A Model for Mental Health Needs and Resourcing in Small Geographic Areas: A Multivariate Spatial Perspective

Article excerpt

This paper considers how small area variations in a set of psychiatric referral outcomes in a London health authority of 750,000 people may inform health need assessment and health resourcing for mental illness based on true need. As well as adopting a multivariate perspective, the spatial interdependence of the outcomes is included in the modelling approach outlined. By contrast, existing studies on mental health need tend to focus on single outcomes, and may not include spatial dependence. The analysis relates to three hospital referral outcomes for psychiatric conditions, and to total community mental health referrals across sixty-seven electoral wards in East London.


Many studies have been made into the associations between health service usage and socioeconomic structure across geographic areas. While these studies may have a mainly epidemiological purpose, they may also provide information on factors determining health need and hence toward establishing equitable health resourcing mechanisms across geographic areas. Health need denotes the "true" needs of a community for mental health service provision, as determined by their illness levels, and may not be truly reflected in their actual usage: the latter is better considered as "health demand" and reflects factors such as willingness of patients to consult, and the referral thresholds of medical practitioners to refer on to acute (hospital inpatient) or community health services.

The ideal frame for analysis of health needs considers individual outcomes within their societal and spatial context, for example, in a multilevel model. Some resource mechanism exercises have used study data from health surveys to this end. However, to construct a uniform-needs measure across nations or regions, many large-scale studies aimed at improving equity in health resourcing have been at the level of small geographic areas (for example, neighborhoods of five thousand to ten thousand population), or at even more spatially aggregated scales. These studies have drawn on aggregated census data and on various indices of actual health service usage. This is true of recent studies in the United Kingdom and elsewhere which have been the basis of health resourcing formulae (Smith et al. 1997; Sheldon et al. 1994; Eyles and Birch 1993).

Often hospital-use data is the most comprehensively and uniformly coded in terms of patient characteristics including details such as age, illness, area of residence, and so on. Accordingly, a particular strand of analysis involves relating small-area hospital admission rates (the outcome R) to area based deprivation scores or census indices of sociodemographic composition (the predictor indicators X). Regression analysis may then be used to provide weights on socioeconomic variables X in an overall health needs score, S. Assuming all the X variables are standardized to the same scale, a predictor variable X with a higher regression coefficient has a higher weight in the overall-needs score. Thus if [[beta].sub.1],..,[[beta].sub.w] denote the W significant regression effects in a linear regression; then a needs score is obtained as S = [[beta].sub.1][X.sub.1] + ... + [[beta].sub.W][X.sub.W].

These analyzes are usually motivated by the absence of measures of true need for mental health care or true morbidity, which might be provided by (say) a comprehensive population measure of mental health status, obtained at recurrent intervals from population surveys. In the absence of such measures one must rely on imperfect sources of information: actual usage is one such source but is affected by demand factors (for example, clinical thresholds) as well as, to some degree, true need. Proxy indices of need provided by social structure measures (for example, deprivation scores) are based on implicit, though firmly evidenced, links between such structural variables and mental health need. …