“The government believes that we need to…encourage behaviour change to help people build healthier lives”.

This is not new –has any government not believed this? – but the prominence that the new coalition government gives this statement and the espoused commitment to public health makes it seem that there will be new energy behind the objective. Almost every week, the Secretary of State for Health mentions outcomes as a gauge of one or other policy goal.

So the government wants to go beyond the rhetoric. It will “give communities greater control over public health budgets with payment by the outcomes they achieve in improving the health of local residents”. But what about the reality?

Traditionally, there has been some debate about what an outcome is, but here the clues are in the phrases ‘behaviour change’ and ‘improving health’. An outcome, then, is likely to be a material change in health status. For example, the incidence of coronary heart disease in a specific Asian community would fall by say, 20 per cent, as a result of a concerted campaign of activity.

But, how do we measure success? It’s relatively easy (though not entirely accurate) to measure the incidence of CHD in a specific community. But there are some real difficulties here, including political ones.

The first problem is that outcomes take a long time to happen (typically five years or more). Politically, governments would need to have some sort of results to talk about in a shorter time-scale. Not to mention the need to make payments based on achieving outcomes. Who would invest in action until the point at which any outcomes are achieved?

This will drive the coalition government as with all previous governments to measure intermediate outcomes, outputs or even activities. All of these exist in a line from inputs, the money, people and know-how that we use to carry out activities designed for example to reduce CHD. Outputs are the things that these activities produce, such as a “look after your heart” campaign, a subsidised healthy food programme or coronary artery bypass graft.

Intermediate outcomes fill the gap between outputs and outcomes to bridge the long time-lag between these. So an intermediate outcome might be, for example, a reduction in the usage of ghee and a corresponding increase in the use of rapeseed oil in the cooking of curries; an increase in exercise and a reduction in waist width. All of these are likely to have some sort of positive impact in reducing CHD. They could be steps along the line from inputs to outcomes, but they are not the outcome itself.

The second problem is that of attributability. There are many factors that lead to one man or woman’s heart disease and there are very many that might lead to its reduction, including genetic/ethnic, cultural, social and economic. In any campaign to reduce CHD, there will be many types of activity, several actors (agencies, private bodies) and the individual person. It is impossible unequivocally to attribute an effect (especially behavioural change), that takes place over at least five years, to any one of these factors or combination of these.  These causal relationships are fundamentally uncertain even if we know from research that some activities are generally more effective than others in achieving a positive impact. The key mental preparation here is first, to recognise that this is an inherently complex situation and second, to be prepared patiently to learn from practice.

The third problem is to do with perception. How do we know that we have an unhealthy lifestyle or that we are unhealthy? There are some ethnic communities where being fat is seen as a sign of wealth and health; where the use of ghee rather than vegetable oil, for example, is seen as a sign of status. Young people’s attitudes to the acceptability of milder forms of obesity may already be changing. Their attitude to alcohol over-consumption is clearly quite different from those of health professionals.

Making people “more aware” may be the tempting response, but research tells us that awareness is probably the weakest of health promoting activities in terms of getting actual outcomes – and even harder to make the attribution link.

Looking to the private sector for help in understanding the outcomes conundrum is not likely to be useful. The nearest thing to outcomes is probably in the advertising sector. Here, agencies generally measure what they easily can –usually the immediate effect of an ad, such as the number of people that recalled seeing it, awareness levels and so on.

Where a sale is a desired outcome, there may be a historical correlation between ad campaigns and increased sales. But where the outcome is more subtle –for example, brand recognition or brand association with a product or feeling- then there is virtually no measurement taking place right now. In fact, the social marketing experience in public health may well be more helpful to the private sector than the other way round.

Health promotion professionals have wrestled for years with these challenges and are well aware of the arguments re-stated here. The political commitment to improving health is welcome; so too is the promise doubling to 4 per cent of the resources to be devoted to improving health. But to sustain this focus will not be easy. Paying only on outcomes seems near impossible. Not only is the area inherently complex but there are also vested interests in the curative side that may be resistant. Experience in WHO demonstration projects in Europe has shown how hard it is to sustain even a tiny shift of resources from curative to preventative –sustained political commitment is crucial.