Professor Tamsin Ford CBE – ‘Supporting children’s mental health as schools re-open’
This was a live webinar recorded on Wednesday 8 July 2020 for ACAMH West Midlands Branch.
Accompanying slides are available to download, and a transcript is below.
Subscribe to the ACAMH YouTube Channel to be the first to see video abstracts and free lectures
About the speaker
Tamsin Ford is Professor of Child and Adolescent Psychiatry at the Cambridge University. She is an internationally renowned Child Psychiatric Epidemiologist who researches the organisation, delivery, and effectiveness of services and interventions for children and young people’s mental health.
She completed her core training in psychiatry on the Royal London Hospital Training rotation and at the Bethlem and Maudsley Hospitals and she completed her PhD at the Institute of Psychiatry, Kings College London. She moved to Exeter in 2007, leading a group of researchers whose work focuses on the effectiveness of services and interventions to support mental health and well-being of children and young people. She was an Editor for ACAMH’s journal CAMH six years, stepping down as lead editor in June 2014.
She recently moved to the University of Cambridge. Tamsin’s research covers the full range of psychopathology and agencies, practitioners and interventions that relate to the mental health of children and young people. Every interaction with a child presents an opportunity to intervene to improve their developmental trajectory. Her work has direct relevance to policy, commissioning and practice.
So, my name is Tamsin Ford. I’m a Child and Adolescent Psychiatrist by background and I’m currently working for Cambridge University. I’m unusual as an academic, in that I did all my clinical training, post-graduate training, before stepping sideways and getting involved in research. So, I worked for 10 years for the NHS full-time and over the past, sort of, well probably decade, if not slightly longer, the interface between mental health and education has become an increasing research preoccupation of mine because children who struggle with their mental health can have a big impact on schools and, likewise, schools can have a big impact on children’s mental health.
My background as a researcher is epidemiology, which means quantify, well, I think people, you know, I don’t have to explain it quite so much now after three months of Covid. I do communicable and non-communicable diseases epidemiology. So, mental health epidemiology. But epidemiology is the quantification of who has problems, what are the risk and protective factors, how those change over times and it’s, kind of, the basic science that underpins public health approaches. And that’s the kind of direction that my work takes.
We have had a series of very large internationally excellent surveys in the UK that we should be really proud of. We perhaps should have had more of them. The adults have been surveyed every five years initially and then every seven years, but the most recent of these surveys involved over nine thousand children between the ages of 2 and 19 and took place in 2017. This showed that of school-age and going up to 19. So, school and college one in eight young people had a diagnosable mental health condition that was impairing their ability to function. And, in fact, 1 in 20 met criteria for two or more conditions and they were more impaired. So, it’s well worth bearing that in mind.
This study was unusual in that it actually went down to the age of two. So these are the first national level data we have for poor mental health in pre-schoolers. And that suggested about 1 in 20 have a diagnosable mental health condition at that point. Now, the question we were answering for government was how many children out there needs services? So, that’s why I’m talking about diagnosable mental health conditions. I’ll come back to that point later. Now, I hope you’ll be able to see my mouse.
So, in 1999 and 2004 there were previous surveys, which because they were done using as close as we could do it, the same methodology. In fact quite a lot of the same team. We’re able to compare across time and what you can see is a deterioration in people’s mental health, in young people’s mental health. The slightly odd age range of 5 to 15 year olds is because that’s the age range in the first survey. So, we have to take the conditions we didn’t study first time round out and we have to take the age range that we’re stuck with. I think, actually, the size of the increase was much smaller than many people thought and you can see it’s there for both boys and girls.
If we look at the type of mental health condition, you can see that the increase is almost completely explained by an increase in emotional difficulties. So, anxiety and depression at clinical levels. And, again, that changes evident in both boys and girls. The prevalence of…or the proportion of young people who meet diagnostic criteria for a mental health condition increases with age, which is not a new finding but it gives us confidence in our findings that it’s there.
And there are differences in the age pattern between boys and girls with there being more boys than girls in preschoolers and primary school-aged youngsters who are struggling. Pretty much even stevens at secondary, and then the switch to girls struggling much more than boys in the later teenage years.
And, in fact, when you look at the type of mental health conditions, now remember that young people can have more than one mental health condition and actually by the time you get to the young people who are seen in Child and Adolescent Mental Health Services, more of them will have two or more conditions than will only have one condition.
So, about a fifth of those who meet diagnostic criteria in previous surveys have comorbidity in the general population. But it’s more like two-thirds once you get to clinics. And you can see in the younger children, it’s primarily behavioral problems and less common disorders in this group will be neuro developmental disorder. So, autism, ADHD, that kind of problem.
Whereas you get a switch to emotional disorders rocketing up in older teenagers, and that’s particularly amongst young women, which I’ll come to later. And by the time you get to the older teenage years, the less common disorders, actually, are more likely to be eating disorders. So, anorexia and bulimia.
The data are not widely available from this survey, which I think is a real problem. But there was led by ONS some analysis of background characteristics of the young person, their family and the neighborhood in relation to having a disorder. We could do this for having any disorder, or having anxiety and depression combined, or a behavior disorder. The other conditions, there are too few of them, to be able to do this kind of analysis.
Where I’ve written something in green, that factor changes with age. So, amongst primary school-aged children, boys were more likely to have particularly behavior disorders. Where I’ve written something in purple, it’s because it keeps cropping up. So, you’ll see poor parental mental health is something that predicts having a disorder, or is associated with having a disorder. Remember both the background characteristics and meeting diagnostic criteria were measured at the same time. But, socio-economic hardship is in there too.
And then there are shifting patterns by the number of children. Now the fact there’s only poor parental mental health related to emotional disorder, I suspect is because there are fewer numbers at this age and there is an issue with what we call statistical power. It doesn’t mean to say that other things aren’t important. It’s just that we don’t have enough to be able to tease out the relationship.
And then this is the same for the secondary school age groups. They didn’t do it. They didn’t do this analysis for the older children. Where you’ll see that for emotional disorder, you’re more likely to be a girl. For behavior disorder, you’re more likely to be a boy. And again socio-economic status is in there and poor mental health across the border.
Now, I think and with the previous surveys I’ve done analysis that have demonstrated this. The relationship between poor mental health in young people and in parents is resprical. So, it works both ways. But there is an implication in that when young people are treated by one service that might be in a different trust, or indeed second, you know, not even secondary care for many adults with poor mental health. But something we could do that might really help young people is to make sure that their parents who have poor mental health get the support that they need.
When I was training as a child physchiatrist, I had many social conversations when, you know, people found out what I did, that usually went along the lines of children, “don’t need a psychiatrist.” Well, I think I’ve just shown you some data, actually, that they may not need a psychiatrist, but they certainly need skilled mental health professionals.
And they need other professionals who work in their circle to be skilled at detecting mental health problems and plugging them into support. When I argued that people would then say, “Oh, but they grow out of it.” But when you look at the data, actually, they don’t. I put quotes around persistence in the title of this slide because actually the 99 survey and the 2004 survey will repeat it three years later. And this mirrors what was found in the very first…one of the first psychiatric epidemiological studies, actually, let alone to do with children.
The Isle of Wight study that Michael Rutter led in the early 1970s. When you go back to the same sample taking youngsters who met diagnostic criteria at baseline. About half of them meet diagnostic criteria three years down the line. We have a little bit of data between those time points, but they’re at, kind of, one year intervals. And, actually, these youngsters SDQ scores, most of them don’t come back into the normal range.
They may decline and I imagine inbetween that people are flipping in and out of that diagnostic boundary. What predicted baseline to having a disorder the second time around at follow up with these factors? So, you’ll see that disruptive…the more externalising problems are more persistent. That’s a very common finding. Peer relationships seem very important. They’re there for ADHD, they’re there for conduct disorder, there also there for anxiety disorder. So, markers of socio-economic status pop up for conduct disorder and for depression and for anxiety.
Now, some of these, also, are associated with having these problems in the first place but it’s not necessarily that what precipitates or predicts a problem. It’s, also, what predicts its persistence and it may be children coming from those backgrounds who need to work particularly hard. So, for example for ADHD, if they also have a neurodevelopmental disorder then perhaps we need to be working particularly hard to help them manage their symptoms of ADHD.
But as many of you know, and I’m getting the sense of quite a lot of people who work within education on this call. There is not such a distinct dividing line between children who meet diagnostic criteria and children who don’t. And, in fact, this is a problem for the whole of medicine. So, there’s lots of fuzziness around when you call wheezing asthma.
If you end up on ITU regularly every time you have a cold, then there’s no doubt you have reactive airways, and I think most people agree, yes, you have asthma. But there will be lots of people in the population who never seek help, who get wheezy when they exercise. For example, blood glucose. We put a dividing line and say at this point you’re called a diabetic. But, actually, the risk of having raised blood glucose increases with each level, it doesn’t suddenly shoot up.
So, it’s not something that’s unique to Mental Health Services. But what you see here is the self-report on the SDQ total difficulty scale, and it’s beautifully normal distribution like height, weight. Most youngsters are in the middle, a very few are absolutely thriving and, in fact, our scale doesn’t quite, you know, there’s a missing bit there. It doesn’t quite cover it all and then, you know, there are some who are doing really badly. The scale goes up to 40. So, very high scores are very rare.
This is the parents and this is the teacher. Now, as an epidemiologist the fact that they are asked exactly the same questions and these curves are very different, I find fascinating. Indeed, the only difference between the parent and the teacher SDQ is the P and the T on the top left-hand of the form. The questions are exactly the same.
The blue arrows indicate the cut point and that we don’t have a distinct population above the cut point and I think clinicians don’t just count problems or count difficulties when they’re making a diagnosis or formulating the predicament that a young person who’s come to them is in. And, in fact, what makes the difference about whether or not you think this is a clinically relevant problem is as with the rest of medicine things that relate to the child’s predisposition, so family history, their family context, their own characteristics, like, you know, how is their level of intelligence, do they have any physical health problems and also their school context can make a big positive or negative difference.
And it’s those things will make a difference between a child whose scoring above the cut point, but you think no they’re not impaired and I wouldn’t count that as a disorder against a child who may be scoring below the cut point, but, actually, is really struggling to function and to do what they have to do.
But that said there are lots of people who will argue that if we get public mental health approaches in terms of prevention and mental health promotion right, maybe we can shift the curve. I think the other thing these graphs tell me is that teachers are reticent in reporting problems. And this ties in with the research I’ve got. They will tell you a problem is there if they’re absolutely sure, but they’re cautious, which means we should listen to them as mental health practitioners.
So that’s where we were before lockdown, and I think there’s…I’m sure you may well have been approached to complete them. I’m sure you’ve seen them going around the internet, there are, you know, people have talked about an epidemic of mental health and other monitoring scales that have gone round. We really lack high-quality data on children.
So, I had a conversation with people at the London School of Hygiene who are doing all the modeling saying where can we get some data on children? We have very little that has baseline samples and the problem with these convenience samples that go out via Facebook or Twitter, or snowball by word of mouth is that they’re often filled in by healthy volunteers. And they’re often filled in by people who are worried.
They’re not a carefully selected sample frame and you don’t know who you haven’t got. You don’t know who saw the post on Twitter or Facebook and didn’t click and do the survey and you’ve obviously excluded anybody who isn’t online. That said, things like the coast based study has given us lots of really useful information. They’ve got a fantastic website and they can tell you about some of the things that young people and parents are worried about.
What the data from previous epidemics suggests is that anxiety depression and for those that have been really badly ill, or traumatically retrieved, trauma are the things that we should be looking for. I think in terms of going back to school, Cathy Cresswell and Polly Waite have asked teenagers and parents of young children, “What are the worries about this?” And they’ve come back. There’s a lots of teenagers who are really worried about being behind at school.
They’re really missing their friends. So, you know, I would encourage anybody who’s interested in that to put co-space into Google and go and get the details of that study and they’re going back, they’re doing waves every month and initially it was every week, but we don’t know if we’ve got the same people each week.
And because of the fact that it’s a convenience sample, because it had to be, we don’t know what it tells us about prevalence. Because we don’t know who the sample is and we don’t have baseline data. The nearest we’ve got to baseline data yet in the UK is two studies, only one of which has any teenagers in it. And it’s this one which is available as a pre-printice coming out in Lancet Psychiatry. So, this is data from April 2020. In a panel of adults from age 16 at the baseline who have been filling in this survey and they’ve been involved between six and nine waves collected over the last five years.
And amongst that group, young people did come up as having more psychological distress. As did women, particularly young women, and also those in adverse circumstances. Now, that last one is boring because actually we’re about to see a lot more people take a financial hit and there is a strong association between debt and poor mental health. I should say that, actually, this gradual deterioration in mental health over the last five years was there for everybody. But in these groups, the deterioration in mental health accelerated and did so significantly.
The other survey used the aspect survey, but the average age of those people is now 28. So, it’s not children. They are beginning to follow their children, but it’s not a systematic sample and they didn’t in April. They didn’t get any data from the children, and also data from a survey in Scotland called, ‘Generation Scotland’, they’re in their 50s. So, it tells you about the parents. It doesn’t tell us about the children.
So, what about going back to school? Well, if children’s mental health is deteriorating and it might well be, we should be worried. This is an association with parent report on the total difficulties questionnaire, and being excluded from school three years later. And what you can see is the more children who are struggling, there are increased exclusions in school.
Now, that’s a story about schools struggling to contain young people and the coast space study is talking about more parents complaining about behavioral disturbances in primary school children. That has implications for schools and for parents and I should say this work was done. The photograph of the young woman in the right-hand corner is Claire Parker, who was my PhD student at the time she did this work. And she’s now a clinical psychologist in Exeter.
We also looked from Baseline in 2004, amongst the children who had a mental health condition that was either recognised by a parent or teacher or both, or not recognised. And what we showed was adjusting for all these different background factors, and sadly the children who have learning disabilities who are facing adversity, you know, all these factors are raised amongst youngsters who are excluded if you look at government statistics.
But having accounted for them statistically and looking forward to 2007, two to four times the number of exclusions. And, in fact, we detected a bi-directional relationship. It works both ways. So, the youngsters who didn’t meet diagnostic criteria, three years later one who’d had an exclusion, three years later were more likely to meet diagnostic criteria. And often there problems are long-standing.
So, this is data from the Asbach Cohort, and what you’re looking at here is SDQ scores, okay, collected at these ages. So, between age 3 and 16. This graph is girls and this graph is boys, and the blue line is children who hadn’t been excluded by age eight, and the red line is children who had. Now, there were very few of these and probably too few girls to study, and the lines don’t look that different.
But the actual estimate could be anywhere on that line, and you’re seeing very big lines and bigger than the boys because there were fewer girls and that’s just a statistical artifact. But what you see with the boys is they come in with slightly raised SDQ scores at three, and, actually, they stay raised all the way up to the…you know, so we can see these young lads coming. That means we should be able to intervene and some of them may need specialist provision. But let’s make that about meeting their needs.
This is the same data. So, again blue is no exclusion, red is exclusion. And these are girls and these are boys, and this is for exclusion by age 16. And what you can see here is we have a picture of deteriorating mental health amongst those who are excluded. So, thinking about mental health, you know, schools sometimes have to exclude pupils for people’s safety, but, actually, it should be a wake-up call to think about mental health issues.
Attendance is associated with, particularly, emotional disorders. So, this is data from the 2004 survey. This was completed by Katie Finnings, this analysis. She’s also done a fantastic series of systematic reviews where she’s reviewed all the literature. And one of which is in the ACAMH Journal of Child and Adolescent Mental Health on anxiety disorders.
There’s one in the Journal of Affective Disorders about depression and one about to be submitted on the, kind of, softer concept of emotional difficulties. But what you see here is, red is secondary schools and blue is primary. And there’s a very strong association with depression. I think the difference by age is that they’re very, very few primary school children who actually meet formal criteria for the depression.
Then we’ve got emotional difficulties and you can see it’s amongst the slightly raised STQ score on the emotional difficulties subscale. It’s not that different. But once you get to high and very high, you get the populations moving apart. That’s related for any absence, and this is for unauthorised absence. And one of the surprising findings, when you think of the old concept of school refusal being related to emotional difficulties and truanting being related to behavior, is actually, we found a stronger relationship between unauthorised absence.
So, this works at an early stage of preparation. There is a hint when we look backwards and forwards of a bi-directional relationship in this sample, but we haven’t managed to replicate it in a primary school sample that used a trial run by the Manchester Education Faculty.
What we did show there was that youngsters who had emotional difficulties were more likely to have poor attendance going forwards. We didn’t find it the other way around, but I think the take-home message is if you have children who are not attending and particularly if it’s a change in attendance, have a think about emotional difficulties, anxiety or depression because they may, you know, it’s difficult to know what’s going on in someone else’s head and emotional difficulties are difficult to pick up in the school context.
I think particularly vulnerable groups to consider as schools go back for all of us working in relation to children. Those with pre-existing mental health conditions have had the most amazing… we’ve all had to adjust massively, but that is going to be harder if you have a pre-existing mental health condition.
Those who live in difficult home circumstances as, you know, we as somebody very eloquently said, “We might all be in the same storm, but we’re not in the same boat,” you know being in a nice house with a garden and your own laptop that you can link with your friends and get your lessons online, is not the same as being in a family of five or six in a three-bedroom flat with no garden and one laptop between all of you.
Some people will have had Covid adverse experiences. There is increasing evidence that infection itself, even if mild, might have central nervous system effects. Very few children have been hospitalised, thank goodness. Very few children have ended up on ITU, but trauma, post-traumatic symptoms after ITU admission is not uncommon.
Some children unfortunately will have been bereaved, and some will be in families where there is sudden financial strain. I think there is a group of children anecdotally, and we’re trying to measure it who, actually, are doing better during lockdown. And I think as I joined there was a discussion about, you know, flexibility and timing being useful for some children.
If school is a hard place for you to be. Not being at school providing home is an alright place to be, might actually be easier for you. And I think, particularly, youngsters with autism spectrum conditions may really struggle to get back into school because of the difficulties handling change, and because also what they’ll be going back to will inevitably not be quite the same and these are a group of youngsters who really struggle with that.
Those who have anxiety based difficulties attending school often struggle after half terms and holidays, let alone six months off and so I think schools and mental health services are going to need to think really carefully about how we support schools and families to make sure that those children don’t miss out more.
In terms of managing behavior, I just want to highlight the education endowment funds practitioners tools. It’s based on three systematic reviews that I was part of, and it’s got very simple practical evidence-based guidelines. I wanted to pull out this diagram from the influences on behavior. So, there are positive and negative influences that were all exposed to all the time, and there are negative and positive behaviors.
And what you hope is the bulk of pupils and children are in this quadrant with positive behaviors and lots of positive influences. But, actually, as you go through life, we all have things we have to deal with and we will all fluctuate about which quadrant your in at any particular time.
Now schools I know sometimes feel like they’re being held responsible for everything, and I think in this systematic review we highlight things that schools can marry and manage directly, and they’re in green. So, relationships in schools, both with between pupils and between pupils and teachers.
The teaching and learning environment, the school context can make a huge difference to children and young people’s well-being. And promoting positive attitudes and self-concept. All of that school can directly influence. There are some things that you can identify and maybe highlight a bit, and then there are other things that really aren’t your business to do anything other than be aware of and perhaps act on. So, sometimes they’ll be cultural clashes between home and school.
Sometimes they’ll be things in the home life that you can’t influence, but you are aware are a stress or are difficult for youngsters. So, for example, young people who are carers or young people who are perhaps having to work outside school in order to bring money into the family, for example. That, you know, is inevitably going to make them more tired, inevitably going to eat into time when they could be working.
So, I think there are six recommendations that I encourage you all to go along and check and I’ll make my slides available at the end, if people want them. I wanted to talk a bit more about this using classroom management behaviors to support good classroom behavior. But, in fact, in the trial that I did, it also promoted mental health in general, particularly for the children who are struggling.
And the program we tested was The Incredible Years Teachers Program. So, this is based on very established empirically demonstrated theory of operant conditioning. What you give attention to, you will get back too. So, trying to ignore as much of the low-level disruption as you can, but also modeling the, kind of, behaviors that you want children to produce and the importance of having warm nurturing relationships with social animals.
So, Carolyn Webster-Stratton who developed this program talks about the piggy bank of goodwill, that will make children more able to hear criticism. It’s six days over six months. So, teachers have a chance to go back and practice in their classrooms. You start from building positive relationships and being proactive and thinking about what certain children struggle with.
Training your attention on the positives, correcting things in a positive way, using incentives and then focussing on decreasing inappropriate behavior with ignoring and misdirecting, following through and then finally thinking about emotional regulation social skills and problem-solving.
So, this is a logic model and I’m going to break it down for you. And I hope you can see it clearly enough. Now, we talked about the manualised curriculum. It’s delivered in groups. The idea being that everybody in the group is an expert in their own context. This is about honing skills and, actually, teachers in our qualitative process study said actually it was very validating. This was stuff they knew, but they hadn’t thought of for a while and it allowed them to, kind of, sit back and reflect.
And that’s exactly the way it’s delivered. And the idea is it’s a safe space, it’s collaborative, the group facilitators we chose all were teachers. And their expert in the program, but they’re not, sort of, aiming to teach teachers how to be better teachers. It’s about coming up with solutions and trying them out.
So, how do we think it works? I think the teachers told us and we expected there to be an improved focus on children’s perspectives and where children are coming from. The teachers move from being reactive to being more proactive, that they felt calmer and more under control and that you then got a virtuous circle with positive impacts on the teacher, on the child and the whole class as a whole that led to shared outcomes with improved relationships and a more positive attitude to learning.
And in our trial, we showed a small but significant improvement in children’s mental health for everybody, but a larger one that was sustained over 30 months for children who struggle. Which I think in itself makes this worth doing. And, in fact, this is the program that our systematic review demonstrated increasing evidence for and a similar pattern. The impact on everyone overall is tiny, but the impact on those who struggle is bigger. And, in fact, we’re halfway through a re-application, which fortunately hasn’t been too badly disrupted by the shutdown. So, there should be more UK-based evidence.
I also wanted to talk about school-based services and counselling. So, I have a link with the charity place to be and this is…I’m sharing their outcome data. So, one of the things, and only one of the things they offer is face-to-face counselling and, so, here we have pre-counselling strengths and difficulties questionnaires from teachers and from parents, and what you see is there’s a big drop down to after the counselling’s finished that, actually, seems to be maintained over the following year by both informants. But that doesn’t give us a control group. We don’t know if those children would have just got better anyway. We’ve talked about how mental health fluctuates over time.
So, what we’ve done is compared children from place to be in red, with children from the British Child and Adolescent Mental Health survey in 2004 looking at follow-ups going out up to, actually, 2007. And what you see, we very carefully matched on as many background characteristics as we had available to us, and what you see is a similar pattern.
There is a bit of a drop in the national survey from children’s scoring about the same at baseline. But those who have counselling, it is bigger and it’s more sustained, and actually it is no longer significant after two years. But it looks like the trend goes down. It’s just our confidence intervals, because we have fewer data points become much wider.
So I wanted to highlight this amazing paper by Mark Greenberg saying that, actually, when you’re dealing with things on a whole school level or any kind of preventative approach, you’re not going to get dramatic effects like you do when you’re dealing with your measuring interventions in the clinic. But, actually, that’s because for some parts of the population, you are promoting mental health. The same program for other more vulnerable children might be preventing them slipping into a mental health condition.
And for those who are already struggling, you might be treating them. So, you’ve got all these sub-groups in there. And I think this will apply to what we’re dealing with post Covid, both in terms of learning and in terms of mental health. So, they’ll be the vast amount of the general population who will be unsettled coming back after such a big gap, as we all have been.
But many of them will be okay with the support that we can put in around them. Then there are a group of children, who because of learning problems, social problems and health problems, both mental and physical health, cannot cope in mainstream and need alternative provision. But there’s this vunerable group, which I think will be bigger because of Covid, and they’re the group that we want to really intervene to try and keep them within the mainstream school and get them over this bump. I’ve put up some useful information about mental health, and then I’ll provide some references and I’ll stop talking there. Thank you very much for your attention.