Pandemic Impacts on Mental, Physical and Educational Outcomes for American Children: Session Two

SPEAKER 1: Good morning, again. We are going to come back together. So I am pleased to welcome my colleague, Kelli Marquardt, economist here at the Federal Reserve Bank of Chicago. And she is going to talk about societal disruptions and childhood ADHD diagnosis during the pandemic. Thank you.

KELLI MARQUARDT: Thank you. Thank you all. I'm going to be presenting this paper, which is together with co-author Seth Freedman, Dario Salcedo, Kosali Simon, and Coady Wing, who are all at Indiana University. And the title of the paper is "Societal Disruptions and Childhood ADHD Diagnosis During the COVID-19 Pandemic."

So the motivation started out very broad. We had all of these news reports, articles coming out about rising child mental health concerns during the COVID-19 pandemic. It looked like things were getting very bad for child mental health.

A lot of survey reports of worsening behavior and mood, perhaps even bigger declines for places that were not going back to school. So we thought maybe school was having a play here. However, there was some other research that came out that showed that there was an initial drop in teen suicides when schools closed, and it actually went back up when schools went back in person, kind of getting at this bullying aspect of schools.

Then there's this other strand of research that Dana has some work on as well, that-- she's going to be my policy reflector-- kind of showing declines in maltreatment reporting, not necessarily that maltreatment of children went down, but that schools are an essential input to identifying cases because teachers and school personnel are a first order in reporting maltreatment, which got us thinking about tying these two together, mental health and the role of schools and teachers in identifying child needs.

What about Attention Deficit Hyperactivity Disorder or ADHD, which is a behavioral mental health condition very commonly diagnosed in children? But also teachers and school personnel are a key input to recognizing symptoms and getting clinical diagnosis.

ADHD recognition in childhood is important. Again, it's one of the most common diagnosed conditions. About 10% of school-aged children are diagnosed with ADHD.

It's subjectively diagnosed. There's not an imaging or a blood test that shows if an individual has ADHD or not. It's based off of a list of symptoms that are present and hindering the children's development.

These key symptoms are impulsivity, hyperactivity, and inattention. And so you can already tell these are probably often identified in school settings.

Research has shown that ADHD has been linked with worse outcomes in multiple domains. So children with ADHD often have lower educational attainment, increase in risky behaviors, and some impacts of downstream labor market success.

However, a lot of these symptoms can be managed. It does require identification and clinical diagnosis in order to get treatment. These are treatment via medication management, behavioral therapy, and academic accommodations, which can help children with ADHD.

So in this paper, we asked, how has the pandemic and the associated societal disruptions affected new ADHD identification in children? So we're looking at the identification and diagnosis of ADHD.

So what we do is we compare the cumulative ADHD diagnosis rate for children exposed to the pandemic to same- age children prior to the pandemic. So think of we're comparing first graders who were in first grade in March of 2020 to children who were in first grade of March of 2019.

We didn't actually kind of know-- I kind of went into this thinking, oh, ADHD diagnosis for sure increased. This was a wild time. It's very easy to get distracted online. It had to have gone up. But wait, teachers are an input. They were very busy, probably not identifying cases, so maybe it went down.

So there was this empirical question. And I'll get into the mechanisms in a little bit. But we do find a decline. So we found a reduction in new ADHD diagnoses. We're going to look at it separately for boys and girls because there is a big difference. Boys are much more likely to be diagnosed than girls.

We found that there was a 8.6% drop in the diagnosis rate for boys and a 11% drop in the diagnosis rate for girls. And we followed that into February of 2021.

We would then use Safegraph mobility data to test for the importance of in-person schooling, particularly in the fall of 2020. We actually compared this to the data that Becky and her co-authors put together. And for the states that we have overlapping information, it's pretty much one to one. I'll talk about some of the nuances if I have time later.

We do find that higher levels of in-person schooling dampened the decline for girls. So girls in states that went back to school did not see as much of a drop in ADHD diagnosis than girls who were in states that did not go back. However, we didn't find actually any moderating effect for boys, so the diagnosis rate for boys pretty much fell universally.

So I want to spend a little bit of time because initially we went into this of just what happened to ADHD. And it's very much more nuanced than how we went into it. In order to get diagnosed and to see you in the data, three things have to happen-- symptom development. You have to develop the symptoms of ADHD.

Behavior recognition-- somebody, typically a caregiver, a parent, teacher recognizes the symptoms and suggests or refers you to a pediatrician or a psychiatrist or psychologist. Those two things have to happen. And then you also have to go to the doctor and get a clinical diagnosis.

Of course, there's a fourth column here that is treatment and symptom management. But that, of course, cannot happen until these first three things do.

So when we're thinking about how were all of these things impacted potentially by the pandemic and particularly due to school closures, starting first with symptom development, we think there's literature showing that increase in stress can exacerbate ADHD-specific symptoms. So we think that, of course, it can make symptoms worse, which would then lead to an increase in ADHD prevalence.

Online learning may have had actually opposite effects, which has come up here a lot. Possible that it does require more attention and increased opportunity for distractions when you're on your iPad and can quickly click away to a game or something. That might increase some symptoms of ADHD.

At the same time-- and I've seen a picture earlier-- there was limited peer distractions. Even if you were in person, you were sitting very far away from other people. And so potentially limits the potential to show symptoms of ADHD.

Behavior recognition-- on one hand, caregivers may have a new view of child behaviors. So they might work from home, and they get a new view of their children's always distracted, whereas normally they were distracted at school. So this could increase.

At the same time, there was limited face-to-face interactions with teachers and school personnel. And even if they were in person, teachers might have been preoccupied with other things going on at the time, trying to increase learning, and so that might reduce referrals. So those may have gone down. So again, opposite effects.

Clinical diagnosis-- of course, early on in the pandemic, they delayed appointments or canceled appointments, really focusing on the emergency and essential care. And so there could have been canceled or delay assessments, which would reduce diagnosis rates.

We are able to control for that, but that's the only thing that we're able to control for. So this decline is not-- the decline that we document of ADHD fall is not explained by the drop in utilization of health care. But other things, all these other things can be leading to it.

Of course, there also could have been a reduction in teacher responses to rating scale requests. Typically, doctors will reach out to teachers or school personnel to ask them to fill in these different rating scales to get the level of ADHD symptoms in the classroom.

And there also could have been some crowd out from other conditions, the idea being that there was a lot of other mental health concerns going on at this time. And so perhaps that's what doctors or clinicians were focusing on.

So again, all of these things go together. And it's an empirical question. Some things may increase ADHD, and some things may drop it. And so we're kind of curious about what's happening in the data.

So what data do we use? We're interested in what's happening nationwide. So one of the data sets we have is Optum's Clinformatics Data Mart database called Optum. It's medical claims for approximately 20% of the commercially insured population nationwide.

And so we can get diagnosis rates of ADHD across all the states. However, this is only for the commercially insured population, so it does not include those on public insurance, such as CHIP or Medicaid. So we complement this with a single state's data set, which is the Indiana Network for Patient Care, the INPC.

This covers health records for providers through the Indiana Health Information Exchange, includes most major health care providers in the state, and importantly, includes all payers or all insurance types, so we can look at what's going on with Medicaid population as well.

Then we look at school openness. So these maps kind of might look familiar to what you just saw earlier. We do use Safegraph data, which is smart device activity by location and date. So it's kind of the cell phone pings. How many cell phone pings are happening in certain locations over time.

We focus on cell phone pings to public elementary schools. And we're going to look at just the openness of schools in the fall of 2020. So did you go back, basically, in person? And we're going to look at the average monthly activity in the fall of 2020 relative to what it was in 2019.

Again, this is very similar to the high, medium, low openings or very similar to what's in the COVID data hub. One of the benefits here is we have all of the different states. And so we group them into low, medium, and high.

Low opening are those that were less than 54% of their activity from last year. And then the higher, those that were at 70 or above percent of where they were last year. So those are mostly in person. And medium's in the middle.

We do a similar thing, but for Indiana, we look at it by counties, so whether counties were high, medium, and low. One thing I want to note is Indiana is a high-opening state. So even within Indiana, the low-opening counties still had up to 76% of activity. So a lot of activity going on.

So we're interested in identifying new cases of ADHD diagnosis, the new identification, which often happens between ages 6 and 11, so in elementary school. So that's what we're going to focus on.

We're going to do a cohort-based comparison separately by gender, again, because boys and girls have different diagnosis rates. What we're going to call the unexposed cohort is we're going to follow children from August 2018 to February of 2020. What did their diagnosis rate increase look like?

And then we're going to compare that to the same-grade kids one year later. So starting in August 2019 to February of 2021, right smack dab in the middle-ish is the start of COVID, a little bit in the beginning. And we're going to look at the change there.

We're going to look at a balanced sample of ADHD-naive children. Essentially, these are children that have not yet been diagnosed with ADHD. And we're going to say you've not yet have ADHD if you don't appear in the medical databases as having a diagnosis in the last six months.

This is kind of our findings in one picture. So what this graph is showing is in calendar time or in month time. So in August of a given year, September of a given year, what's the cumulative diagnosis rate for boys and girls separately and then for those that were exposed to the pandemic versus what happened in the year prior.

And what you can see is in a typical year-- look at that. The black squares filled in are the unexposed cohort boys. So in a typical year, about 6% of boys who started out as not having ADHD in August will be diagnosed with ADHD by the following February.

So that's what you think about typically. In the pandemic, so starting in August of 2019, boys were following the same increase in diagnosis rate. But then right when the pandemic started, you kind of see this drop off. So there was a drop in the rate of new kids, new boys being diagnosed with ADHD. So this pandemic cohort has a persistent shortfall in the new diagnosis.

The orange is, again, for girls, it's a little bit harder to see because girls have a lower diagnosis rate. So maybe putting them on the same graph here is not ideal. But there is this drop there as well.

Now, this is nationwide across all boys. So when I get to the empirical analysis in the next slide, we're going to be comparing-- we're going to add some additional controls for changes in health care utilization. We're going to be having fixed effects for grade, state-- I didn't specify-- grade, state, race, cohort, and month.

So what we're doing is an empirical strategy. What we're doing is difference in event study designs. So essentially what we're doing is we're calculating the relative difference in the cumulative diagnosis rate between the two cohorts.

So we're not measuring that drop in the previous graph. We're changing the difference from one month to the next to the one month to the next in the open dots.

We're going to use February as our reference month. So how you interpret this is starting in February of 2020, we test if the monthly ADHD diagnosis rate increased at the same rate as it did in the prior year. You can kind of tell from the picture the answer is no. We actually see a drop.

Again, we can definitely shut down the channel of whether or not you were going to the doctor to get a diagnosis. We control for changes in health care utilization during these months, looking at changes in child well visits and adult evaluation and management visits to control for general utilization.

And then within Indiana, we can also control-- so in the nationwide, again, it was the commercially insured. So within Indiana, we can also control for health insurance composition. So within a county, what percent are on Medicaid and what percent are on private.

We're going to run the analysis separately by gender. And then we're also going to test for outcome differences based on the fall 2020 in-person school groups. So we're going to compare states that-- what was happening to the diagnosis rate in states that were open versus states that were closed.

So this is what those event studies look like in picture form. So focus first on the boy graph, which is on the left. What this is plotting is essentially the difference that I was saying in the month from month in the exposed cohort, from the month to month in the unexposed cohort, the change in those two rates.

So what this is showing is that leading up to February of 2021, diagnosis rates were similar in the sense that there's-- if they cross the zero line with the confidence intervals. So they were trending at a similar rate. And then after the pandemic started, the exposed cohort had a much slower increase in ADHD diagnosis than what was happening in the year before. And that you can see by that drop.

This is true for both boys and girls. And what the green triangle is showing is what happens when you control for the changes in health care utilization during this time. It does explain a very tiny bit, but it's still much below zero. So there was a big drop regardless.

So again, there was a very sudden fall-off in diagnosis rates following the start of the pandemic not explained by the fall in health care utilization during this time. When you put these all together in a difference in difference framework, we get an 8.6% drop for boys and a 11% drop for girls.

And interestingly, they do not appear to rebound by the end of the sample period. It's consistently lower by February of 2021. Of course, we can't go into the next month because once you go into March of 2021, you're now comparing it to March of 2020, which was the start of the pandemic.

Then we look at this differently by school opening, by those states that were high, medium, and low. So black circle is the low-opening states. Blue diamond is the medium-opening states, and the orange triangle is the high- opening states.

And you can see, for boys, it's pretty consistent drop regardless of the school opening. For girls, there was a big increase. There was initial drop, but then girls in high-opening states kind of returned back to the diagnosis rate that it was pre-pandemic. So back to that zero line. So it does suggest that perhaps in-person schooling is an essential input for ADHD diagnosis in girls.

We do a bunch of additional analyses in the paper. I just want to highlight some of them here. Again, when we look within Indiana, which is a high-opening state, things get a little bit more noisy, but qualitatively, things are very similar.

Big drop for boys, not as much of a drop for girls, again, because Indiana is a high-opening state. So it's exactly one of those orange triangles on this graph here. The benefit is that we can show that it's not driven by differences in the commercially insured versus Medicaid population.

One thing that we do do in the paper is look at differences by school stability, which is whether you're always in person or always closed, which might matter. And one thing that we found here is that the decrease in ADHD diagnosis is smaller in states with higher school stability for boys.

So it looks like stability of the schooling matters. So it doesn't matter if they're always closed or always open, but you need to have higher stability of the schooling mode.

We also look at differences by race and ethnicity. And we find that the biggest drops in diagnosis rates were among white, non-Hispanic and Hispanic children. However, the decrease in ADHD diagnoses is smaller for non- Hispanic Black children. And while not statistically significant, there was even potentially point estimates of an increase in the ADHD diagnosis rate among Black children.

But I want to specify that there's potentially some differences in the pre trends, what was happening in the diagnosis rate before COVID, which can kind of be seen here. This is the percent of children that are diagnosed with ADHD over time. And you can kind get a sense that within these different race and ethnicity categories, rates were picking up or slowing down over time.

So what have we learned? In this paper, we documented that there were significant changes in the cumulative ADHD diagnosis rate during the early times of the pandemic. We examined how the lack of in-person schooling may have contributed to the fall in the diagnosis rate. We think that it did contribute.

In-person schooling appears important for ADHD identification in girls, but not for boys. And, in fact, it might be the stability of the schooling mode that matters for boys in terms of getting identified as having ADHD.

Of course, we stop in February of 2021. So I want to use the additional five minutes I have here to talk about can we go past that. We can't with our specific data sets that we use in the paper. But what we do have-- it just got released, I think a month ago, maybe less-- is the 2023 survey results for the National Health Interview Survey, which is a national survey. So we can look at responses to this survey in 2019 all the way to 2023.

And we're going to use this to look at have ADHD diagnosis rates rebounded since the drop that we recorded? And then what about overall mental well-being during this time in general? So we're going to look at the answer to two questions on the survey. The first is about ADHD, and it just asks, has a doctor or other health professional ever told you that your child has ADHD?

And then we're also going to look at overall mental well-being, which is the response to the question, how often do you feel worried, nervous, or anxious? Would you say daily, weekly, monthly, a few times a year, or never?

So starting with the first question, have ADHD diagnosis rates rebounded? Again, this is not an exact comparison to what we did because we were looking at cumulative new diagnosis rates. This is just overall percent that have said they're diagnosed.

What we see is we're splitting it by age group 5 to 11, which was the group that we focused on in our paper, but then the older kids as well, again, by differences for male and female. And you can kind of see, especially for males in the lower age group, there was that drop and then some evidence that it is rebounding back up.

This is even true for the older kids as well. There was this drop, but then kind of going back up and maybe even potentially above trend.

Females in the older group, actually looks like there's maybe been a longer time period of a fall and then kind of back up in 2023. So it'll be important to look at what's going on in the upcoming years to see what's going on with ADHD here.

And then just thinking about overall mental well-being. So what this graph is showing is over time, what percent of kids report feeling worried, nervous, or anxious, weekly or monthly and then those that have severe or daily feelings.

And you can see, at least for those that report feeling these anxiety weekly or monthly, there was this big increase in 2022. It did go back down. And the levels of severe remained pretty flat.

One thing I want to highlight, though, is there is a difference by age and by gender in this category as well. So, in fact, the lower age group, we saw this increase, but it was pretty similar for boys and girls. This older age group, so the 12 to 17-year-olds, you can see that females had a very big uptick. Almost up to 40% of females in this age group were reporting feeling anxious weekly or monthly.

And then even this severe daily worry went up as well. So maybe potentially some promising things that it's going back down, but definitely something to think about and look at.

So I documented what our paper showed, what we can learn from additional information, but I have more questions to consider. How have test scores and other academic outcomes been impacted by the decline in new ADHD diagnoses? I showed ADHD diagnoses fell. Rebecca showed that test scores fell.

Is there some relationship between these two things if ADHD fell? It's not that we necessarily think that prevalence fell. It's that they're not being identified and getting academic accommodations or treatment.

Are there other long-run impacts of this decline that will be realized in adulthood? Potentially, if we don't see that kids that really do have ADHD are getting accurately identified, this could have long-term impacts, particularly on labor market participation and productivity. So it'll be important to think about that.

How were parents and other adults impacted by these changes in child mental and behavioral health? We know that parents were impacted by the school closures and the pandemic more broadly. But they also may have been impacted by changes in their children's well-being.

And then finally, which can get us motivated to the policy reflection is, well, what do we do with this information? It kind of depends. There is some evidence and anecdotal stories out there that ADHD is potentially overdiagnosed, especially among boys. So this might actually be a good thing.

It depends. Are we correcting an overdiagnosis problem or adding to an underdiagnosis problem? And that might differ between boys and girls, age groups, race, and ethnicity as well.

Depending on that, that will help us better understand what policy tools, both in schools and out, can help facilitate both accurate and timely identification of behavioral and mental well-being in children. So I think that tees up nicely to my policy reflector. So thank you.

KRISTIN BUTCHER: I'm not the policy reflector, but I'm going to introduce the policy reflector. I'm Kristen Butcher. And I'm vice president and director of microeconomic research here at the Federal Reserve Bank of Chicago. I have the great fortune of having Kelli on my team.

And I think we have a fantastic policy reflector for this discussion. It's Dana Weiner, who has a PhD in clinical psychology and is a senior policy fellow at Chapin Hall. And she has 20 years of working at the nexus of implementation and research on behalf of children.

So I think she's the perfect person to do this. And in 2022, Governor Pritzker named her as the director of children's behavioral health transformation, which is an unassailably cool title. So please join me in welcoming Dana Weiner.

DANA WEINER: As others have said, it's an honor to be here today, especially to converse. I had a couple slides that aren't included there in the deck I sent, the later deck. But that's OK. I can just talk about them. I sent a subsequent deck that had seven-ish slides, but we'll just use these.

Anyway, it's thrilling to be here to have the conversation with other disciplines thinking about these things. And I'm going to start by highlighting what I think are the key contributions in Kelli's work. And then I'm going to talk about the implications.

And I have the luxury of talking about concrete example of the policy implications from where I stand in the governor's office working on transforming children's behavioral health.

But first, I just want to spotlight the things that I thought were really important about this paper. One is the differentiation between the actual occurrence of the phenomenon itself, the degree to which kids actually have ADHD, distinguishing that from the ability to detect them, which depends largely on school personnel and distinguishing that from the degree to which it's documented.

Because those three things-- we love it as researchers when those are all the same number and we have data that is reliable and valid and we know that we can count the syndrome from the reports. But it's often not the case. And I'll refer to an example from child welfare in a moment about that.

But I just want to highlight the other two things that I really loved about this paper. One, the importance of disaggregating data, which I think Dr. Harris talked about earlier, the idea that lumping things together obscures our view on what people need, and that it's really important to look separately at gender, to look separately at race, to try to understand what are the different dynamics.

With ADHD in particular, we say it all as one word, ADHD, but it's Attention Deficit Hyperactivity. It's usually the behavior part, that second part, that causes teachers to notice ADHD. They're not usually walking around the classroom noticing that a kid is quietly zoning out or not paying attention. They're noticing when a kid is disruptive, not sitting in their seat.

That often happens more with boys than with girls. Or if it happens with girls, it's kind of-- you have to get to a higher threshold. So understanding that dynamic, I think, is really important, too.

And then the third thing I want to highlight is the possibilities of understanding things from this rebound. We had this unprecedented experience of the pandemic. And we start to see things come back. And there's so much more to understand about it.

I was particularly struck by the very last slide you showed about the other mental well-being issues that arise and the degree to which girls 12 to 17 years old experience this kind of spike in anxiety around 2022. Lots of us can relate to that.

All of a sudden we're put back into social situations that we, some of us, were not really in as much. And I think among adolescent girls, paired with the kind of overexposure and dependence on social media-- I think we ended up, really, with an epidemic of a different kind for social-emotional issues among teenage girls.

I will talk about the implications. I'm first going to take a little detour and talk about a child welfare example. And then I'll come back and talk about the things that we're concretely doing in policy in Illinois to try to address some of the issues that have been raised.

So first, I just want to go back to this separating out the presence of the syndrome from the detection, from the documentation. Use an example from the work that we did early in the pandemic, because what we were hearing-- and I should back up and say, so I've been providing data, analytic consultation, technical assistance, program evaluation to child welfare, juvenile justice, and mental health service systems across the country.

And child welfare directors were in a panic because they were saying, well, we're out of school. Teachers make up the biggest single chunk of child abuse and neglect hotline reports. And they were saying, we're not getting those reports now. Rates have plummeted.

But when we all go back to school, which they imagined would happen all in the same day-- when we all go back to school, we're going to get this deluge of reports catching up from what we didn't see. And as we know, that's not what happened at all.

But what we did as researchers was we tried to dig into this very issue to figure out, well, what could we learn about what school personnel typically report? We have seasonality. Kids aren't in school in the summer, and reports drop. Do they catch up? Is there some cumulative rebound or do they just return to the previous?

And we were able to learn a lot. But one of the most illuminating things we learned was that the school personnel reporters are the least accurate. So in child welfare, we have this check and balance because somebody reports something to the hotline.

The child welfare system says, OK, we'll check it out. And then they do an investigation where they either indicate the child abuse and neglect, say, yes, there was abuse or neglect. Or they unfound it, and they say, no, there wasn't.

We know what the rates of indication and unfounded are. And we know that among school personnel, only 11% to 15% of what they report ends up being abuse or neglect. A lot of what they report is secondary to poverty, things like inadequate food, shelter, clothing, things that could be addressed in another way.

But that was a really important thing to learn for the field because it started this whole other conversation around decoupling poverty from neglect, figuring out what economic supports families need to avoid coming into contact with the child welfare system.

So the reason I tell that story is because in the child welfare context, we were then able to apply a correction factor to say, OK, so we're missing this big chunk of reports, but we know that only a small proportion of those are actually reflective of the syndrome we're looking at.

You don't have that with ADHD. We don't know to what degree teachers are accurately reporting. And what I would flag is that ADHD-- many disorders are often misdiagnosed trauma. ADHD is probably top of the list, where kids who've experienced trauma often have dysregulated behavior, problems paying attention, and teachers may not be sensitized to that enough. But that's one of the things that we're working on from a policy perspective.

So moving into the policy implications-- what I took from this is, OK, how do we support school personnel in doing a better job? Because they are the people who have the most interaction with our kids. How do we help them do the best job they can, recognizing what they're seeing, knowing what to do about it, and possibly, where they can, link people to referrals and resources?

The language in child welfare has changed from mandated reporters, because all school personnel are mandated reporters of child abuse and neglect. How do we give them the tools to shift them from mandated reporters to be mandated supporters, people who, if they notice something, they have the tools to quickly make a linkage or connect someone to what they need?

OK, so the other slides I prepared are not in this deck, so I'm just going to tell you about them. I'm going to tell you a little bit about the work that I'm doing out of the governor's office. So coming out of the pandemic, Governor Pritzker and others recognized that youth mental health problems had been on the rise, but we were in a crisis that was compounded by the fact that we had kind of a shrinking mental health service workforce so that if we had less capacity and more demand, we were headed for trouble.

And in Illinois, in particular, we have six state agencies that are responsible for children's behavioral health, not one. We have the Child Welfare department, the Medicaid agency, the Department of Human Services, the Illinois State Board of Ed, Juvenile Justice, and the Department of Public Health. And they each oversee a piece of delivering services to kids.

So in the beginning of 2022, for about 10-ish months, I worked on a blueprint, the blueprint for transformation. And then to my amazement, the governor turned around and issued an executive order and said, we're going to follow this plan. And we would like you to lead the implementation of it. And careful what you wish for. That's a big job in Illinois across six agencies.

But how often do they ask the researcher to redesign things in a way where the incentives are aligned and we can maybe get some transformational change? So there are a number of recommendations in that blueprint. There are 12 in total. But the work I want to flag coming out of the blueprint are a few things that I think address the issues that Kelli raised.

One is working toward, on a very long runway, universal mental health screening in schools. We screen kids for vision problems. We screen kids for hearing problems, since I was a kid. I did those screenings. Because kids can't learn in school when they can't hear and they can't see.

We're at a point where kids can't learn in school if they are dealing with mental health issues. So I would argue both for that reason and because it will help us on the data end of things, it's really important that we have-- that we at least offer universal mental health screening.

That doesn't mean people have to do it. Parents can opt out just like they opt out of other screenings. But there's work to be done to get all of our school districts ready to do that. There are policies that need to be put in place, technology that's needed.

What we started with was a landscape scan in Illinois to say who's doing universal mental health screening? And what we learned was a pleasant surprise, which is that 2/3 of school districts are conducting mental health screening. And 28%, it's universal, meaning they offer every year to every child.

So that's number one, is providing a screening opportunity in school that can, using a standardized and tested tool, provide a reliable and valid assessment of who has ADHD, among other problems.

The second thing we're doing, and it's kicking off next week, is a TA psychoeducational effort to engage everyone who comes into contact with kids in an understanding of the impacts of trauma. Much of the time when I go out and speak, I'm asked, oh, and can you also give us a little trauma 101?

And that's from attorneys and judges, people who are making decisions about the lives of kids who want to understand more about what does trauma look like, and how can we distinguish it from some of the other problems we're seeing? So that's a second thing that we're doing.

And the third thing that we're doing in Illinois is working on how do we put referral service information in people's-- in their hands. We all carry around these smartphones. I could make a dinner reservation in Atlanta from my phone right now, but I can't find an open slot for outpatient therapy for a young person who needs it.

And that's a totally doable thing. We have the technology to do it, and we have the information. So happy to say, just six weeks ago, we launched a centralized care portal that Google built for our state. And it's the first of its kind. It provides a single place for parents, teachers, and others to go to get information about the state-funded services kids are eligible for.

But it also offers the potential to link families with other kinds of benefits and resources that they need to get them to the right system. I'm also interested in diverting them from the child welfare system if that's not what they need.

The other one-- I just had a few thoughts I jotted down. Am I almost out of time? I wasn't keeping track. The one other thing that I wanted to say was the conversation about the rebound and the work you can still do.

I also think a component of talking to teachers, convening some focus groups of teachers to hear from them, because I feel like we need to learn in the ADHD space what we learn in the child welfare space, what actually does cause you to flag an ADHD problem? And then what tools do you have at your disposal? And what would you need to support you in order to do a better job at that?

Because while I don't want teachers to be necessarily the front door, they kind of are, with our kids. So I'll stop there and let you lead us in some other questions. But thanks.

KRISTIN BUTCHER:

No, no, no. You can be there. Oh, OK. I'm just gonna do it.

[SIDE CONVERSATION]

All right. Well, thank you both very much for those great presentations. I'm going to start with a question to Kelli.

So I'm, as I said, director of microeconomic research here at the Fed, which sometimes I joke I'm director of the, hmm, I didn't know the Fed would be interested in that research, which includes a lot of Kelli's work, fine work on child well-being and child mental health. I'm very lucky to have followed in Dan Sullivan and Dan Aaronson's footsteps, who've always been robust advocates for the fact that everything is economics.

But I wanted to give, Kelli, you a chance to talk about how this fits in with Fed policy, not just-- and then we can talk about other types of policies.

KELLI MARQUARDT: Sure. So I think I kind of tried to hint at this a little bit at the very end. So the Fed has this dual mandate, stable prices and full employment. And I think where this really ties in is in that employment group, in particular, kind of two ways.

One is what are the caretakers and teachers-- how does their employment and labor market outcomes change as a result of ADHD and behavioral well-being in children? We know that parental labor market attachment declines with a child health shock.

This is not necessarily a big child health shock in the way that it's often documented in the existing literature. But there is some changes in child mental health, and this could have impacts on caregiver or teacher labor market outcomes.

The other thing which I think is super interesting, but we won't for a very long time, is how does this drop in ADHD identification have long-run impacts on what goes on in the educational attainment, graduation rates, which majors are selected, what jobs people go into, and they're tied to the labor market and productivity when they are adults. So that's a question we don't quite have an answer to yet.

KRISTIN BUTCHER: Before we talk more about your policy role, Dana, could you talk to us a little bit about how you were in a position to have this great title? And what were some of the things that you, as a researcher, found super helpful in leading this kind of transformational change?

DANA WEINER: OK, so first of all, I've always been a researcher that follows the policy issues and questions. So even when I was in a university setting for six of those years, I was also embedded in the Department of Children and Family Services, not only to witness firsthand the types of problems and questions that arise that can be answered with data, but also to get some firsthand experience on what will and won't work.

Because when we as researchers are formulating recommendations-- and I think it's great that we dream big. But sometimes the things we come up with, if we don't understand how the people do their jobs and what their business processes are, we might recommend things that are never going to fly.

And so it's been really important in my career to have that balance of seeing what actually happens on the ground and what motivates people and then being able to inform with data reasonable strategies to address things.

How I came to this position was in 2019, there were some very high-profile child deaths among kids who had been known to the child welfare system but remained in their family's homes. And our governor-- it really got our governor's attention. And he wanted to know what's going on with this Intact Family Services Program and DCFS in general, that we would have missed these things.

I did a quick turnaround report then. And so I think when we found ourselves in crisis again-- and the "tip of the iceberg" symptom of the crisis was that there were about 55 kids psychiatrically hospitalized beyond medical necessity who were just sitting in the hospital. They were in custody of DCFS.

And everybody wanted to know why are they sitting-- why are they not in a treatment setting? And so that was the tip of the iceberg. The iceberg is much bigger. It includes all of the kids that are not diagnosed with problems that have problems or the kids that aren't reflected in the health care utilization data because there's such a high number of kids who have problems that don't get the services they need.

So digging into the data meant getting data from all six of those systems, from the Illinois State Board of Ed and from the Illinois Department of Public Health and others, while building some near-term technological strategies to get those agencies to work together to get those kids unstuck.

So did that answer your question. That's kind of how I got here?

KRISTIN BUTCHER: Awesome. Thank you. So I have to confess, not only am I an economist, I'm also the parent of three young-adult children who have significant learning disabilities. And all of them were from a junior in high school down to last year in middle school during the pandemic.

So this conference gets me where I live. I feel several internal organs are involved in my reaction. And I'm really pleased that we're doing the academic impact and the mental health impact, as well as the physical health impact on children and the role of the safety net here.

So as a parent of kids who went through with IEPs, one of the things that always struck me was how hard it was to get them the services that they were on paper entitled to. And so this idea you have of universal screening is really intriguing to me, also as a researcher.

KELLI MARQUARDT: I wrote that down.

Universal screening.

KRISTIN BUTCHER: Yeah, I'd like to not have missing information. That's awesome. So what do you think it would take to really roll that out to 100%? What would it take to get researchers more access to data?

DANA WEINER: The graphic I wish you could see is-- I prepared the 12 strategies in the blueprint and our goals for improving our system, which are adjust capacity, improve accessibility, allow us to intervene earlier, improve agility and accountability and transparency, those five things.

And the reason I show this slide is because it illustrates that there is no one silver bullet. I can't implement universal screening without doing at least five other things to make sure that the kids we identify have a place to go.

That's everybody's fear is-- the school districts who don't do screening are like, well, what do we do when we find them? Are we liable for getting all of them services? Now, that's not a reason not to screen them, because the kids who are there, who need help, they're there regardless of whether we detect and link them with services.

But to answer your question, let me talk about a couple of other things. And Kelli's comment made me think of these too. So some of those other strategies are building in-- and this relates to the workforce issue. You mentioned parents impacted by a child health crisis.

So one of the things-- we looked across the country at strategies that other states were using. And one of the things in place in a few places is something called in-home behavioral health aides. So some of our kids-- and I am the parent of a 21- and 23-year-old. The 23-year-old had an IEP and at one time had a one-on-one aide in kindergarten.

Those kids that have a one-on-one aide in school because everybody realized like, this kid is not going to be able to learn without some-- they still need help at home. And we're expecting their parents to do that, in addition to maybe parenting siblings and other responsibilities.

So building this ability to have in-home behavioral health aides in Illinois is another of the 12 strategies that we're working on. And I'm happy to say that not only have we now applied for a pilot exemption to do a two-year pilot of this to make sure it's feasible to train the people and get them in the homes, but we have them waiting in the wings at HFS pursuing a state plan amendment to ensure that once we finish this pilot, this could be a Medicaid- billable service.

But get it-- It's like conducting an orchestra to get because this requires so many state agencies to work together on these solutions. It doesn't live in just one place.

KELLI MARQUARDT: Can I add to that real quick? It kind of brings you back to the bigger macro economy. We do know also during this time that the biggest occupation that did see drops and kind of drops out of the occupations are school personnel, teachers, as well as some health providers.

DANA WEINER: That dropped out.

KELLI MARQUARDT: That dropped out of that-- not necessarily out of the labor market entirely, but perhaps changed careers. There was this big burnout. And so maybe thinking about not just we need people to help the children and the families, but also we need those people. So how to get those people and benefits of the job to stay in that. That'll also be something that's important.

AUDIENCE: [INAUDIBLE] Also, are you considering training for the personnel that are in school? Teachers and teacher parents aren't trained mental health professionals. And ADHD and [INAUDIBLE] is a mental health crisis.

So we're educators and we're teachers and we have this epidemic that's happening with the mental health piece. But we don't have the training. So part of that [INAUDIBLE] the training so that the accommodations that your children so desperately needed are accessible by someone.

DANA WEINER: Yeah. And so the training I mentioned, the trauma-awareness one, I'm also advocating for mental health first aid, which is more of like a how do you recognize when there's a mental health problem and respond to it.

We also need training in our schools. And NAMI is actually doing a great job at this in Illinois of ending the silence to reduce stigma to make it more-- but I think to your point, training and understanding the difference between trauma and ADHD, how to know when something is-- we don't want to put-- I don't want to add too much to what teachers are doing. But I think a certain amount of support and education could make their jobs easier, probably.

I was going to say one other policy thing, which is I mentioned that Google built this centralized portal. One of the things we've done is included parents in all the conversations we have when we're planning the thing, not just at the end when we've done it and we're like, hey, we built this thing, what do you think of it? But what functions would you want a new technology to have?

And what parents told us was a couple of things that we incorporated. One, we'd love a place to be able to create an account where we could upload all these evaluations and screenings and things so that other people could get to it.

But two, to your point, navigational help. How can we get a human being-- and so now we have a team of people at the division of mental health who see the queue every day. They can respond to concerns. And if someone says, I need a navigator to help me get an IEP, we have another team that can then step in and give at the elbow support.

KRISTIN BUTCHER: Awesome. So all of this sounds expensive. And so I'd like to know a little bit about who had to say yes to start making these things happen. Was it all executive branch? Do you need the legislature? How does that work?

DANA WEINER: Turns out you need all three branches of government to do these things. But we are at a kind of unique moment in time where we have a governor who feels strongly about this, a General Assembly that has given bipartisan support on mental health issues.

Every time I've gone to testify in support of-- we've had a few-- now we're on our third proposed bill. There's been one each year of this initiative to build some of the things that require statutory foundation. And we've had a lot of support for it.

Now, this past budget year was not as-- there were cutbacks in some things. But we have a group of leaders right now who recognize that it's less expensive to provide preemptive supports on the front end than it is to remediate problems on the back end.

And I think the case is the most clear when you look at diverting kids from child welfare. I mean, bringing kids into state custody in order to get a psychiatric service paid for, that's a perversion. That's not what the child welfare system was intended or resourced to do.

And it's expensive. So I think there are some savings in addition to-- but I don't want to minimize-- the system that Google built is a $4.8 million system. Now, Google kicked in a million and a half dollars because they wanted to build a system like this and have an example of what it could do.

But it's taken commitment both from the executive branch and the legislative branch. I mentioned the judicial branch because it's important-- judges and hospital social workers are some of the hidden decision-makers that can send a kid one way or another down a path. And it doesn't seem like-- they're not health professionals, but yet they have a lot of power to make decisions that impact what people have access to.

KRISTIN BUTCHER: But that's the cost side. I'm sure everybody in this room believes there's a benefit to a good diagnosis and intervention. But do we actually have a-- if you aren't diagnosed in kindergarten, then you have how much learning loss or test scores? Or does anybody else in the room have some idea of what are the benefits of--

KELLI MARQUARDT: For ADHD, it's-- I've done a lot of research in this area. It's particularly hard because, kind of back to what you said, you need to know who has it. And just having the condition itself is even hard to identify because of the subjective nature of diagnosis.

So in order to see the benefits of the different types of treatment, you need to know, was the person-- you would have to have a counterfactual world where someone that you know for sure has ADHD did not get diagnosis. And so it's really hard to tell.

You can have very large numbers. I one time did this big-- I don't remember the actual numbers-- calculation of the costs and benefits of getting a diagnosis and treatment. And the balance on it was just way too large to really, truly say anything.

We definitely need-- it would be helpful if we had a kind of a universal screening or some sort of data on the prevalence of ADHD for sure.

KRISTIN BUTCHER: Dr. Harris--

BRIAN JENSEN: Yeah, actually, if I can-- psychiatrist or pediatrician, if you'd like to weigh in.

KRISTIN BUTCHER: Let's let Dr. Harris because I saw her first--

BRIAN JENSEN: Yeah. Sounds good.

KRISTIN BUTCHER: --then go to you.

DR. HARRIS: I just wanted to add the nuance and just great conversation around the diagnosis is that, unfortunately, we see implicit bias rearing its ugly head sometimes. And you will get young Black boys labeled with a conduct disorder versus ADHD.

I have often received the question, are we over-diagnosing or under-diagnosing? And I would say, it depends. We've just seen a very horrible situation where folks were just-- and these are folks who, in my opinion, abdicated their professional and ethical obligation. But we've seen just using online. And many folks were more interested in diversion than treatment.

That does happen. But then we've seen in other instances where folks don't even have the access to the screening. And the diagnoses get labeled as a conduct disorder, which really puts them on a path. So just further nuance to the conversation.

KELLI MARQUARDT: That's true. It's not ADHD versus nothing. It's typically, you're maybe mistaking one condition for another and getting treatment for one that you should be getting treatment for another. So yeah.

DR. HARRIS: And trauma [INAUDIBLE] PTSD [INAUDIBLE].

KRISTIN BUTCHER: Yeah. So do you want just--

BRIAN JENSEN: Hi, I'm Brian Jensen. I'll be speaking later. I'm a physician scientist, so I'm a practicing primary care pediatrician at Children's Hospital of Philadelphia and a health services researcher.

So first off, kudos for all of his great work. It speaks to the need for this interdisciplinary workforce working to help improve the lives of children. So kudos for this work.

I think what you spoke to is around ADHD is a good example of the complicated interplay of parents wrap around resources for child in a school setting. I think you're speaking to this big issue in terms of the lack of-- the loss of the support systems in place that really help protect children.

Now, that other question that came up around ADHD, there are well-designed, randomized controlled trials that show how effective treatment can be. Treatment, whether it be pharmacotherapy or the additional, all mental health needs kind of three big things. And I'll close after this is. And we kind of undersell each of them in different ways.

But pharmacotherapy sometimes is first, but it actually probably third in terms of what we should be doing. The mental health support, usually through some form of therapy, cognitive behavioral therapy, and then structure. We shouldn't be surprised that when we threw off structure, all these bad things happened across our culture and society, especially to children, our most vulnerable.

Regardless of their underlying demography, all children are vulnerable. And so there is really good, compelling data that treatment can help over the long run. When you mentioned-- and what we're seeing in clinical practice and I think is going to catch up in the research and I hope you continue to contribute to this, is that all these delays in diagnoses are we're not catching up.

And so just anecdotally, but then backed up by-- I think, more research that will come out. I've been-- despite my hopefully youthful visage, I've been in practice for a while. And this is the first time we're seeing children who are getting diagnosed in middle school where we usually were intervening in kindergarten, elementary school.

And I can't tell you how many kids I'm seeing who are in fifth, sixth, seventh grade who can't read. And I'm only in practice one day a week. But this is a huge issue for all pediatricians in our large care network that we're seeing. So, again, kudos for all of this work to help protect children.

KELLI MARQUARDT: That's true. And I want to add, I wasn't saying that-- we do have a lot of evidence of the treatment of ADHD via pharmaceuticals and behavioral therapy. I was more kind of getting to the more economic part as well of do we know the benefits on that side of things as well, not just managing symptoms, but the carryover into productivity in the labor market and society overall as well.

DANA WEINER: And bringing us back to that, I think another place we can look for that-- and I'm awaiting an analysis from the Illinois Department of Public Health epidemiologist on hospitalizations. Because I see the inpatient hospitalization is kind of like, we missed it all along.

And so that's like, just like a child welfare system entry is kind of a fail in my mind. So is a hospitalization, particularly one that lasts for a long time. So we're looking to see whether those hospitalization rates, how they're changing over time.

Not that ADHD is a condition that lands people in the hospital. But some of the kids that we see in the hospital had undetected mental health problems for a very long time, sometimes trauma-related. And ultimately, they get worse and worse until they're treated in acute setting, which is very expensive.

KRISTIN BUTCHER:

We're almost at time. So any final thoughts from the audience? Yep.

AUDIENCE: I actually just had a question. If we go back to the data set, were you able to tease out any cause to why the boys didn't recover? Is it just that there's an expectation that boys will be boys and maybe our threshold is higher for their behavior?

Because the girls, they seem to be kind of right back where they were pre-pandemic. But the boys have continued to stay low. So--

KELLI MARQUARDT: Yeah, the girls were only back up if they were in a high-opening state. So for 2/3 of the country, at least by February of 2021, they were still lower diagnosis rates. We can't say why.

We can say it's not in-person levels of schooling. We did find some evidence that the stability-- I didn't show the figures. But if you were all in-person or all virtual, the diagnosis rates went back up.

Perhaps there's something about boys are much more likely to have the externalizing behaviors. Might be easier to notice either if you're always looking at them on Zoom or always looking at them in person. But that's just some suggestive evidence. I don't-- I'm definitely open to other ideas. But yeah.

KRISTIN BUTCHER: Do my panelists have any final thoughts they want to leave us with? All right. Thank you very much.