Using behavioural insights to reduce non-urgent demands on child protection services by school-based designated safeguarding leads (DSLs)

Warwickshire County Council worked on the LGA Behavioural Insights Programme and commissioned Social Engine to reduce the volume of non-urgent calls made by school-based safeguarding leads (DSLs) to their Multi Agency Safeguarding hub (MASH) by adopting a behaviourally informed intervention.

The project sought to provide clear and accessible guidance highlighting more appropriate sources of support for DSLs, provide clarity over decision making, in order to reduce non-urgent contacts.  The intervention took the form of training and a poster. DSLs receiving the intervention reduced non-urgent calls by 65 per cent compared with other DSLs. This reduction in contacts would mean 1,398 fewer non-urgent calls over the course of a year, realising savings of over £370,000 a year.  

The challenge

Since the MASH facility opened in 2016, it has received higher levels of referrals compared to statistical neighbours and, against a backdrop of increasing demand, a drive for efficiency and review attention is focused on securing improvement to interactions with referrers and stakeholders in order to ensure it can operate most effectively. MASH receives a high volume of calls, with a significant proportion of these being identified as non-urgent matters. These calls, which could more appropriately be directed to other services, such as Early Help and Family Information Service, place MASH under considerable pressure, constraining capacity to deal with urgent safeguarding issues. WCC calculate that non-urgent calls to MASH cost over £570,000 each year.

This project sought to reduce the volume of non-urgent calls to MASH by adopting a behaviourally informed intervention.

The solution

The initial scoping phase of the project helped to identify a number of key areas to explore in order to understand the drivers of the excessive demand placed on MASH by school professionals raising non-urgent matters. Insight was gathered through primary and secondary research and statistical analysis. Interviews and workshops were conducted with key stakeholders including school professionals, designated safeguarding leads and the MASH team.

Our intervention framework was based on the extensive evidence and insight gathering phase, drawing on the following hypothesis:

Providing clear and accessible guidance highlighting more appropriate sources of support would reduce calls by DSLs to MASH.

With this backdrop, we sought to develop an intervention that would support and guide DSLs to make more appropriate decisions about how and when to contact the MASH. By the spring of 2020, design of an intervention intended to influence the behaviour of DSLs within schools was complete. However, the COVID-19 pandemic interrupted the start of the trial.

Subsequently, the project shifted to focus on ensuring that behavioural insights informed the messaging contained in safeguarding training. This decision was based on earlier findings of the importance of DSL training as a key influence on safeguarding leads’ behaviour as presenting significant opportunities to positively influence behaviour.

By reviewing and redesigning the training materials together with WCC colleagues, we aimed to make more salient the appropriate way to access advice and consultation without regard to the MASH helpline.

The principal influences which were incorporated into our intervention were:

  • Simplification - When something is simple, clear and easy to understand (salient) it increases the likelihood of the reader responding positively. In practice this meant editing content, using accessible language and increasing resonance.
  • Social norms - When someone thinks that everyone else is doing something, they are far more likely to do it themselves. The use of a descriptive social norm – pointing out that that the majority of DSLs achieve satisfaction from following a particular process, or that higher numbers of children/families get back on track by accessing Early Help is a useful way to emphasise the ‘take action early’ messages.
  • Framing - Documents were framed to emphasise the positive gain (‘help received; case resolved!’) of taking a particular route.
  • Intrinsic incentives - Messages that resonate with the core ambition of the DSL to resolve the situation and to achieve this with minimal effort were seen to appeal to their intrinsic motivation.

The intervention was delivered to new and existing DSLs attending training in Autumn 2020. In total 165 DSLs participated in training courses and received our intervention.

Our goal was to ensure that training and guidance materials provided for DSLs was accessible, useful and provided appropriate information. The aim was to ensure that DSLs were clear in their responsibilities, confident in their decision making and understood the actions they must undertake.

A quasi experimental method was used to evaluate the intervention effect as we could not randomly assign who attended the intervention training without the sample size becoming too small. All DSLs attending new DSL or refresher safeguarding training were exposed to the intervention and their behaviour was compared with DSLs who did not attend the training (which formed a comparison group). In total, 165 school DSLs (14 per cent) attended the training and were in the treatment group, and the remaining 984 (86 per cent) of DSLs formed the comparison group.

The impact (including cost savings/income generated if applicable)

DSLs in the treatment group made on average 65 per cent fewer contacts compared to DSLs in the comparison group. The proportion of DSLs who called MASH to raise a non-urgent issue was considerably lower in the treatment group (5 per cent) than those in the comparison group (21 per cent). Therefore, DSLs who were not exposed to our intervention were four times more likely to call with a contact compared to DSLs who were exposed to our intervention. All these findings point towards a significant and positive effect on behaviour.

The intervention effect – a reduction in non-urgent contacts of 65 per cent - would mean 1,398 fewer non-urgent calls over the course of a year. Based on WCC’s cost calculations, this would realise savings of over £370,000 per year. Despite the limitations of a quasi-experimental method, the strength of the effect size leads us to be confident that the intervention was effective at reducing non-urgent calls to MASH.

How is the new approach being sustained?

Having proved to be effective, the intervention poster is now to be used more widely as a key resource with DSLs and partner agencies. Work is also underway to ensure messaging across WCC are aligned to reduce the risk of conflicting or inconsistent information being received by DSLs.

Efforts are also underway to engage DSLs more effectively and regularly in order to ensure activity and messaging is informed by their perspectives and insights. These include the Head Teacher Forum feeding in to training design and embedding the new head teacher coach role which aims to specifically target support where needed. Further, the implementation of a DSL information line has provided a useful channel for DSLs to explore concerns and develop confidence in decision-making prior to taking action. Improvements to data collection and integration has also taken place as a result of learning from the project, with the development of a new data dashboard to ensure more timely, accurate and comprehensive evidence are driving decision-making.

Lessons learned

The project has shone a light on the risk of adverse effects arising from inconsistent or contradictory messaging.  The trial has shown that clearer communication – removing ambiguity and potentially confusing messaging – is likely to have a positive impact and reduce undesired behaviours.

Understanding the attitudes, behaviours and influences on decision-making among DSLs was key to informing the design of our intervention. Actively engaging DSLs and Head Teachers to understand their perspectives is crucial to designing an efficient and relevant service.

There is considerable benefit from understanding how it feels for DSLs when their demands of MASH are not met – and why – as this is likely to influence their subsequent behaviour. This insight-led approach offers learning not just to the design of framing messaging and services for DSLs but also as an approach that can be equally applied to other services.

There is still potential to further reduce unnecessary contacts by DSLs. Whilst the intervention has demonstrated progress, there is still further iteration and experimentation to be done.