Giving child well-being data a work-out

Giving child well-being data a work-out

Article Type: Editorial From: Journal of Children's Services, Volume 9, Issue 1.

Lots of child well-being data is collected but it is rarely exploited fully. Often this is because the data is unfit for purpose, or packaged unhelpfully, or because the tools for exploiting it are unavailable. Several papers in this edition address this issue, focusing respectively on the development of a new measure of well-being for children under eight (Thompson et al.), the need to review progress in terms of outcomes (Hutchings et al.) and a new approach to routinely monitoring outcomes (Hurst et al.).

We use this editorial to reflect on how data can help leaders of children's services to perform important functions that they often struggle with[1]. We suggest that performing these functions better will strengthen the accountability of service providers to service users and taxpayers. Our focus is mainly on high-quality child well-being data gathered for representative samples in given geographic areas.

The first function involves counting the prevalence and distribution of impairments or other outcomes – for example, the numbers of children living in poverty, being maltreated, displaying behaviour problems, or misusing drugs or alcohol. Here, the aim is to chart the pattern of child well-being in a given population, be that a community, a country, or an administratively-defined sub-group, such as children served by a social services agency. This can provide an empirical grounding for decisions about service provision.

Second, such data can be used to prioritise which aspect of well-being to focus on. There are many dimensions of children’s health and development that services may seek to improve, and many potential influences on children’s health and development – risk or protective factors – that services may seek to change (e.g. parenting, school environment, neighbourhood). It is neither possible nor desirable to address all of these. When robust data on the well-being of local children is used alongside appropriate banks of comparison or norm data it is possible to prioritise outcomes and associated risk and protective factors. The data may indicate outcomes where children are developing on a par with or better than children in comparable populations, and also highlight deficits to address.

Third, child well-being data should inform decisions about what interventions to invest in. The data may indicate a large tail of the distribution comprising children with likely impairments to their health and development. Such children may benefit from targeted interventions. Alternatively, the data may point to the need for universal services. These approaches are not mutually exclusive.

As well as highlighting areas for innovation, the data may guide the selection of specific evidence-based interventions. Numerous online databases can be searched to identify interventions demonstrated to improve a particular outcome, reduce specified risks or enhance identified protective factors. Thus, if the data indicates elevated rates of adolescent drug misuse among adolescents, a programme such as Life Skills Training may be suitable. Greater specificity may be achieved by filtering programmes according to the risk and protective factors targeted. The databases commonly provide information on the implementation requirements of interventions, including human resource requirements, training packages and costs.

Fourth, child well-being data can help commissioners to estimate the size and demographics of target populations. For example, data on the “tail” of the distribution can be used to estimate the number of children with actual or likely impairments to their health and development. Beyond this broad category of “children in need”, it is possible to calculate the number of children who would benefit from a given intervention – let's say parents of three to four year-olds with serious conduct problems, which would be a suitable focus for the Incredible Years parenting programme. This might be defined as children who fall into the “high need” category of the SDQ “total difficulties” score.

Fifth, child well-being data can be used to help estimate the likely or realistic magnitude of improvement. There is a tendency here to be overoptimistic, which sets interventions up to fail. A more accurate estimate is possible. Let's say, for example, that the behaviour of the average child in a given area is significantly worse than that of the average child in the country, and that nearly twice as many children have a likely conduct disorder. Drawing on prior knowledge of the size of effect that given interventions can have over a specified period, it is possible to plan a portfolio of interventions that stands a good chance of making up some of this ground – as long as they are implemented for long enough, well and on a large enough scale.

Sixth, well-being indicators may be measured over time to show whether aspects of child well-being are stable, deteriorating or improving. Without a control group it is difficult to attribute change to a specific intervention, but these indicators can cautiously indicate the direction of change and be used the “take the pulse” of child well-being in a given context.

Making data work even harder

How can well-being data be made to work even harder? Developments in a handful of areas are ongoing.

One involves comparing the well-being of children from community-wide representative samples with that of children involved in public systems. This requires using the same measures in both cases. It can show whether some children in contact with services have low level needs or are receiving the wrong services for their specific needs. It can also show how many children in the broader population who could benefit from services are missing out. The analyses should inform decisions about where to invest scarce resources.

A connected development involves strengthening the links between data on child well-being and data on service-use. Until now, measures of service-use have generally been inadequate, making it hard to know exactly what children and families in a given system or geographical area actually receive. But if meaningful service-use data is collected in community-wide surveys, or if existing administrative identifiers are used to match community and agency data, it becomes possible to assess the match between needs and services.

We think that service commissioners will also benefit from knowing how changes in child well-being affect outputs. For example, how does reducing the rate of child maltreatment affect the proportion of children taken into care? Ultimately, this information should help commissioners to invest in interventions known to improve outcomes because they will see the likely beneficial effects for the output indicators against which agencies are traditionally judged. The information will also help, potentially, to reallocate resources towards prevention and early intervention activity premised on anticipated future savings on heavier-end interventions.

Lastly, the gap between child well-being and service data and the people who need it most – directors and commissioners but also managers and practitioners – needs to be closed. This can be achieved by making data “live”: getting it into the hands of these people rapidly and in a format that is accessible and easy to manipulate and analyse. These so-called “continuous feedback systems” can be used for several purposes. One is to monitor whether outcomes are improving, or if services are being implemented as designed. Building on this, managers might offer additional training to staff who are struggling to implement an intervention well. Or a practitioner might observe a client's progress and notice that they respond better to some treatments than others. Lastly, if the data are made widely available in anonymised form, potential users could use it to help them select services. Now that would be radical.

Nick Axford and Michael Little

Note

1.We draw on the paper Axford et al. (2013).

Reference

Axford, N., Hobbs, T. and Jodrell, D. (2013), “Making child well-being data work hard: getting from data to policy and practice”, Child Indicators Research, Vol. 6, pp. 161-77