Subscribe
Share
Share
Embed
Podcast series from the MRC/CSO Social and Public Health Sciences Unit, University of Glasgow.
Ruth Dundas:
Welcome to 15 Minutes on Health Inequalities and a podcast discussing an ongoing project, which is about building a community to understand and evaluate policies across the UK that affect maternal and infant health using administrative data. I'm Ruth Dundas from the MRC/CSO Social and Public Health Sciences Unit at the University of Glasgow, and today our podcast is a little bit different. Instead of interviewing one person, there's three of us discussing MatCHNet, and I'm the lead for the Maternal and Child Health Network and with me are
Alastair Leyland:
Alastair Leland, also from the Social and Public Health Sciences Unit and the co-lead of the Inequalities in Health programme. And
Anna Pearce:
Anna Pearce, also from SPHSU.
Ruth Dundas:
So, I think we can start the discussion today by acknowledging the importance of early years and how MatCHNet fits into that. Pregnancy and infancy are known to be important periods for both living a healthy life and a longer life, and Scotland has committed by law to ensuring that every child has the best start in life. But despite this, health inequalities emerge early in life and track across the life course, and also from one generation to the next. So, that produces avoidable economic and social costs. We lead the Maternal and Child Health Network, MatCHNet for short, which is a community of researchers, decision makers, third sector organizations, service providers and data controllers. And what MatCHNet is striving to do is to make better use of data to test the impacts of policies through making comparisons across the four UK countries. And this is an important approach because many of the policies, which are most likely to improve children's lives, or the upstream ones, for example, early years childcare provision, welfare payments and grants to pregnant women. And although these types of upstream policies can't be tested in sort of traditional trial settings, because usually because they apply to many people in the population and the way that policies are implement can make it difficult to do these evaluations, but it's still really important to evaluate the policies because these policies or these evaluations can support effective decision making and direction of resources.
Alastair Leyland:
So, I guess the question is, why is it important to evaluate these policies? The, the good news is that there is a will to spend money to improve the lives of infants in particular. Just a question over what works. An example is Scotland's Baby Box. The Scottish Government commissioned an evaluation of how the policy was rolled out, how midwives found the process, and whether parents liked the box and its contents. However, they didn't commission an outcome evaluation. They didn't think it important to find out if the baby box impacted on maternal and child health outcomes. Our colleague, Ronan McCabe has undertaken such an evaluation and you can watch his webinar on the MatCHNet website, and he's also done a podcast in this series about it. The policies that were prioritized by the stakeholder group were early years childcare provision, welfare payments, and grants for pregnancy and infants. Now, childcare provision for preschool children can enhance cognitive, behavioural, social, and physical development. And evidence shows that early years childcare can improve school readiness and reduce inequalities in development, which may track into later life. Since the 1998 National Childcare Strategy, the UK early years education childcare provision has expanded. But there's also been a divergence of policy across the four UK nations. Provision of childcare differs by eligibility for universal childcare, whether parents are working or whether they're disadvantaged. Provision also differs by the number of hours children are entitled to and the aged children can access childcare. So one question is, does giving extra childcare provision to low income families support mothers or parents into work and in this manner improve their own health and their children's health? I guess the broader question though is the money being spent on the right policies, which have a positive impact?
Ruth Dundas:
And I think MatCHNet really can support these evaluations that you're talking about there Alastair coz MatCHNet is, the way it's set up, is it focuses on three challenge areas. So we've got the challenge area of policies, the challenge area of using administrative data and the challenge area around methods. And the first challenge area, as I said, is around policies. So what are the key policies that we need to evaluate? And to do this, we needed to work with stakeholders to identify these key policies that vary across the four UK countries that should be a priority evaluate in terms of maternal and child health. And then once we've done that, we need to identify the key administrative data sources that can be used to create longitudinal maternal and child health cohorts. And then third, what methods can be used to evaluate the policies with administrative data? We did consult back to the challenge area of the policies. We did consult on the policy priority areas, which identified research gaps and we've co-produced policy briefs for these deprioritized areas of universal credit, maternity grants, and earliest childcare provision. And we really use these to highlight the cross UK country differences in policies and show where evaluation of such policies will be helpful for decision makers.
Anna Pearce:
Maybe Ruth, you could briefly explain how that policy prioritization was carried out.
Ruth Dundas:
Yeah, we had different kind of like stages. The first stage we went through was hand searching of the UK government and the devolved government websites looking for policies that act on the social determinants of health. And once we had found that list of policies, which was quite long – I think there was over 300, we then did a kind of initial screening to make sure they met criteria that we had envisaged as being important. So they had to be population wide, they had to be able to be evaluated using administrative data and also likely to have a large effect on maternal or child health outcomes. And once we'd narrowed that list down, that screening shorter list, we then prioritized that list with some experts first of all, and then we had an online stakeholder, online consultation with our stakeholders. And once again, that list had been shortened, we then had meetings with stakeholders to make sure that, that it was the right three or the right policies, and we came up with those three policies.
Anna Pearce:
So I think that leads us to the second challenge area that you mentioned, Ruth. So administrative data and seeing what data we have and how comparable is it across the four UK countries. So, just to say, when we talk about administrative data, we are meaning data that are routinely collected in the course of services. So we have a reasonably good understanding of the power of administrative data like these, including from some of the work that we've been carrying out here in Scotland. So for example, we've used these data to understand where and why some children in Scotland have worse health than others by linking across different types of health records of mothers and children from pre-pregnancy and across the early years, we've created what's an, essentially now an administrative cohort. So in that cohort, we've shown that mothers from less advantage backgrounds are almost three times as likely to be prescribed medications for anxiety and depression, and that their children are seven times as likely to be exposed to tobacco, both in utero and post-birth. There are also large inequalities in other outcomes like overweight and obesity, cognitive and socio-emotional development. By linking to birth records, we've also been able to show that inequalities are larger when looking at family level circumstances. So things like occupational status and family structure. Then when looking at patterns of health by neighbourhood deprivation and by tracking records over time, we can get a better picture of how poor health and health related behaviours persist. And it's here where we often see especially large inequalities. I guess thinking more generally, the beauty of administrative records are in contrast as opposed to survey data is that they pretty much capture the whole of the population. This is important if you're interested in inequalities or the burden of ill health in particular groups, such as those that are most disadvantaged as they're often less likely to be captured in surveys. And these data are also really big. So in Scotland, there're around 50,000 births every year. So these large numbers allow us to look at rarer outcomes, or for example, the co-occurrence of multiple aspects of health. These data are also really well placed for looking at change over time and making comparisons across countries with different policy contexts. Essentially allowing us to evaluate the sorts of policies that Ruth previously mentioned that can't be tested in trials.
Alastair Leyland:
Yeah. And that leads on to the third challenge area, which concerns methods. Uh, currently there are few evaluations, of the sort we've been talking about, because these types of evaluations can be hard to plan and conduct due to challenges, a number of challenges, for example, comparison groups. That is finding comparable people who are not subject to the policy, data which need to measure both exposure and outcome for both the intervention and comparison groups and methods, ideally making our comparison groups as comparable as possible, meaning that we are getting as close as possible to a randomized controlled trial. MatCHNet is working to overcome these difficulties by mapping out the data, uh, by collating the methods and working with policy makers and practitioners to identify and understand the policies to prioritize for evaluation. Anna, perhaps it's worth clarifying why there is this cross UK angle to MatCHNet. Do the different nations want to learn from each other?
Anna Pearce:
Well in MatCHNet we certainly are getting that sense that, yeah, the UK countries are interested to see a) how things are being done differently across the different parts of the UK, but then to also see how that reflects in terms of health differences. So, as you mentioned Alastair, you can utilize those differences across countries to gain a better understanding of where the most fruitful places to invest might be. And in other cases, we might have a fairly good understanding of which policies are most likely to be beneficial in the broadest sense, but we might still be lacking information about who is most likely to benefit and in what situations. It's a variation between countries, even in similar policies, can help us to explore some of those questions if they're implemented in slightly different ways. So Ruth had already mentioned the example of free early years learning and childcare, or perhaps it was Alistair. But it’s available right across the UK to children of a particular age. But there are variations sometimes in terms of eligibility. So for example, free places kick in at different ages in the different countries, and sometimes this is available universally or sometimes it's targeted according to whether parents are working, for example, or whether families live in disadvantaged areas. And so dose and intensity can also vary according to these different eligibility criteria. So for example, in Scotland, all three to four year olds are entitled to 30 hours of free early childcare every week. Whereas in England, only children whose parents are working are entitled this much childcare. If the parents aren't working, then they're only getting around 15 hours a week. So we can exploit these variations to learn which features of childcare provision deliver the greatest and most equitable impact on child and maternal health outcome.
Ruth Dundas:
I think we've, you know, just listening to our conversation today, we have done quite a bit of learning while addressing these three challenge areas of MatCHNet as well as the will of governments to spend money and introduce policies in the early years, which does mean there's a lot of potential for evaluation of the policies. There's also a lot of really good quality administrative data that's collected that can be used to evaluate these policies. We do need to do more work about accessing the administrative data and harmonizing and standardizing definitions and time series across the four UK nations. In some cases, to do this harmonization really well it's better to use a specific example or focus on a particular topic area. And we have done this with a project to understand the data available in England and Scotland that can be used for perinatal mental health outcomes. Just to summarize about the take home message, when new policies are being proposed, there should be an outcome evaluation using a suitable comparison group to assess the impact on maternal and child. Administrative data sets offer the most potential for undertaking such evaluations, including linking data sets across different sectors, which can then identify good comparison groups. But we do need support from data owners and other government departments to permit this use of routine data so that we can evaluate upstream population level policies. We're coming to the end now, and usually we're finished by asking what are the implications for health inequalities.
Alastair Leyland:
Yeah. Uh, we haven't directly discussed inequalities too much in this podcast, although a lot of what we've been talking about is relevant. But inequalities are a big focus for MatCHNet. It's important to understand the extent that policies act on social inequalities and also ethnic inequalities are considered increasingly important for maternal and child health. When we do undertake the policy evaluations, we will examine the existence of differences by social position. We can definitely look at small area deprivation, but social class and educational attainment are also possibilities. And I've, I've enjoyed this conversation. It's been good talking like this. It's given an overview of MatCHNet and where we've got to to date, but what about the next steps? What's MatCHNet going to focus on next?
Ruth Dundas:
Well, MatCHNet was funded, as you know, with the express purpose of forging new collaborations. And we've certainly made good progress in this area, we now are a multidisciplinary community of public health researchers, methodologists, policy makers, advocacy groups. When we started, we did set out to establish the approaches and the tools to evaluate the upstream policies that impact on maternal and child health using administrative data. We now have the policies that should be evaluated first, and we've made progress on the methods to do the evaluations. I think we can now then focus on undertaking evaluations of the policies, but what we also need to do is ensure the longevity of MatCHNet and that involves continually expanding the network with new members who have an interest in policy evaluation in the early years. So if any of our listeners want to join, please sign up. Information on how to do that will be in the notes for the podcast. Is that right, Anna?
Anna Pearce:
Yeah, that's right. Details of that and also more information about MatCHNet more broadly can be found in the podcast notes. And also just to plug the series more broadly by searching 15 Minutes on Health Inequalities on most major podcast platforms, you should be able to find us. Thanks for listening.