The Analysis of Small Area Statistics and Planning for Health

Abstract

Recent years have seen a focussing of attention on the social and geographic variations that exist in the need for health care in the U.K., the social and geographic variations that exist in the provision of health care and the tendency for these variations to coincide inversely. Of the two sides of the needs-provision equation it has, necessarily, been the former that has proved most difficult to determine with any measure of general agreement. Over the last year, the British Medical Journal has published two competing census-based indices, each of which claims to measure the . level of need for primary care. This paper reviews and assesses the merits of each and suggests that a more illuminating analytic approach might be found by applying clustering techniques.

Deprivation

“…Slightly or substantially different lists of indicators, with different weightings, have been preferred in other studies (see, for example, Scott-Samuel, 1983; Department of Planning and Design, Sheffield, 1983;Davies, 1984;Townsend, Simpson and Tibbs, 1984;GLC, 1985;Thunhurst, 1985aThunhurst, , 1985b. It is evident from the various studies that more attention needs to be given to the justification of the indicators chosen-or their 'coherence'.…”

Need and Provision in the National Health Service: an inverse care law?

“…'It is, we believe, mistaken to treat being black, old and alone, or a single parent as part of the definition of deprivation, Even if many among these minorities are deprived, some are not, and the point is to find out how many are deprived rather than operate as if all were in that condition'. Thunhurst (1982Thunhurst ( , 1985 makes a similar point and shows how partial correlation analysis can help to overcome the problem. More generally, there needs to be more discussion in the studies over the choice of indices rather than indices being produced out of thin air, with little or no attempt at justification.…”

“…The measurement of need. There are both conceptual and empirical problems in measuring need (Plant et aI, 1980;Thunhurst, 1985;Townsend et aI, 1988). Briefly, the conceptual issues revolve around the question of whether need is to be ascribed by a third party, for example normative need, or whether need is to be expressed by the people concerned, for example felt need (Plant et aI, 1980).…”

Regional socio‐economic variations and deprivation indices

“…6 In 1985 Thunhurst had made the distinction between 'direct' and 'indirect' measures of deprivation. 7 However, when using the census it cannot be certain that even apparently 'direct' measures -such as the proportion of unemployed or the proportion of households lacking a car -are always measuring disadvantage or deprivation. It can be argued that any dichotomous variable available in the census has dimensions which mean that the categories cannot simply be taken to fit with a deprived/not deprived dichotomy.…”