Browse By

Systems modelling a Joint Health and Wellbeing Strategy

In my last blog, I outlined the broad shape of what a Joint Health and Wellbeing Strategy might look like.  I included a rough diagram to illustrate it but wanted to produce a slightly more detailed systems map, showing some of the connections and feedback loops.

I started drafting something in Vensim (powerful systems modelling software, with a free version for personal and educational use).  I soon decided, though, that I wanted more than just causal loop diagrams.  I wanted to put it to the test – to put some figures against it.  Vensim allows such simulations, with values and variables, stocks and flows and equations and functions used to show the relationships between the different elements.  What I have done is fairly basic and only covers part of the strategy but still took some time since I was fairly new to the software (I had briefly used it a few years ago) and don’t have all the necessary mathematical and modelling expertise.

The part I have modelled is a positive feedback loop where work on prevention reduces demand on health services, releasing resources that can then be reinvested in further prevention.  Since a significant proportion of ill health comes from unhealthy lifestyles (in terms of diet, exercise and substance use and abuse) the idea was to make a transformational change in people’s behaviour.  This is based on the idea that people tend to be influenced by what others do rather than the information they are given.   The aim was to produce an ‘epidemic of behaviour change’ by creating the conditions where the ‘healthy living’ people influence the ‘unhealthy’ until a tipping point is reached where healthy behaviour becomes the new normal.  This was to be a ‘whole system’ change involving a partnership, and active support, of public, private and voluntary sectors and the public as a whole.  Influencing behaviour to be more healthy is supported by a campaign and work to bring people together in ways that make influence more likely.  As people become more healthy, the demand on the health service reduces, releasing resource that can be reinvested.

The picture below shows my attempt at modelling this part of the strategy (too small to read in detail but hopefully providing a general idea of what’s there).

I will give the main assumptions underlying the model at the end of this blog, but it shows, out of an overall population of 500k, the number of people becoming healthy increasing from 50k to 330k after three years, 417k after five, and reaching its maximum of 422k between seven and eight years.  There is a massive increase in the pooled fund produced from savings arising out of reduced demand, but this seems infeasibly large and suggests further changes to the model are needed to make it more realistic (which could then affect the other figures).

So, what have I learned from this exercise?  First of all, as the current crisis has shown, modelling can be very useful but is not without its problems.  These are amplified when it is being used fairly crudely by a non-expert.  However, what it did for me, was to make the situation more real.  Rather than thinking vaguely that people could influence each other and that would produce savings which could be reinvested, actually having to put figures on that tested my assumptions.  As well as testing individual assumptions, it also enabled the interrelationships between the various factors to be tested.

Another lesson was how lots of connections and feedback loops can create strange and extreme behaviour of the system.  To be fair, some of that came from the particular equations I used (and occasionally the mistakes I had made!).  But some were real.  You could easily find a small change making virtually no difference for several years, then shooting up exponentially.  Which is exactly the sort of graph we have seen with coronavirus infections.  However, while such exponential change is possible and does occur in the real world, there are often dampening effects with social change so I had to try and build those in (usually in a fairly crude way).

So, this has been a worthwhile exercise in terms of testing my particular outline strategy but are there lessons for Joint Health and Wellbeing Strategies more generally?  Here are a few possible ones:

  • Exponential change can make a huge difference but it can take some time to take off. So you need to be patient.
  • In terms of the model, a lot depends on the assumptions you make. That suggests there may, in reality, be lots of variables you can tweak to try and improve things.
  • Making people healthier can also save money at the same time, which can feed back into the system.
  • The success of this strategy relies on having innovative ways to make change without costing too much. In practice, this is unlikely to be one single ‘big idea’ but probably lots of smaller, but still worthwhile, ventures.  When you look at the whole system (public, private and voluntary sectors and the public) and think long term, there are many more options available to you.

So, does this model show that this strategy is viable?  It is by no means definitive, and is certainly not a prediction, but it does give me some confidence that it might work, it can’t definitively be ruled out, and it is worthy of further research, particularly on the underlying theory.  And this just covers one part of the outline strategy.  There is even more hope for it if such things as the social determinants of health, working with children and young people and the use of technology are included.  Even then, there would remain lots of uncertainty and it would need to be tackled gradually with frequent feedback and revision.  But there is some hope that by working together and looking long term, radical change could be achieved.

Of course, as I have noted before, even more could be achieved if something like this was part of a coherent regional and national strategy.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.