Can administrative data save your life?

Good health is something we all need – but it doesn’t simply start and end with the body. Increasingly, we’re beginning to understand how different parts of our lives and circumstances can combine and interact to shape our health, including from before we’re even born. Economic circumstances, social inequalities and childhood experiences all have a role to play in public health.   

Because administrative data contains a wealth of information about these different areas of our lives, it can be a valuable tool for research to help us understand and improve wellbeing. By linking this data together, we can build a better picture of how different aspects of public life affect our health.  

In the first episode of series 2, co-hosts Shayda Kashef and Mark Green are joined by Helena Benes Matos da Silva. Helena is an epidemiologist working with the Nutrition Group at Cidacs, the Centre for Data and Knowledge Integrations for Health in Brazil. The conversation draws on her varied career to explore the complex human experiences underlying our health, and how data can shed light on these connections. 

Want to know more about administrative data? Visit: adruk.org. 

Read more about the work of Cidacs here: cidacs.bahia.fiocruz.br   

Co-host Shayda Kashef mentions the ECHILD dataset; read more about the data and how to access it here: adruk.org/data-access/flagship-datasets/education-and-child-health-insights-from-linked-data-england 

Connecting Society is presented by Shayda Kashef and Mark Green, our producers are Eleanor Collard, Holly Greenland, Laura Mulvey and Shayda Kashef.

This podcast is brought to you by ADR UK (Administrative Data Research UK), a partnership transforming public sector data into research insights and policy evidence to improve lives. We are an investment by the Economic and Social Research Council, part of UK Research and Innovation.

Transcript

Shayda (00:10)
Hello and welcome to Connecting Society, a podcast about how everyday data can shape our lives. I'm Shayda Kashef, Senior Public Engagement Manager for Administrative Data Research UK, or ADR UK to me and you.

Mark (00:25)
And I'm Mark Green, Professor of Health Geography at the University of Liverpool. We are your co-hosts and guides around the wonders of administrative data.

Shayda (00:34)
In this podcast, we're exploring all the different ways in which the information that's collected about our everyday lives, from our interactions with health services, our voting behaviors, police and crime reporting, educational achievements, and more, is used by researchers and policymakers to make better decisions to support society and make the world a better place. And we're back for season two.

Mark (01:01)
Yeah, we're back and much like David Bowie talked about in Changes, we're doing things differently this season. We're on YouTube, so you can now see us as well as hear us. Apologies for people having to look at me. And we've got a new format as well.

Shayda (01:18)
That's right, we are going interview style, in depth investigations, kind of, not really, but we're having more personal one-to-one conversation where we can really dig deep and learn a little bit more about the person or the topic that we're talking about. Speaking of, I think you had your first interview today, Mark.

Mark (01:41)
Yeah, let's take a listen.

Mark (01:47)
So today we are discussing whether administrative data can save your life. It's a bit like that song by The Fray, but hopefully with less depressing melody and a bit more uplifting chat about data. Here to help us discuss this a bit more is Helena Benes Matos da Silva, who is an epidemiologist and nutritionist at CIDACS, which is the Center for Data and Knowledge Integration for Health, based in Salvador, Brazil.

Welcome to the show, Helena.

Helena (02:18)
Thank you so much. Thank you for having me here.

Mark (02:21)
It's great to have you. I saw your talk at the ADR UK annual conference. I thought it was absolutely brilliant. So I'm very excited to be doing this interview myself. Helena, would you like to introduce yourself and tell our listeners a bit about who you are and why you are so awesome?

Helena (02:38)
Wow. Thank you. Well, you say my name. You say the place where I'm from. I'm from Salvador, a city in northeast in Brazil. The same place where CIDACS is based. My background is in nutrition. I did some things on nutrition related to biochemistry, lab stuff. And later on, clinical nutrition.

And after some time, I started going to epidemiology. And my Master's and my PhD was in epidemiology related health. That's how I build up my expertise, going through different areas and at the same time, sticking to the nutrition field.

Mark (03:31)
So in this episode, we're talking about how administrative data can save your life. And I don't mean quite like it's going to show up suddenly if you're dangling off a cliff and kind of raise you to safety.

I'm more thinking about how it can understand the factors that influence and shape our health and wellbeing. So when I teach my year one students, we often talk about how there's a lot of emphasis on the kind of biomedical, maybe a bit technical term, but things like genetics or biology, or even just the random bad luck that certain people might have. And we sort of forget that these wider factors about society, about the context in which we live our lives in. Really matter. You know, how our experience is during childhood, if we have a well-paid and secure job, do we live in a green area? You know, time and time again research has shown that these things matter and it's about who we are and how we interact with the world around us. And in sort of technical speak or in academic speak we often call this the social determinants of health, so how societal factors affect our health. 

Helena, you're an expert, you do lots of research in this area. For you, what do you see as the social determinants of health? Tell us a bit about how you kind of see this wonderful field.

Helena (04:48)
Yeah, this wonderful field, when I go through it, I see my reality. I see like how okay, because Brazil is a very diverse place and it's also very famous for its inequalities unfortunately. So when I think about social determinants of health and there's the condition of how we, before even we...

We are in the fetus moment, like our parents, how they had their experience, the pregnancy of our mothers, how we were grown, and the education we had access, and all those things. So for me, it's impossible to see research out of thinking the social determinants of health. And I was even thinking you were telling this now and came to my mind one example when I was in my early career as a nutritionist and we are very biology-oriented. Vitamins and minerals and micronutrients and then I was talking to one patient and she was like, I'm doing everything correct but my high blood pressure doesn't come down, doesn't reduce at all.

Mark (06:00)
Yeah, macronutrients.

Helena (06:13)
And we started talking about her living conditions. And she was from a very poor neighborhood with a lot of violence. And we came to the conclusion that it was so hard for her blood pressure to go low. It was impossible because it was like shooting in front of her house. And then I had that feeling like, okay, I cannot study just the nutrients itself. I need to go deeper and have a broader view of life to be able to understand and to be able to do research.

Mark (06:51)
Yeah, I mean, I think that's a great example. So I used to work with this guy, was originally a medical doctor, a clinician, and he said the reason he got into research was because he got bored, not bored, but frustrated is probably the right word, of constantly seeing the same patients, treating them, them going away and coming back with the same issues. For him, was they were sending them back into the environment that was unhealthy. And then that was the problem.

All he was doing was treating the kind of issues that resulted from that rather than the solutions to it. So can see how that kind of parallel to that similar story to yourself works really well. Can you tell us a bit about kind of, because you do a lot of research in this area around teasing out how our kind of life and society influences our health. Can you tell us a bit about the work that you do?

Helena (07:43)
Yeah, this for me started being a reality, my work since I joined the CIDACS team because we do have this centre to do the linkage of the administrative data and the main view of the centre itself is to work with this data to investigate how different strategies like policies or how it impacts life of the people.

Otherwise, we can also see different social determinants of health as we have information about their living conditions, education, marital status, access to healthcare. So when I joined the team, this started in my life as a reality. So like now, okay, now I can do something. And it is happened during my PhD. And I did conduct one for me, it was one.

Of my most important work during my PhD where we were investigating childhood growth. Having a background that in Brazil, just for you to understand, we still have the consequence of racial inequality. So we have racism and the structural racism and this is something that we need to fight and deal with.

So we were investigating how was the child growth up to five years of age in the different racial categories. And I was a bit surprised, even though we know it exists, but when we see in the data, it's something that shocks us. And we could see a clear difference between the groups.

We have in Brazil. And the indigenous people as well, they were like in a most vulnerable situation. And also the child growth give us this idea of quality of life and how things are going. And seeing this difference for me during that research was a shock and also just a portrait, like a picture.

Of the inequalities, racial and ethno inequalities as well as we have in Brazil. And since then I've been trying to dig more and dive in this area. And recently, but this is something...

More recently have been working with childhood cancer. And this is not published yet. And we are still working on this manuscript. But again, just having a broader look of the social determinants of health and how it impacts hospitalization for cancer in children and also their mortality. We can have an idea like poor people and most vulnerable situations, they will have probably poorer outcome. But again, when we can extract this information and see the reality through data coming from real world, from the real life and...

Thousands of people because we do have this nationwide data is still something that impresses me.

Mark (11:24)
Yeah, and I think data is that point in which you see all of these complex factors in society and all the ways that they imprint on our life. It's where things become real, they become visible. And even though when you look at something like hospitalisation, it is just a single event, but actually it tells you such a rich diversity.

Or, you know, even we see things like intergenerational. So things that even before they were born were starting to influence, you know, we see that particular childhood growth, you know, across many different countries as well.

What's the main drivers do you see in the data when you study it? What are the key factors that you pick out from within Brazil?

Helena (12:10)
It's very clear to see access to healthcare. So even though we use as a proxy of the attendance for prenatal care, if it was adequate or not. There are some limitations when we use administrative data so we don't have everything, but we can take a proxy of it. So this is a good sign like when they have adequate prenatal care in the womb.

This makes a difference later in life because we can see like they were follow up or something happened later on. This is some important data. Also education. We usually have collected in this administrative system the schooling of the mom. So the maternal education, we usually have this information. And it's clear, crystal clear.

Higher education, better outcomes. It's very clear in our data and for most of the things. And usually we have the housing conditions. So if you have a sewage, appropriate sewage disposal, which also give us an idea of the water they have access, the quality of the water.

So these again, it's very clear the housing conditions, but I would say I would stick if you say like Helena choose two, I will choose maternal education. We always see something there, a gradient and also the prenatal care if they had adequate prenatal

Mark (13:56)
I agree. Really these are the modifiable we can change. Like we can improve educational or access to education. We can improve housing quality. And I think that's what's really powerful about this work is it's identifying what we can, what we can improve. And that will have very big impacts on health both today, but also, like you say, knock on effects later through people's lives as well. Which just makes

What you're doing is so important really.

So I was really interested because earlier you said you have this nutrition background and I think sometimes when we think about administrative data we think about the of the nerds behind the maths and the statistics, understanding and playing around with data. But actually from a lot of my work that some of the best data scientists actually also have a nutrition or dietary background, that's where they come from. So I think it's a bit of a superpower in a sense.

Can you tell us a bit about what led you towards the kind of data side of things and how you think your kind of nutritional background informed the way that you've worked with data?

Helena (15:09)
I think the most important for me is I still remember who is in the other side. I can still remember if I close my eyes, I still remember the face of some patients. I still remember them.

And sometimes I'm analyzing data, I feel like, oh my goodness, this is boring, you know, the math stuff, these things. But I remember them. You know, it's because sometimes it's like, you know, it takes ages for things to get in the correct path. And we do, and we do it again, and we need to repeat many times. But I remember them. And this actually helps me to have a more, a better approach to my work nowadays. Only seeing data, only seeing the other side of the forms. And I feel like this experience at the hospital I had, dealing with the health professionals, dealing with the patient itself and their families, and having these memories of their history and how it actually works in the field.

This gives me the background to nowadays, when I went to CIDACS, we do have access there to different kind of health data from the birth system, mortality system, the nutritional surveillance system, infectious disease system. So we do have this from the government. They do have access and they have this place for doing the analysis and anonymization and everything. But when we receive it after the linkage, I'm still able to see and imagine like, okay, so this came from that region or this community itself and it helps a lot to give a different power to my work. Sometimes I feel like, I don't want to.

Mark (17:11)
Yeah.

Helena (17:16)
But no, no, no, I remember them. I can do it. I will do it until the end. I think it combines.

Mark (17:24)
Yeah. And I think it centers you back down to what really matters. Always remembering those people who those data represent. I think it's great. I really do.

I think it's quite a nice segue as well. Because you work on the 100 Million Brazilian Cohort Study, which blows my mind, a hundred million. Like my laptop struggles to open up one email, let alone, you know, a hundred million bits of data. I mean, it'd be chaos here to be honest. Can you tell us a bit more about that study and what it's been used for?

Helena (17:56)
Yeah, yes for sure. And so then we have this baseline that now it's 143 million.

The baseline of the cohort it's the Cadastro Unico, is the unified register.

That is used in Brazil for people who want to apply for social programs. So if they fit in the poor or very poor part of the society in Brazil, you can apply for social programs. And to apply, you will go through the Cadastro Único. It's filling the forms where you give information from your...

Personal, individual information and from the household and your family. This is, unfortunately again, I would say we have this inequality and in Brazil is so huge that more than half of our population is in this unified register. And that's how we have our main big cohort at baseline with this 140 plus million people and because we have access also for all the nationwide level surveillance.

And with the now the math brains from CIDACS they do perform the linkage and with that we can have access. So then we build the cohort. We have the baseline with the social demographic data and all this information we are talking here from the social determinants of health usually we go and extract from the main base of the cohort. And according to what we want to study, according to our research questions, the linkage is going to be performed. So hospitalization data, mortality data, birth data, nutrition data, they will be linked in adding these to fill the gaps, let's say like that, and build the history. So, yeah.

I think it's a valuable tool. We usually see administrative data for economic purposes. Sometimes they want to measure things. But as we can use it for research and to actually see the invisible, we can really see the invisible because it's very expensive and very hard to collect data. And some communities, they are in rural areas, Quilombolas community, Indigenous communities. So with this kind of data, administrative data, we can make them visible. We can actually look what is going on with them. And that's the main, what I call it, the gold, the main gold of the data.

Mark (21:09)
The kind of public view in Brazil around it? Are people like it? Are people quite happy that their data are being used like this? What has the sort of response been?

Helena (21:17)
It's been viewed in a good way. So, there is one main study about this from CIDACS. It was a qualitative study where people were asked about how being used their data and those things, how would they feel.

It has a positive view and actually I would give you an example as an answer because recently, I'm telling you the experience of CIDACS recently people from the community, from the Black community we call in Brazil, Quilombolas, they reached the CIDACS researchers asking for some information.

Mark (21:44)
Excellent.

Helena (22:06)
They needed information from their communities because it's quite rare, very low studies, like low quality studies or low studies about. And they were asking for this as they were making an appeal for the Minister of Health for having better healthcare access and other things. And one of the policies as well for their communities. And then...

Using the 100 Million Cohort and using this linked information, we could have a broader view that they do need. We could show it. We could show that homicide was higher for men from Quilombola community. Suicide was higher when we compared to the general population for men in their Quilombola community. And many other things affecting their health related to preventable disease, so malnutrition.

Influenza mortality for these kind of things that we are not expecting to see anymore in 2026. They have higher rates than when we compare to the general population in Brazil. So they could come out with this report. It's like a policy brief, we call it bulletin in Brazil. They came out with this document and it was one of the tools they used to show to the Minister of Health that their health care, their health care facilities, the ones that goes to their communities, the health care practitioners, they would need a reinforcement. They would need more investment. And they got it. I would say, of course, it's not just because of the report, but the report was something that they could use.

And it was produced from this data from researchers in communication with the community. This was one recent thing that happened. And for me, it's quite important as we could see the use of the data changing people's lives, changing your policy.

Mark (24:32)
Yeah, I mean, it's great. It's such a good example showing how data goes straight into action. Even if we come back to "Can administrative data save your life?" I mean, these are great examples of how data are being used to help people and improve health and indeed, hopefully save lives to some extent.

So unfortunately we're at kind of end of time now, so just to say a really big thank you for taking the time to come and talk about everything, you've been fantastic. So thank you very much, Helena.

Helena (25:08)
Thank you, it was great to be here. Thank you so much.

Mark (25:15)
Okay, we're back with the boss, Shayda. What did you make of all of that?

Shayda (25:19)
Thanks, Mark. I thought it was really interesting conversation and I loved the conversation between both of you, especially on social determinants of health. I find that really, really fascinating. I loved hearing Helena's story about how listening to people sort of coming in and talking about how they're doing all the right things.

Think she mentioned this lady, she was doing all the right things and yet her blood pressure couldn't go down. And it was only in sort of looking at where she lived in this neighborhood with a lot of crime and poverty that it became clear that actually the environment was what was affecting her health rather than exercise or food or whatever it was that she was doing that she felt like was within her control. And I think you also shared a similar story of how somebody you worked with was similarly influenced and inspired by looking at social determinants of health from seeing people sort of coming in with these sort of the same issues and tracing it back to the environment. I'm a huge fan of Michael Marmot and similarly, it's his story too. It's just, I feel like social determinants of health is one of those things where it seems really obvious, but it's really easy to ignore and not really consider until you properly stop and think about it how everything is kind of linked. Yeah, I was just really, really fascinated by the whole chat.

And it made me think of a lot of our datasets as well to just kind of bring it back to ADR UK. Helena mentioned how the maternal health and education seems to be so influential to, and sort of health during pregnancy seems to be so influential to the outcomes of the child. And it just made me think of a dataset that we have at the moment called ECHILD, which is

Which has the information on like the maternal child link.

But how did you find it? What were some of the things that stood out to you?

Mark (27:41)
I think a few things stood out to me. So one would be always having people at the forefront of what you're doing and whether that's thinking about the real people behind the data or thinking about why this matters and how it can come back to that everyday person. I think that's just such a great approach to research, to thinking about how we solve a lot of these problems because data can be just numbers or...

Bits of text and you lose out the person in that and if you forget the person then you're never really going to deal with these issues.

There's a really wonderful study that came from the US a few years back that showed that even if we estimate things like medical progress or having new drugs, the benefits on our population are quite small compared to if we say tackle educational inequalities or racial inequalities. And so really just kind of points towards tackling these social determinants of health should be our focus because the real benefits to our population health are just so great.

Shayda (28:43)
Definitely. And I think something you said so rightly on the pod was the things that need to change are modifiable. So I think hopefully with the more research that comes out on these linked datasets, highlighting the invisible, which Helena so beautifully stated, can be super, super powerful.

But yeah, I'm afraid we're out of time. I could talk about this all day. Maybe I'll catch you after this recording and we can hash this out a little bit more. For now, that's everything from our episode. Thank you so much for listening and watching if you're watching us on YouTube.

Mark (29:29)
Until next time, stay curious about how your everyday data might shape society.