Using Cost-effectiveness Analysis to Inform Policy Responses to Childhood Obesity

Using Cost-effectiveness Analysis to Inform Policy Responses to Childhood Obesity


Good day, and welcome to the Self-Sufficiency
Research Clearinghouse�s Using Cost-effectiveness Analysis to Inform Policy Responses to Childhood
Obesity Webinar. Today�s call is being recorded. At this time, I�d like to turn
it over to Dr. Kristin Moore, Senior Scholar at Child Trends. Please go
ahead. Thank you, and welcome, everyone. I�m Kris
Moore, moderator for this Webinar on Cost-effectiveness in Childhood Obesity,
with Michael Long, our Emerging Scholar, and Michael Bono,
discussant. I�m going to begin with some quick updates on the SSRC.
The Self-Sufficiency Research Clearinghouse is, as I hope most of you know, a virtual
portal of research and other resources related to self-sufficiency.
It functions as an online community for researchers, for practitioners, and for other stakeholders
interested in self-sufficiency, employment, and family and child
well-being. The SSRC�s purpose is to disseminate quality research, and we currently have over
4,600 items in the library, and we are constantly adding
new resources. The library�s materials are organized into 12 topical
areas that are listed on the right-hand side of your screen. Every item included in the
library is reviewed for relevancy. Users may search by keyword or
use filters by topic area, target population, geographic location,
or research methodology to browse the collection. Every topic area page under the Browse Topics
tab includes Our Librarian Recommends box that
highlights research and resources recommended by the SSRC
library team. Each topical area also includes relevant federal laws and regulations.
To make it easier for SSRC users to include and accurately cite research in their work,
the SSRC has a new citation assistance feature that enables
users to obtain citations for any entry in the SSRC library. Simply
select the checkbox next to the entry of your choice from the library filter screen and
select �download publication� at the top of the page. You
will then be redirected to a screen similar to what you�re seeing on
the slide, which is an example of a publication co-authored by our Emerging Scholar, Dr. Long.
Once you select �submit,� a separate file will
load with the formatted citations. Visit the SSRC library to generate other
citations and browse research related to today�s Webinar.
Well, turning to the Emerging Scholars program of the Self-Sufficiency Research Clearinghouse,
we welcome your nominations of future Emerging Scholars. Our selection criteria are shown
here. An Emerging Scholar can be a graduate student
or a degree recipient, but with no more than 10 years of
experience, currently doing research on self-sufficiency issues related to those 12 SSRC topic areas,
with conducting high-quality research that fills
a knowledge gap or that addresses a self-sufficiency issue that
warrants greater visibility. And I want to highlight that they can be working in an academic
setting, or in a program setting, a think-tank, or a public
agency. Michael Long is the thirteenth Emerging Scholar,
and previous Emerging Scholars have come from varied backgrounds and covered varied
interesting topics, as shown here. Our speakers today are Michael Long, Science
Doctorate, who is an assistant professor at the
Department of Prevention and Community Health at Milken Institute School of Public Health
at George Washington University. And our discussant
is Michael Bono, Ph.D., Human Services Administrator in the Los
Angeles County Department of Public Social Services. And, I, Kristin Moore, am the moderator.
You can submit questions any time through the Q&A feature, the Question and Answer feature,
just at the bottom right of your screen. The questions will be answered after the presentation
or, if we run out of time, via responses that are posted
on the SSRC, with other Webinar materials afterwards.
And, finally, we encourage you to join today�s conversation on Twitter using the SSRC Webinar
hashtag displayed on the screen. Tweets using this hashtag will display on the left side
of the Webinar platform. Thank you. And now, I will turn
the mic over to Dr. Michael Long. Thank you, Kristin. I�m very excited to
join this audience, and very grateful to the SelfSufficiency
Research Clearinghouse for inviting me to give a talk today on Using Cost-effectiveness
Analysis to Inform Policy Responses to Childhood Obesity.
And, we know that childhood obesity is high on the policy
agenda, and I don�t know that the link between childhood obesity and adult obesity and self-sufficiency
has been made clear. So, I think this is a great
opportunity to really bridge the gap between two different
research and practice communities. Our work was funded by a range of foundations,
the Robert Wood Johnson Foundation, the Pritzker Foundation, the CDC [Centers for
Disease Control and Prevention], and primarily ongoing funding
from the JPB Foundation. I want to talk first about, why is childhood
obesity a threat to self-sufficiency? And then, why is
cost-effectiveness analysis essential to informing policy responses to childhood obesity? Then,
I want to go over our Childhood Obesity Intervention Cost-effectiveness
Study, and implications for research, practice, and policy.
So, again, why are we talking about childhood obesity in a self-sufficiency research context?
So, what is obesity? In adults, obesity is defined by having a body mass index of 30
units or greater. Currently, 34.9 percent, or 78.6
million adults in the U.S. have obesity, and there are substantial
disparities in the obesity burden by race/ethnicity and by sex.
So, how does this relate to self-sufficiency? Well, narrowly defined, self-sufficiency could
mean that individuals and families are able to
provide for all their needs without public or private assistance. But
more broadly, self-sufficient individuals, families, and children are achieving their
highest potential with active engagement in work, school, and civil
society. So, how is adult obesity a barrier to self-sufficiency?
Well, primarily, obesity leads to early death, and severe obesity leads to a loss of life
similar in scale to smoking. And so, it�s a major threat to life
expectancy. Obesity also leads to substantial disease burden and to related healthcare costs.
One study estimated that obesity accounted for 9 percent
of all healthcare costs in 2008. And, as the obesity epidemic
continues to grow, that cost will grow. And I think this is of primary importance to Medicaid
at the state level dealing with cost containment and obesity
is a major challenge for states trying to manage their healthcare
cost responsibilities. More closely aligned with our interest in
self-sufficiency, obesity is associated with missed work. It
accounts for between 6.5 and 13 percent of total absenteeism costs in the workplace.
So, we�re talking about loss of life and healthcare costs, but
we�re also talking about missed work. And I think, when we�re
focused on self-sufficiency, understanding the barrier that obesity creates to really
engaging with work, it�s very important.
So, I also want to talk about weight bias, and in no way am I suggesting that in order
to get rid of weight bias that we need to reduce obesity.
In fact, we need to reduce bias and reduce discrimination.
However, adults with obesity face discriminatory hiring practices, are denied promotions, and
face wrongful termination. Sixty-nine percent of women with
overweight and obesity report stigmatization based on
weight from their doctor. And experience of weight-based stigma leads to overeating and
reduced physical activity. So, I think we can�t have an honest
conversation about the effects of obesity if we ignore stigma,
and we also should never argue that the effects of obesity are stigma, but that we need to
focus on both reducing obesity and reducing stigma.
So, why have I been talking about adult obesity, when this presentation is really about childhood
obesity? Well, currently, there is no scalable medical strategy to reverse the obesity epidemic
during adulthood. You can have bariatric surgery,
which has dramatic individual effects, and there are some clinical
behavioral interventions that have shown modest success, but, right now, we see no medical
strategy that will turn around the epidemic among adults.
In one recent study, fewer than one in five adults were able to
sustain a BMI [Body Mass Index] reduction over nine years. So, reducing adult obesity
over the long term will require prevention of childhood obesity. That
is the link between childhood obesity and the really serious
healthcare, health, and employment effects that we see due to adult obesity.
So, what are we talking about with childhood obesity? Currently, 17 percent of youth 2
to 19 years of age have obesity in the United States,
and that�s defined as having a BMI (body mass index) greater than
or equal to the 95th percentile for the same age and sex. And, I think it�s encouraging
that we see that prevalence has remained stable over the last
decade, and yet it�s at an all-time high. And so, if we focus
really only on the stabilization as a success and forget the fact that we�re still seeing,
really, the highest level of this epidemic in the history of the United
States, we�re missing the point. Dr. Long: And while we focus on, and celebrate,
the fact that the epidemic has stabilized, we�re starting to
see emerging and increasing, very striking racial/ethnic disparities in childhood obesity
rates: Hispanic youth with 22 percent obesity rates during
childhood, non-Hispanic black 20 percent, compared to 14
percent for non-Hispanic white youth. And so, while we focus on the overall epidemic,
we need to also pay particular attention to what these disparities
mean and how they interact with the other barriers to selfsufficiency
that are facing minority groups in the United States, and really focus our efforts on reversing
these disparities, while we also focus on reducing the overall epidemic.
So, there are problems, health problems and other problems, during childhood due to obesity,
but the biggest health risk of childhood obesity is actually the increased likelihood of adult
obesity; and the older children get and the more they have
obesity during adolescence, the more likely it is that they will
maintain obesity during adulthood. And, as I already mentioned, it is very difficult
to reverse obesity during adulthood. So, really, we should be focusing
on getting children through childhood at a healthy weight, so
that they can have a chance at having a healthy weight during adulthood. And I think this
is consistent with the broader self-sufficiency community�s
focus on child development as, really, a core component of longterm
self-sufficiency development at the population level.
Dr. Long: But there are also near-term negative effects of childhood obesity. There�s increased
risk of prediabetes, sleep apnea, joint problems,
and 37 percent more sick days leading to missed school. And here,
again, is another barrier to self-sufficiency. If children are missing school, they�re
not getting education, and we know that education is a key component
of creating a well-developed child that will develop into a selfsufficient
adult. As I mentioned with adults, I do want to highlight,
again, the negative effects of weight bias in our
society. Youth with overweight and obesity are subject to pervasive victimization, to
teasing, and bullying. Weight-based stigma can lead to severe negative
emotional outcomes, including suicidality. And stigma
also leads to poorer academic outcomes. So, again, I�m not arguing that reducing obesity
is the way that we will reduce stigma, but that we should always
be focusing our efforts on both reducing stigma and reducing
obesity. So, in summary, obesity poses a serious threat
to self-sufficiency across the life course. Racial/ethnic disparities exacerbate existing
disparities in barriers to self-sufficiency. And although not
caused by obesity, weight bias and discrimination prevent full engagement in work and school.
And that a long-term strategy will require prevention
of childhood obesity and efforts to reduce weight bias and
stigma. So, I think we�ve talked about why obesity
is a problem, why it�s a barrier to self-sufficiency. So,
why do we need to use cost-effectiveness analysis? And why is that essential to informing a policy
response to childhood obesity?
We know that reversing childhood obesity will require sustained effort from all levels of
government and civil society, and that it will require many different strategies. And
there have been decades of research on intervention effectiveness,
but very limited research on cost-effectiveness of the
strategies to reduce childhood obesity. In order to provide the best value to society,
decision-makers need to integrate the best evidence on effectiveness,
where there is more information, but also on population
reach and on cost. So, here you see a cost-effectiveness plane.
In the upper left-hand quadrant in the red, you see
some things that we do have higher cost and worse outcomes. And you might think, why are
we studying that? But in fact, many of the things we do
in medical care cost money and harm patients, and there�s an
active field of research trying to identify things that we should stop doing that have
higher cost and worse outcomes. In the lower right-hand corner,
in green, things that save society money and that improve health
and well-being. These are the things that, in decision science and cost effectiveness,
we would say, �You should definitely do these things, if there
are no other barriers to doing them, that we should always do
things that save us money and improve outcomes.� But, in reality, most of our decisions fall
into the upper right-hand quadrant, where we�re trying
to invest in better outcomes, but there are costs. And so, then the
question is, how much are we willing to pay to have better outcomes?
So, what is the risk of ineffective prioritization, of not using cost-effectiveness as a tool
to guide the investments that we�re making? There
are budget constraints and limits of political capital that require
decision-makers to choose a limited basket of obesity prevention strategies. We�re
not going to do everything. And failing to consider cost-effectiveness
can lead to inclusion of strategies with high cost and
limited impact on the epidemic. And so, using cost-effectiveness allows us to get the right
things into the basket, and knowing that we can�t put everything
in. I want to talk now about a project that we�ve
been working on over the past five years, a
Childhood Obesity Intervention Cost-effectiveness Study, or CHOICES [Childhood Obesity Intervention
Cost Effectiveness Study] project. And I encourage
you to visit our website: www.choicesproject.org. So, this project is a collaboration led by
Professor Steven Gortmaker at Harvard, and researchers at
the Harvard T.H. Chan School of Public Health, Columbia University, here at the Milken Institute
School of Public Health in the U.S., as well as researchers
at Deakin University and Queensland University in Australia.
And, over the last few years, we�ve been assessing the comparative effectiveness and
cost-effectiveness of more than 40 interventions aimed at reducing
childhood obesity. It�s a large team led by Steve Gortmaker,
and I just want to acknowledge all the work that
everyone on the team has done on this project over the last few years.
So, what is our process? I�m going to walk through six steps. First, we selected and
recruited a group of stakeholders who guided our overall
project. We selected interventions that we wanted to
evaluate, specified the interventions, implementation scenarios and their costs, and then, evaluated
intervention effects, reach and cost[s], and modeled, then, the short-term and 10-year
cost-effectiveness of doing each of these 40 interventions. And
then, included qualitative understanding of what are the
implementation[s] and equity considerations of doing each of these 40 interventions.
So, first, the stakeholder group. We recruited stakeholders to represent U.S. policymakers,
policy researchers, and programmatic experts. And
they provided us with advice concerning the specification of
the interventions, data sources, how to do the technical analyses, as well as really
grounding our qualitative analysis of what are the implementation and
equity considerations. And I think it is important to highlight
the value of bringing stakeholders into a cost-effectiveness analysis process early
on, at the very beginning, to guide the whole project, but also to make
sure that the results that are coming out of a research endeavor
are really relevant to decision-makers. So, we selected 40 interventions by the stakeholders
to evaluate, and I�ll present today the first
four interventions, which are published in the July 2015 issue of the American Journal
of Preventive Medicine. And, these are all available for
free download on the website www.choicesproject.org. So, these
four are: evaluating the impact of an excise tax of a one cent per ounce tax on sugar-sweetened
beverages [SSB]; eliminating the tax deductibility of
TV advertising of nutritionally-poor foods to children; state policy
requiring 50 percent of existing PE (physical education) time to be moderate to vigorous
physical activity; and a state policy to make early childcare
settings healthier, targeting physical activity time, improving
nutrition, and reducing screen time. So, what do I mean by �specify an intervention�?
So, to give an example of the active PE specification, this means that states would
implement a policy directing states� boards of education to
include curriculum requirement that 50 percent of PE time be devoted to moderate to vigorous
physical activity. And this is based on policies that
have already been passed in state legislatures in Texas and
Oklahoma. And implementation of the active PE policy during existing PE classes would
include providing schools with new PE curricula, with portable
equipment, and with teacher training. And also, I think critical,
we costed out and included a monitoring component that we considered necessary for implementation.
And here, when we�re talking about policy change, when we don�t include the cost of
ongoing monitoring, I think we are either over-assuming the effect
or undercounting cost. And so, incorporating an ongoing
monitoring and continuous improvement component to all policy change is important, both from
an evaluation and modeling stance, and also for
enacting real policies. So, this specification led us to develop a
logic model, where we can go from what would happen
when the state changes an active PE policy to what kind of PE practices would change,
to what we think the estimates of physical activity level change
would be, and then how this affects BMI, obesity, and healthcare
costs. And, in each stage in this logic model, we try to find the best evidence available
to link this change, so that we can really get the best evidence linking
the state policy, all the way through to obesity and
healthcare costs, and I�ll give you an example on the next slide.
Well, first, how do we get this evidence? We used an evidence review process evaluating
study quality in agreement with Cochrane and GRADE
approaches. And so, we conducted a broad range of
systematic evidence reviews using these logic models to link behavioral changes to shifts
in energy balance and obesity. And in the four interventions
that I�ll talk about today, all of them had direct evidence linking
the behavior changes targeted by the policy to changes in BMI.
So, here, the example for active PE linking changes in PE practices to changes in physical
activity levels, we used a systematic review and meta-analysis
that was already published of a broad range of PE
interventions that found that if you do these kinds of interventions, you will increase
activity time during class by 6.24 percent. We then did another
systematic review and found two studies, one a randomized
controlled trial and one an observational study that found that for every additional
minute per day that a child is moving, it leads to a 0.023 BMI unit
reduction. Again, all four of the interventions had direct evidence
linking the kind of behavior changes that I am describing here to changes in BMI.
So, once we have the effect size, we need to know, how many people do these interventions
reach? So, we estimate the number of individuals reached by each intervention, assuming that
they�re implemented nationally. And so, for the active
PE intervention, the intervention would reach children age 6
to 11, attending public elementary schools in the 47 states without an existing policy,
who are attending schools with PE and who regularly attend PE.
And then, we only estimate benefits to individuals in schools
that would implement policies, and there�s some evidence that only about 70 percent of
teachers would actually implement these kinds of policy changes.
So, at every step, we start with the national population of
children 6 to 11, and use the best evidence we can find to really get a good estimate
of what would be the number of people reached by this kind of a
policy change. And then, cost. What would it cost to implement
this kind of intervention? Well, we use a modified societal perspective on cost. That
means we don�t care who is paying for the intervention. We want
to know, what does it cost society to do this? So, if some of the costs accrue to school
districts and some to local government, both of those would be included,
or if they accrue to industry to comply to regulatory
changes, we count those costs to industry. We don�t include the cost of intervention
participants� time. So, we don�t count the cost of a child�s time
in PE. We just count the cost of actually running these kinds of
programs. So, then we take the information on the per
person effect size, how many people we think it will
reach, and how much it will cost to do these things, and we integrate those in a simulation
model of the 2015 U.S. population age two years and older.
And so, what we do is we follow the existing population
forward, assuming that we do nothing, and then we follow the same population after we�ve
implemented each of the interventions. And so, we then
simulate the health and healthcare cost experience of the U.S.
population over 10 years, from 2015 to 2025. And, in the short term, we estimate, what
are the effects of the intervention on BMI compared to doing nothing?
And then, what are the longer-term effects on BMImediated
reductions in disease incidence and quality of life, life expectancy, [and] healthcare
costs? And so, we think it�s important to focus both on
the short-term outcomes, where we have strong evidence on BMI
change, and then also to try to estimate and quantify what we think the longer-term benefits
to society would be to doing these interventions.
So, how do we get savings in healthcare costs? So, there�s a lot of literature on increased
healthcare costs due to obesity. And there�s a published analysis of data from the Medical
Expenditure Panel Survey, which found that for every year, a
person with obesity has higher healthcare costs than a person
with normal weight � and for youth 8 to 19, obesity is associated with an excess of
about $240 per year. And this rose dramatically as people aged, as
their overall healthcare costs grow, but also as the longer-term
disease impact of obesity becomes more apparent. And so, every year that we, through our interventions,
have reduced the level of obesity in society, we then count that reduction in healthcare
costs by preventing obesity.
So, finally, and I mentioned previously how important it is to do a qualitative analysis
on top of�if we only show these quantitative results and
say, �You should do these four interventions because they save
society money,� it misses the complexity of actual policy-making and the political
process of getting changes through. And so, we look at the quality
of evidence that we have in our models, and what impact
doing these things would have on equity, on whether these interventions would be acceptable
to stakeholders, whether they�re actually feasible
to do, how sustainable they are, whether there are positive or
negative side effects, and whether they have broader social and policy norm effects. Whether
doing an intervention may not change obesity levels
that much, but it might create a signal that would lead to
broader behavioral changes in society. So, again, we�re looking at four interventions
right now: An excise tax of one cent per ounce of
sugar-sweetened beverages; eliminating the tax deductibility of TV advertising of nutritionally-poor
foods to children. And so, here, this is � companies
can claim an ordinary business expense for marketing costs, and
there are, and have been in the past, in Congress proposals to remove ordinary business expenses
for advertising nutritionally-poor foods to children
on television. And so, that�s what we�re modeling, is to
remove that ordinary business expense exemption. And then, a state policy requiring 50 percent
of existing PE time to be moderate to vigorous physical
activity; a state policy to make early childcare settings healthier,
targeting physical activity time, improving nutrition, and reducing screen time. And all
four of these interventions are detailed at length in the
papers that are available at the www.choicesproject.org website.
So, I think this first result is what a lot of people want to know. Is doing these things
going to save us in healthcare costs? And for three of them,
not too much. Active PE and ECE policy changes, between $50
and $60 million in healthcare savings nationally over 10 years, slightly larger with the TV
advertising change, $350 million over 10 years. And there are
a few reasons for this. One, that they have relatively small effects
per person. They reach a smaller part of the population, but they�re also targeting children.
And, as I showed earlier, there really aren�t substantial
differences in healthcare costs at very young ages. We do see, and we
estimated a savings of $23.6 billion in healthcare costs nationally from doing the sugar-sweetened
beverage excise tax. And I think this is part of the
reason that the sugar-sweetened beverage excise tax is on top of the
policy agenda in childhood obesity now, is that really we think would result in near-term
healthcare cost savings, primarily due to the fact that it
doesn�t just reach children, it also reaches adults. And by reducing
slightly adult obesity levels the same time that we�re targeting childhood obesity,
we have a chance to cut near-term healthcare costs and to reduce childhood
obesity. So, this next result, the cost per BMI unit
reduction. So, we�re talking about over the two years that
it takes to reach full effect on BMI of these interventions, how much would it cost to run
the intervention for that two years? And then, how much would it
cost per unit of BMI reduced? And so, I think everyone on the
call can think, maybe they want to lose some weight personally, and you could think what
would you pay for a one unit BMI reduction? And, I think
all of these interventions show that they�re far less expensive than
medical interventions and surgical interventions, and they all, I think�there�s no established
threshold for what society�s willing to pay for a BMI
unit reduction. But all of these, I think, are acceptable when compared
to clinical interventions. And here, again, you see a big range, where active PE has cost
$400 per unit, childcare policies $60, and then TV advertising
only $1 per each BMI unit reduced. And so, if you asked
around who would be willing to pay $1 for a BMI unit reduction, I think a lot of people
would find that to be an acceptable value.
How do we get these results? Well, first, what is the effect per person of a BMI unit
reduction? And, in all of the 40 that we looked at, there
really aren�t � there�s not one policy that�s going to reverse the
epidemic. We�re actually going to need to put a suite of policies together that will
actually reverse the epidemic, but there are differences in the
effect size. And one of the reasons that the sugar-sweetened
beverage tax is so important is that there�s strong evidence that sugar-sweetened beverage
intake is one of the key driving factors of the epidemic, and
that a tax would lead to a substantial reduction in consumption.
And so, here, that�s why you see fourfold almost over all the other per-person effect
of the sugar-sweetened beverage excise tax.
So, reach, again, and we talked about setting specific policies that only are targeting
only children in school, or only children in childcare,
aren�t going to reach as many people. But broad tax policies, you
know, if we�re targeting advertising to children that would reach all children in
the country. If we�re targeting changes in prices in sugar-sweetened
beverages, that would reach the entire country. So, you
again see a big range in the reach of different policies.
And a range in the costs, where active PE includes training, materials, curricula, that
costs money nationally, but I think $71 million is really
not a major expense compared to many of the programs that we
talk about today. And TV advertising at only $1 million, we�re really talking about a
tax policy change with only a few companies affected, that would
require limited time for auditing, on both the industry side and
the government side. And I think it�s important to note that we don�t count the tax revenue
as a cost, because this is considered in the field to
be a transfer cost. So, if industry or consumers are giving
government tax revenue, and then government receives the revenue, the money wasn�t used,
and there was no resource consumed. So, that is not
included in the cost of these interventions. And, in general, I think
doing policy changes here, and in doing tax policy changes specifically, have much lower
cost than implementing very intensive programmatic interventions.
So, equity considerations. Well, we know that the SSB tax is a regressive tax, and I think
it�s incumbent on public health researchers who
are talking about implementing this kind of a policy to be
upfront that this is a regressive tax, but to clarify that the health benefits may accrue
to more to lowerincome populations. There�s good evidence that
you�ll see a greater effect of the tax in lower-income
populations. And, if the tax revenue were earmarked to offset the regressive nature,
we do believe that progressively earmarking tax revenue to invest
in improving the health of low-income populations would
outweigh the regressive direct effect of the tax.
The TV advertising, it has the potential, actually, to reduce inequality. We know that
minority children watch more TV, and there�s some
evidence that they�re targeted more for nutritionally-poor food.
In the early care and education policy change, were considered that there may be some increase
in disparities if family-based care implementation varies by income. So, we can get to all the
centers and make change policy, but if lower-income households
are more likely to be in family-based care where it�s
harder to monitor compliance with policies, we could see a widening of disparities due
to this policy. And, similar with the active PE intervention,
there is a potential negative effect if we�re only
talking about changes in existing PE policy. So, if there are disparities in access to
PE currently, then we could see a widening of disparities due to
access to PE. Across all the interventions, we see wide
variation in the impact, the reach, the cost, and the cost
effectiveness, and we�re only looking at four interventions here. The SSB tax would
reach the entire population, have the largest per capita effect
on BMI. Removing the tax subsidy for TV advertising would
reach all children at a lower cost, but have smaller effects. And, while we�re seeing
smaller effects in the setting-specific interventions and they reach
a smaller number of people, they potentially have lower
political barriers. And so, in the 40 interventions that we�re evaluating, we focused on broadening
the menu of setting-specific programmatic and policy
changes that we think have a lower political barrier to some of
the other policies, like a sugar-sweetened beverage tax.
So, what are the implications for this kind of research for researchers, for practitioners,
and for policymakers?
The first is that collecting and reporting cost is critical. There�s a lack of cost
or resource utilization reporting in most obesity prevention research,
and across public health. And the lack of this information on
what it costs to do an intervention really limits our ability to make decisions based
on the potential costeffectiveness of prevention. Changing this practice will
require broader dissemination and adoption of
existing best practices. However, it is important to note that collecting costs of programs
is much less difficult than measuring effectiveness, and
we want to make a strong argument to the audience and to the
general field that we should be collecting costs, we should be reporting them, and it�s
really not that hard to do.
Second, is that we need to embed cost-effectiveness evaluation into the planning and evaluation
process. So, cost-effectiveness analysis has the greatest impact when it�s used to help
decision-makers prioritize investments to achieve goals. If
we�re just publishing papers and talking to the research
community, we�re really not making change in practice. So, what we need to do is to
really build a bridge between the research and the practice community,
and to partner with the practice community in order to
build agency capacity to use cost-effectiveness as a standard planning and evaluation tool.
Finally, I think we need to broaden the self-sufficiency lens to include child health promotion
more broadly. The obesity epidemic is a barrier to achieving the narrowly defined self-sufficiency
definition, and to this broader vision for a society where
we have healthy families and healthy communities, and where
we have individuals who can actively engage in their school, in their work, and in their
community activities. And so, efforts to promote self-sufficiency
may need to better align poverty reduction efforts with obesity
prevention goals, and there are a number of examples where I think we can achieve the
goals of poverty reduction and obesity prevention with smarter
policy, informed by the type of analyses that I�ve talked
about today. So, I will end there, and look forward to
questions to Michael�s comments. Thank you. Thank you very much. A very cutting-edge analysis.
I want to remind listeners that they can submit questions through the Q&A feature,
which is in the bottom right of your screen, and we�ll answer
them as soon as we hear from our discussant, Dr. Michael Bono. And he is a member of the
Self-Sufficiency Research Clearinghouse Technical Workgroup.
So, we are doubly grateful to him for being the discussant
today. Michael? Thanks for the opportunity to comment, Kristin.
And thanks, Michael, for a very good presentation. Working in the self-sufficiency
area of public policy myself, I like how you conceptualize
obesity as a threat to achieving potential, and the broad definition that includes engagement
in productive life. I want to also add, we�ve been thinking
about the definition of self-sufficiency a lot in local government,
and we�ve actually expanded to actually include receiving public assistance as a way
for people to reach their potential. An example that comes to
my mind and in part of my life is that fact that my 90-year-old
mother is able to remain at home and live independently, because she gets public assistance
from her local government for domestic services, like doing
the laundry, doing housecleaning, and doing a little cooking.
And so, the connection still eludes me a bit, because I�ve been working in this area for
about 12 years, and I�m still not sure what self-sufficiency
is. I don�t have a good operational definition. And, in fact,
last month I was given the opportunity comment on the TANF (Temporary Assistance for Needy
Families) reauthorization bill, and that�s one of
the things I ask Health and Human Services to do, because they use
self-sufficiency throughout this bill language, is to just give us a definition of what it
is, so we can use it in our research practice.
I�d like to � well, let me just say, the cost-effectiveness analysis is really, I agree,
is especially valuable to policymakers, and it�s been
my experience that policymakers are typically focused on
effectiveness, without even considering the cost of getting there. I agree that there�s
a communication gap between the academic community and practitioners
and government policymakers, and I fear that this
approach will not be considered in local government agencies because, first of all, you know,
they�re not going to be reading these articles. We don�t
subscribe to the myriad number of journal articles, journals out
there. And, secondly, a lot of people, as you mentioned about capacity building, a lot
of agencies don�t have professional researchers like me in their
departments who can champion such an approach. And that�s
something that I�m aware of, and I feel, you know, unique in being in my position,
but it�s really a challenge to get research into practices. And maybe
that�s probably � maybe that�s not news to anybody, but it�s just
when something like this comes along, there�s going to be a lag in adopting it.
Now I want to turn to the 40 interventions under Michael�s investigation by his team,
and clearly, your models required documentation about each
intervention and health indicators. And I suspect some
interventions were not well-documented. And I think you made a good case that we need
to do a better job documenting how we�re doing the interventions,
and especially capturing the cost. And, I wanted to ask
you, Michael, what are the challenges you faced in conducting this analysis in the context
of evidence requirements?
Well, I think one is just the sheer number of studies that are published. And so, in
looking at 40 interventions, we, you know, had to read a
hundred thousand articles to find just a few, what we considered
to be articles that have good, strong evidence, good study design, good reporting. And so,
I think what you mentioned just about one barrier outside of
the field is that there are so many journals. I think in the field
there�s so much being published. And so, a big challenge is to really standardize that
search for evidence and to be transparent about it. In doing that,
we actually looked at more than 70 interventions recommended by the stakeholders, and of those
30, or so we didn�t think, had good enough evidence to
even model an effect, even making strong assumptions that many of the policies that are under
consideration really don�t have evidence of their effectiveness. And so, the limit
down to 40 is in part due to the scope of our work, but in part that many
of the things people are talking about really need more work
on the frontend to develop the evidence. So, I think that there are a number of places
where we�re excited about policies, but just don�t have anything
to model. The last is that using the logic model that
I talked about allows us to sort of bridge the gap
between the ideal evidence, where you have a specific policy change and you have measured
changes in BMI and healthcare cost. That would be ideal,
and there are almost no cases where we directly link a policy
change to the eventual outcome. And so, what we had to do was to make assumptions about
linking across that chain of a logic pathway, and that requires
readers to understand the limitations of the quality of
evidence at each piece of that chain. And also, I think it requires us to broaden our
search for what we consider to be evidence. And really, that
link between behavior changes to BMI often has randomized
controlled trials, but some of the links between policy changes and setting-level adoption
might be in nonpeer reviewed published reports.
And so, it�s taken, I think our group, time to expand the definition of what we consider
evidence, and then to really focus on having a core
piece of our argument, that if you change a specific behavior, it will
result in a reduction in BMI and obesity, to be the strongest scientifically, and then
to broaden our search out from there. But I think communicating
why we think that logic model, with that broad use of evidence,
is still a compelling piece of evidence is one of the core challenges. Whenever you have
multiple change, it gets more complicated, and communicating complexity
to a broad audience is very difficult. Well, I think you just made a compelling case
for the Self-Sufficiency Research Clearinghouse, because, you know, Kristin mentioned thousands
of articles and resources in there, and a lot of them aren�t
published research. They�re, you know, government documents and government projects that are
being shared. And so, that�s another source for
people to get information about interventions is actually going to
the Clearinghouse itself. And, I know I�ve been pushing our work there, and it also is
expanded in including presentations from conferences. So, I really
have been using that myself. So, just a little plug in there for the
SSRC. I want to follow up on your strong case for
the tax policy changes. That was very kind of
compelling, and in government I�m aware that other governments, outside of California,
have proposed legislation to curtail certain foods from
diets, especially amongst low-income families. For example, earlier
this year, Wisconsin proposed legislation � I�m not sure of the [unclear] � preventing
people from buying junk foods with their SNAP (Supplemental Nutrition
Assistance Program) benefits. You know, SNAP is
usually, well, it was formerly called food stamps. And Wisconsin did, similarly, wanted
to restrict beef and seafood, calling them luxury foods. And, in
my city, in Los Angeles, we had an ordinance passed in certain
areas that would prohibit new fast food restaurants from coming into the neighborhood. And a recent
report conducted by RAND documented this ordinance�s failure to do anything. It was pretty striking.
The report, in fact, documented that overweight
and obese/obesity rates, I should say, increased faster in these
areas than it did in other parts of Los Angeles, and nobody could really explain why at that
point. One of the arguments I heard was that there needed to
be more of a two-pronged approach, where we�re reducing the
number of fast food restaurants, but to nudge people in the right direction we needed to
make available to them fresher foods, like through farmers�
markets, and have other stores move in, and that didn�t happen.
And so, the neighborhood basically was just kind of left under � undeveloped over that
period of time. So, Michael, I want to ask you, what are your
thoughts about intervention by legislation that reduces access to
certain foods, as the examples that I just mentioned?
Well, I think we should look first at what has worked, and the Healthy Hunger-Free Kids
Act of 2010 was a sea change in the quality of the
school meals that are served and it is the first meaningful federal
legislation that required schools to stop selling junk food to kids outside of the school
meal. And, we think that�s one of the biggest successes in recent
policy change targeting childhood obesity. So, I think what you
could learn from that is that when policies are targeting specific settings that are already
under government control, particularly when we�re talking
about children, that there are changes that can be made that have a
real effect on the quality of children�s life, on the quality of their diets, and on
obesity rates. And so, I don�t think every policy will work, but I do think
there are examples where we can have real changes, and I think
the best place to start is in settings where we already have major government involvement.
And you brought up the SNAP program, and I think here is where it�s very important
to take that qualitative perspective as well, and to think
about, really carefully, what are the broader implications of
changes to the SNAP program. Why would you focus on it? It�s the largest food assistance
program in the country and it really is supposed to be targeting
nutrition. So, I think there�s a good argument to think
carefully about what we can do to improve the nutritional impact of the SNAP program,
and potentially use the program to reduce obesity rates. You�ve
seen meaningful changes in the WIC [Women, Infants and
Children] food package since the last revision, and a lot of people, researchers believe that
those changes in the WIC food package actually have led to
reductions in early childhood obesity. So, I do think it�s possible to
make changes to food assistance programs in ways that can still maintain their ability
to lower the negative impacts of poverty, while also promoting good
nutrition. The SNAP program I think is more complicated.
There was a Healthy Incentives Pilot in Massachusetts that found that incentivizing healthier food
purchases did work to an extent, and that people were
buying healthier foods if they were incentivized to do so. I think
restricting the kinds of foods in food stamps or in the SNAP program is a challenge, and
I think it�s been criticized in the food insecurity world, because
there�s been a move to remove restrictions from income
transfer programs. And this program is really one of the core federal policies that provides
income to families who really need it. So, I think taking
our eyes off of that goal and really making sure that children
aren�t living in poverty that�s so destitute that we�re not cutting off benefits by making
changes. But I do think there�s room for the USDA and for
the federal government to do demonstration projects to test
whether there are ways to improve the nutritional impact of the program. And there, I think,
there is a movement, and I think we would encourage this
in our modeling world, to not just make policy changes,
but to do pilot studies, to do good study designs within government to test whether
the changes you�re making actually would have an impact. And
so, that�s a hot-button political issue, but I do think it will remain
a focus because of the scale of the program and how important it is to families.
Thank you. Go ahead. I�m sorry.
There�s just a couple of questions that have come in from the listeners, and one is
whether it�s possible to have a copy of the slides. And
yes, the slides will be posted in a week or so. Another one is
around the monitoring. So the question is, in the assessment of improved childcare were
costs of monitoring, educating childcare workers, etcetera;
were they included in the costs? Yes. So, there was a training cost component,
as well as there�s already existing monitoring structures, and I think a portion of that
existing time was allocated to this intervention. So, yes, and I think I
highlighted in the active PE policy that we thought it was very important to include training
of principals to evaluate and monitor whether these changes
are actually happening in the tax policies. Of course, that�s
one of the primary costs of the interventions is to do monitoring for compliance through
auditing. So, I believe that in all of our 40 interventions,
we�ve taken consideration of whether there are added costs to
government or to industry, or other stakeholders for compliance activities; and definitely,
in early childcare, there are.
There is a related question on that. What would monitoring of an intervention look like?
Well, in the case of active PE, there are principals who would be evaluating teacher
performance and including this evaluation in their ongoing
evaluation of teacher performance, and they�d be trained on
how to evaluate whether teachers are able to do the practices that we think lead to
higher physical activity during existing PE. In the case of the tax
policy, that�s really the standard auditing practice that tax agencies
do. So, I think it really varies intervention by intervention, but, in general, there has
to be somebody who�s responsible for checking that these things
are actually being done. Thank you. And, I encourage listeners to submit
additional questions. I have another one here. Are there other content areas where this kind
of cost-effectiveness evaluation has been effective in
impacting or influencing policy development or change what self-sufficiency researchers
can look to for a model or an example?
I think there is work in early child development, and trying to evaluate what are the long-term
gains to society for improving early childcare, from a child development and self-sufficiency
perspective. So, you see those kinds of evaluations, and in
the Moving to Opportunity study, you see long-term evaluations
of what this kind of housing investment would do. I�ve been most struck by the political
impact that evaluations of changes in environmental standards
have. So, I think making a case for policy change at such
a broad level really requires this kind of thinking, and I think there�s been a lot
of good work on the environmental side. In the field of public
health, the vaccine planning and evaluation process using costeffectiveness
has really been built into the international funding structure, so that governments around
the world are actually using these kinds of methods
to make decisions about what kind of vaccines they want to
invest in. So I think it�s growing. It has a longer history in clinical medicine, and
it�s really growing rapidly in public health prevention, where NIH (National
Institutes of Health) and other funders are starting to request
that cost analyses are done, and that this thinking really has made the transition from
clinical medicine now to focus on how are we investing our very
limited public health dollars. Interesting. And, Michael Bono, it sounds
like you might have a question. Oh no. I was just going to comment on the
SNAP issue, and just to say that the USDA (United
States Department of Agriculture) is actually now getting ready to do a national survey
of SNAP participants, to ask them about the content of the food
that they�re purchasing, to kind of have some evidence to analyze
the nutritional value of what people are buying and moving towards, what Michael was talking
about. Any comment on that?
Yes. I think this is going to remain a focus for the USDA, is how do they really leverage
this very large investment in the SNAP program to make
sure they�re doing all they can to promote healthy diets for
participants. And I think always it�s balanced against the risk that a focus on potentially
poor nutritional choices by participants is used as a judgment
on them overall. And, I think we should stay away from a
judgment-based perspective and really focus on what can we do to improve the program to
help people to eat healthier diets, as opposed to judging
what people are doing. So, I think that there is active focus from
the USDA and others on what we can do that supports families, as opposed to what judges
them for what they�re doing now.
Kristin, I do� Go ahead.
I was going to say, I do have another question that�s totally different. So, Michael, you
were talking about the link between childhood obesity
and adult obesity, and I was wondering if you�re familiar
with any of the studies about what part of childhood is obesity, where the link is the
strongest? Is it in early childhood? Is it the case that early childhood
obesity that children are obese the remaining years, or is it
when it emerges in teenage years? Are you familiar with that literature?
Yes, and I think it�s actually that there�s a lot of change in childhood, particularly
for very young children, that the weight status does change
a lot and it�s less predictive. But as you get into adolescence
and later adolescence, that, of course, you�re already almost an adult, but the changes in
weight status that you see in later adolescence are very strongly
correlated with long-term adult BMI categories. So, the
question is, do you just intervene on adolescents, or are you trying to take a life-course perspective
and make sure that, at every transition point,
you�re making sure that when children leave early childcare they�re
getting into school at a healthy weight? When they�re in elementary school and transitioning
to middle school, they�re still at a healthy weight?
And I think it�s really about tying together interventions across the
life course that will allow us to have an effect on the long-term. If we just intervene
in early childcare where we can see some good effects, and then send
children off into environments that we know will make them
sick, then in one sense we�ve wasted an investment. But if we don�t do anything
early and we wait until children already have obesity, it�s much
harder to reverse. So, really, I would focus on taking a life-course
perspective and on intervening across�even prenatal, all the way through childhood into
adulthood, we should be thinking how can we tie together
across settings and across ages to make sure that children can
get to adulthood healthy, and then once they�re adults, that they also have healthy environments
and can maintain a healthy weight as they age.
I want to encourage listeners to do the survey before the end of our questions. We have time
for one more quick question. And I would be interested,
and this is my question, in asking how, from your
different perspectives, you feel that this kind of information can be presented and shared
to affect policy at the local level or the state level or the
national level? That it�s not generally, I imagine, publishing it in a
journal article. From the CHOICES perspective, I think what
we�re really focused on is how can we partner with
decision-makers, try to better understand what are the data inputs that they need to
make good decisions, and then work with them to improve their existing
decision-making process. And, I really do think that that
partnership model is how we can move the field forward the fastest.
Thank you. And, Michael Bono? Well, I think one of the ways that we get
information is by meeting people like Michael at
conferences. I actually met him for the first time a few weeks ago in Atlanta. But we get
a lot of information through the state government too, because
we�re a [unclear] administered state. And, to the extent that,
you know, the states learned about these things and are pushing it down to the local government,
you know, I think that that would be probably
the most effective strategy. For example, there�s been a lot of
interest kind of locally about trying to incorporate executive functioning research into changing
some of our policies. We�ve been also involved with
some behavioral economic interventions, and those are, you know,
again, those are kind of coming down with state support for us to kind of explore those.
So, and yes, again, we�re not reading journal articles. So,
to the extent that we can match up researchers like Michael with
people who are the policymakers in other ways is, I think, the way that we can communicate
better. Great. Well, I�d like to thank both of you
very much for this very interesting session. I encourage
everyone to fill out the survey. And, also, if you know of someone who would be a wonderful
Emerging Scholar and who would do a Webinar, please
nominate them as well. Thank you very much. Thank you. Thanks. That concludes today�s presentation. Thank
you for your participation.

Leave a Reply

Your email address will not be published. Required fields are marked *