large British Medical Research Council trial comparing usual care with augmented health checks has been established but results have yet to be published.
Evaluations of case-finding in primary care have been systematically reviewed as part of a larger review of comprehensive geriatric assessment (Stuck et al. 1993). A total of seven randomized controlled trials (five performed in the United Kingdom and two in Scandinavian countries) were reviewed. The pooled results showed a statistically non-significant improvement in mortality (odds ratio 0.87, 95 per cent confidence intervals 0.71, 1.07), no difference in the chances of living at home at 12 months (odds ratio 1.07, 95 per cent confidence intervals 0.87, 1.31), and no difference in functional improvement at 12 months (odds ratio 0.96, 95 per cent confidence intervals 0.71, 1.29). Hospital readmissions were significantly reduced in those receiving health checks (odds ratio 0.84, 95 per cent confidence intervals 0.73, 0.96) but there was also significant heterogeneity between the trials suggesting that differences between the trials in the intervention studied contributed to this finding.
Since this review was published two further American trials, an Italian trial, and a Dutch trial have been reported. One American trial demonstrated that a home-based screening programme for people over 75 years was effective in reducing disability (Stuck et al. 1995). However, this analysis took no account of deaths and was based only on survivors. Reanalysis of the data using an intention to treat approach and combining deaths and disability produces a non-significant odds ratio of 0.73 (95 per cent confidence intervals 0.48, 1.12) which is consistent with the pooled data from the other seven trials. A Medicare preventive health care intervention offered to people aged 65 years and older found only ‘modest health benefits’—a 25 per cent reduction in mortality (8.3 versus 11.1 per cent)—which was achieved at no increased cost to Medicare (Burton et al. 1995b; German et al. 1995). The Italian trial examined the effects of an integrated social and medical care and case management package among frail elderly people already in receipt of social support and found improvements in disability and reduced health service use over the period of a year (Bernabei et al. 1998). The Dutch trial failed to find any benefits associated with intervention, although a subgroup analysis showed some improvement in those people rating their health as poor (van Rossum et al. 1993). It is unlikely that findings on preventive health care can be generalized from systems of health care where financial barriers to consultation may exist to ‘free at the point of use’ services.
Screening for specific problems does not appear to be any more beneficial than attempting to deal with several problems together. For example, screening for visual problems is not associated with improvements in visual impairment (Smeeth and Iliffe 1998, 1999). Screening for mental health problems does not appear to improve quality of life or mood (Cole 1998). Screening for factors that might increase risk of falls, followed by targeting interventions at intrinsic and environmental factors may be of value but single interventions such as exercise or health education appear to have little value (Gillespie et al. 1999). An exercise intervention evaluated as part of a preplanned data pooling has demonstrated reductions in the risk of falls (Province et al. 1995). In each of these areas, trials are beset with methodological problems common to the evaluation of any complex intervention; of these problems, small sample sizes are the most common.
High blood pressure
Treatment of high blood pressure as a means of preventing stroke among relatively fit people up to the age of 80 years is supported by strong evidence from randomized controlled trials (Insua et al. 1994; Mulrow et al. 1994). Treatment of people over the age of 80 years has been examined in a subgroup analysis of the seven randomized controlled trials of drug treatments. Reductions in non-fatal cardiovascular events were found with no overall reduction in total mortality (Gueyffier et al. 1999). However, increased efforts to screen for high blood pressure, either by case-finding or screening, have not shown improvements in population coverage, in adherence to treatment, or better blood pressure control (Ebrahim 1998).
Involving elderly women in routine breast and cervical cancer screening programmes is controversial. For women aged between 65 and 69 there is consistent evidence of benefit from breast cancer screening with a reduction in mortality of 30 per cent (95 per cent confidence intervals 50 per cent, 90 per cent) (Fletcher 1996). Among women over the age of 65 who have had a series of three negative cervical smears, further routine cervical screening is of no benefit (Fahs et al. 1992).
In summary, efforts to improve disability and to prevent disease through systematic case-finding are not supported by currently available evidence. It seems likely that failure to achieve benefits is due to inadequately specified actions to be taken in the light of problems detected by case-finding. Further work in this area needs to ensure that interventions are better planned and should evaluate how well the processes of care operate in addition to examining outcomes. Reliance on trials performed in different health care systems is unwise owing to the very different barriers to care, expectations of people, and differences in need. More data are required to guide policy on the application of trial evidence on treating hypertension and screening for cancers among elderly people.
Community and social care
The wide range of community and social services provided for elderly people (Box 1) are of potentially great importance in maintaining elderly people in their own homes and reducing inappropriate use of long-term institutional care or acute hospital care. Considerable resources are spent on social services for elderly people but there is no culture of evaluation comparable to that in health services. Community and home care services are characterized by problems of mismatch between needs and resources, inadequate assessment processes, lack of monitoring and review, and inflexibility of provision and lack of integration (Challis 1996).
In the United Kingdom, the use of care management was intended to bring greater coherence to planning by putting a single social worker in charge of building a ‘package of care’ for a patient. This package would be based on the patient’s needs and would reduce the bias towards institutional care for frail elderly people. The research that underpinned this innovative approach had shown the importance of care manager training and experience and the need for discretionary budgets that would be controlled by care managers. Under these circumstances the need for institutional care was reduced, elderly people and their families were more satisfied, and the overall cost was no greater than usual patterns of care (Wolman 1994; Tseng et al. 1995). A before-and-after evaluation of the impact of community care reforms in Northern Ireland demonstrated that the use of institutional care was more targeted towards people with greater physical and mental disability after implementation of the reforms (Crawford et al. 1999).
Unfortunately, implementation of care management by social services departments in the United Kingdom has been difficult with many managers ill suited to the task and strapped for sufficient cash to make up viable packages of care. The principles of care management are sound but implementation requires greater commitment from senior managers.
Inevitably, the pattern of community and social services available in a locality will vary. Consequently, maintaining older people in their own homes may require available resources to be used in innovative and flexible ways. Public health medicine has a major role in ensuring better communication between service users, and the sectors paying for and providing services. In this way, local plans may be developed which would aim to cover the majority of needs. Research and development in community and social services is desperately needed to provide stronger evidence to support continued investment in services.
The variation in the organization of hospital services for physically ill elderly people, reflects differing beliefs and historical developments. In the United Kingdom, geriatric medicine grew out of the NHS and was initially a custodial service with responsibility for large long-stay care hospitals (Warren 1946). The first cohort of geriatricians managed to assess, rehabilitate, and discharge many of their long-stay patients, demonstrating the inadequacy of existing general medical services that had put patients in such places in the first place. These services grew into needs-based services which aimed to reduce the chances of elderly people being inappropriately treated by providing a rapid ‘take-away’ service for acute medical specialties. As the specialty of geriatric medicine grew, alternative models developed. Firstly, age-related services arose as a mechanism for obtaining better quality facilities in major, rather than peripheral, hospitals. These services agreed to manage all patients over a certain age and in return received a quota of acute beds. Next, integrated services arose as it became apparent that geriatric medicine was simply another subspecialty of medicine and should be organized no differently. In this model, all physicians take a share of all the acute medical admissions and only those elderly people with difficult problems are referred on to the physician with an interest in geriatric medicine (Evans 1983; Evans and Graham 1984).
As no good evidence exists to defend or support any of these three major models, considerable debate has occurred over which model best meets the needs of older people (Young 1989). As their needs are so variable, it is possible to make a compelling case for any model of care. And, with good communication and sensible referral of difficult cases, any model of care can be made to work to the benefit of patients and the hospital.
In developing countries where geriatric services seldom exist even in teaching hospitals, elderly patients are beginning to be experienced as ‘problems’ in acute wards and ‘bed-blocking’ has begun (Thamprechavai et al. 1992). The American hospital services have not had a history of geriatric medical services but have been in the forefront of performing randomized controlled trials to find out what is effective management for elderly patients (Stuck et al. 1993). This work, while not necessarily directly of general applicability to British or European services, does demonstrate the feasibility and importance of performing trials in circumstances of uncertainty.
Comprehensive geriatric services
A systematic review of comprehensive geriatric assessment examined the effects of hospital geriatric evaluation and management units and inpatient geriatric consultation services on mortality, place of discharge, and change in functional status (Stuck et al. 1993). In total, six trials of geriatric evaluation and management units and eight trials of inpatient geriatric consultation services were included in this review but one of the trials in each group was of patients with fractured neck of femur managed in orthopaedics or by geriatric services. As the focus was on comprehensive geriatric assessment applied to elderly patients admitted to hospital, it is debatable whether these two trials should be included when trials of specialist management of stroke patients were excluded. The review demonstrated short-term (6 months) improvements in mortality: odds ratios were 0.65 (95 per cent confidence intervals 0.46, 0.91) for geriatric evaluation and management units and 0.77 (95 per cent confidence intervals 0.62, 0.96) for inpatient geriatric consultation services, which had attenuated by 12 months. Interestingly, geriatric evaluation and management units were very successful at discharging people to their own homes and improving functional ability at 6 months (odds ratios of 1.80 and 1.63 respectively,) whereas inpatient consultation services did no better than usual care. This suggests that geriatric services that have direct responsibility for patient management do better than those that simply give advice.
A recent trial comparing inpatient assessment alone, assessment together with postdischarge home support, and usual care for elderly people admitted to hospital found that disability, readmissions to hospital, and institutionalization (but not mortality) were only improved by the combined assessment and home support service (Nikolaus et al. 1999).
Although the conclusions of this meta-analysis of trials of comprehensive geriatric assessment and the recent trial appear to support current working methods, there is always the possibility that publication bias has resulted in small negative trials remaining unpublished. This was examined in a study comparing the effect sizes derived from meta-analyses of small trials with definitive ‘mega-trials’ (Egger et al. 1997). The small trials appeared to overestimate the effect of geriatric care when compared with a large health maintenance organization trial of 2353 patients which found no improvements in survival or functional status (Reuben et al. 1995). The authors of the original meta-analysis suggest that one reason for the difference between their findings and those of the mega-trial was that patients in the usual care control arm of the mega-trial received greater geriatric medicine expertise than would typically be found outside of health maintenance organizations (Stuck et al. 1998). This explanation seems unlikely but does suggest that better characterization of the process of care in trials of complex interventions is essential for interpretation and implementation of findings (Beck and Stuck 1996).
Specialist services for patients suffering stroke have been evaluated in small randomized controlled trials and appear to be valuable, reducing mortality and institutionalization by about 20 per cent at 12 months (Stroke Unit Trialists’ Collaboration 1999). However, these trials showed considerable statistical heterogeneity when meta-analysed which may be explained by differences in the precise nature of the stroke services or the types of patients included in the trials (Ebrahim and Harwood 1999). Although this evidence of benefit of organized stroke care is widely believed to be true, it is quite possible that the pooled evidence is misleading as it is derived from small, often poorly designed trials. In such circumstances it would be logical to perform a large, uniform protocol, multicentre trial to determine which patients benefit from organized stroke services and at what cost. Specialist services for fractured neck of femur patients (orthogeriatric services) have also been examined in systematic reviews of small randomized controlled trials; pooled effects show a non-significant trend towards improved survival and less dependence. However, the trials are very heterogeneous in terms of the interventions used and do not provide secure evidence for or against such services (Cameron et al. 1999).
Hospital-at-home and day hospitals
Various ‘hospital-at-home’ and early discharge schemes have been promoted in efforts to reduce demand on acute hospital services and several evaluations have been performed. Non-randomized evaluations were encouraging and several schemes were established in the United Kingdom, but systematic review of randomized controlled trial evaluations do not provide strong enough evidence to either support further development or discontinue existing schemes for elderly medical patients, patients who have had elective surgery, or those with a terminal illness (Shepperd and Iliffe 1999). A more recent trial has reported reductions in length of stay with associated cost-savings but no differences in health outcomes (Jones et al. 1999; Wilson et al. 1999).
Day hospitals that attempt to facilitate early discharge or to manage patients as an alternative to hospital admission show no survival or disability benefits when compared with comprehensive geriatric inpatient care and costs may be higher (Forster et al. 1999a,b).
Both comprehensive geriatric services and specialist hospital services are not well supported by available evidence and practice varies widely between countries. It is unlikely that evidence from trials of these complex interventions will be transferable from one health care system to another owing to the differing age distributions and comorbidities of patients and the selection effects introduced by different systems of social and health care. As the costs of investing in such services are high, evaluation of their cost-effectiveness should be a high priority.
Health services for mentally ill people
Unlike geriatric services, health services for elderly people suffering from functional mental illnesses (largely depressive illness) and organic brain syndromes (largely dementia) have followed the general trend of moving out of the psychiatric hospitals and into the community. In the United Kingdom, considerable energy and vision have been used to create a new specialty of psychogeriatrics as a direct response to the needs of an ageing population with an increased burden of dementia (Arie and Jolley 1998). Old age psychiatry services follow different models ranging from intergrated with geriatric medicine, separate old age services, or linked with adult psychiatry. Most services adopt common principles of attempting to be community orientated, comprehensive, available, and flexible (Lindesay 1996). Most have a common core of services including an acute assessment unit, a rehabilitation unit, a day hospital, and community liaison teams. As with hospital services for physically ill people, convincing evidence to support particular styles of practice in terms of costs and outcomes is not available. Current evaluations have focused on examining the processes of care and a disproportionate amount of research endeavour in the specialty is now devoted to drug trials rather than organization of service trials.
Most rehabilitation received by elderly people is provided in the context of their acute admission to hospital and in day hospitals. The randomized controlled trials referred to above generally included some component of rehabilitation. This is often essential in the management of acute illness because of rapid deterioration in physical capacity. However, much less attention is given to rehabilitation outside hospital despite the importance to patients of rehabilitation in their own environments rather than those of the hospital (Young et al. 1999).
In developing countries that lack investment in hospital services for elderly people, community-based rehabilitation offers an alternative model that requires urgent evaluation. Establishing rehabilitation services outside hospitals is feasible and services are certainly more accessible and are probably cheaper for patients but probably not for health service providers (Jitapunkul et al. 1998). Services given in people’s homes may be more effective in reducing handicap, optimal use of aids and appliances, and in ensuring that prevention of future problems is tackled. For stroke patients not admitted to hospital, community occupational therapy reduces disability and handicap (Walker et al. 1999). Specific issues that need study are the role of families in the process of rehabilitation, local low-cost production of disability aids and appliances, and education and training in rehabilitation for primary care teams. While there are good reasons for expecting community rehabilitation to have promise, only through systematic study of its costs and benefits will sound policy be developed.
Long-term care is often equated with institutional care rather than a more comprehensive definition of the care required by people to permit them to achieve their potential and maintain abilities in the face of chronic and often progressive disability (Kane and Kane 1991). In the United Kingdom, many long-term care institutions arose out of Poor Law workhouses that were originally set up to ensure that the destitute of every parish had somewhere to go (Townsend 1962). The level of provision had to be worse than that existing outside the workhouse to deter people from entering the workhouse. Standards of care in long-term care institutions are among the major concerns of elderly people and their carers (Counsel and Care 1991; Avebury 1996) (Table 7).
Table 7 Age and prevalence of institutionalization in different countries
Costs of long-term care
The issue of who pays for long-term care is important as protection of health and social services budgets may not result in the optimal use of resources or good standards of care. Community care reforms have been introduced in many countries as a response to growing public concern over the costs of long-term care (Fig. 10). Projections suggest that long-term care costs will outstrip other health sector costs owing to the very large number of very elderly people who lack support (Laing et al. 1991). However, there is greater than fourfold variation between countries in the proportion of the elderly population aged 75+ receiving long-term institutional care which demonstrates that substitution of alternatives to institutions is feasible (Table 7). In all countries, the increasing trend of institutionalization with older age is apparent. In developing countries, there is a growth in institutions for elderly people and this is most marked in large towns and among more affluent sections of society. A move towards institutional care, rather than care at home, is likely to consume more resources and result in upward cost pressures and aspirations for more institutional care among less affluent people in the future. Unfortunately, much urban planning has ignored the needs of older people and is designed inappropriately (e.g. high-rise public housing), and is suited to small nuclear families rather than larger extended families.
The relationship between severity of dependency and cost of care is relatively flat for institutional care but increases approximately exponentially for care at home (Evans 1993b). Anyone can be looked after at home if sufficient resources are available; packages costing over £100 000 (US$165 000) a year have been set up for children and young adults with learning difficulties. Promoting care at home is a desirable policy but there comes a point where, within a cash limited budget, provision of home care for an individual becomes unethical if it is accepted that the ethical perspective of purchasing is to achieve the greatest good for the greatest number.
Need for long-term care
The future need for long-term care will be determined by demographic trends, the strong relationship between age and disability, the extent to which community care permits people to stay at home, and the willingness of relatives to provide continued unpaid help. Most of these factors will tend to result in more rather than less need for long-term institutional care. In many countries, current levels of provision will be inadequate to meet future needs. Ignoring the problem is likely to have predictable consequences for waiting lists and admission of acutely ill elderly people as beds become clogged with people waiting for some alternative to an acute hospital bed.
Economic evaluations of private versus public sector residential homes suggest that the private sector is cheaper (Laing et al. 1991). Making these judgements is difficult and it is all too easy to make biased comparisons (Challis 1992). Variation in the costs of care depend on the following factors: levels of dependency, the number of residents and number of short-stay admissions, nursing qualified staff and supervisory staff, good physical standards, especially the proportion of single rooms, and local socio-economic status (Judge 1986). The cost advantages of private sector homes are probably attributable to the following points: small business enterprise, low rate of return on capital investment acceptable, lower wage rates, dependency may be managed more efficiently (i.e. with fewer staff), and a lower proportion of single rooms.
Evaluation of long-term care
Institutional care has had a chequered history, limiting its acceptability among many policy-makers. Goffman’s early work was influential in raising public awareness of the inadequacies of institutions (Goffman 1961). He defined the total institution as working for itself rather than its inmates with characteristics of group rather than individual treatment, routines and depersonalization, and emotional distance between staff and residents. Consequently, most current philosophies of care have emphasized the need to avoid use of institutions, and if this is not possible, to defend the autonomy of individuals, promote choice, and the use of person-centred styles of care.
The majority of people who require and currently use long-term care are demented and need 24 h supervision. This has led to attempts to evaluate outcomes in a partial way, considering reduction of carer stress as a major outcome of long-term care, together with maintenance of dignity and provision of personal care as additional important aims of services (Melzer et al. 1992). While maintaining ‘quality of life’ is an obvious outcome of good long-term care, most of the indicators used in evaluations are insensitive to variation in the quantity and quality of care. Examining changes in these quality of life measures is complicated by deaths; the most severely ill die ensuring that improvements in survivors’ abilities will be found. Satisfaction among relatives and staff may be far more important outcomes to consider.
Two randomized controlled trials have been conducted which aimed to compare care provided in purpose-built NHS nursing homes with more typical care in NHS long-stay wards. These trials illustrate the problems of applying ‘objective’ standardized measures of outcome. The first study randomized 464 residents and used the Crichton Royal Behavioural Rating Scale, a psychiatric assessment schedule, semistructured interviews of quality of care and self-rated health, life satisfaction index, and survival as outcomes (Bond et al. 1989a,b; Bond and Bond 1990). By 1 year, just under half the residents were dead and no differences were found in the outcomes measured. Non-participant observation demonstrated that there were differences in more subtle aspects of life in the two settings: calling staff by first names, later waking times, more positive feelings towards staff, residents’ interaction, choice, and a more flexible day were hallmarks of NHS nursing home care. These findings were confirmed in the second, smaller trial (Clarke and Bowling 1990; Bowling et al. 1991).
Both trials demonstrated better processes of care and outcomes in terms of the residents’ and staffs’ everyday life in NHS nursing homes and that conventional outcomes (i.e. disability and behaviour scales) were not very useful in this context. Unfortunately, it is difficult to generalize from these trials to the wider world, and it would be wrong to conclude that any nursing home is better than a hospital ward. For health service commissioners, the message is clear—good quality care can be provided in state-run sectors and non-participant observation offers a powerful means of assessing standards of care.
The joint dangers of attempting to purchase sufficient low-quality long-term care places or of purchasing insufficient numbers of high-quality institutional places are self-evident. Neither option will do. The object of commissioning in this area is to ensure that both amount and standards are considered.
Donabedian (1989) has suggested that in evaluating health care for elderly people it is sensible to focus on processes of care rather than improvements in health (i.e. outcomes) because for elderly people, the cardinal principles of comprehensiveness, co-ordination, and continuity are more important than the outcomes themselves. Comprehensiveness means that care is not fragmented into social and health domains and implies an individual and holistic approach. Co-ordination is largely concerned with ensuring that everyone knows about and accepts responsibility for management and implies good communication. Continuity of care is achieved by maintaining the same staff and keeping the resident in the same place. Therefore, in the absence of efficacy studies (which are unlikely to be mounted because of cost and problems in defining sensible outcomes), it is appropriate to consider recommended standards in terms of their contributions to comprehensiveness, co-ordination, and continuity as goals in their own right.
Social support and informal carers
Observational epidemiological studies have demonstrated that social support is associated with improved survival, avoidance of institutionalization, reduced disability, and improved quality of life (Bowling and Browne 1991; Steinbach 1992; Mendes de Leon et al. 1999). The way in which social care resources are allocated shows considerable unexplainable variation. For example, in Sweden differences in allocation of resources such as home helps, sheltered housing, and institutional care places were found which could not be explained by differences in need for services (Lagergren and Johansson 1998). Evidence from controlled trials of the impact of social services schemes to provide social support are fairly rare and tend to be small scale. For example, a trial of outreach management of elderly people discharged from hospital failed to find any difference in quality of life or functional performance (Curtis et al. 1998). Similarly, social support interventions with stroke patients failed to demonstrate any effects (Friedland and McColl 1992). By contrast, an Italian trial of integrated social and medical care with case management did find reduced institutionalization and functional decline (Bernabei et al. 1998). Interpretation of small, often methodologically unsound, and underpowered negative trials is beset with problems. Much more work is required to develop both theoretically sound interventions and better methods of evaluation of these complex interventions.
Informal unpaid carers of frail elderly people provide a contribution to maintaining them at home that dwarfs state provision in any country (Travers 1996). It is estimated that there are at least 6 million carers in the United Kingdom (i.e. about one in ten of the population) (Smith 1994). Many caregivers are themselves elderly people. The causes of caregiver stress and breakdown have been well documented and include the type of care provided, living with the person cared for, coping with combined mental and physical problems, and ill health of the carer. Respite care (i.e. brief periods of admission to institutional care) does not appear to have any consistent effects or to moderate burdens (McNally et al. 1999). Other methods of carer support have been evaluated in specific contexts, particularly dementia, showing that an accurate diagnosis may help, as may family and individual counselling, and training (Schofield et al. 1998; Mittleman et al. 1996; Brodaty et al. 1997; Bamford et al. 1998).
The common belief—’families do not want to care for elderly relatives’—is not supported by European surveys which show that families do want to provide support to their parents and grandparents (McGlone and Cronin 1994). Willingness to give care is not related to employment status of women (Robison et al. 1995) but, perhaps inevitably, unmarried daughters tend to provide more direct care than married or divorced daughters (Brody et al. 1994) or sons. Informal carers, unlike formal sector services, are capable of providing very complex caring arrangements and are more flexible (Vetter et al. 1992). Therfore informal carers may substitute for inadequate formal sector provision or may act as a bridge, bringing elderly people into contact with services (Logan and Spitze 1994).
Health and social policy implications of population ageing
The WHO has highlighted the importance of ‘active ageing’—the process of optimizing opportunities for physical, social, and mental well being throughout the life course in order to extend healthy life expectancy—as its major policy (Kalache 1999). Recognition of the contributions made by elderly people themselves to families, communities, and national wealth is an essential step in improving their situation and avoiding ageism. For frail sick elderly people, the major policies of care in the community, avoidance of long-term institutional care, and support for informal carers are essential elements. In developing countries, policy based on the assumptions that families will cope without support and that a tradition of respect for elders will be maintained seem doomed to failure. Greater efforts are needed to understand the best ways of maintaining health in old age, equipping primary care teams with appropriate skills in the management of health problems of elderly people and building support for families. Policy in the areas of education, employment, transport, social welfare, housing, and environment should be examined for its potential impact on elderly people so that the general goals of maintaining older people in their own homes and improving their health expectancy can be achieved.
Age Concern (1999). Turning your backs on us. Age Concern, London.
Anonymous (1999). Statistical yearbook 1999 for the Federal Republic of Germany. Statistisches Bundesamt, Wiesbaden.
Arie, T. and Jolley, D. (1998). Psychogeriatric services. In Brocklehurst’s textbook of geriatric medicine and gerontology (ed. R. Tallis, H.M. Fillit, and J.C. Brocklehurst) (5th edn). Churchill Livingstone, Edinburgh.
Avebury, K. (1996). A Better home life. A code of good practice for residential and nursing home care. Report of an Advisory Group. Centre for Policy on Ageing, London.
Balarajan, R., Yuen, P., and Raleigh, V.S. (1989). Ethnic differences in general practitioner consultations. British Medical Journal, 299, 958–60.
Bamford, C., Gregson, B.A., Farrow, G., Buck, D., Dowswell, T. and McNamee, P.B.J. (1998). Mental and physical fraility in older people: the costs and benefits of informal care. Ageing and Society, 18, 317–54.
Beck, J.C. and Stuck, A. (1996). Preventing disability. Beyond the black box. Journal of the American Medical Association, 276, 1756–7.
Bernabei, R., Landi, F., Gambassi, G., et al. (1998). Randomised trial of impact of model of integrated care and case management for older people living in the community. British Medical Journal, 316, 1348–51.
Bond, S. and Bond, J. (1990). Outcomes of care within a multiple case study in the evaluation of the experimental NHS nursing homes. Age and Ageing, 19, 11–18.
Bond, J., Gregson, B.A., Atkinson, A., and Newell, D.J. (1989a). The implementation of a randomised controlled trial in the evaluation of the experimental NHS nursing homes. Age and Ageing, 18, 96–102.
Bond, J., Gregson, B.A., and Atkinson, A. (1989b). Measurement of outcomes within a multicentred randomised controlled trial in the evaluation of NHS nursing homes. Age and Ageing, 18, 292–302.
Bone, M., Bebbington, A., Jagger, C., Morgan, K., and Nicolaas G. (1995). Health expectancy and its uses. HMSO, London.
Bowling, A. (1999). Ageism in cardiology. British Medical Journal, 319, 1353–5.
Bowling, A. and Browne, P.D. (1991). Social networks, health, and emotional well-being among the oldest old in London. Journal of Gerontology, 46, S20–S32.
Bowling, A., Formby, J., Grant, K., and Ebrahim, S. (1991). A randomized controlled trial of nursing home and long-stay geriatric ward care for elderly people. Age and Ageing, 20, 316–24.
Brazier, J., Jones, N., and Kind, P. (1993). Testing the validity of the Euroqol and comparing it with the SF-36 health survey questionnaire. Quality of Life Research, 2, 169–80.
Brodaty, H., Gresham, B., and Luscombe, G. (1997). The Prince Henry Hospital dementia caregivers’ training programme. International Journal of Geriatric Psychiatry, 12, 183–92.
Brody, E.M., Litvin, S.J., Albert, S.M., and Hoffman, C.J. (1994). Marital status of daughters and patterns of parent care. Journal of Gerontology, 49, S95–103.
Buchner, D.M. (1997). Preserving mobility in older adults. Western Journal of Medicine, 167, 258–64.
Bunker, J.P. (1994). Medicine’s core values. Medical care does add to life expectancy. British Medical Journal, 309, 1657.
Burton, L.C., Paglia, M.J., German, P.S., Shapiro, S., and Damiano, A.M. (1995a). The effect among older persons of a general preventive visit on three health behaviors: smoking, excessive alcohol drinking, and sedentary lifestyle. The Medicare Preventive Services Research Team. Preventive Medicine, 24, 492–7.
Burton, L.C., Steinwachs, D.M., German, P.S., et al. (1995b). Preventive services for the elderly: would coverage affect utilization and costs under Medicare? American Journal of Public Health, 85, 387–91.
Cameron, I., Finnegan, T., Madhok, R., Langhorne, P., and Handoll, H. (1999). Co-ordinated multidisciplinary approaches for inpatient rehabilitation of older patients with proximal femoral fractures (Cochrane Review). In The Cochrane Library, Issue 4. Update Software, Oxford.
Centers for Disease Control and Prevention (1997). Missed opportunities for pneumococcal and influenza vaccination of Medicare pneumonia inpatients—12 western states 1995. Journal of the American Medical Association, 278, 1307–8.
Challis, D. (1992). Providing alternatives to long-stay hospital care for frail elderly patients: is it cost-effective? International Journal of Geriatric Psychiatry, 7, 773–81.
Challis, D. (1996). Community care. In Epidemiology in old age (ed. S. Ebrahim and A. Kalache). BMJ Publishing, London.
Charlton, J. and Murphy, M. (1997). The health of adult Britain 1841–1994. HMSO, London.
Chenet, L., McKee, M., Fulop, N., et al. (1996). Changing life expectancy in central Europe: is there a single reason? Journal of Public Health Medicine, 18, 329–36.
Chenet, L., McKee, M., Leon, D., Shkolnikov, V., and Vassin, S. (1998). Alcohol and cardiovascular mortality in Moscow; new evidence of a causal association . Journal of Epidemiology and Community Health, 52, 772–4.
Clarke, P. and Bowling, A. (1990). Quality of every day life in long-stay institutions for the elderly: an observational study of long stay hospital and nursing home care. Social Science and Medicine, 30, 1201–10.
Cole, M.G. (1998). Impact of geriatric home screening services on mental state: a systematic review. International Psychogeriatrics, 10, 97–102.
Connor, J., Rodgers, A., and Priest, P. (1999). Randomised studies of income supplementation: a lost opportunity to assess health outcomes. Journal of Epidemiology and Community Health, 53, 725–30.
Counsel and Care (1991). Not such private places. Counsel and Care, London.
Crawford, V., Beringer, T.R., and Stout, R.W. (1999). Comparison of residential and nursing home care before and after the 1993 community care policy. British Medical Journal, 318, 366.
Curtis, J.L., Millman, E.J., Struening, E.L., and D’Ercole, A. (1998). Does outreach case management improve patients’ quality of life? Psychiatric Services, 49, 351–4.
Department of Health (1998). Our healthier nation: a contract for health. HMSO, London.
Devore, D. (1998). 1996 disability survey. Follow up to the family resources survey. Technical report, Office for National Statistics, London.
Doll, R., Peto, R., Wheatley, K., Gray, R., and Sutherland, I. (1994). Mortality in relation to smoking: 40 years’ observations on male British doctors. British Medical Journal, 309, 901–11.
Donabedian, A. (1989). Quality of care and the health care needs of the elderly patient. In Care of the elderly patient. Policy issues and research opportunities (ed. J.A. Barondess, D.E. Rogers, and K.N. Lohr), pp. 3–13. National Academy Press, Washington, DC.
Downie, R.S., Fyfe, C., and Tannahill, A. (1990). Health promotion. Models and values. Oxford University Press.
Dowse, G.K., Gareeboo, H., Alberti, K.G., et al. (1995). Changes in population cholesterol concentrations and other cardiovascular risk factor levels after five years of the non-communicable disease intervention programme in Mauritius. Mauritius Non-communicable Disease Study Group. British Medical Journal, 311, 1255–9.
Drever, F. and Whitehead, M. (1997). Health inequalities. Decennial supplement DS No. 15, HMSO, Office for National Statistics, London.
Ebrahim, S. (1995). Clinical and public health perspectives and applications of health-related quality of life measurement. Social Science and Medicine, 41, 1383–94.
Ebrahim, S. (1996a). Ethnic elders. British Medical Journal, 313, 610–13.
Ebrahim, S. (1996b). Migration and ethnicity. In Epidemiology in old age (ed. S. Ebrahim and A. Kalache), pp. 201–9. BMJ Publications, London.
Ebrahim, S. (1998). Detection, adherence and control of hypertension for the prevention of stroke: a systematic review. Health Technology Assessment, 2, 1–78.
Ebrahim, S. (1999). Demographic shifts and medical training. British Medical Journal, 319, 1358–60.
Ebrahim, S. and Davey Smith, G. (1996). Health promotion in older people for the prevention of coronary heart disease and stroke. Health Education Authority, London.
Ebrahim, S. and Harwood, R.H. (1999). Stroke: epidemiology, evidence and clinical practice. Oxford University Press.
Ebrahim, S. and Davey Smith, G. (1997). Systematic review of randomised controlled trials of multiple risk factor interventions for preventing coronary heart disease. British Medical Journal, 314, 1666–74.
Ebrahim, S., Hedley, R., and Sheldon, M. (1984). Low levels of ill health among elderly non-consulters in general practice. British Medical Journal, 289, 1273–5.
Ebrahim, S., Brittis, S., and Wu, A. (1991). The valuation of states of illhealth: the impact of age and disability. Age and Ageing, 20, 37–40.
Ebrahim, S., Thompson, P.W., Baskaran, V., and Evans, K. (1997). Randomized placebo-controlled trial of brisk walking in the prevention of postmenopausal osteoporosis. Age and Ageing, 26, 253–60.
Egger, M., Davey Smith, G., Schneider, M., and Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. British Medical Journal, 315, 629–34.
Egger, M., Schneider, M., and Davey Smith, G. (1998). Spurious precision? Meta-analysis of observational studies. British Medical Journal, 316, 140–4.
Evans, J.G. (1983). Integration of geriatric with general medical services in Newcastle. Lancet, i, 1430–3.
Evans, J.G. (1990). How are the elderly different? In Improving the health of older people: a world view (ed. R.L. Kane, J.G. Evans, and D. Macfadyen), pp. 50–68. Oxford University Press.
Evans, J.G. (1993a). Hypothesis: Healthy Active Life Expectancy (HALE) as an index of effectiveness of health and social services for elderly people. Age and Ageing, 22, 297–301.
Evans, J.G. (1993b). Institutional care and elderly people. British Medical Journal, 306, 806–7.
Evans, J.G. (1997). A correct compassion. The medical response to an ageing society. Journal of the Royal College of Physicians of London, 31, 674–8.
Evans, J.G. and Graham, J.M. (1984). Medical care of the elderly: five years on. Journal of the Royal College of Physicians of London, 18, 18–20.
Fahs, M.C., Mandelblatt, J., Schechter, C., and Muller, C. (1992). Cost effectiveness of cervical cancer screening for the elderly. Annals of Internal Medicine, 117, 520–7.
Fentem, P.H. (1994). ABC of sports medicine. Benefits of exercise in health and disease. British Medical Journal, 308, 1291–5.
Fine, M.J., Smith, M.A., and Carson, C.A. (1994). Efficacy of pneumococcal vaccination in adults: a meta analysis of randomized controlled clinical trials. Archives of Internal Medicine, 154, 2666–77.
Fletcher, A. (1996). Breast cancer. In Epidemiology in old age (ed. S. Ebrahim and A. Kalache), pp. 317–23. BMJ Publishing, London.
Fletcher, G. (1999). Physical inactivity as a risk factor for cardiovascular disease. American Journal of Medicine, 107, 10S–11S.
Forster, A., Young, J., Langhorne, P. and the Day Hospital Group (1999a). Medical day hospital care for the elderly versus alternative forms of care (Cochrane Review). In The Cochrane Library, Issue 4. Update Software, Oxford.
Forster, A., Young, J., Langhorne, P. and the Day Hospital Group (1999b). Systematic review of day hospital care for elderly people. British Medical Journal, 318, 837–41.
Friedland, J.F. and McColl, M. (1992). Social support intervention after stroke: results of a randomized trial. Archives of Physical Medical Rehabilitation, 73, 573–81.
Fries, J.F. (1980). Aging, natural death, and the compression of morbidity. New England Journal of Medicine, 303, 130–5.
Fries, J.F. (1989). The compression of morbidity: near or far? Milbank Quarterly, 67, 208–32.
Fries, J.F., Green, L.W., and Levine, S. (1989). Health promotion and the compression of morbidity. Lancet, 1, 481–3.
German, P.S., Burton, L.C., Shapiro, S., et al. (1995). Extended coverage for preventive services for the elderly: response and results in a demonstration population. American Journal of Public Health, 85, 379–86.
P>Gillespie, L., Gillespie, W., Cumming, R., Lamb, S., and Rowe, B.H. (1999). Interventions for preventing falls in the elderly (Cochrane Review). In The Cochrane Library, Issue 4. Update Software, Oxford.
Goffman, E. (1961). Asylums. Penguin, London.
Golini, A. (1997). Demographic trends and aging in Europe. Genus, 53, 33–74.
Gonner, E.C.K. (1913). The population of England in the eighteenth century. Journal of the Royal Statistical Society, 58, 261–96.
Gray, J.A.M., Bassey, E.J., and Young, A. (1985). The risks of inactivity. In Prevention of disease in the elderly (ed. J.A.M. Gray), pp. 78–94. Churchill Livingstone, Edinburgh.
Gross, P.A., Hermongenes, A.W., Sachs, H.S., Lau, J., and Levandowski, R.A. (1995). The efficacy of influenza vaccine in elderly persons: a meta-analysis and review of the literature. Annals of Internal Medicine, 123, 518–27.
Gueyffier, F., Bulpitt, C.J., Boissel, J.P., et al. and for the INDIANA Group (1999). Antihypertensive drugs in very old people: a sub-group meta-analysis of randomised controlled trials. Lancet, 353, 793–6.
Harwood, R., Rogers, A., Dickinson, E., and Ebrahim, S. (1994a). Measuring handicap: the London Handicap Scale, a new outcome measure for chronic disease. Quality in Health Care, 3, 11–16.
Harwood, R., Jitapunkul, S., Dickinson, E., and Ebrahim, S. (1994b). Measuring handicap: motives, methods and a model. Quality in Health Care, 3, 53–7.
Haussler, B., Hempel, E., and Reschke, P. (1995). Changes in life expectancy and mortality in East Germany after reunification (1989–1992). Gesundheitswesen, 57, 365–72.
Hillsdon, M. and Thorogood, M. (1996). A systematic review of physical activity promotion strategies. British Journal of Sports Medicine, 30, 84–9.
Hillsdon, M., Thorogood, M., Anstiss, T., and Morris, J. (1995). Randomised controlled trials of physical activity promotion in free living populations: a review. Journal of Epidemiology and Community Health, 49, 448–53.
Himsworth, R.L. and Goldacre, M.J. (1999). Does time spent in hospital in the final 15 years of life increase with age at death? A population based study. British Medical Journal, 319, 1338–9.
Hooper, L., Ness, A., Higgins, J.P.T., Moore, T., and Ebrahim, S. (1999). GISSI–Prevenzione trial. Lancet, 354, 1557–9.
Hooper, L., Summerbell, C.D., Higgins, J.P.T., et al. (2001). Dietary fat intake and prevention of cardiovascular disease: systematic review. British Medical Journal, 322, 757–63.
Hoxie, R.E. and Rubenstein, L.Z. (1994). Are older pedestrians allowed enough time to cross intersections safely? Journal of the American Geriatrics Society, 42, 241–4.
Iliffe, S., Haines, A., Gallivan, S., Booroff, A., Goldenberg, E., and Morgan, P. (1991a). Assessment of elderly people in general practice. 1. Social circumstances and mental state. British Journal of General Practice, 41, 9–12.
Iliffe, S., Haines, A., Gallivan, S., Booroff, A., Goldenberg, E., and Morgan, P. (1991b). Assessment of elderly people in general practice. 2. Functional abilities and medical problems. British Journal of General Practice, 41, 13–15.
Insua, J.T., Sacks, H.S., Lau, T.S., et al. (1994). Drug treatment of hypertension in the elderly: a meta-analysis. Annals of Internal Medicine, 121, 355–62.
Jenkins, R.D. (1999). Gerontology on the world wide web. Reviews in Clinical Gerontology, 9, 289.
Jenkinson, C., Gray, A., Doll, H., Lawrence, K., Keoghane, S., and Layte, R. (1997). Evaluation of index and profile measures of health status in a randomized controlled trial. Comparison of the Medical Outcomes Study 36-Item Short Form Health Survey, EuroQol, and disease specific measures. Medical Care, 35, 1109–18.
Jitapunkul, S., Bunnag, S., and Ebrahim, S. (1998). Effectiveness and cost analysis of community-based rehabilitation service in Bangkok. Journal of the Medical Association of Thailand, 81, 572–8.
Jones, J., Wilson, A., Parker, H., et al. (1999). Economic evaluation of hospital at home versus hospital care: cost minimization analysis of data from randomised controlled trial. British Medical Journal, 319, 1547–50.
Judge, K. (1986). Residential care for the elderly: purposes and resources. In Residential homes for elderly people (ed. K. Judge and I. Sinclair), pp. 5–20. HMSO, London.
Judge, K. (2000). Testing evaluation to the limits: the case of English Health Action Zones. Journal of Health Services Research and Policy, 5, 3–5.
Kalache, A. (1999). Active ageing makes the difference. Bulletin of the WHO, 77, 299–305.
Kane, R.L. and Kane, R.A. (1991). Special needs of dependent elderly persons. In Oxford textbook of public health medicine (ed. W.W. Holland, R. Detels, and G. Knox) (2nd edn), Vol. 3, pp. 509–21. Oxford University Press, New York.
Khaw, K.T. (1999). How many, how old, how soon? British Medical Journal, 319, 1350–2.
Kinsella, K.G. (1996). Demographic aspects. In Epidemiology in old age (ed. S. Ebrahim and A. Kalache), pp. 32–40. BMJ Publishing, London.
Kirkwood, T. (1999). The time of our lives: the science of human ageing. Weidenfield and Nicolson, London.
Koivula, I., Sten, M., Leinonen, M., and Makela, P.H. (1997). Clinical efficacy of pneumococcal vaccine in the elderly: a randomized, single-blind population-based trial. American Journal of Medicine, 103, 281–90.
Lagergren, M. and Johansson, P.-A. (1998). Are there differences in standard of care for the elderly? A comparative study of assistance decisions in Stockholm. Scandinavian Journal of Social Welfare, 7, 340–9.
Laing, W., Hall, M., and Lumley, P. (1991). Agenda for health. The challenges of ageing. Association of the British Pharmaceutical Industry, London.
Leon, D.A., Chenet, L., Shkolnikov, et al. (1997). Huge variation in Russian mortality rates 1984–94: artefact, alcohol, or what? Lancet, 350, 383–8.
Lindesay, J. (1996). Health service use in mental illness. In Epidemiology in old age (ed. S. Ebrahim and A. Kalache), pp. 96–105. BMJ Publishing, London.
Logan, J.R. and Spitze, G. (1994). Informal support and the use of formal services by older Americans. Journal of Gerontology, 49, S25–S34
McCormick, A., Fleming, D., and Charlton, J. (1995). Morbidity statistics from general practice. Fourth national study 1991–1992. HMSO, London.
McGlone, F. and Cronin, N. (1994). A crisis in care? The future of family and state care for older people in the European Union. Family Policy Studies Centre, London.
McNally, S., Ben-Shlomo, Y., and Newman, S. (1999). The effects of respite care on informal carers’ well-being: a systematic review. Disability and Rehabilitation, 21, 1–14.
Mahoney, F.I. and Barthel, D.W. (1965). Functional evaluation: the Barthel Index. Maryland State Medical Journal, 14, 61–5.
Manton, K.G. (1980). Sex and race specific mortality differentials in multiple cause of death data. Gerontologist, 20, 480–93.
Manton, K.G. (1982). Changing concepts of morbidity and mortality in the elderly population. Milbank Memorial Fund Quarterly—Health and Society, 60, 183–244.
Manton, K.G. and Stallard, E. (1991). Cross-sectional estimates of active life expectancy for the U.S. elderly and oldest-old populations. Journal of Gerontology, 46, S170–S182.
Manton, K.G., Corder, L., and Stallard, E. (1997). Chronic disability trends in elderly United States populations: 1982–1994. Proceedings of National Academy of Science of the United States of America, 94, 2593–8.
Marmot, M., Adelstein, A.M., and Bulusu, L. (1984). Lessons from the study of immigrant mortality. Lancet, i, 1455–7.
Martin, J., Meltzer, H., and Elliot, D. (1987). The prevalence of disability among adults. OPCS surveys of disability in Great Britain. HMSO, London.
Mathers, C.D. and Robine, J.M. (1997). How good is Sullivan’s method for monitoring changes in population health expectancies? Journal of Epidemiology and Community Health, 51, 80–6.
Melzer, D., Hopkins, S., Pencheon, D., Brayne, C., and Williams, R. (1992). Epidemiologically based needs assessment: dementia. NHS Management Executive, London.
Mendes de Leon, C.F., Beckett, L.A., Fillenbaum, G.G., Brock, D.B., Branch, L.G., Evans, D.A., and Berkman, L.F. (1997). Black–white differences in risk of becoming disabled and recovering from disability in old age: a longitudinal analysis of two EPESE populations. American Journal of Epidemiology, 145, 488–97.
Mendes de Leon, C.F., Glass, T.A., Beckett, L.A., Seeman, T.E., Evans, D.A., and Berkman, L.F. (1999). Social networks and disability transitions across eight intervals of yearly data in the New Haven EPESE. Journal of Gerontology. Series B, Psychological Sciences and Social Sciences, 54, S162–72.
Mittleman, M.S., Ferris, S.H., Shulman, E., Steinberg, G., and Levin, B. (1996). A family intervention to delay nursing home placement of patients with Alzheimer’s Disease. Journal of the American Medical Association, 276, 1725–31.
Mullooly, J.P., Bennett, M.D., Hornbrook, M.C., et al. (1994). Influenza vaccination programs for elderly persons: cost-effectiveness in a health maintenance organization. Annals of Internal Medicine, 121, 947–52.
Mulrow, C.D., Cornell, J.A., Herrera, C.R., Kadri, A., Farnett, L., and Aguilar, C. (1994). Hypertension in the elderly. Implications and generalizability of randomized trials. Journal of the American Medical Association, 272, 1932–8.
National Center for Health Statistics (1999). Health, United States 1999. National Center for Health Statistics, Hyattsville, MD.
Nichol, K.L., Margolis, K.L., Wuorenma, J., and Von Sternberg, T. (1994). The efficacy and cost effectiveness of vaccination against influenza among elderly persons living in the community. New England Journal of Medicine, 331, 778–84.
Nicholson, K.G. (1993). Immunisation against influenza among people aged over 65 living at home in Leicestershire during winter 1991–2. British Medical Journal, 306, 974–6.
NIH Consensus Development Panel on Physical Activity and Cardiovascular Health (1996). Physical activity and cardiovascular health. Journal of the American Medical Association, 276, 241–6.
Nikolaus, T., Specht-Leible, N., Back, M., Oster, P., and Schlierf, G. (1999). A randomised trial of comprehensive geriatric assessment and home intervention in the care of hospitalised patients. Age and Ageing, 28, 543–50.
Norman, A. (1985). Triple jeopardy: growing old in a second homeland. Centre for Policy on Ageing, London.
Normand, C. (1998). Ten popular health economic fallacies. Journal of Public Health Medicine, 20, 129–32.
Nouri, F. and Lincoln, N.B. (1987). An extended activities of daily living scale for stroke patients. Clinical Rehabilitation, 1, 301–5.
Office of Population Census and Surveys (1991). Census for Great Britain 1991. HMSO, London.
Olshansky, S.J., Rudberg, M.A., Carnes, B.A., Cassel, C.K., and Brody, J.A. (1991). Trading off longer life for worse health: the expansion of morbidity hypothesis. Journal of Aging and Health, 3, 194–216.
Ortqvist, A., Hedlund, J., Burman, L.A., et al. (1998). Randomised trial of 23-valent pneumococcal capsular polysaccharide vaccine in prevention of pneumonia in middle-aged and elderly people. Swedish Pneumococcal Vaccination Study Group. Lancet, 351, 399–403.
Potter, J., Stott, D.J., Roberts, M.A., et al. (1997). Influenza vaccination of health care workers in long-term-care hospitals reduces the mortality of elderly patients. Journal of Infectious Diseases, 175, 1–6.
Province, M.A., Hadley, E.C., Hornbrook, M.C., et al. (1995). The effects of exercise on falls in elderly patients. A preplanned meta-analysis of the FICSIT Trials. Frailty and Injuries: Cooperative Studies of Intervention Techniques. Journal of the American Medical Association, 273, 1341–7.
Razzell, P. (1998). The condundrum of eighteenth century English population growth. Social History of Medicine, 11, 469–500.
Reuben, D.B., Borok, G.M., Wolde-Tsadik, G., et al. (1995). A randomized trial of comprehensive geriatric assessment in the care of hospitalized patients. New England Journal of Medicine, 332, 1345–50.
Rippe, J.M. and Hess, S. (1998). The role of physical activity in the prevention and management of obesity. Journal of the American Dietetic Association, 98, S31–S38.
Rivlin, M. (1999). Should age based rationing of health care be illegal? British Medical Journal, 319, 1379.
Robine, J.M. and Romieu, I. (1998). Healthy active ageing: health expectancies at age 65 in the different parts of the world. REVES Paper 318, INSERM, France: Network on Health Expectancy and the Disability Process (REVES).
Robine, J.M., Blanchet M., and Dowd, J.E. (1992). Health expectancy. First workshop of the International Healthy Life Expectancy Network (REVES). HMSO, London.
Robison, J., Moen, P., and Dempster-McClain, D. (1995). Women’s caregiving: changing profiles and pathways. Journal of Gerontology. Series B, Psychological Sciences and Social Sciences, 50, S362–S373.
Royal Commission on Long Term Care (1999). With respect to old age: long term care—rights and responsibilities, HMSO, London.
Rubenstein, L.Z. and Nasr, S.Z. (1996). Health service use in physical illness. In Epidemiology in old age (ed. S. Ebrahim and A. Kalache), pp. 106–25. BMJ Publishing, London.
Ruta, D.A., Garratt, A.M., Leng, M., Russell, I.T., and MacDonald, L.M. (1994). A new approach to the measurement of quality of life. The Patient-Generated Index. Medical Care, 32, 1109–26.
Schofield, H., Murphy, B., Herrman, H.E., Bloch, S., and Singh, B.S. (1998). Carers of people aged over 50 years with physical impairment, memory loss and dementia: a comparative study. Ageing and Society, 18, 355–69.
Shah, S., Vanclay, F., and Cooper, B. (1990). Efficiency, effectiveness and duration of stroke rehabilitation. Stroke, 21, 241–6.
Shah, S., Vanclay, F., and Cooper, B. (1991). Stroke rehabilitation: Australian patient profile and functional outcome. Journal of Clinical Epidemiology, 44, 21–8.
Shepperd, S. and Iliffe, S. (1999). Hospital-at-home versus in-patient hospital care (Cochrane Review). In The Cochrane Library, Issue 4. Update Software, Oxford.
Sisk, J.E., Moskowitz, A.J., Whang, W., et al. (1997). Cost-effectiveness of vaccination against pneumococcal bacteremia among elderly people. Journal of the American Medical Association, 278, 1333–9.
Smeeth, L. and Iliffe, S. (1998). Effectiveness of screening older people for impaired vision in community setting: systematic review of evidence from randomised controlled trials. British Medical Journal, 316, 660–3.
Smeeth, L. and Iliffe, S. (1999). Community screening for visual impairment in the elderly. In Cochrane Review. Cochrane Library, Issue 4. Update Software, Oxford.
Smith, J.P. (1994). Informal carers in Great Britain. Journal of Advanced Nursing, 20, 787.
Steinbach, U. (1992). Social networks, institutionalization, and mortality among elderly people in the United States. Journal of Gerontology, 47, S183–S190.
Stroke Unit Trialists’ Collaboration (1999). Organised inpatient (stroke unit) care for stroke (Cochrane Review). In The Cochrane Library, Issue 4. Update Software, Oxford.
Stuck, A.E., Siu, A.L., Wieland, G.D., Adams, J., and Rubenstein, L.Z. (1993). Comprehensive geriatric assessment: a meta-analysis of controlled trials. Lancet, 342, 1032–6.
Stuck, A.E., Aronow, H.U., Steiner, A., et al. (1995). A trial of annual in-home comprehensive geriatric assessments for elderly people living in the community . New England Journal of Medicine, 333, 1184–9.
Stuck, A.E., Rubenstein, L.Z., and Wieland, D. (1998). Bias in meta-analysis detected by a simple, graphical test. Asymmetry detected in funnel plot was probably due to true heterogeneity. British Medical Journal, 316, 469–1.
Thamprechavai, S., Somerville, K., Jitapunkul, S., Bunnag, S., and Ebrahim, S. (1992). Elderly bed-blockers in a Thai teaching hospital: is it a problem? Journal of the Medical Association of Thailand, 75, 418–22.
Tonks, A. (1999). Medicine must change to serve an ageing society. British Medical Journal, 319, 1450–1.
Townsend, P. (1962). The last refuge. Routledge and Kegan Paul, London.
Travers, A.F. (1996). Caring for older people. Carers. British Medical Journal, 313, 482–6.
Tseng, B.S., Marsh, D.R., Hamilton, M.T., and Booth, F.W. (1995). Strength and aerobic training attenuate muscle wasting and improve resistance to the development of disability with aging. Journal of Gerontology.Series A, Biological Sciences and Medical Sciences, 50, 113–19.
van Rossum, E., Frederiks, C.M., Philipsen, H., Portengen, K., Wiskerke, J., and Knipschild, P. (1993). Effects of preventive home visits to elderly people. British Medical Journal, 307, 27–32.
Vetter, N.J. (1999). The impact of smoking in elderly people. Reviews in Clinical Gerontology, 9, 273–80.
Vetter, N.J. and Ford, D. (1990). Smoking prevention among people aged 60 and over: a randomized controlled trial. Age and Ageing, 19, 164–8.
Vetter, N.J., Charny, M., Farrow, S., and Lewis, P.A. (1988). The Cardiff Health Survey. The relationship between smoking habits and beliefs in the elderly. Public Health, 102, 359–64.
Vetter, N.J., Lewis, P.A., and Llewellyn, L. (1992). Supporting elderly dependent people at home. British Medical Journal, 304, 1290–2.
Vita, A.J., Terry, R.B., Hubert, H.B., and Fries, J.F. (1998). Aging, health risks, and cumulative disability. New England Journal of Medicine, 338, 1035–41.
Walker, M.F., Gladman, J.R., Lincoln, N.B., Siemonsma, P., and Whiteley, T. (1999). Occupational therapy for stroke patients not admitted to hospital: a randomised controlled trial. Lancet, 354, 278–80.
Ware, J.E.J. and Gandek, B. (1998). Overview of the SF-36 Health Survey and the International Quality of Life Assessment (IQOLA) Project. Journal of Clinical Epidemiology, 51, 903–12.
Warren, M.W. (1946). Care of the chronic aged sick. Lancet, ii, 841–3.
Watkins, J., Rogers, C., and Evans, J. (1999). Implications of age-based policies for influenza immunisation. Lancet, 353, 208–9.
Weiner, J. and Illson, L.H. (1994). Private insurance options for long term care financing. In Community care: new agendas and challenges from the UK and overseas (ed. D. Challis, B. Davies, and K. Traske), pp. 264–81. University of Kent Personal Social Services Research Unit, Canterbury.
Williams, A. and Evans, J.G. (1997). The rationing debate. Rationing health care by age. British Medical Journal, 314, 820–5.
Williams, E.I. and Wallace, P. (1993). Health checks for people aged 75 and over, pp. 1–30. Occasional Paper, Royal College of General Practitioners, London.
Williamson, J., Stokoe, I.H., and Gray, S. (1964). Old people at home: their unreported needs. Lancet, i, 1117–20.
Wilson, A., Parker, H., Wynn, A., et al. (1999). Randomised controlled trial of effectiveness of Leicester hospital at home scheme compared with hospital care. British Medical Journal, 319, 1542–6.
Wolman, R.L. (1994). ABC of sports medicine. Osteoporosis and exercise. British Medical Journal, 309, 400–3.
WHO (World Health Organization) (1980). International classification of impairments, disabilities and handicaps. WHO, Geneva.
WHO (World Health Organization) (1984). The uses of epidemiology in the study of the elderly. Technical Report Series 706. WHO, Geneva.
WHO (World Health Organization) (1998). World atlas of ageing. WHO, Kobe, Japan.
WHO (World Health Organization) (1999). International classification of impairments, activities and participation. Beta version. WHO, Geneva.
Wrigley, E.A. and Schofield, R.S. (1981). The population history of England, 1547–1871. Cambridge University Press.
Young, A. (1989). There is no such thing as geriatric medicine, and its here to stay. Lancet, ii, 263–5.
Young, A. (1992). Strength and power. In Oxford textbook of geriatric medicine (ed. J.G. Evans and T.F. Williams), pp. 597–604. Oxford University Press.
Young, J., Brown, A., Forster, A., and Clare, J. (1999). An overview of rehabilitation for older people. Reviews in Clinical Gerontology, 9, 183–96.