Participants and consent procedures

All adults with intellectual disability aged 60 years and older, currently resident in any of the five boroughs, were identified from Social Services electronic databases and intellectual disability team healthcare records (current and past recipients of care who have been recorded at any time to have intellectual disability), and from all local residential and day service providers (voluntary or government sector) for adults with intellectual disability.

Informed consent was obtained from participants with intellectual disability who had capacity, but for those who lacked capacity we sought agreement from carers. We also gained consent from informants for their own participation in the survey. Informants were family members, social workers or care staff who had regular contact with the participants. If necessary, further informants or historical records were sought.

Severity of intellectual disability

Intellectual disability was defined according to ICD–10 criteria for mental retardation as global developmental delay, IQ below 70 and impairment of social functioning. ⁹ Those in whom the intellectual disability status was uncertain at screening underwent an assessment and were excluded if they did not meet these ICD–10 criteria. Each participant's severity of intellectual disability was rated to be mild or more severe, according to their early and adult abilities (including IQ score if available).

Physical health

Information on physical health was obtained from medical records, or from informants in the small number of cases for whom records were not available. We grouped participants into those with one or more chronic health problems and those with none. (Chronic health problems were conditions requiring long-term treatment such as cardiovascular disorders, including hypertension; lipid metabolism disorders; respiratory diseases such as chronic obstructive airway disease; arthritic conditions; gastrointestinal disorders such as peptic ulcers and chronic constipation; endocrine disorders such as diabetes; urinary or renal disorders such as chronic urinary tract infection and renal failure; haematological conditions such as anaemia; solid tumours; and neurological conditions such as epilepsy.) Mobility was rated independently of other physical health problems using the Instrumental Activities of Daily Living scale. ^{Reference Lawton and Brody10} Mobility problems were defined as the need to use a railing, cane, frame or wheelchair, or being bedridden.

Hearing ability was assessed with the whispered voice test. ^{Reference Pirozzo, Papinczak and Glasziou11} Participants using hearing aids were tested with their aids and hearing was rated as follows: no hearing impairment; can hear speech of normal volume; can only hear loud speech; deaf or near deaf. The screening version of the Kay picture test was used to screen for deficits in visual acuity. ^{Reference Kay12} It was designed for children 2–3 years old, and has also successfully been used in adults with intellectual disability. ^{Reference Prasher13,Reference Woodhouse, Griffiths and Gedling14} Participants using correction wore their glasses during testing and visual acuity was rated as follows: normal; impaired; blind or almost blind (even after correction).

Mental health and dementia

Participants were divided into those with and those without a current history (obtained from informants or health records) of one or more serious mental illnesses, including psychoses, affective disorders, anxiety disorders, severe behavioural problems and personality disorders. We did not include a rating for autism-spectrum disorders.

All participants were screened for cognitive or functional decline and memory function. Those who screened positive had a detailed assessment, and the participants were divided into those with and those without a diagnosis of dementia (according to ICD–10, DSM–IV or DC–LD criteria; diagnosis using these criteria has been described in a previous report). ^{Reference Strydom, Livingston, King and Hassiotis7}

Service use

All informants were interviewed using the Client Service Receipt Inventory (CSRI) adapted for use in intellectual disabilities. ^{Reference Beecham, Knapp and Thornicroft15} This collects comprehensive data on a range of demographics and services including usual place of residence, accommodation charges, income and sources of income, and receipt of all healthcare, social care and other services (day care, domiciliary care, respite care and nursing care, in-patient stays, primary and community care contact). Time spent caring by family, other unpaid carers and external providers and aids and adaptations used by study participants were also recorded. We collected data for the preceding 3 months, except for hospital services and aids and adaptations, which covered the previous 6 months. We recorded regular medication use (excluding as needed medication) for the preceding month. Agreement between information collected with the CSRI and general practitioner records has been shown to be relatively high. ^{Reference Byford, Leese, Knapp, Seivewright, Cameron and Jones16}

Costs

Statistical analysis

Ordinary least squares analysis was used to examine the associations between baseline characteristics and total costs. Bivariate analysis was used to investigate associations between each of the baseline characteristics and the total costs of care, using simple linear regression. Associations between costs and continuous variables were computed. All explanatory variables that had a bivariate association with costs were initially included in the model. Variables that did not have bivariate associations with costs were then included one at a time and were kept if they added significantly to the model. The decision to retain or discard a variable was based on significance at the 10% level. This approach was chosen as it allowed us to investigate the impact of various potential predictors of costs independently and when all were included in the model. This approach avoids bias due to omitted variables by including not only variables associated with costs in the preliminary analysis, but exploring the importance of other variables.

Study limitations

We excluded older adults with Down syndrome, as their needs are better researched than those of people without this syndrome. Very few people with Down syndrome survive beyond the age of 60 years; our estimated costs should therefore be representative of the older intellectual disability population. Although we set out to identify all people with intellectual disability aged 60 years or more, it is possible that we missed individuals not known to intellectual disability services. However, such people are likely to have mild intellectual disability and are unlikely to be using many formal services, and therefore should incur relatively low intellectual disability-associated service costs. A further limitation is our reliance on informant reports and clinical records for information concerning physical and mental health status, rather than undertaking a detailed assessment of participants. Nevertheless, the rates of physical and mental illness in our study were similar to those found in previous studies such as that by Cooper et al. ^{Reference Cooper, Smiley, Morrison, Williamson and Allan6} Although we made a detailed assessment of costs and service use, it is possible that some elements relevant to overall costs were not as accurate as we would have liked; for example, we used average costs in residential care settings, not the actual individual cost of care.

Cost of intellectual disability and ageing

We did not demonstrate a relationship between age and cost, possibly because the lifelong, complex disability associated with intellectual disability means that people have longstanding care needs and the additional problems of ageing do not add significantly to this. Furthermore, those with most severe intellectual disability tend to die younger, ^{Reference Patja, Mölsä and Iivanainen24} so that those who are older might be relatively healthy. ^{Reference Janicki, Davidson, Henderson, McCallion, Taets and Force25} In addition, there are several other possible contributors to costs that we have not accounted for, such as provider characteristics, ^{Reference Matosevic, Knapp and Le Grand26} market structure, price-cost mark-up or subsidisation of charges, and the purchasing power of local authorities and local NHS commissioners. ^{Reference Knapp, Hardy and Forder27} Costs for older people with mental health needs in the general population rise with age, probably owing to increasing physical disability; but it has been found that, if those above and below age 65 years are considered, the support costs in relation to need are lower for the older people. ^{Reference Beecham, Knapp, Fernandez, Huxley, Mangalore and McCrone28} As in our study, higher costs and service use in younger people in care accommodation in England are associated with more severe intellectual disability and more challenging behaviour, ^{Reference Knapp, Comas-Herrera, Astin, Beecham and Pendaries29} but vary widely according to models of accommodation, individual organisations, settings and service users. Direct comparison of specific accommodation types have shown that semi-independent living may offer cost and quality of life benefits compared with fully staffed homes. ^{Reference Felce, Perry, Romeo, Robertson, Meek and Emerson30}

Dementia and mental illness costs in older adults with intellectual disability

If dementia is included, psychiatric morbidity rates are raised in older, compared with younger, adults with intellectual disability. ^{Reference Cooper31} Dementia did not predict higher costs in this study. In the general older adult population, dementia is the most expensive psychiatric disorder in older adults owing to high personal Social Services costs. ^{Reference Nelson, Fernandez, Livingston, Knapp and Katona32} However, having dementia was a negative predictor of health service use, ^{Reference Nelson, Fernandez, Livingston, Knapp and Katona32} and therefore not always associated with high healthcare costs. As with age, dementia may not lead to increased costs in the intellectual disability population as people are often already living in supported accommodation with personal Social Services and so do not use additional resources. Alternatively, they may have unmet needs because dementia is often undiagnosed and unmanaged. ^{Reference Strydom, Livingston, King and Hassiotis7} Previous studies have found that mental illness, especially depression, significantly increases service costs in older adults in general. ^{Reference Livingston, Manela and Katona4} Mental health problems occurred frequently in our sample of older adults with intellectual disability, and a diagnosis of mental illness was the most prominent clinical predictor of costs. However, there was a relative lack of healthcare, suggesting ample scope to improve healthcare without making major additional demands on overall costs, thus possibly leading to reduced needs for long-term care and care costs.

Comparison with personal Social Services budget in England

During 2006–7, £3.3 billion of the £20 billion (1 billion = 10 ⁹ ) spent on personal Social Services in England was used to support adults aged under 65 years with intellectual disability. ³³ Similarly, in The Netherlands, nearly 10% of the care budget is used for people with intellectual disability. ^{Reference Polder, Meerding, Bonneux and van der Maas3} Older adults with intellectual disability make up only a small proportion of the older population (0.15–0.25%); ^{Reference Emerson and Hatton34} using our figures, the cost of providing social care to them is between £595 million and nearly £1 billion, which is up to 5% of the total budget. Even small increases in the population of older adults with intellectual disability may therefore have considerable overall cost implications.

Implications

Older adults with intellectual disability are a growing population who consume a significant, disproportionate and increasing proportion of resources, although perhaps less than their morbidity levels would suggest is equitable according to need. Their frailty in old age may be underestimated, or subsumed into existing provision. We have demonstrated that mental illness, severity of disability and health problems are important predictors of costs; these characteristics should therefore be taken into consideration when planning resource allocation and service development at a local level. Services and funders should ensure equal access to care and support, regardless of age, and consider family carers. Nevertheless, there may also be opportunities to improve care without large increases in overall spend, which could result in cost reduction in the longer term, by early identification and management of mental illness or behavioural problems, which could lead to placement in less restrictive and less costly environments. This is in keeping with the principles of the UK government strategy for the care of adults with intellectual disability, ^35,36 which emphasises the need to reduce costly placements (often out of area) by ensuring access to appropriate specialist support, local accommodation options, education and day opportunities.