Data provision

The figure for expenditure is that given by individual health authorities to the Audit Commission. It does not necessarily translate into direct provision on the ground because of different interpretations of which facilities, staff, resources and administrative costs accurately come under the ‘learning disability’ umbrella. In interpreting the results, therefore, this may be an issue for clinicians to clarify directly if spend figures seem out of step with actual resources.

The data used to calculate the total burden were derived from social services, and there may well be regional differences in the accuracy of the figures obtained. However, there is no other available source of such data at present. We must assume the accuracy of the data collected but there is no independent means of validation.

Excluded categories

A section of the population with learning disability is, by definition, excluded from this study; these are adults with learning disabilities who live at home alone or with their families and do not receive any form of help from health or social services, including welfare benefits. To include this group was beyond the means of this study. A larger group of people with mild learning disabilities (IQ=55-69) living independent lives and not identifying themselves as having disabilities are also excluded.

The IQ test with mean 100 (s.d. 15) is normally distributed in populations, giving a predicted 2.27% of the population falling below IQ=70 (used to define learning disability). However, there is also a small effect below IQ=50 from specific pathological conditions. Some data confirm these figures (Reference Rutter, Tizard and YuleRutter et al, 1976; Reference Szymanski, Kaplan and WeinerSzymanski & Kaplan, 1997). However, the percentages identified in other prevalence studies of 0.3% to 0.65% (Reference FryersFryers, 1993; Reference Van Schrojenstein Lantman-De ValkVan Schrojenstein Lantman-De Valk, 1997) correlate with the figures obtained here, illustrating the difficulty of including all who qualify. Few reliable sources for prevalence estimation exist because of varying selection criteria and case-finding methods, as well as variations in age and in temporal and social factors (Reference Fryers and RussellFryers, 1997).

Financial complexities

Unusual forms of purchasing arrangements can cloud accurate spend figures. There are funds known as ‘Section 28 transfers’, which allow the movement of funds from health authorities to local authorities. These are applied to people who were resident in long-stay hospitals before April 1971. This money follows the individual until movement out of National Health Service (NHS) care or death, and counts against health authority expenditure. Owing to the complexity of care provisions, with different health authorities making different arrangements for the continuing care of former asylum residents, exact expenditure has become difficult to untangle. Other funds, not tied to individuals, may also count against spend; these are now under investigation by the Department of Health.

Further complication may arise from health authorities using ‘health’ funds to provide what is effectively social care, funding which should be provided by social services. Old long-stay beds are an example for which we have accounted, but other residential placements of people with high-level needs may have a percentage of their specialist residential care paid for by the health budget. The definitions for these provisions and standard percentage contributions for such cases may vary across the country, and we have no means of tracking them individually.

Provision for autism and children's services

A further complicating factor is that of provision for children with learning disabilities. Although spend figures are supposed to represent adult expenditure only, there are very different practices across England in how child and adolescent mental health services liaise with learning disability services. In addition, some regions provide services for autistic-spectrum disorder under the learning disability services umbrella, whereas others do not. This could well be mirrored by different financial arrangements allowing for transitional services and dual working.

Regional differences

Clarification of data enabling analysis of confounders such as comparator grouping does not seem to explain the spend/burden ratio discrepancies. However, in looking for regional patterns the results show more rural areas with lower expenditure relative to higher burden.

‘Exporter’ v. ‘receiver’ areas

There is significant correlation between spend/burden ratios and the proportions of placements funded by local authority social services. Given the rarity of privately funded placements, results suggest that lower-ratio health authorities (‘receiver areas’) have more imported placements from other regions. Higher-ratio health authorities are significantly more likely to be net exporters. It is notable that all clear exporter health authorities are smaller, urban regions. It may be that it is the geographical size of the region rather than the urban—rural distinction that increases the likelihood of export, with smaller authorities having fewer internal residential resources. It may be cheaper to build residential facilities in rural areas than in cities, and this too could account for greater use of exports to rural regions. ‘Out of area’ placements are often used where higher levels of specialist services are needed (implying greater health needs). Unless service level agreements are arranged with the receiving health authority, the receiver area funds any future mental health needs. In effect, this can mean that the health burden of such placements is particularly high, this being the experience locally in Gloucestershire. There is an implication that accurate enforcement of service level agreements for all exported people with specialist mental health needs might redress the imbalance of present expenditure patterns. If this is unworkable, overhaul of present allocation methods may be necessary.

Although recent research has looked at the extent of ‘out of area’ admissions to psychiatric beds in general psychiatry (Reference Glover and BindmanGlover & Bindman, 2001), there is no similar research in the learning disability sector, either for hospital or residential ‘out of area’ placements.

Long-stay hospital data

The lack of correlation between numbers of long-stay (asylum) placements and spend/burden ratios is surprising, given that these figures would be an obvious source of discrepancies.

Effect of asylum closure

The fact that expenditure on learning disability services is not distributed equitably between health authorities mirrors the findings of research into general mental health service expenditure (Reference Bindman, Glover and GoldbergBindman et al, 2000). The noted high level of expenditure on general mental health services in London (Reference Chisholm, Lowin, Knapp, Johnson, Ramsay and ThornicroftChisholm et al, 1997) has been attributed to need greater than that accounted for by the York index (Reference Ramsay, Thornicroft, Johnson, Johnson, Ramsay and ThornicroftRamsay et al, 1997). Whether this translates into spending on learning disabilities services is unknown. Crucially, the movements of this vulnerable population over the past decade as asylums have closed have not been accounted for in a coherent national approach with regard to relative levels of expenditure. This has created ethical dilemmas (Reference GreavesGreaves, 1997).

This study suggests that the export of people out of their home county has not been paralleled by an increase in spending by the receiving counties facing the consequent higher burden. The closure of asylum provision may therefore have occurred at the expense of such receiver areas without adequate follow-through of finances to fund the needed community care.

Political initiatives

The impact of political initiatives on spending is not documented in learning disability research. The formation of ‘health action zones’ in areas such as Stockport and Sandwell may have affected their expenditure. Similarly, some areas have pioneered effective joint working between health and social services, as has been done in Hillingdon, and this may have had a beneficial effect on improving the efficiency and profile of the service area.

The White Paper Valuing People: A New Strategy for Learning Disability for the 21st Century (Department of Health, 2001) sets out proposals for improving the lives of all people with learning disabilities. A new learning disability development fund of £22 million revenue and £20 million capital is being introduced in 2002, to ensure implementation. The observed discrepancies suggest that individual health authorities should address potential inequalities, particularly in the light of future budgetary increases, to ensure that people with learning disabilities are optimally and equitably served.