April 15, 2025 • 27 mins
Let us know what you think about Health Affairs podcasts at communications@healthaffairs.org. If you have 30 minutes to spare, let us know and we'll set up a 30-minute chat for the first 20 listeners that reach out. Coffee will be on us.
Health Affairs' Senior Deputy Editor Rob Lott interviews Kurt Hager of the University of Massachusetts to discuss his recent paper that explores how Medicaid nutrition supports were associated with reductions in hospitalizations and emergency department visits in Massachusetts through 2020–23.
Order the April 2025 issue of Health Affairs.
Currently, more than 70 percent of our content is freely available - and we'd like to keep it that way. With your support, we can continue to keep our digital publication Forefront and podcast
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Rob Lott (00:31):
Hello, and welcome to a health podocy. I'm your host,
Rob Lott. The food is medicinemovement uses food based
interventions to address dietrelated health conditions. Think
prescription from your doctorfor fresh produce to be redeemed
(00:54):
at your local farmers market orthink about medically tailored
meals for patients withconditions like diabetes or
heart disease. These models areamong the concepts covered in
the April 2025 issue of HealthAffairs, a new theme issue
dedicated entirely to theintersection of food, health,
(01:15):
and nutrition.
Now although food is medicinetypically operates through or
partners with clinicians in thehealthcare system, not all
healthcare payers cover theseinterventions. Medicaid, for
example, typically does not,unless you're in a state that
has a waiver to test it, toexplore how it might actually
(01:37):
work and to study if food ismedicine can make a difference
in people's health andwell-being. Well, one of those
states that does have a waiveris Massachusetts and one of
those people studying its impactis Doctor. Kurt Hager, an
associate professor at the UMassChan Medical School. We're
(01:59):
delighted to have Doctor.
Hager here with us on today'sepisode of A Health Odyssey. He
has a paper in the April 2025issue whose title is also its
main general finding, quote,Medicaid nutrition supports
associated with reductions inhospitalizations and ED visits
(02:19):
in Massachusetts from 2020 to2023. We're so lucky to have
doctor Hager here with us.Welcome to this episode of
Health Odyssey.
Kurt Hager (02:30):
Thank you so much. It's wonderful to be here today
to talk about our study and whatwe found in Massachusetts
Medicaid.
Rob Lott (02:37):
So let's start with just a little bit of background.
I just described food asmedicine. I'd love your take
today. How'd I do? Did Idescribe it okay?
And how would you answer someonewho asked what exactly is the
food as medicine movement?
Kurt Hager (02:52):
Yeah. You you did a great job. I will say
specifically, how we use it andand conceptualize it in the
context of this study areprograms that are providing
healthy food to patients who'vebeen identified by their health
care provider, to be foodinsecure and having a diet
(03:12):
related, illness that would bemuch likely to be, improved
through with improved nutritionand getting access to healthy
food. So the main connection isa referral from a health care
provider who's identified apatient who would benefit from
this food, and it's actually,free provision of this healthy
food for an extended period oftime. So for example, in our
(03:36):
studies, many participantsreceive this food for about six
months.
So a substantial interventionthat's much more than just
nutrition education.
Rob Lott (03:45):
How widespread is this? Are most health care
systems engaging in this kind ofintervention, or is it still
relatively new?
Kurt Hager (03:53):
It is still relatively new, and I think it's
important to kind of ground thisconversation today,
acknowledging that theseprograms are unavailable to the
vast majority of Americans whomight benefit from them. So,
there's still it's an emergingconcept. There's more studies
happening. There's certainly alot more interest in this at a
(04:14):
policy level. But the challengeright now is that historically,
these programs have been paid bygrants and donations.
So often, the studies prior thatI've been involved in, for
example, have been funded by,let's say, a a food company or a
foundation, and they wanted toto pilot kind of a one off
(04:34):
program that that lasted sixmonths. But even, you know, if
the participants' healthimproved, by the end of that six
months, there really wasn'tthere was no opportunity to
continue the program becausebecause the funding had had run
short. So I think thathighlights some of the
challenges in integrating theseinto health care in a meaningful
way. And that, part of whatwe'll discuss with our study is,
(04:58):
in Massachusetts Medicaid, therewas indeed a new pathway for
that allowed more sustainablefunding for these programs and
allowed them to become a morereliable and integrated part of
clinical care for selectMedicaid members.
Rob Lott (05:11):
Okay, great. I want to hear a little more about that
program, but one more questionfirst is sort of about the
evidence. How do we know orrather what do we know about the
effectiveness of programs likethis? I would say prior to
talking about your study and thefindings of your study, what was
sort of the general consensusabout how these programs work
(05:34):
with the caveat obviously thatthere sort of different versions
and different kinds of programsthat might be effective in
different ways or to differentdegrees?
Kurt Hager (05:44):
So there's a handful of evidence to suggest that
these programs can improve,first and foremost, food
insecurity and and dietaryintake, we know, food insecurity
and suboptimal diet are areassociated with, higher rates of
diet related chronic illness,higher health care utilization,
(06:07):
and higher health care costs. Sothere's there's a there's a
wealth of evidence that theseprograms can impact those
important, measures. There'salso increasing evidence that
these programs can improve, keyclinical outcomes. Their
strongest evidence that they canimprove markers of
cardiometabolic health. So thismeans things like, blood
(06:29):
pressure, hemoglobin a one c, orblood sugar levels among those
with diabetes.
And there's even some, researchthat has shown prior to this
study, reductions in acutehealth care utilization,
including reductions inhospitalizations and ED visits
and even health care costs. ButI will say that this evidence is
(06:50):
still emerging. There are a lota lot of these studies have been
quasi experimental studies, andwe are now just seeing more and
more randomized trials that areseeking to confirm the findings
from previous studies.
Rob Lott (07:05):
Great. Well, that's a perfect segue to talk about your
research. Can you describe theprogram that you were studying?
Kurt Hager (07:12):
So as you mentioned in your intro, Medicaid at a
national level, largely does notcover, food as medicine
services. So, it cannot directlypay for food to, individuals
enrolled in Medicaid. However,there is a pathway that has been
(07:33):
used increasingly by statescalled the eleven fifteen
waiver. And at a high level,these waivers allow states to
pilot innovative idea in theirMedicaid population that they
believe is likely to improvehealth outcomes and also, is
cost neutral. So it doesn't addcost to the state or the federal
government.
(07:53):
And then centers for Medicareand Medicaid services can
approve the waivers for up tofive years. And so Massachusetts
was the first state to get aneleven fifteen waiver that
included direct provision ofhealthy food to select members.
Rob Lott (08:09):
Once Massachusetts had a waiver, what kind of specific
program did they create withthat opportunity?
Kurt Hager (08:15):
One component of Massachusetts eleven fifteen
waiver was the creation of theflexible services program. So
the flexible services programoffered both nutrition and
housing supports to high riskMedicaid members. And this is a
key thing to keep in mind isthat this program was not
available to all Medicaidmembers and was highly targeted
and tailored to specificmembers. So the eligibility
(08:39):
criteria for the nutritionprogram, and this was the focus
of our study, members had to beenrolled in a newly created
accountable care organizationand the creation of those
accountable care organizationsor ACOs was also a part of the
eleven fifteen waiver. They hadto be under age 65 plus meet at
(08:59):
least one health needs basedcriteria.
And so MassHealth, theMassachusetts Medicaid
organization set, a range ofhealth needs based criteria.
This included a behavioralhealth need or a persistent
physical medical condition likediabetes or heart failure, need
(09:20):
for assistance with activitiesof daily living. So we can think
of this as someone needing helpwith everyday tasks and living
independently, high emergencydepartment use, or a high risk
pregnancy. So you can see theseare a pretty high risk, high
need member population, and plusthe final eligibility criteria
(09:40):
was, also having foodinsecurity.
Rob Lott (09:44):
Okay. Great. So you've got this qualified population.
And can you describe a littlebit what intervention or what
kind of interaction thispopulation had with clinicians
or some sort of supportivenetwork, I presume, implementing
this food is medicineintervention?
Kurt Hager (10:04):
Yes. So when MassHealth designed the flexible
services program, they designedit in a way to give accountable
care organizations a lot offlexibility to create their own
programs that they thought wouldbe best suited to meet their
members' needs and to choosetheir own, partnerships with
community based organizations intheir service area, to to design
(10:28):
and implement those programs. Sowhat we found was that there was
a wide mix of program designs,and we included all of those in
our studies. So you mentionedsome of them at the start of the
episode. This includes medicallytailored meals that are often
home delivered to, to people andare tailored by registered
dietitians to to meet the veryspecific medical conditions of
(10:51):
individuals.
These included produceprescriptions. They were often
implemented on electronic cardsthat people could use to
purchase produce of their choiceat retail grocery stores. Other
program models included foodboxes, so kind of like raw
ingredients that people couldthen prepare meals on their own,
and others included applicationassistance to other federal
(11:15):
nutrition programs and referralsto food pantries in their in
their area.
Rob Lott (11:21):
Well, I can't wait to hear what you found studying
those programs. But first, we'regonna take a quick break. And
(11:45):
we're back. I'm here talkingwith Doctor. Kurt Hager about
his study in the April 2025issue of Health Affairs about
food is medicine.
Kurt, can you tell us a littlebit about what you found when
studying the programs underMassachusetts Medicaid waiver?
Kurt Hager (12:04):
Our study focused on acute health care utilization
and cost. So the outcomes wewere interested in were changes
in hospitalizations, emergencydepartment visits, and health
care costs. And we also lookedat primary care visits as a
secondary outcome. Ourcomparison group included
(12:24):
Medicaid members who wereeligible for flexible services
but did not receive thoseservices. And we looked at the
change in, healthcareutilization and costs in the six
months prior to startingenrollment in flexible services
to the period during whichindividuals receiving those
services.
So we and we compared thatchange from the baseline period
(12:47):
to the program period and lookedat that difference between the
treatment group and thecomparison group. And we
accounted for, a long list ofpotential confounders, things
like diagnoses, baseline healthcare utilization, where someone
lived, their sociodemographics,and the health systems that were
referring them. So after,combining all that data together
(13:10):
and comparing the two groups, wefound that receiving food from
the flexible services program,resulted in a twenty three
percent reduction inhospitalizations and a thirteen
percent reduction in emergencydepartment visits. And this was
among just over 20,000 Medicaidmembers who received these, food
(13:33):
programs from January 2020 toMarch 2023. So we also looked at
health care costs, as Imentioned, and across all of
these 20,000 individuals, therewas a change in health care
costs that favored participants,but it was not statistically
significant.
(13:53):
But we because this is such alarge group and, a long time
period of several years. Andimportantly, as I mentioned,
these programs started inJanuary 2020. So this is right
before the pandemic hit. So I dowanna give a huge kudos to
MassHealth, the accountable careorganizations, and the community
based organizations forlaunching this huge program
(14:17):
during such a tumultuous time.And so but because of that, we
thought it was really importantto look at the effects of the
program and the immediateoutbreak of the COVID nineteen
pandemic versus later on, oncevaccines are widely available
and the economic shutdowns hadlargely stopped.
So what we found is in the firsttwo years of the program in 2020
(14:40):
and 2021, there was actually nochange in health care
utilization or costs amongrecipients of the flexible
services nutrition programs.However, in years 2022 and 2023,
we found even greaterreductions, in hospitalizations
and ED visits, specifically aforty seven percent reduction in
(15:03):
hospitalizations and a twentyone percent reduction in
emergency department visits. Andduring this period, we did see a
statistically significantreduction in health care costs
of about $1,700 per person whenthey were receiving their food
supports.
Rob Lott (15:21):
So if I can just drill down on that for a minute, I
think, generally, the assumptionwas that utilization and
spending decreased in the earlystages of the pandemic broadly.
Kurt Hager (15:32):
Correct.
Rob Lott (15:33):
And it sounds like what you're saying is that the
reductions here came later sothat presumably the cause or the
driver of those reductions wasnot sort of the general pandemic
effects, but really moreattributable potentially to this
program. Is that a fair take?
Kurt Hager (15:55):
So, you bring up a great point. And if I understand
your question correctly, it's,kind of like what it might have
been driving this kind of nullfinding in those early years.
Yeah. So we can't stateconcretely what it was, but I
suspect it's a combination ofmultiple factors. So as you
(16:16):
mentioned, people were scared ofgetting COVID nineteen, and
there it was an observedphenomenon across The US that
there were actually feweremergency department visits.
People were, you know there wasoverall less health care
utilization except, of course,among those who are getting very
sick with COVID nineteen. Sojust statistically, that does
(16:37):
make it, more challenging todetect an effect when there's
fewer events occurring. But,also, let's think about that
time period. There is suchmassive disruptions in clinical
care, as we mentioned, hugedownturns in the economy. And
we're talking about a Medicaidpopulation that, on average, was
probably much less likely tohave jobs in which they could
(17:00):
work remote.
There was probably largefinancial upheavals for many of
these members. Schools wereclosed for for quite some time,
so increased childcare demands.There were, you know, increases
in food costs during this timeperiod. And on the flip side,
there also was a really robustfederal response to the
(17:20):
pandemic, did increase federalnutrition benefits. There were
large stimulus checks that wentout, tax credits to family, and
members in both the treatmentgroup and the comparison group
would have received those otherlarge support.
So it's also possible therecould have been some kind of
washing out of the, effects ofthe program during this time. So
(17:41):
it's impossible to know, but Ithink it's, important to think
through these these multipleinterconnected factors. And, it
was an interesting finding tosee in the early years of the
program. We didn't see thischange in health care
utilization, but that it was infact even larger in a period of
greater stability once the worstthe pandemic was winding down.
Rob Lott (18:04):
That's fascinating, and I'm sure it'll be really
interesting to continue to trackthe impact in the years ahead as
well as the pandemic is furtherand further in the rearview
mirror. You alluded to some ofthe other sort of pandemic era
benefits, but I wonder if youcan say a little more about
(18:26):
programs like SNAP and WIC thatMedicaid enrollees are also
typically or often eligible for.And I'm wondering how you think
about the interaction betweenthis program and its benefits
and the benefits of nutritionsupport programs like SNAP and
WIC.
Kurt Hager (18:47):
So within the context of our study, we did not
have access to SNAP and WICenrollment. In a related project
we were hoping to do, we weretrying to link or trying to see
if it was possible to link theSNAP and Medicaid data at the
(19:09):
state level. And turns out it'sincredibly challenging thing to
do, actually, to to merge someof these these programs, that
are, you know, run by differentstate departments. So that
proved challenging. So thatcould be certainly a limitation
of our study.
We didn't have access to thatdata. But I think it's, one
thing to keep in mind is that,most people in Medicaid would be
(19:31):
eligible for SNAP and, ofcourse, you know, pregnant or
postpartum women and theirnewborn children would be
eligible for WIC. InMassachusetts, there is a common
application for SNAP andMedicaid, so someone can apply
for both programs at once. Sowhile we don't know for sure, I
would expect there to be similarenrollment rates in SNAP in the
(19:53):
treatment group and thecomparison group, but we don't
we don't know that, for surebecause that was data we we
didn't have access to.
Rob Lott (20:00):
I saw that your study included both children and
adults. Did you find anydifferences between those two
populations?
Kurt Hager (20:06):
Yes. We did. So we did another stratified analysis
looking at outcomes separatelybetween children and adults.
Interestingly, but perhaps notsurprisingly, we found no change
in health care utilization orcosts among children. Children
tend to be, you know, healthieroverall and have fewer
hospitalizations and emergencydepartment visits.
(20:27):
But that is not to say theseprograms aren't, wouldn't be
beneficial to childhooddevelopment. There are many
other outcomes. Like, we knowthat increased food insecurity,
increased dietary quality are soimportant for childhood
development and well-being. Wejust didn't analyze those within
this study. So, that could besomething that we look at in the
(20:50):
future.
For the adults, those resultswere generally quite similar to
the overall finding, but we alsolooked at the impact of
enrollment length. And a reallykey finding from our study was
that among adults who wereenrolled for three months or
longer, had slightly largerreductions in hospitalizations
(21:11):
and ED visits. But mostimportantly, the also had a
reduction in health care costs.And this reduction in health
care costs was greater than thecost of paying for the program.
So this resulted in about $200of cost savings per adult member
for the Medicaid program.
And these costs included thefood costs and also the
(21:34):
administrative costs at the ACOsand for MassHealth in
implementing the program.
Rob Lott (21:38):
Gotcha. Okay. So a really important finding for
potential decisions about futureprojects and and more widespread
implementation, which maybebrings us to the fact that there
are some food as medicineskeptics out there, folks who
sometimes point to challengeslike low levels of adherence, or
(22:01):
maybe the unnecessarymedicalization of nutrition as
reasons to approach theseprograms with caution. And I'm
curious, how your findings fromthis study maybe inform that
debate.
Kurt Hager (22:15):
Great. Well, both are reasonable critiques, and,
I'll start with the adherencecomment first. So first, I think
we all absolutely who areworking in the space should
continually work to refineprograms to maximize participant
engagement and adherence, andthat is much more likely to
(22:36):
improve health outcomes. I wannaapplaud the American Heart
Association who's launched ahuge effort to, conduct trials
that are aimed at improvingprogram implementation and
design so that we can begin,really moving forward in an
important and meaningful way, tocome perfect these programs, to
(22:59):
increase health outcomes.Personally, I believe any
program that maximizesparticipant choice and the foods
that they receive and maximizesengagement, will be most likely
to improve health outcomes andhave higher adherence levels.
Second, regarding themedicalization, this is also a a
(23:20):
a reasonable comment. I wouldsay, you know, to me, there are
high food insecurity rates amongthose with diet related chronic
illness in Medicaid. And to me,this reflects that existing
federal nutrition programslikely aren't sufficient to
adequately address thenutritional needs of these
members to prevent, treat, andmanage diet related illness. And
(23:42):
thus, health care providers andstate Medicaid programs are
really eager to find additionalsolutions to improve clinical
care, health and well-being, andhealth care utilization costs as
our as our study showed. I'vealways personally found it
interesting that food insecurityis grouped into this bucket of
social determinants of health,and I understand, some academics
(24:06):
like to debate terminology thatfood insecurity now reflects
food access.
But to me, I feel like sometimeswe are forgetting that,
nutrition is perhaps the mostcore foundational biological
determinant of health. And Ibelieve our results in part
reflect that reality and show,this is true and and that these
(24:30):
types of programs really canhave meaningful impacts on
health outcomes as reflected byfewer hospitalizations,
emergency department visits, andfor, many participants in our
study, fewer health care costsas well.
Rob Lott (24:46):
Okay. What we're talking, in early April. Your
paper just came out, and I'mwondering if you've gotten any
early feedback. What's theresponse been, and how are you
feeling about it?
Kurt Hager (24:56):
We are certainly very excited. Our partners, at
Massachusetts Medicaid whoworked so hard for years to
design and implement thisprogram are very pleased with
the results. We are also quiteencouraged by a very similar
study that was published justseveral weeks ago looking at the
eleven fifteen waiver in NorthCarolina, which also authorized
(25:19):
very similar food as medicineprograms. This was actually,
published in JAMA and led bydoctor Seth Berkowitz, who I
know was on your podcastrecently. And, I'm quite
encouraged by very similarfindings between the two states
and the two studies.
So there are a few keydifferences in the study designs
(25:41):
in the program implementation,but largely very similar,
program models were in both,studies in states. But
Massachusetts, North Carolinaare very different state
context. And the fact that wenow have two, state eleven
fifteen Medicaid waivers thatare covering these services in
(26:02):
slightly different capacities,but are both finding reductions
in emergency department visitsand, in many cases, reductions
in health care costs, to me, isan encouraging finding that,
this could be replicated inother states and that there is,
in fact, bipartisan support tocontinue programs like this.
Rob Lott (26:23):
Well, that's a great encouraging note to end on.
Thank you so much, doctor KurtHager, for taking the time to
chat with us today.
Kurt Hager (26:33):
Yeah. Thank you so much.
Rob Lott (26:34):
To our listeners and our readers, if you enjoyed this
podcast, please recommend it toa friend, smash that subscribe
button and tune in next week.Take care all.
Kurt Hager (26:49):
Thanks for listening. If you enjoyed
today's episode, I hope you'lltell a friend about a health
policy.
Hello, and welcome to a health podocy. I'm your host,
Rob Lott. The food is medicinemovement uses food based
interventions to address dietrelated health conditions. Think
prescription from your doctorfor fresh produce to be redeemed
(00:54):
at your local farmers market orthink about medically tailored
meals for patients withconditions like diabetes or
heart disease. These models areamong the concepts covered in
the April 2025 issue of HealthAffairs, a new theme issue
dedicated entirely to theintersection of food, health,
(01:15):
and nutrition.
Now although food is medicinetypically operates through or
partners with clinicians in thehealthcare system, not all
healthcare payers cover theseinterventions. Medicaid, for
example, typically does not,unless you're in a state that
has a waiver to test it, toexplore how it might actually
(01:37):
work and to study if food ismedicine can make a difference
in people's health andwell-being. Well, one of those
states that does have a waiveris Massachusetts and one of
those people studying its impactis Doctor. Kurt Hager, an
associate professor at the UMassChan Medical School. We're
(01:59):
delighted to have Doctor.
Hager here with us on today'sepisode of A Health Odyssey. He
has a paper in the April 2025issue whose title is also its
main general finding, quote,Medicaid nutrition supports
associated with reductions inhospitalizations and ED visits
(02:19):
in Massachusetts from 2020 to2023. We're so lucky to have
doctor Hager here with us.Welcome to this episode of
Health Odyssey.
Kurt Hager (02:30):
Thank you so much. It's wonderful to be here today
to talk about our study and whatwe found in Massachusetts
Medicaid.
Rob Lott (02:37):
So let's start with just a little bit of background.
I just described food asmedicine. I'd love your take
today. How'd I do? Did Idescribe it okay?
And how would you answer someonewho asked what exactly is the
food as medicine movement?
Kurt Hager (02:52):
Yeah. You you did a great job. I will say
specifically, how we use it andand conceptualize it in the
context of this study areprograms that are providing
healthy food to patients who'vebeen identified by their health
care provider, to be foodinsecure and having a diet
(03:12):
related, illness that would bemuch likely to be, improved
through with improved nutritionand getting access to healthy
food. So the main connection isa referral from a health care
provider who's identified apatient who would benefit from
this food, and it's actually,free provision of this healthy
food for an extended period oftime. So for example, in our
(03:36):
studies, many participantsreceive this food for about six
months.
So a substantial interventionthat's much more than just
nutrition education.
Rob Lott (03:45):
How widespread is this? Are most health care
systems engaging in this kind ofintervention, or is it still
relatively new?
Kurt Hager (03:53):
It is still relatively new, and I think it's
important to kind of ground thisconversation today,
acknowledging that theseprograms are unavailable to the
vast majority of Americans whomight benefit from them. So,
there's still it's an emergingconcept. There's more studies
happening. There's certainly alot more interest in this at a
(04:14):
policy level. But the challengeright now is that historically,
these programs have been paid bygrants and donations.
So often, the studies prior thatI've been involved in, for
example, have been funded by,let's say, a a food company or a
foundation, and they wanted toto pilot kind of a one off
(04:34):
program that that lasted sixmonths. But even, you know, if
the participants' healthimproved, by the end of that six
months, there really wasn'tthere was no opportunity to
continue the program becausebecause the funding had had run
short. So I think thathighlights some of the
challenges in integrating theseinto health care in a meaningful
way. And that, part of whatwe'll discuss with our study is,
(04:58):
in Massachusetts Medicaid, therewas indeed a new pathway for
that allowed more sustainablefunding for these programs and
allowed them to become a morereliable and integrated part of
clinical care for selectMedicaid members.
Rob Lott (05:11):
Okay, great. I want to hear a little more about that
program, but one more questionfirst is sort of about the
evidence. How do we know orrather what do we know about the
effectiveness of programs likethis? I would say prior to
talking about your study and thefindings of your study, what was
sort of the general consensusabout how these programs work
(05:34):
with the caveat obviously thatthere sort of different versions
and different kinds of programsthat might be effective in
different ways or to differentdegrees?
Kurt Hager (05:44):
So there's a handful of evidence to suggest that
these programs can improve,first and foremost, food
insecurity and and dietaryintake, we know, food insecurity
and suboptimal diet are areassociated with, higher rates of
diet related chronic illness,higher health care utilization,
(06:07):
and higher health care costs. Sothere's there's a there's a
wealth of evidence that theseprograms can impact those
important, measures. There'salso increasing evidence that
these programs can improve, keyclinical outcomes. Their
strongest evidence that they canimprove markers of
cardiometabolic health. So thismeans things like, blood
(06:29):
pressure, hemoglobin a one c, orblood sugar levels among those
with diabetes.
And there's even some, researchthat has shown prior to this
study, reductions in acutehealth care utilization,
including reductions inhospitalizations and ED visits
and even health care costs. ButI will say that this evidence is
(06:50):
still emerging. There are a lota lot of these studies have been
quasi experimental studies, andwe are now just seeing more and
more randomized trials that areseeking to confirm the findings
from previous studies.
Rob Lott (07:05):
Great. Well, that's a perfect segue to talk about your
research. Can you describe theprogram that you were studying?
Kurt Hager (07:12):
So as you mentioned in your intro, Medicaid at a
national level, largely does notcover, food as medicine
services. So, it cannot directlypay for food to, individuals
enrolled in Medicaid. However,there is a pathway that has been
(07:33):
used increasingly by statescalled the eleven fifteen
waiver. And at a high level,these waivers allow states to
pilot innovative idea in theirMedicaid population that they
believe is likely to improvehealth outcomes and also, is
cost neutral. So it doesn't addcost to the state or the federal
government.
(07:53):
And then centers for Medicareand Medicaid services can
approve the waivers for up tofive years. And so Massachusetts
was the first state to get aneleven fifteen waiver that
included direct provision ofhealthy food to select members.
Rob Lott (08:09):
Once Massachusetts had a waiver, what kind of specific
program did they create withthat opportunity?
Kurt Hager (08:15):
One component of Massachusetts eleven fifteen
waiver was the creation of theflexible services program. So
the flexible services programoffered both nutrition and
housing supports to high riskMedicaid members. And this is a
key thing to keep in mind isthat this program was not
available to all Medicaidmembers and was highly targeted
and tailored to specificmembers. So the eligibility
(08:39):
criteria for the nutritionprogram, and this was the focus
of our study, members had to beenrolled in a newly created
accountable care organizationand the creation of those
accountable care organizationsor ACOs was also a part of the
eleven fifteen waiver. They hadto be under age 65 plus meet at
(08:59):
least one health needs basedcriteria.
And so MassHealth, theMassachusetts Medicaid
organization set, a range ofhealth needs based criteria.
This included a behavioralhealth need or a persistent
physical medical condition likediabetes or heart failure, need
(09:20):
for assistance with activitiesof daily living. So we can think
of this as someone needing helpwith everyday tasks and living
independently, high emergencydepartment use, or a high risk
pregnancy. So you can see theseare a pretty high risk, high
need member population, and plusthe final eligibility criteria
(09:40):
was, also having foodinsecurity.
Rob Lott (09:44):
Okay. Great. So you've got this qualified population.
And can you describe a littlebit what intervention or what
kind of interaction thispopulation had with clinicians
or some sort of supportivenetwork, I presume, implementing
this food is medicineintervention?
Kurt Hager (10:04):
Yes. So when MassHealth designed the flexible
services program, they designedit in a way to give accountable
care organizations a lot offlexibility to create their own
programs that they thought wouldbe best suited to meet their
members' needs and to choosetheir own, partnerships with
community based organizations intheir service area, to to design
(10:28):
and implement those programs. Sowhat we found was that there was
a wide mix of program designs,and we included all of those in
our studies. So you mentionedsome of them at the start of the
episode. This includes medicallytailored meals that are often
home delivered to, to people andare tailored by registered
dietitians to to meet the veryspecific medical conditions of
(10:51):
individuals.
These included produceprescriptions. They were often
implemented on electronic cardsthat people could use to
purchase produce of their choiceat retail grocery stores. Other
program models included foodboxes, so kind of like raw
ingredients that people couldthen prepare meals on their own,
and others included applicationassistance to other federal
(11:15):
nutrition programs and referralsto food pantries in their in
their area.
Rob Lott (11:21):
Well, I can't wait to hear what you found studying
those programs. But first, we'regonna take a quick break. And
(11:45):
we're back. I'm here talkingwith Doctor. Kurt Hager about
his study in the April 2025issue of Health Affairs about
food is medicine.
Kurt, can you tell us a littlebit about what you found when
studying the programs underMassachusetts Medicaid waiver?
Kurt Hager (12:04):
Our study focused on acute health care utilization
and cost. So the outcomes wewere interested in were changes
in hospitalizations, emergencydepartment visits, and health
care costs. And we also lookedat primary care visits as a
secondary outcome. Ourcomparison group included
(12:24):
Medicaid members who wereeligible for flexible services
but did not receive thoseservices. And we looked at the
change in, healthcareutilization and costs in the six
months prior to startingenrollment in flexible services
to the period during whichindividuals receiving those
services.
So we and we compared thatchange from the baseline period
(12:47):
to the program period and lookedat that difference between the
treatment group and thecomparison group. And we
accounted for, a long list ofpotential confounders, things
like diagnoses, baseline healthcare utilization, where someone
lived, their sociodemographics,and the health systems that were
referring them. So after,combining all that data together
(13:10):
and comparing the two groups, wefound that receiving food from
the flexible services program,resulted in a twenty three
percent reduction inhospitalizations and a thirteen
percent reduction in emergencydepartment visits. And this was
among just over 20,000 Medicaidmembers who received these, food
(13:33):
programs from January 2020 toMarch 2023. So we also looked at
health care costs, as Imentioned, and across all of
these 20,000 individuals, therewas a change in health care
costs that favored participants,but it was not statistically
significant.
(13:53):
But we because this is such alarge group and, a long time
period of several years. Andimportantly, as I mentioned,
these programs started inJanuary 2020. So this is right
before the pandemic hit. So I dowanna give a huge kudos to
MassHealth, the accountable careorganizations, and the community
based organizations forlaunching this huge program
(14:17):
during such a tumultuous time.And so but because of that, we
thought it was really importantto look at the effects of the
program and the immediateoutbreak of the COVID nineteen
pandemic versus later on, oncevaccines are widely available
and the economic shutdowns hadlargely stopped.
So what we found is in the firsttwo years of the program in 2020
(14:40):
and 2021, there was actually nochange in health care
utilization or costs amongrecipients of the flexible
services nutrition programs.However, in years 2022 and 2023,
we found even greaterreductions, in hospitalizations
and ED visits, specifically aforty seven percent reduction in
(15:03):
hospitalizations and a twentyone percent reduction in
emergency department visits. Andduring this period, we did see a
statistically significantreduction in health care costs
of about $1,700 per person whenthey were receiving their food
supports.
Rob Lott (15:21):
So if I can just drill down on that for a minute, I
think, generally, the assumptionwas that utilization and
spending decreased in the earlystages of the pandemic broadly.
Kurt Hager (15:32):
Correct.
Rob Lott (15:33):
And it sounds like what you're saying is that the
reductions here came later sothat presumably the cause or the
driver of those reductions wasnot sort of the general pandemic
effects, but really moreattributable potentially to this
program. Is that a fair take?
Kurt Hager (15:55):
So, you bring up a great point. And if I understand
your question correctly, it's,kind of like what it might have
been driving this kind of nullfinding in those early years.
Yeah. So we can't stateconcretely what it was, but I
suspect it's a combination ofmultiple factors. So as you
(16:16):
mentioned, people were scared ofgetting COVID nineteen, and
there it was an observedphenomenon across The US that
there were actually feweremergency department visits.
People were, you know there wasoverall less health care
utilization except, of course,among those who are getting very
sick with COVID nineteen. Sojust statistically, that does
(16:37):
make it, more challenging todetect an effect when there's
fewer events occurring. But,also, let's think about that
time period. There is suchmassive disruptions in clinical
care, as we mentioned, hugedownturns in the economy. And
we're talking about a Medicaidpopulation that, on average, was
probably much less likely tohave jobs in which they could
(17:00):
work remote.
There was probably largefinancial upheavals for many of
these members. Schools wereclosed for for quite some time,
so increased childcare demands.There were, you know, increases
in food costs during this timeperiod. And on the flip side,
there also was a really robustfederal response to the
(17:20):
pandemic, did increase federalnutrition benefits. There were
large stimulus checks that wentout, tax credits to family, and
members in both the treatmentgroup and the comparison group
would have received those otherlarge support.
So it's also possible therecould have been some kind of
washing out of the, effects ofthe program during this time. So
(17:41):
it's impossible to know, but Ithink it's, important to think
through these these multipleinterconnected factors. And, it
was an interesting finding tosee in the early years of the
program. We didn't see thischange in health care
utilization, but that it was infact even larger in a period of
greater stability once the worstthe pandemic was winding down.
Rob Lott (18:04):
That's fascinating, and I'm sure it'll be really
interesting to continue to trackthe impact in the years ahead as
well as the pandemic is furtherand further in the rearview
mirror. You alluded to some ofthe other sort of pandemic era
benefits, but I wonder if youcan say a little more about
(18:26):
programs like SNAP and WIC thatMedicaid enrollees are also
typically or often eligible for.And I'm wondering how you think
about the interaction betweenthis program and its benefits
and the benefits of nutritionsupport programs like SNAP and
WIC.
Kurt Hager (18:47):
So within the context of our study, we did not
have access to SNAP and WICenrollment. In a related project
we were hoping to do, we weretrying to link or trying to see
if it was possible to link theSNAP and Medicaid data at the
(19:09):
state level. And turns out it'sincredibly challenging thing to
do, actually, to to merge someof these these programs, that
are, you know, run by differentstate departments. So that
proved challenging. So thatcould be certainly a limitation
of our study.
We didn't have access to thatdata. But I think it's, one
thing to keep in mind is that,most people in Medicaid would be
(19:31):
eligible for SNAP and, ofcourse, you know, pregnant or
postpartum women and theirnewborn children would be
eligible for WIC. InMassachusetts, there is a common
application for SNAP andMedicaid, so someone can apply
for both programs at once. Sowhile we don't know for sure, I
would expect there to be similarenrollment rates in SNAP in the
(19:53):
treatment group and thecomparison group, but we don't
we don't know that, for surebecause that was data we we
didn't have access to.
Rob Lott (20:00):
I saw that your study included both children and
adults. Did you find anydifferences between those two
populations?
Kurt Hager (20:06):
Yes. We did. So we did another stratified analysis
looking at outcomes separatelybetween children and adults.
Interestingly, but perhaps notsurprisingly, we found no change
in health care utilization orcosts among children. Children
tend to be, you know, healthieroverall and have fewer
hospitalizations and emergencydepartment visits.
(20:27):
But that is not to say theseprograms aren't, wouldn't be
beneficial to childhooddevelopment. There are many
other outcomes. Like, we knowthat increased food insecurity,
increased dietary quality are soimportant for childhood
development and well-being. Wejust didn't analyze those within
this study. So, that could besomething that we look at in the
(20:50):
future.
For the adults, those resultswere generally quite similar to
the overall finding, but we alsolooked at the impact of
enrollment length. And a reallykey finding from our study was
that among adults who wereenrolled for three months or
longer, had slightly largerreductions in hospitalizations
(21:11):
and ED visits. But mostimportantly, the also had a
reduction in health care costs.And this reduction in health
care costs was greater than thecost of paying for the program.
So this resulted in about $200of cost savings per adult member
for the Medicaid program.
And these costs included thefood costs and also the
(21:34):
administrative costs at the ACOsand for MassHealth in
implementing the program.
Rob Lott (21:38):
Gotcha. Okay. So a really important finding for
potential decisions about futureprojects and and more widespread
implementation, which maybebrings us to the fact that there
are some food as medicineskeptics out there, folks who
sometimes point to challengeslike low levels of adherence, or
(22:01):
maybe the unnecessarymedicalization of nutrition as
reasons to approach theseprograms with caution. And I'm
curious, how your findings fromthis study maybe inform that
debate.
Kurt Hager (22:15):
Great. Well, both are reasonable critiques, and,
I'll start with the adherencecomment first. So first, I think
we all absolutely who areworking in the space should
continually work to refineprograms to maximize participant
engagement and adherence, andthat is much more likely to
(22:36):
improve health outcomes. I wannaapplaud the American Heart
Association who's launched ahuge effort to, conduct trials
that are aimed at improvingprogram implementation and
design so that we can begin,really moving forward in an
important and meaningful way, tocome perfect these programs, to
(22:59):
increase health outcomes.Personally, I believe any
program that maximizesparticipant choice and the foods
that they receive and maximizesengagement, will be most likely
to improve health outcomes andhave higher adherence levels.
Second, regarding themedicalization, this is also a a
(23:20):
a reasonable comment. I wouldsay, you know, to me, there are
high food insecurity rates amongthose with diet related chronic
illness in Medicaid. And to me,this reflects that existing
federal nutrition programslikely aren't sufficient to
adequately address thenutritional needs of these
members to prevent, treat, andmanage diet related illness. And
(23:42):
thus, health care providers andstate Medicaid programs are
really eager to find additionalsolutions to improve clinical
care, health and well-being, andhealth care utilization costs as
our as our study showed. I'vealways personally found it
interesting that food insecurityis grouped into this bucket of
social determinants of health,and I understand, some academics
(24:06):
like to debate terminology thatfood insecurity now reflects
food access.
But to me, I feel like sometimeswe are forgetting that,
nutrition is perhaps the mostcore foundational biological
determinant of health. And Ibelieve our results in part
reflect that reality and show,this is true and and that these
(24:30):
types of programs really canhave meaningful impacts on
health outcomes as reflected byfewer hospitalizations,
emergency department visits, andfor, many participants in our
study, fewer health care costsas well.
Rob Lott (24:46):
Okay. What we're talking, in early April. Your
paper just came out, and I'mwondering if you've gotten any
early feedback. What's theresponse been, and how are you
feeling about it?
Kurt Hager (24:56):
We are certainly very excited. Our partners, at
Massachusetts Medicaid whoworked so hard for years to
design and implement thisprogram are very pleased with
the results. We are also quiteencouraged by a very similar
study that was published justseveral weeks ago looking at the
eleven fifteen waiver in NorthCarolina, which also authorized
(25:19):
very similar food as medicineprograms. This was actually,
published in JAMA and led bydoctor Seth Berkowitz, who I
know was on your podcastrecently. And, I'm quite
encouraged by very similarfindings between the two states
and the two studies.
So there are a few keydifferences in the study designs
(25:41):
in the program implementation,but largely very similar,
program models were in both,studies in states. But
Massachusetts, North Carolinaare very different state
context. And the fact that wenow have two, state eleven
fifteen Medicaid waivers thatare covering these services in
(26:02):
slightly different capacities,but are both finding reductions
in emergency department visitsand, in many cases, reductions
in health care costs, to me, isan encouraging finding that,
this could be replicated inother states and that there is,
in fact, bipartisan support tocontinue programs like this.
Rob Lott (26:23):
Well, that's a great encouraging note to end on.
Thank you so much, doctor KurtHager, for taking the time to
chat with us today.
Kurt Hager (26:33):
Yeah. Thank you so much.
Rob Lott (26:34):
To our listeners and our readers, if you enjoyed this
podcast, please recommend it toa friend, smash that subscribe
button and tune in next week.Take care all.
Kurt Hager (26:49):
Thanks for listening. If you enjoyed
today's episode, I hope you'lltell a friend about a health
policy.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.