Metrics and measures in tackling the social determinants of health - the example of mental health and housing

This paper is an account of the discussions and recommendations by the expert advisory panel on potential metrics and 'sentinel indicators' for improved outcomes in housing and mental health, as part of an inter-agency seminar called to advise on the development of metrics and measures for community mental health, for Fair Society, Healthy Lives: The Marmot Review (Marmot, 2010). The seminar covered all aspects of mental health in both its broadest and narrower senses. Much of the background material for these discussions, therefore, cuts across familiar knowledge silos between the fields of health and housing. Where it is necessary to elucidate the text, references are included to relevant research and policy frameworks that may be unfamiliar to the general reader. This paper is not, however, intended as a general literature review nor is it an evaluation of the available research. A paper on this subject will feature in a future issue of the Journal. The conclusions from the panel discussion are presented in four main areas, reflecting the need to specify metrics across the wide-ranging interface between housing and mental health, while still keeping the task manageable. Five current or potential health service metrics were proposed as having particular value as signal indicators. Two of these (relating to primary care prevention and public health) have no precision as yet, partly as new services and approaches are still evolving. Among existing health datasets, the Mental Health Minimum Dataset (MHMDS) (NHS Information Centre, 2009a), SITuation REPorts (SITREPS) (Department of Health, 2003), and the Summary Care Record data were singled out, though each is thought to need more work to improve the current data categories as well as data collection. One rather more fundamental point made was that the identifying, assessing and encouraging of effective inter-sector partnership work will be the key to tackling health inequalities. The use of other, non-health services data therefore holds great potential for a better recognition both of needs and of outcomes in successful partnership work, especially where this can be interpreted at local level. These wider comments are elaborated in the context of housing, but may be applicable to all efforts to evidence and work with the social determinants and the social outcomes of mental health. For the future, a combination of well-crafted nationally sanctioned metrics and the 'soft intelligence' of locally identified meaning may be most effective. Subsequent developments confirm the potential in cross-sector development work, and indicate the potential for further collaboration via the local performance framework. As policy frameworks continue to evolve rapidly, the article ends with a Codex, updating the relevant policy frameworks context since the seminar (in Spring 2009) and especially in the context of a new coalition government with aspirations to articulate and promote public health in the context of the local performance framework and the 'new localism' agenda. This final section and comments therein are therefore entirely the responsibility of the author.