From data to delivery
Chris Robinson reports on learning from Supporting People Programme Grant outcomes in Monmouthshire
Outcome monitoring was introduced in April 2012. There were concerns around the data collation methodology and the ability to extract useful information from the data, but data collation was implemented whilst the data sub-group looked at resolving the collation and analysis issues.
In Monmouthshire we are conducting a fundamental review of our SP services – we need to consider how best to balance meeting need with ensuring we optimise the ‘bangs we are getting for our bucks’. We were convinced that our unanalysed 36Mb of outcome data would contain some ‘golden nuggets’ of information – evidence to influence the range and scope of services we commission in the future and ensuring that our spend plan becomes more evidence based.
Before discussing what we learned, it is appropriate to highlight some of the weaknesses in our analysis.
Analysis weaknesses
Firstly, we were only able to measure outcome distance travelled when the person was in a service for two successive periods. This made us consider average performances once we had established, in very short-term services, that the results for those spanning two periods were representative of all people in the service.
Secondly, our initial results were skewed by our providers’ responses that seemed to show that in the two periods of data collation, older people achieved remarkable progress. In many cases achieving full competency of an outcome had been recorded as being achieved after the implementation of outcome monitoring, when in fact this had been achieved prior to implementation.
Thirdly, we had not been as effective as we thought in cleansing and validating data when it was submitted.
All that being said, we did learn some very interesting and useful information about the effectiveness of our services and in understanding our client categories better.
From a fair access perspective, we provided service for roughly twice as many women as men – a gender disadvantage.
And we noted that women achieved approximately 40 per cent more in terms of outcomes than men. We commissioned some research to help us understand the gender access skew and this has led to a number of recommendations that represent positive action.
Broad range of categories
Our spend plan appears to show that we don’t provide services for a number of client categories. However, analysis of the needs that generic floating support services support shows that we do actually provide services for the full broad range of categories. The analysis of the effectiveness of generic floating support as compared with specialist services showed that the average length of stay in generic services was probably too little to ensure a longer-term positive impact. This led us to devise a pilot programme aimed at ‘primary prevention’ – preventing the service users having to go through the full rigours of the SP processes when they only needed very short-term support in one particular area (normally benefit-related). This reduced demand on the generic services and enables the services to provide support over a longer period.
The outcome analysis showed that we provide insufficient support for the young, especially care leavers. Again, this led to a pilot service being developed that focuses on this client group.
And lastly, it appeared that the most difficult outcome to achieve was feeling safe for the elderly – obviously a very ‘fragile’ outcome for which a single incident can immediately undo the good work done over a considerable period.
In terms of understanding our client categories better, our analyses enabled us to convert homelessness into its component needs – for example mental health, vulnerable single parents and substance abuse being the greatest three needs from the range of 15 identified.
Also, we were able to analyse the complexity of support needs in the various client categories. Domestic abuse and vulnerable two-parent families were the most complex.
Finally, we have used the outcome information to model funding based on outcome delivery. This model will be included in a range of models that will need to be considered to ensure that our future provision is appropriately focused and funded. However, this modelling also confirms the need to apply different approaches to services that involve prevention as opposed to the longer-term maintenance services.
Despite the limitations in the quality of the initial data and our experimental data analysis, the analyses have provided evidence that changed our short-term priorities and addressed some significant inequalities – well worth the effort!
Chris Robinson is the supporting People lead at Monmouthshire County Council