| Summary: | This thesis explores the role economic analysis can play in policy formulation in the British National Health Service. The historical allocation process in the NHS is based on the idea of meeting fully the 'need' for health care of each individual. However, 'need' has never been comprehensively defined nor universally applied but has been interpreted by numerous, isolated judgements, often by doctors and others at the point of delivery of the service. These decisions have been strongly influenced by local availability of resources and since these have been distributed unevenly, the result has been a wide variation in standards of service, eg hospital cases per head of population and costs per case. While the perceived objective was to give everyone all the health care 'needed', it was not necessary to consider problems of distribution, between individuals and geographical areas, or problems of effectiveness, that is the relative efficiency of different forms of intervention. However, in reality the NHS budget is constrained and a rational decision-making mechanism must incorporate a rationing device which contains a consideration of distribution and effectiveness and these aspects comprise the two parts of the thesis. PART A THE ALLOCATION OF RESOURCES BETWEEN AREAS It has been the declared objective of the NHS since its creation that health resources should be equitably distributed. However, there has always been considerable variation in the per capita expenditure of the 14 regions because the system of incremental budgeting meant that the existing level of services was always financed. Even in 1971 when a formula was introduced for hospital expenditure, 50% of revenue funds was still distributed according to factors influenced by historical provision. Moreover no consideration was given to the allocation process below regional level though it was suspected that variations sub-regionally were greater than between regions. The objective of this study is to explore the concept of an 'equitable distribution' at the disaggregated level of Area Health Authorities, to see if it can be defined and quantified in operational terns suitable for policy recommendations. 'Equal treatment opportunity for patients of similar risk' is taken as the initial definition. The re-organisation of the NHS in 1974 made it feasible to relate services to geographical Areas. Data for the Oxford region in 1971-2 were analysed and applied to the post 1974 structure. Dividing general hospital expenditure by the population of each Area gave co-efficient of variation of 20%. Tests were then applied to see if there were factors, consistent with a 'equitable distribution', which explained part of the variation. The most important factor was found to be the flow of patients across boundaries. These flows were used to derive notional catchment populations and when these were divided into expenditure, the co-efficient of variation fell to 10%. An attempt was made to see if the variation was explained by differences in the morbidity characteristics of the populations. Various indicators were considered but the age-sex structure was found to be the only discriminator for which it was possible to obtain data on differential health service use. While nationally old people make greater use of services, there was no evidence that in the Oxford region more resources had been made available to the Areas with a greater proportion of elderly. Others factors considered include the cost of regional clinical specialties, the cost of teaching hospitals, psychiatric hospital provision and community health care. (These costing exercises themselves have been a useful spin-off). At each stage of the analysis an attempt was made to relate the expenditure differences between Areas to indicators of quantity and quality of services. It was found for example that the Area with least expenditure had the lowest acute-hospital provision but not the lowest provision for the chronic sick. It was concluded that while the variation was not as great as originally suggested, nevertheless taking into account all the factors, a re-allocation of 3% of the regions expenditure would be necessary to bring about 'equality of opportunity'. However, even at this level some differences would remain. For example, if cross-boundary flows were perpetuated, some people would travel further than others to receive care, and while the special funding of teaching hospitals and regional specialties nay be justified on efficiency grounds, this conveys special benefits for local residents. Despite these reservations, in operational terms much can be done to reduce 'basic' inequalities. No work had previously been undertaken on sub-regional variations and, already, at the request of the Minister of State for Health, this analysis has been extended to all 90 Areas in England and the results have added impetus to the formation of the Resource Allocation Working Party, on which the author has served as technical adviser. PART B COST-EFFECTIVENESS ANALYSIS OF THE COMMUNITY HOSPITAL PROGRAMME The objective of this part is to examine how far analysis can help the choice of efficient methods of delivering health care. It comprises an appraisal of the Oxford region's Community Hospital (CH) Programme, a system of peripheral acute hospitals surrounding a District General Hospital (DGH). Originally a full cost-benefit study was planned but the practical problems of quantifying, valuing and aggregating benefits could not be overcome. Preliminary results suggested that differences in benefits were not measurably significant and so a cost-effectiveness approach was used. Firstly a regression analysis of the costs of 525 existing small hospitals showed that average costs vary with size, the curve being a tilted 'L' shape. The minimum cost size was 25 beds though when the sample was later disaggregated to remove hospitals which treat more complex cases this rose to 35 beds. This result was important for subsequent analysis since the two experimental CHs are in the higher cost range. Secondly the capital costs of CHs were compared with hypothetical DGH ward equivalents, arguing that with a rising population, building a CH is an alternative to building an additional DGH ward of the same size. CH construction costs were lower because less space per bed was used and certain features simplified. Land costs per bed, on the other hand, were higher because a low plot ratio was used which more than offset the relatively cheaper land on peripheral sites. The cheapest way, hov/ever, of providing CH services is to convert existing hospitals which have no alternative use and hence low opportunity cost. The main part of the cost-effectiveness exercise entailed a comparison of each component service, considered in terms of the type of patient, with the service which would have been provided without the CH. For example, surgical patients transferred after their operations would otherwise have been retained in the DGH. Some medical patients would have been treated in the DGH, others at home. No evidence was found to suggest that length of stay in the CH differed from the DGH, though this could not be proved and the conclusions were contingent on this. The CH had higher resource costs than the DGH alternative because of high nursing levels, and this was partly the result of the small scale. Also for the surgical patients the transfer by ambulance increased the cost. No account was taken of possible differences in benefits. It was concluded that the CH in-patient services could be cost-effective only if nursing staff were reduced or if converted hospital buildings with a low opportunity cost were used. A comparison of the CH service with domiciliary care entailed a detailed study of the cost of domiciliary services.
|
|---|
