Health Policy to Address Hunger: All Carrots, No Sticks

Name: Health Policy to Address Hunger: All Carrots, No Sticks
Uploaded: 2026-01-29T20:14:20.2833333Z
Duration: 58 min 58 s
Description: Elena Byhoff, MD, MSc, UMass Chan Medical School October 16, 2025 Yale GIM “Research in Progress” Meeting Presented by Yale School of Medicine’s Department of Internal Medicine, Section of General Internal Medicine

January 29, 2026

Elena Byhoff, MD, MSc, UMass Chan Medical School

October 16, 2025

Yale GIM “Research in Progress” Meeting Presented by Yale School of Medicine’s Department of Internal Medicine, Section of General Internal Medicine

About the speakers

Patrick G. O'Connor, MD, MPH, MACP
Dan Adams and Amanda Adams Professor of General Medicine; Chief, Section of General Internal Medicine; Head of Advisory House, Yan House, Office of Student Affairs

Information

ID: 13791
To Cite: DCA Citation Guide

Download Transcript

00:04Hi. Good morning, every well,
00:06good afternoon, everyone.
00:09Welcome to our noon research
00:11in progress.
00:13Excited to get started,
00:16as people are making their
00:17way in.
00:18First,
00:19some intro slides. This is
00:21our CME code du jour.
00:24Our research and scholarship retreat
00:26is coming up soon. Please
00:27keep your eyes open for
00:29an email from us, that'll
00:30have not only signed just
00:32RSVPing for the retreat, but,
00:35we are really looking for
00:36volunteers for people who wanna
00:37share abstracts,
00:39grant ideas, specific games, and
00:41so we can workshop them.
00:42So soon that we'll be
00:44distributing that. So then on
00:45the
00:46December ninth day,
00:49we'll have plenty to talk
00:50about. We have some good,
00:52good special guests coming as
00:53well.
00:57Speaking of, special guests,
00:59actually from the inside,
01:01doctor Seth, Schaff will be,
01:03speaking about cervical cancer screening,
01:05next week in the general
01:07medicine grand rounds, Thursday morning.
01:09And Sarah Biggerstaff will be
01:11speaking about conflict competence,
01:15as part of
01:16the combination
01:18medical ed grand rounds speaker
01:20series next Thursday at noon.
01:26Disclosure.
01:28More disclosure.
01:29Okay. So without further ado,
01:31really,
01:32it's a great treat to
01:34welcome,
01:35doctor Elena Byhoff here in
01:37person,
01:39to join us,
01:40today.
01:42Elena,
01:43went to UC Berkeley for
01:44undergrad and then Penn for
01:46her medical school and and,
01:48residency in primary care. Then
01:50she was a clinical scholar
01:52out of the University of
01:53Michigan.
01:54She did a lot of
01:55bouncing around different
01:56parts of the country, is
01:57now at at UMass,
02:00on faculty,
02:01and
02:03has really
02:04been able to navigate the
02:06complex funding environment when it
02:08comes to
02:10working toward understanding and addressing
02:12health related social risks, and
02:13how can we better integrate
02:15that into,
02:16our primary care? How can
02:18we integrate that into the
02:19way our health systems are
02:20functioning?
02:21So,
02:23her specific focus is on,
02:25food insecurity,
02:26but I'm really excited today
02:27because she's also going to
02:28be broadening that out a
02:29bit to talk about how
02:30how should we be thinking
02:31about health related social risks.
02:33And in the current funding
02:35environment, to your credit, you're
02:36keeping the train moving, as
02:38far as,
02:39working to identify solutions. So
02:42it's something that,
02:43I think we can all
02:44learn from and looking forward
02:45to learning more.
02:51Alright. Thanks, everybody. I'm I'm
02:53extra bad at computers, so
02:55we'll see if this works.
02:57Yes. Okay. Great. It is
02:58so nice to see folks
02:59in person. I was just
03:01talking to Carrie about how
03:02I've only seen Zoom talks,
03:04and it's just thousands of
03:05little black boxes. So this
03:07is great. I heard you're
03:09an interrupting crowd, which means
03:10my talk is infinity too
03:11long. So feel free to
03:13interrupt, and I'll skip over,
03:15what doesn't need to be
03:16said. But,
03:17all right. So I just
03:17want to acknowledge,
03:19the work I'm presenting today
03:20is kind of the tail
03:21end of my k and
03:22launching into the beginning of
03:23my first r o one.
03:24And I have two main
03:25collaborators,
03:26that I work with in,
03:28all of my research. One
03:29is the Greater Lawrence Family
03:30Health Center. I'll talk a
03:31lot about them. Wendy, I
03:32know from her time there,
03:34and the Greater Boston Food
03:35Bank, who does a lot
03:37of food insecurity work in
03:38Massachusetts. So
03:40the outline for today, we're
03:41gonna talk about me, how
03:42I got into this space,
03:44because I do a lot
03:45of non hospital based health
03:46care.
03:47A lot of it is
03:48social care with community partners.
03:50And then we're gonna this
03:51is a backstory, the health
03:53care design solutions to food
03:54insecurity. That is not where
03:55my research lives, but I
03:57feel like it's an important
03:58topic to cover such that
03:59I can set up what,
04:01my research has been so
04:02far, which are community designed
04:03solutions to food insecurity. And
04:05I'm going to dive into
04:06two specific test cases
04:08around policy, payment,
04:10and testing of food insecurity
04:12programs, and then lessons learned
04:14and next steps.
04:15So about me. So I
04:17did not go to medical
04:18school right out of residence
04:20or right out of college.
04:21I actually worked at the
04:23New York City District Attorney's
04:24Office
04:25for two and a half
04:26years. This is a picture
04:27of the folks from Law
04:28and Order. And the reason
04:30it's important to include keeping
04:31in mind this is not
04:32the first season. This is
04:33the Angie Harmon season.
04:35But I they were actually
04:37filming in my office every
04:38single day. I saw them
04:39all the time.
04:40And the New York City
04:42district attorney is more than
04:43just a prosecutor. A lot
04:44of it is a prosecutor,
04:45but he, is an elected
04:47I don't think it's been
04:48a woman. It's been a
04:49man.
04:49He, to this day, is
04:51an elected official. So he
04:52has a political arm, and
04:53I worked in the political
04:55side of the DA's office
04:56portfolio.
04:57New York City is the
04:59home this is this is
04:59a great place to give
05:00this intro because I'm gonna
05:01tell you things you don't
05:02need to know. New York
05:03City is the home of
05:03the Rockefeller drug laws, which
05:05is a three strikes, you're
05:06out policy. So if you
05:08commit
05:09three drug offenses, regardless of
05:11what they are, right, carrying
05:13a dime bag of marijuana,
05:15if you get caught the
05:16third time, you're in jail
05:17for the rest of your
05:17life. And it turns out
05:19we were putting a lot
05:20of people in jail who
05:21were not violent criminals, who
05:23maybe were not living their
05:24best lives being thrown away
05:26after their third, nonviolent criminal
05:29or nonviolent drug offense. And
05:30so we'd set up a
05:31drug treatment alternative to prison
05:33program with drug treatment courts
05:35that kept nonviolent
05:36repeat drug offenders, essentially, outside
05:39of the prison system got
05:41them into drug treatment. But
05:43what we were seeing is
05:44folks would complete these drug
05:46treatment programs
05:47and then go back to
05:48use drugs. And I felt
05:50like as much as I
05:51enjoyed seeing the cast of
05:53Law and Order all the
05:53time,
05:55that a big root cause
05:56of what was happening was
05:57really outside of the criminal
05:59justice system. It was a
06:00lot of behavioral health issues,
06:01a lot of mental health
06:02issues, sometimes a lot of
06:03comorbid chronic disease. And I
06:05felt like where I needed
06:07to be was not in
06:08the criminal justice system, but
06:09instead in health care. And
06:11so I moved to West
06:12Philadelphia.
06:13So this is
06:15Will Smith from the Fresh
06:15Prince of Vallejoire. I went
06:16to, medical school and,
06:19at Penn, which is located
06:20in West Philadelphia.
06:22I
06:23was a primary care
06:25tracked person in residency there
06:27as well, meaning that my
06:29clinic was in an FQHC
06:30in West Philadelphia. It was
06:31health center number four.
06:34But even more so than
06:35my clinical experience was my
06:36hospital experience as a resident
06:38where patients were just this
06:39is also I should caveat,
06:40this is before the Affordable
06:41Care Act. This is in
06:43Pennsylvania where we had tons
06:45of uninsured folks. And people
06:47would come into the hospital,
06:48get their COPD tune up,
06:50get their CHF tune up,
06:51be discharged right back home,
06:52or they couldn't afford their
06:53medicines, they couldn't afford their
06:55food, they couldn't afford their
06:56housing. And we were discharging
06:58people back into negative circumstances
06:59that led them right back
06:59to the hospital. And I
06:59felt that my experience at
07:01the
07:04DA's office and my experience
07:05in the hospital in West
07:06Philadelphia had a lot in
07:08common.
07:09And it was that structural
07:10forces outside
07:12of those,
07:13kind of, spheres of influence
07:14were
07:15impacting people's lives more so
07:17than I could in as
07:19an individual provider or an
07:20individual kind of essentially grant
07:22writer at the DA's office.
07:23So I did an RWJ
07:24Clinical Scholars Program. This is
07:25back when it was RWJ
07:26funded. I was the penultimate
07:28class, which is my favorite
07:29word, second to last,
07:31at Michigan, where I did
07:32a lot of policy work
07:33similar to, I'm sure, a
07:34lot of the stuff you
07:34guys are doing here at
07:35NCSP and through Yale. And
07:37then I moved to Boston.
07:38And so for those of
07:39you with small children, this
07:40is Make Way for Ducklings.
07:42Right? Two ducks moves to
07:43Boston to raise their family.
07:45So that's essentially what happened
07:46to me. So I originally
07:47landed at Tufts, which was
07:49an interesting place to be,
07:50right, as the Massachusetts
07:52Medicaid system
07:53was launching an Affordable Care
07:55Act demonstration pilot,
07:57that was meant to address
07:58social determinants of health, which
08:00was something I was very
08:00interested in.
08:02But now I'm at UMass,
08:04which has a direct ties
08:05to MassHealth Medicaid,
08:08where the sausage gets made
08:09in policy.
08:11So the entirety
08:12of my
08:13experience in the work world
08:15is really kind of highlighted
08:17by these two kind of
08:19comparable slides where we're sending
08:20people out into circumstances that
08:22don't allow for health
08:25and then they just come
08:26right back. And so so
08:27much of about
08:29where I
08:30do my work, do my
08:31thinking, want to partner with
08:32community partners is because so
08:34much matters outside of the
08:36four walls of the hospital,
08:38just like so much matters
08:39outside the four walls of
08:40the jail or the prison,
08:41that if we don't focus
08:43on what's in essentially the
08:45soil, the environment, the ether,
08:47like, we're missing the forest
08:48through the trees, and we're
08:49just not gonna fix anything.
08:50It's all Band Aids.
08:52Alright. So this,
08:54I've been putting in my
08:55slides since time immemorial.
08:57This is something that should
08:58be familiar to all of
08:59you. It's what goes into
09:00your health. Twenty percent, right
09:02down here, this is accepted
09:03wisdom, is health care and
09:06the rest of it is,
09:06you know, social determinants of
09:08health. During COVID, because I
09:09had time on my hands,
09:11I tried to figure out
09:12where this number came from.
09:14So it's been published in
09:15the New England Journal of
09:16Medicine. It's been published in
09:17Health Affairs since two thousand
09:18two.
09:20It is based on an
09:21expert panel who who felt
09:23that this was right.
09:26So this is just vibes.
09:27We don't actually know how
09:29much of health care goes
09:30into health and how much
09:31of social determinants are going
09:33into health, but suffice it
09:35to say, it's a lot.
09:37So we really I try
09:38to move away from those
09:39quantitative
09:40models into a more qualitative
09:42model. This is from a
09:44health affairs blog that came
09:45out in, I think it
09:46was twenty nineteen, it might
09:47have been twenty eighteen. I
09:48feel like everything pre COVID
09:49is just pre COVID now.
09:51But it's, you know, thinking
09:52about the structures of social
09:53determinants of health being upstream.
09:54So that's
09:56policies, opportunities,
09:58laws, regulations that either create
10:00opportunities for advantage or don't
10:03create opportunities for advantage. And
10:04I think the most important
10:05thing that always needs to
10:06be explicitly stated
10:08is there are positive social
10:10determinants of health. Like my
10:11daughter has all of the
10:12positive ones. We live in
10:13a nice neighborhood outside of
10:15Boston, good public schools, good
10:17libraries, safe streets, green spaces,
10:20equidistant between a Whole Foods
10:21and a Trader Joe's, right?
10:23She has all the advantages
10:24of a healthy life.
10:26You know, but policies that
10:28create an environment where you
10:29don't have good choices,
10:31right? Where can you afford
10:32to live? Well, redlining and
10:34mortgage policy influences where you
10:36can afford to live. What
10:38is the quality of the
10:39air you breathe? Well, environmental
10:40policy, transportation
10:42policy, all of those things
10:44impact you and your community.
10:46And then downstream is social
10:47needs, so that's how things
10:49kind of get manifested at
10:50the individual level. I'm food
10:52insecure because I don't make
10:53a living wage, so I
10:54can't afford food. I'm food
10:55insecure because I live in
10:57a food desert and can't
10:58really find any place to
11:00buy healthy food for myself.
11:01I'm housing insecure because there
11:03isn't enough affordable housing being
11:05built. Right? It's personal experiences
11:06of larger policies and structures.
11:08And then all that stuff
11:09shows up in the hospital
11:11for us as health care.
11:13Right? Medical care. This used
11:14to say health disparities, but
11:15I'm afraid of who might
11:17be watching this now. So
11:18it's medical care. So when
11:20someone comes to my clinic
11:21and has asthma because they
11:22live in a highly polluted
11:24part of town that's right
11:25by a freeway on ramp
11:27that's backed up with congested
11:28traffic. I treat the asthma,
11:30I don't treat the traffic.
11:31Right? So they're going back
11:33home, they're getting exposed to
11:34all the pollution, and they're
11:35coming right back. And so
11:37the shorthand, right, we have
11:39government systems,
11:40we have people who live
11:41in the communities
11:42designed by those government systems,
11:44and then all of those
11:46unequal choices trickle down to
11:48health or not health.
11:51Alright. So health care design,
11:52solution to testing, food insecurity.
11:54So this is an actual
11:55patient
11:56that I actually took care
11:58of that couldn't have done
11:59a better job setting me
12:00up for a career. All
12:01right. So this is Mrs.
12:02Y. She was a patient
12:03of mine that I knew
12:04from residency
12:06into attending this.
12:08Sixty two years old, came
12:09in to see me for
12:10follow-up.
12:11She has a history of
12:12hypertension, type two diabetes, obesity.
12:14I had seen her the
12:15month before for dizziness, and
12:17I told her, I'm like,
12:17oh, we'll stop your blood
12:18pressure medicine. You're probably over
12:20medicated.
12:21If it's not better, come
12:22back next month and we'll
12:23figure it out. And I,
12:24you know, I see patients
12:25like this all the time,
12:26and more often than not,
12:27they don't come back.
12:29So she came back. Her
12:30blood pressure was fine. Her
12:31heart rate was fine. No
12:32orthostatics.
12:33She was only taking the
12:34hydrochlorothiazide
12:35because she swore up and
12:36down that she stopped her
12:37amlodipine that I told her
12:38to stop last time. And
12:39essentially, she had a fully
12:41negative physical exam. I could
12:42not figure out what was
12:43wrong with her. Nothing clinically
12:45jumped out to me.
12:46I was fairly certain it
12:47wasn't vertigo based on her
12:49description. She had a negative
12:50Dix Hallpike. So I do
12:51what I do as, you
12:52know, the excellent clinician that
12:54I am am, is when
12:55something isn't obvious to me,
12:56I ask her, I said,
12:57Mrs. Y, I don't know
12:59why you're dizzy.
13:00What do you think? Like,
13:01do you have any guesses
13:02as to why you're dizzy?
13:03And she said to me,
13:04she's like, Doctor. Bajoff,
13:06do you think it's because
13:08I'm only ever eating
13:11breakfast? And I said, well,
13:13maybe.
13:14Tell me more about that.
13:15And so the story was
13:17that missus Y, I've known
13:18her, as I said, for
13:19a while,
13:20she worked part time in
13:21a call center, like twenty
13:23hours a week, which qualified
13:25her based on income to
13:26get food stamps, SNAP.
13:28And she was doing fine.
13:29She was living in,
13:31like, section eight housing,
13:33making ends meet with her
13:34food stamps and her part
13:35time job, going to church,
13:36lovely lady.
13:38But her grandson
13:39lost his job and couldn't
13:41afford rent with his roommates
13:42anymore, and she let him
13:44move in with her. And
13:45she loved having him live
13:46with her. He was so
13:47helpful around the house. He
13:48fixed the leaky sink. He
13:49got things down off of
13:50the high shelf she couldn't
13:52reach. He came with her
13:53to church. She loved having
13:54the company. But he was
13:56a twenty four year old,
13:57twenty six year old growing
13:58I guess not growing, boy,
14:00who ate all of her
14:01food. And she felt so
14:03bad telling him that she
14:04he couldn't eat all of
14:05it that she would just
14:06cut back her self. So
14:07she would just have oatmeal
14:08and coffee in the morning
14:10and skip lunch and skip
14:11dinner or just have a
14:12piece of toast. And so
14:13she was barely eating anything.
14:15And I'm like, yeah.
14:17That sounds like a reason
14:19to be dizzy. And so
14:21for those of you just
14:21to show you, this is
14:23what the SNAP benefits are
14:25currently. So I saw this
14:26woman,
14:27many years ago when I
14:28transitioned from residency,
14:30to faculty, but,
14:32it was much lower than
14:33this ten years ago. But
14:35now if she was getting
14:36the maximum amount of SNAP
14:38benefits per month, she would
14:39be getting a little bit
14:40less than fifty dollars a
14:41week for essentially covering all
14:42of her groceries. And so
14:44for one person, sure.
14:46For two people, not really
14:49enough. And so, you know,
14:50this is the problem with
14:51them as we really focus
14:53on how health care is
14:54delivering
14:55food,
14:56is that we kind of
14:58ignore it in general and
15:00treat health problems as health
15:01problems, and I could have
15:04easily ordered a CAT scan,
15:06an echo, a halter, a
15:08stress test. Any of those
15:10studies would have been covered.
15:11I could have spent one
15:13hundred thousand health care dollars
15:15on this woman that would
15:16have solved no problems,
15:18that would have made her
15:19feel not one iota better.
15:21And if anything, she might
15:22have thought I was a
15:23pretty crappy doctor because I
15:24kept ordering these tests that
15:25didn't show anything.
15:27And so when we focus
15:28on the downstream and we're
15:29not focusing on the patient
15:33context, this is where we
15:34end up. Right? And so
15:34we're here,
15:36but the agenda is set
15:38here. Right? And so everything
15:39that we do is influenced
15:41by the way policy is
15:42designed. And so just like
15:44one kind of side is,
15:47there have been a lot,
15:48you know, a lot of
15:49ink has been spilled about
15:50the integration of health care
15:51and social care in two
15:52thousand eighteen and nineteen. This
15:53was published in two thousand
15:54nineteen. The meetings happened at
15:56the National Academy of Medicine
15:57on integrating health care and
15:58social care, and this came
15:59up with,
16:00an overarching framework for how
16:02we need to think about
16:03the integration of health care
16:04and social care. And so
16:05just to go through all
16:06of them oh, I'll stay
16:07by the microphone.
16:09Adjustment
16:10is adjusting your care plan.
16:11Right? Like my patient can't
16:13afford this very expensive medicine.
16:15I will find a cheaper
16:15alternative so they don't pay
16:15for medicine instead of food.
16:15Assistance is getting
16:20patient, you know, assistance
16:22with social needs.
16:23Alignment is, working closely with
16:26community based organizations so that
16:27the health care system and
16:29the community based system are
16:31working together,
16:32and sharing patients and sharing
16:33information.
16:34And advocacy is kind of
16:36real deal, call your senator,
16:37let's get more money for
16:38social supports and social care.
16:40But underpinning all of it
16:42is awareness. And so this
16:43is where I've done a
16:44good amount of publishing here,
16:45which is screening for social
16:47needs in clinical
16:48environments.
16:49But where health care tends
16:51to focus is on both
16:52of these. So we're screening
16:54for social needs. I'm not
16:55going to dive into the
16:56value or low value of
16:56whatever that has brought us.
16:56But what I am going
16:56to dive into is, like,
16:56how health
17:04assistance
17:05needs to be delivered,
17:06when we identify social needs
17:08in our patient population.
17:10And so to set the
17:11stage,
17:12because we agree that screening
17:14for social needs and addressing
17:15those social needs is important,
17:17money has been made available.
17:19And I'm so I'm using
17:20the passive voice. Right? Money
17:22has been made. And so
17:23it's insurers, it's CMS,
17:25it's philanthropy
17:27to design programs to identify
17:29food insecurity. I'm gonna focus
17:31on food insecurity now, although
17:32this expands to all social
17:33needs,
17:35and do something to address
17:37it. But when we give
17:39health care
17:40dollars
17:41to do something like this,
17:42it often comes with a
17:43lot of strings as it
17:43does in this case. And
17:44the strings are often quality
17:46metrics. Right? How many of
17:47your patients are you screening?
17:49Beyond screening, how many of
17:51them are getting referred out?
17:52Beyond referred out, how many
17:54of them are actually having
17:55their needs met? And beyond
17:57having your needs met, how
17:58many of them are actually
18:00showing improvement in their health
18:01outcomes? Is their diabetes getting
18:03better? Is their obesity getting
18:05better?
18:06Are they coming to the
18:07hospital less? And so hospitals
18:09are not,
18:10despite what some may think,
18:12are not,
18:14real engines of tremendous innovation.
18:16They tend to stick with
18:17what they know. And so
18:18rather than going out into
18:20the community, finding patients, or
18:21not finding patients, finding people
18:24who may be food insecure,
18:25who may be at risk
18:26for bad things, they actually
18:28turned inside and looked for
18:30patients. Right? So the first
18:31pass of requirement of SDOH,
18:34social determinants of health screening
18:35from CMS, from JACO,
18:38from all these,
18:39other regulatory bodies are in
18:41the emergency room and are
18:43inpatient hospitalizations.
18:45And so we find patients
18:46who not only have chronic
18:49disease,
18:50but are experiencing
18:51exacerbations
18:52of their chronic disease. Right?
18:53The horse is so far
18:55out of the barn, it's
18:56like down the street and
18:57around the corner.
18:58But those are the folks
18:59we're gonna choose to follow-up
19:01on because we have them.
19:02They're in our hospitals. We
19:03don't have to go out
19:04looking for them, and then
19:05we can give them information,
19:06prescribe them medicine, and prescribe
19:08them some food. And so
19:10we treat patients
19:12with food the same way
19:13we treat patients with medicine.
19:15We give it to you
19:15during your hospitalization,
19:17enroll you in a program,
19:18send you on your way,
19:20and then, you know, watch
19:22all the lower a one
19:23c's
19:24roll in. So this is
19:25what has led us to
19:26the Food Is Medicine movement.
19:29A lot of Food Is
19:30Medicine, so focuses up here.
19:32Right? So medically tailored meals,
19:33medically tailored food packages, which
19:35is medically tailored groceries, nutritious
19:37food referrals, which is actually
19:39produce prescriptions.
19:40And then notice that prevention
19:42down here, population healthy level
19:44population level healthy food programs.
19:45I don't actually quite know
19:46what this is. It's very
19:47vague in how they describe
19:48it. I think it's like
19:50school meals. And then here,
19:53actually kind of cash transfers
19:55and benefits and emergency food
19:56programs. So,
19:58what hospitals have doubled down
19:59on is up here. Right?
20:01Medically tailored. So we find
20:02you. We know you have
20:03a disease. We're gonna give
20:04you food to treat your
20:05disease. We're not giving you
20:07food because you're hungry, mind
20:08you. We're giving you food
20:09to treat your disease. And
20:10let me tell you something,
20:12there's a lot of it.
20:16So that is just a
20:18few of of the medically
20:19tailored meal studies. If you
20:20wanted to produce prescriptions, I
20:22could give you another fifteen
20:23headlines.
20:25But the question I'm sure
20:26you're all asking right now
20:27is does it work?
20:29And I would argue
20:31no
20:32based on data.
20:34And so this
20:35is a meta analysis and
20:37systematic review done by Carlos
20:38Aronce, a fellow clinical scholar
20:40at UCLA. This was published
20:42in JAMA Health Forum in
20:43two thousand twenty one. It
20:44has studies through two thousand
20:46twenty,
20:47looking at hemoglobin A1C, health
20:49related quality of life, BMI,
20:50and depressive symptoms. And as
20:52you can see, they cross
20:54zero every single one. The
20:56quality of evidence is mixed,
20:57but there are some RCTs
20:58in here.
20:59The thing I will make
21:00note of, and you'll see
21:02why in a bit, is
21:03that the depressive symptoms were
21:04not a primary outcome in
21:05any of these studies. They
21:07were all a secondary outcome.
21:08The primary outcomes were biometric
21:10endpoints, BMI,
21:12survey related health related quality
21:13of life, and hemoglobin A1C.
21:15If this does not convince
21:16you, you're like, but, Elena,
21:18there has been so much
21:19more additional evidence since then.
21:21You're right. And it shows
21:22the same thing. So this
21:23was a food and medicine
21:25program.
21:26I read this article very
21:27closely. They don't say what
21:28it is. I think it's
21:30Geisinger because it's a, like,
21:31integrated health care system in
21:33the mid Atlantic, and I
21:33think that's the code word
21:34for Geisinger.
21:35But it's a medically tailored
21:37meal intervention for patients with
21:38diabetes, and as you can
21:39see, no difference from control.
21:41This was two years after
21:43Carlos published
21:44his,
21:45the systematic review.
21:47And so, you know, that
21:48brings us to like how
21:49is how are how did
21:50we get here? Who are
21:52making these decisions? Why are
21:53we treating food as medicine?
21:56And it's because of who's
21:57making the policy. Right? So
21:58we have hospitals getting money
22:00and saying, well this is
22:00what we can do. This
22:02is what is feasible with
22:03the data we have, the
22:04patients we serve,
22:05and this is what we're
22:06willing to kind of
22:08design and develop.
22:09And the government's saying, okay,
22:10that sounds great.
22:12But I want to say,
22:13if you actually focus on
22:14the community organization, so that's
22:16what this schematic is, the
22:17community based organization,
22:19and listen to what they
22:20have to say, if we
22:21gave if we hypothetically
22:23gave them money. I could
22:24give a lot of talks
22:25on how we're not giving
22:26them money. I've done a
22:27lot of research on that
22:28too, but that's not this
22:28talk. This talk is what
22:29happens if we do give
22:30them money. So Okay. So
22:31let's talk about two,
22:33test cases when we focus
22:35on the community based organizations
22:37and take the the power
22:38away from the hospitals and
22:39the government.
22:41So first test case is
22:42we're gonna leverage existing community
22:43food programs
22:44to show value beyond medicine.
22:46Right? So rather than
22:48an RCT
22:49where we're enrolling patients with
22:51uncontrolled diabetes and we're giving
22:52them a diabetically
22:54designed meal and then measuring
22:56their A1C,
22:57we're just gonna give food
22:59to people and see what
23:00happens. And then the second
23:02case is, increasing participation in
23:04an existing food program that
23:05has pretty strong evidence to
23:07support its use. So the
23:08first one, our community partners.
23:09So this is a map
23:10of Massachusetts.
23:12This is Boston. And this
23:14is Lawrence, Massachusetts.
23:16I have worked very closely
23:17with the Greater Lawrence Family
23:18Health Center and the Greater
23:18Boston Food Bank on, implementing
23:21and evaluating food programs in
23:22this community.
23:24I'm going to talk a
23:24little bit more about them
23:25in detail later for the
23:27second test case, but the
23:28Greater Lawrence Family Health Center
23:30is an FQHC
23:31serving
23:32a very large Latino immigrant
23:34community
23:35up in Lawrence.
23:37I always say the Lawrence
23:38claim to fame is back
23:39in the before Trump won,
23:42Elizabeth Warren announced her candidacy
23:45for president
23:46in Lawrence
23:47because it was as a
23:48sanctuary city. It was kind
23:49of representative of all the
23:51wonderful things America has to
23:53offer. And then right across
23:54the border in New Hampshire,
23:56there was like an anti
23:58immigrant protest at the same
23:59time.
24:01And so it
24:02is a wonderfully diverse city,
24:04with a lot of socioeconomic
24:06need.
24:07And so what is the
24:08existing food program,
24:10that we want to see
24:10if brings benefit? It's a
24:11monthly mobile market. So the
24:13third Saturday of every month,
24:14the Greater Boston Food Bank
24:15delivers just a semi tractor
24:17trailer full of fresh fruits
24:19and vegetables and protein and
24:21rice and so forth, And
24:23creates it in a farmer's
24:24market style,
24:26environment where their table is
24:27set up, the produce is
24:28set up, you can come
24:29and everyone gets the same
24:30thing, but you get about
24:31somewhere between twenty and forty
24:33pounds of fresh produce for
24:36everybody.
24:36Right? And And so this
24:37is this is the highlight.
24:38Low barrier to attend. I'm
24:40not screening you for diabetes.
24:42I'm not asking if you're
24:43depressed. I'm not asking if
24:45you have been hospitalized with
24:46heart failure in the past
24:47six weeks. You can come
24:49if you want fruits and
24:50vegetables. And people come. People
24:52like it. We've done a
24:53lot of research. It's acceptable.
24:55And so the research question,
24:57was if we don't restrict
24:58access to the mobile market,
24:59right, we don't design it
25:00to treat a disease, but
25:02let anyone come, can we
25:03see health improvements beyond cardiometabolic
25:05bio markers? And so because
25:07we do research and because
25:08we have to define
25:10what outcomes we want to
25:11see improvements on, so we
25:13wanted to know if use
25:13of a monthly food security
25:15program, the mobile market, is
25:16associated with changes in depressive
25:18symptoms.
25:19All right. And so the
25:21mobile market
25:22requires
25:23registration. You can register at
25:25any point. You can register
25:26same day. You can register
25:27beforehand in the clinic. And
25:28that generates a mobile market
25:30ID number. Because it's so
25:32closely linked with the Greater
25:33Lawrence Family Health Center, the
25:34mobile market ID number actually
25:36ends up in your EMR
25:37as well.
25:38And so we could link
25:39all of the data they
25:40collect at time of registration
25:41with your mobile market registration
25:44to the electronic medical record.
25:46At time of registration, we
25:47screen you for things. We
25:48screen you for food insecurity.
25:49It turns out it's a
25:50bad idea to screen people
25:51for food insecurity right before
25:53they're about to get free
25:54food. They all say they're
25:55food insecure because they're worried
25:56if they don't, they won't
25:57get food.
25:58We ask you if you're
25:59enrolled in SNAP and WIC.
26:00We ask you how many
26:01children are in your household.
26:04And then you have this
26:05mobile market ID that you
26:06use every time you attend
26:08so we can track your
26:08attendance over time. And as
26:10I mentioned, link you to
26:11the American,
26:12electronic medical record. So what
26:14did we do? So we
26:15linked
26:16EMR data from the Greater
26:17Lawrence Family Health Center. So
26:19PHQ nine scores, all kind
26:21of socio demographics,
26:23all your comorbidities,
26:24and your type of insurance
26:26with mobile market data. So
26:27yes. Go ahead. Oh, so
26:29I knew it was the
26:29Yeah. Go ahead.
26:39Yeah. Yeah. Well, no, that's
26:41true. So we didn't so
26:42what we wanted to do
26:43is, like, collect some real
26:44world data. So we didn't
26:46wanna start
26:47adding survey data collection to
26:49the mobile market as a
26:50barrier of attendance or anything
26:53to kind of dissuade people
26:54to come who felt like
26:54they needed it. And we
26:56were largely informed
26:58by the Oregon
27:00that the first thing to
27:01move when people newly got
27:03health insurance,
27:04their utilization was higher, their
27:06costs were higher, their health
27:07outcomes didn't move so much,
27:08but their mental health outcomes
27:10did move.
27:11And so the thinking was
27:12that,
27:13depression is something that
27:15as you turn eighteen, everyone
27:16should be universally screened, so
27:17there would be a lot
27:18of PHQ-9s in the electronic
27:20medical record. And that it
27:21was something that would be
27:22tracked over time, and theoretically
27:24we could see an association
27:26between depressive symptoms and this
27:27like kind of oh, like,
27:29community market that wasn't designed
27:31specifically for anything else,
27:33except to improve hunger. But,
27:34yeah, that's a great question.
27:36It was it was kind
27:37of strategic to pick PHQ-nine,
27:40because the other thing is
27:41is you don't necessarily have
27:42to be diagnosed with depression
27:44to have PHQ-9s,
27:46at least at baseline,
27:47but for A1C and
27:49BMI, like, a lot of
27:50those things you're not gonna
27:52get as frequently if you
27:53don't have the underlying diagnosis.
27:54Also, it's a highly prevalent
27:56comorbidity,
27:57and so we thought for
27:58all of those reasons, that's
27:59why we picked it. But
28:00it's a valid question.
28:04Okay. And then for mobile
28:05market data, we linked, total
28:06number of visits during the
28:08study period.
28:09So we had four years
28:10of data of EMR data,
28:11but, the mobile market actually
28:13didn't exist in its usual
28:14state from March twenty twenty
28:17through December.
28:18So we used,
28:19those last, I think, nine
28:21months of just EMR data
28:23without mobile market attendance data
28:24because they weren't collecting anything.
28:25They were just giving away
28:26free bags for everybody.
28:28SNAP enrollment, children in the
28:29house, and number of household
28:30numbers.
28:32And so we built two
28:34models. So first we wanted
28:35to look at change in
28:36PHQ-nine
28:37score associated with mobile market
28:39participation over time, and more
28:41specifically because there are a
28:42lot of folks who get
28:43a PHQ-nine, as I mentioned,
28:44that don't have depression, it's
28:45just part of routine screening,
28:46We wanted to focus on
28:47those who did actually happen
28:48to have a diagnosis of
28:49depression and see what happened
28:51over time for those as
28:52they attended the mobile market.
28:54And so, the table one
28:55in the paper is actually
28:56like a thousand rows long,
28:58but, both, so this is
28:59the group of everybody who
29:02attended the mobile market and
29:03had
29:04complete EMR data
29:06and had a baseline PHQ
29:08nine. So overall, about three
29:09thousand patients attended the mobile
29:11market in this four year
29:12time span.
29:13Of those, we had, complete
29:15EMR data on twenty seven
29:17hundred.
29:18So,
29:19about eighty nine percent of
29:20all mobile market attendees were
29:22patients in the health center
29:23that we could identify.
29:26But of those twenty seven
29:27hundred, only
29:29fifteen sixty six had a
29:30baseline PHQ nine, so a
29:32PHQ nine score before they
29:34attended their first mobile market.
29:36And then of those, four
29:37hundred and thirty two had
29:38a diagnosis of moderate to
29:40severe depression.
29:42The overall, the groups were
29:43very similar except for the
29:45baseline PHQ
29:46nine.
29:47And you'll see, like, there's
29:48quite a range of how
29:49many, mobile market visits the
29:52folks had. In general,
29:54there's a lot of people
29:55who register and don't attend
29:56or attend only once, but
29:58the tail is quite long
29:59long and some people were
30:00attending every single mobile market.
30:02Alright. So,
30:04what did we find? So
30:05for each
30:07attendance of the mobile market
30:09for all participants,
30:11there was a zero point
30:13zero five decrease in their
30:15PHQ nine score, which is,
30:16like, clinically insignificant.
30:18There was a larger decrease
30:19among those with moderate to
30:20severe depression.
30:22But interestingly,
30:23if you think about the
30:24patients who are going five
30:26times, that's a one point
30:28decrease. If you're going five
30:29times over four years,
30:31now we're talking about a
30:33four point decrease, and that's
30:34enough to get you from
30:35moderate depression to mild depression
30:37or severe depression to moderate
30:39depression. And so there is
30:40some statistically significant
30:42small
30:43effect of a program that
30:45was not designed to do
30:46anything for depression at all.
30:48It was just designed to
30:49give people food.
30:51All right. So what does
30:52that mean? So when you
30:53design a program for everybody,
30:55you're gonna see unintended benefits.
30:57Right? Stuff you didn't think
30:58to measure in the first
30:59place. Yes.
31:02I mean, it's interesting.
31:03Did you was there quality
31:04work that went along with
31:05this? And just to to
31:06ask, like, was
31:08it related to, like,
31:10improving their mood symptoms or
31:12their food is improving their
31:13treatment engagement, which improves their
31:15mood symptoms? Or were you
31:15able to, like, tease out?
31:17Yeah. So there's so we
31:18did qualitative work, but there's
31:19there's a good amount of
31:20qualitative work in this space.
31:21And a lot of it
31:22has to do with self
31:22efficacy. Like, if I can
31:24provide food for my family
31:25and, like, the vast majority
31:26I don't know if I
31:26I can go back.
31:29Oh, yeah.
31:30It's a lot of women,
31:32right, who are the people
31:33who do the grocery shopping.
31:34So there's a lot of
31:35if I can provide food
31:37for my children, healthy food
31:38too, because it's fruits and
31:39vegetables and produce, then I
31:41feel like I'm doing something
31:43good. I feel less stressed.
31:44I feel less kind of
31:45burdened,
31:46and overall better. So it's
31:48a it's like self efficacy
31:49is the piece on the
31:50pathway between food insecurity and
31:52mental health. At least that's
31:52the hypothesis.
31:54Good question.
31:55All right.
31:57Whereas, you know, when the
31:58health care system is designing
32:00their food programs,
32:02there's a high barrier of
32:03entry, and we're not seeing
32:05a lot of benefit. And
32:06it turns out if there's
32:07a low barrier of entry,
32:08we can maybe see some,
32:10better outcomes despite the fact
32:11that none of the this
32:12food program was not in
32:13any way designed to show
32:15any sort of health outcome.
32:16It was literally just designed
32:17to get food. So let's
32:19get on to the,
32:21second test case. Oh, I
32:23saw the slide. Okay. Oh,
32:25yes, go ahead. I'm not
32:28gonna leave Lawrence.
32:30Yeah. Very hard hit by
32:31the opioid epidemic. Like, a
32:33lot of overdoses.
32:35Mhmm.
32:36Was there any
32:38any hints that it might
32:39impact, like, sort of substance
32:40use related outcomes? We didn't
32:42look. We didn't even look.
32:43Yeah.
32:45You're right. I think in
32:46a lot of economically depressed
32:47areas, you see high rates
32:48of opioid use. There is
32:50a lot of work being
32:50done on the ground for
32:51opioid use.
32:52I don't intersect with that
32:53world very often,
32:55so I don't know.
32:57But it's a good question,
32:58and possibly worth exploring.
33:00Okay. So test case number
33:01two, how can we increase
33:03participation in an existing
33:05food program? Okay. And so
33:07what is the existing food
33:08program we want to increase
33:09participation in? And it's it's
33:10SNAP. It's a supplemental nutrition
33:11assistance program. It's a federal
33:13nutrition program that's also known
33:14as food stamps.
33:16In twenty twenty four, forty
33:18two million
33:19Americans were enrolled in SNAP.
33:21And of that, one third,
33:23so I think it's like
33:24sixteen million,
33:25were children. So one out
33:27of every three children, or
33:28one out of every three
33:29people receiving SNAP is receiving,
33:31is a child.
33:32There's a national income eligibility
33:34limit. So unlike Medicaid, where
33:35every state has kind of
33:37different rules as to who
33:38qualifies for it,
33:40everyone across the country qualifies
33:42for SNAP if they're at
33:43one hundred and thirty percent
33:44of FPL.
33:45And just there is no
33:46direct causal link showing that
33:48food stamps
33:50buys you health, but there's
33:52a lot of evidence to
33:53show that food stamps improve
33:54food security,
33:56which then improves health.
33:58And so there's a big
34:00push, especially in Massachusetts,
34:02to get everyone who is
34:04eligible on SNAP. There's a,
34:05you know, as as food
34:07insecurity was on the cover
34:08of the New York Times
34:09during COVID,
34:11it became widely appreciated that
34:12people cannot afford food,
34:14and we should do everything
34:16we can to reduce food
34:18insecurity. And that includes making
34:20use of excellent federal programs,
34:23that not everyone is making
34:24use of. So
34:25there was a big push
34:26in Massachusetts to address the
34:28snap gap. So what is
34:29the snap gap? I will
34:30tell you that the snap
34:31gap is not a universal
34:32definition because when I Googled
34:33it, the
34:35AI summary of what the
34:36snap gap is
34:38is not what the snap
34:39gap is.
34:41And so I think this
34:42is a Massachusetts,
34:43like, not only in Massachusetts,
34:45but it's pretty, specifically defined
34:47as the folks who are
34:49eligible for SNAP in the
34:50state of Massachusetts but are
34:51not enrolled. And so how
34:53they determine the denominator is
34:54with Medicaid. So because Medicaid
34:56is up to one hundred
34:57and thirty eight percent of
34:58the federal poverty level and
34:59SNAP is one hundred and
35:00thirty percent of the federal
35:01poverty level, If you are
35:03on Medicaid, you should probably
35:04also be on SNAP.
35:06But in Massachusetts,
35:07forty four percent
35:09of Medicaid enrollees are not
35:11enrolled in SNAP. And so
35:12there was a big policy
35:14push to get folks on
35:16SNAP who are on Medicaid.
35:18So let's take a little
35:19detour into Connecticut data. So,
35:21there is no great Connecticut
35:23data on the SNAP gap,
35:24but we are here in
35:25Connecticut and I, navigated your
35:28Connecticut Department of Social Services.
35:30That website was not easy
35:31to navigate. I'm just gonna
35:32throw that out there. They
35:33do not make it easy.
35:34I had to download the
35:35data and put it in
35:36my own Excel spreadsheet to
35:38make this graph, just so
35:39you know.
35:41But here we have SNAP
35:42enrollment over time since two
35:43thousand twelve.
35:45Importantly,
35:46I didn't know this until
35:48I dove into it, but
35:49Connecticut instituted work requirements in
35:51January of twenty sixteen in
35:53eighty seven towns across the
35:55state. And you can see
35:56what happened.
35:57Interestingly, in twenty seventeen they
36:00repealed work requirements in half
36:02of those towns.
36:03So as far as I
36:04can tell, there's
36:05forty six towns in Massachusetts
36:08or Connecticut that have work
36:09requirements still, but you can
36:11see that SNAP enrollment never
36:12really fully rebounded. So, you
36:14know, you can see what's
36:15coming down the pike as
36:16work requirements get rolled out
36:17in the big beautiful bill.
36:20Connecticut Medicaid,
36:22This is, Medicaid enrollment over
36:24time. This is Affordable Care
36:25Act bump. This is the
36:26COVID bump. So what was
36:28happening during COVID is no
36:29one got kicked off. Right?
36:30So if you were enrolled
36:31in Medicaid, you stayed in
36:32Medicaid, and as people lost
36:33their jobs and enrolled in
36:34Medicaid, this is a huge
36:36jump up, and it's only
36:37now starting to come back
36:38down.
36:39I assume it'll get worse
36:40before it gets better.
36:41But this this was the
36:42chart that I could find,
36:44from your Department of Social
36:45Services. And so this is,
36:47red is is all medical,
36:49but largely Medicaid.
36:51And then green is food,
36:53which includes, like, WIC and
36:55and HIP, so non SNAP
36:56and other similar programs. But
36:58as you can see, it's
36:59becoming a smaller slice instead
37:00of a bigger slice of
37:02the pie. And so does
37:03Connecticut have a SNAP gap?
37:04So based on my calculations,
37:06there are almost five hundred
37:07thousand individuals enrolled in SNAP
37:09and over a million folks
37:10enrolled in Medicaid. So you
37:11have quite a large SNAP
37:12gap. I calculated it to
37:13be a fifty four percent
37:14SNAP gap.
37:16And so what is the
37:17policy approach to address this
37:18snap gap? So in twenty
37:19twenty two, Massachusetts
37:21legislature
37:22passed the act providing for
37:24the development and implementation of
37:25a secure common application portal
37:28for individuals to simultaneously
37:30apply for state administered needs
37:31based benefits and services.
37:34In other words, there was
37:35a single common application. And
37:36this did not occur to
37:37legislators until twenty twenty two
37:40that doing it once instead
37:41of twice would make sense.
37:43Code for America has this
37:44really amazing website where you
37:46can see how hard or
37:47easy it is to apply
37:48for benefits by each state
37:50in all fifty states. And
37:51Massachusetts, despite being incredibly blue,
37:54is one of the worst
37:55ranked states in ease of
37:57applying
37:58for benefits.
37:59And so, they have this
38:00common application. It was passed
38:02into law in twenty twenty
38:03two. At the end of
38:04twenty twenty three, they saw
38:06a zero point nine percent
38:08increase in SNAP enrollment. So
38:10it did not work
38:12at all.
38:14Does Connecticut have something similar?
38:16So I tried to figure
38:17it out, and what you
38:18will see, so this is
38:19the application page for the
38:20Department of Social Services.
38:22As you can see here,
38:22I don't know if you
38:23can see really well, you
38:24can click this box. So
38:25I tried to click through.
38:26I got stuck. I don't
38:27know if it's a common
38:28application because they wanted, like,
38:29my Social Security number, and
38:30I'm like, no. I'm going
38:31to stop here. But it
38:33appears,
38:33based on food assistance, SNAP,
38:35and cash assistance, which is
38:36TANF, Temporary Assistance to New
38:38Families,
38:39and medical, which is Medicaid.
38:40And I'm assuming Huskies are
38:43Connecticut. Yeah.
38:46So I'm guessing if I
38:48had a Connecticut address and
38:49Social Security number and continue
38:50down, Connecticut would also have
38:52a common application, despite the
38:54fact that you have a
38:54fifty four percent SNAP gap.
38:56So okay. So the research
38:57question after watching the state
38:59fail spectacularly,
39:01was can we design an
39:02implementation strategy to increase SNAP
39:04participation? And so the diverse
39:06Latina community is Greater Lawrence
39:08Family Health Center again. They
39:09are my,
39:12favorite partners. And so more
39:13details on who's in Lawrence.
39:15So, in two thousand twenty
39:16four, sixty one thousand patients
39:18were served in the health
39:19center. They have quite cap
39:21good capture of the entire
39:22Merrimack Valley.
39:24Data shows that approximately eighty
39:26to eighty five percent of
39:27patients or people who live
39:29in that area get their
39:30primary care at the health
39:31center.
39:32It's a majority minority population,
39:34mostly Hispanic Latino, mostly Spanish
39:37speaking. Fifty percent have Medicaid.
39:39A large chunk is uninsured
39:41because it is a large
39:43immigrant community, both documented and
39:45undocumented,
39:46and very much living below
39:48poverty.
39:49So this is where I'm
39:50going to everyone put on
39:50your implementation science hats, because
39:52we're going to use some
39:53implementation science jargon. And we
39:55did an implementation science study
39:57to understand how best to
39:58enroll patients in SNAP
40:00from the perspective of the
40:02health center. So this is
40:03the PRISM framework, the practical
40:04robust
40:05implementation science model. The reason
40:07I picked this one, not
40:08that it matters, is because
40:09it really focuses on context.
40:10So what's happening outside
40:12that influences what's going on
40:14inside.
40:15And so we had pre
40:16existing data.
40:18This known patient level barrier
40:19so I work, as I
40:20said, with the Greater Boston
40:21Food Bank. They do a
40:22statewide survey every year of
40:24food assistance utilization, food insecurity,
40:28and, you know, use that
40:30feedback to better better design,
40:32Greater Boston Greater Boston Food
40:33Bank programs. So we knew
40:34from the survey patient level
40:36barriers across the state of
40:37Massachusetts were eligibility concerns,
40:40application
40:41difficulties,
40:42and then benefits. Like, is
40:44the juice worth the squeeze?
40:45If I'm gonna spend three
40:46days trying to get my
40:47SNAP application in, am I
40:49gonna get enough money
40:50to make it all worth
40:51it? And then what documents
40:52do I need? So,
40:54that was largely,
40:56in terms of financial documents,
40:57like a lot of cash
40:59jobs, under the table jobs,
41:01is that going to have
41:02enough documentation
41:03to prove that I don't
41:04make enough money to get
41:05SNAP? And then we knew
41:07that health centers, as I
41:08mentioned, they, the state passed
41:10this law. They just let
41:11it happen.
41:13No one enrolled in SNAP.
41:15And so what the state
41:16did is say, Okay. Well,
41:17if you're a Medicaid provider,
41:19you now have to address
41:20the SNAP gap. And you,
41:21as Medicaid providers,
41:23have to increase SNAP enrollment
41:25for all your Medicaid enrollees.
41:26So the organizational perspective from
41:28the Greater Lawrence Health Center
41:30point of view is that,
41:31well, this is something we
41:32have to do, and so
41:33everyone's on board to do
41:34it. Okay.
41:36The external environment, as I
41:38mentioned, immigration status and immigration
41:40policies were top of mind
41:41as we're asking people if
41:43they're eligible for SNAP that
41:45includes
41:46questions about immigration and documentation
41:47status.
41:48And after Trump one point
41:50zero,
41:51there was a lot of
41:52concern about public charge, which
41:54is
41:55you lose the pathway to
41:57get your green card if
41:58you're considered a burden on
42:00the system. Historically, it had
42:02just been for cash assistance.
42:03There were discussions about making
42:05it for SNAP. It never
42:06actually they never changed the
42:08law. So it was never
42:09that if you were enrolled
42:10in SNAP, you could be
42:11considered a public charge, but
42:12just the fact that the
42:13idea was out there was
42:15enough to really scare folks.
42:16And then there was a
42:17huge burden of food insecurity
42:18in the community. And then
42:19organizational
42:20characteristics, you know, who's gonna
42:22do it? How are we
42:23gonna do it? How are
42:24we gonna prove that we
42:25did it? And then who's
42:26gonna help us along the
42:27way? Because in Massachusetts, I
42:29actually don't know if this
42:30is true in all other
42:30states, you can get paid
42:32to be a SNAP enroller.
42:33So if my organization
42:35goes through
42:37three hours of training courses
42:38with the state of Massachusetts,
42:40I will get reimbursed for
42:42every person that submits a
42:44SNAP application.
42:45It doesn't even have to
42:45be a successful application, just
42:47the process of doing it
42:48gets me reimbursed. And so
42:49we were getting the Greater
42:50Lawrence Family Health Center up
42:51and running on this, but
42:52they wanted to be able
42:53to refer out if they
42:54felt overwhelmed.
42:55Okay. And so this is
42:57why community
42:58work is so hard. I
42:59did this for three years
43:00and got one paper out
43:01of it.
43:03But using PRISM to increase
43:05SNAP enrollment first, we started
43:06with weekly stakeholder meetings. We
43:08had food bank SNAP specialists
43:09so these are folks from
43:10the Greater Boston Food Bank
43:11who knew all the ins
43:12and outs of SNAP enrollment.
43:13Like, what the questions were
43:14back and forth, what the
43:15eligibility criteria was, forwards and
43:17backwards, and can help us
43:18as we set up our
43:19enrollment plan.
43:20We had health center, community
43:21health workers, and doctors who
43:23were trying to identify workflow.
43:24We had Medicaid program staff,
43:26that
43:27provided oversight and data on
43:29the SNAP gap as we
43:29went along. And then I
43:31was your smiling academic partner.
43:33And then we reviewed all
43:34the relevant data. So the
43:35food bank survey data and
43:37attendance data to give us
43:38an idea of where people
43:39with food insecurity
43:40were showing up and likely
43:41to be so that we
43:42could enroll them in SNAP.
43:44Health center data in terms
43:45of who among the health
43:47center patients were SNAP enrolled
43:48in food insecure, and then
43:50state SNAP data to identify
43:52pockets of high levels of
43:53the SNAP gap in Lawrence.
43:55And so the overarching goal
43:56was to identify and pilot
43:58feasible strategies to increase SNAP
44:00enrollment. And so we ended
44:01up with two strategies, and
44:02I'll dive into them. One
44:04is the local strategy, similar
44:05using the mobile market, and
44:06one is the policy strategy
44:08leveraging the public health emergency
44:09unwinding. So the local strategy,
44:11as I mentioned before,
44:13there's a monthly mobile market
44:14that's well attended in Massachusetts
44:16or in Lawrence. We collected
44:18data on size and SNAP
44:19enrollment. And then for patients
44:21who
44:22reported that they were not
44:24already enrolled in SNAP, we
44:26referred them to the Greater
44:27Boston Food Bank SNAP enroller
44:29so that they could get
44:29enrolled the very next working
44:31day. So the mobile market
44:32happened on Saturday.
44:34On Monday, they would receive
44:35a phone call saying, we
44:36got your data. You've been
44:38referred to enroll in SNAP.
44:39Let's complete the application now.
44:43We knew in general that
44:45there's a large undocumented
44:47and recently arrived immigrant population
44:49at the mobile market. So,
44:50this was kind of key
44:52for the data we were
44:53collecting because as you know,
44:55or maybe you don't know,
44:56many families have mixed documentation
44:59status where
45:00the parent, the grandparent is
45:02undocumented
45:03or recently arrived,
45:05but the child was born
45:07in the US. And if
45:08the parents and grandparents are
45:09not eligible for SNAP, often
45:10the child is, and we
45:12wanted to make sure that
45:14anyone in the household who
45:15is coming to the mobile
45:16market
45:17was enrolled in SNAP who
45:18is eligible.
45:21The second strategy was the
45:22policy strategy. So at the
45:23end of the COVID public
45:24health emergency, people started getting
45:25kicked off their Medicaid for
45:26the first time in three
45:27years.
45:28Because of that, health care
45:30organizations
45:31had a big push to
45:32reenroll everyone in Medicaid. So
45:35the staff that does reenrollment
45:36are outreach and enrollment staff.
45:38And so we got them
45:39on board to say, listen,
45:41when you're enrolling in Medicaid,
45:43make sure you use the
45:45common application.
45:47All it is now I
45:48cannot stress this enough. All
45:49you have to do is
45:50check a box on the
45:51top of the Medicaid enrollment
45:53form. That is it. You
45:54have to collect no additional
45:55data. You have to do
45:56no additional work. You just
45:58have to check a box
45:59at the top. And we
46:00train the outreach enrollment staff
46:02on this, and then we
46:03had to set up our
46:04own unique data collection
46:06infrastructure for this, because outreach
46:09and enrollment would just submit
46:10Medicaid applications without tracking what
46:12they were doing or for
46:13whom they were doing it.
46:14And in order to see
46:15if we were doing it
46:15well, we wanted to start
46:17tracking it.
46:18And that that was a
46:19really interesting adventure.
46:21And so it was meant
46:22to leverage the common application,
46:23as I mentioned. All right.
46:24So what happened? So the
46:25local strategy,
46:26despite the fact that over
46:28four thousand
46:29patients
46:31have attended the mobile market
46:32in the past five years,
46:34we managed to enroll fifty
46:36eight patients.
46:38Well, not enroll. We referred
46:39fifty eight patients to the
46:40Greater Boston Food Bank. Of
46:42the fifty eight patients who
46:43said that they were not
46:44SNAP enrolled that were likely
46:45to be eligible, Thirty nine
46:47answered the phone two days
46:48later when the Greater Boston
46:49Tribune called.
46:50Of those thirty nine, twenty
46:51eight were ineligible.
46:53The vast majority of ineligible
46:55folks were due to income.
46:56They made slightly too much
46:58money to be eligible for
46:59SNAP.
47:00The,
47:01the other
47:02most second most common reason
47:04was,
47:05documentation status.
47:07Of the twenty of the
47:09thirty nine who answered the
47:10phone, eleven applied for SNAP
47:11and only four enrolled.
47:13So it was not it
47:14was not a great success.
47:16The policy strategy was even
47:18worse.
47:19We collected no data. The
47:20outreach and enrollment staff did
47:23not like that they were
47:24asked to collect additional data.
47:25They did not feel like
47:26that was part of their
47:27job.
47:28They collected no data. So
47:29we have no idea how
47:31many folks that they submitted
47:33applications to the common application
47:34for.
47:35The other feedback that we
47:36got from the outreach enrollment
47:38staff
47:38is because
47:40we had tried to
47:41leverage the electronic medical records
47:44so they didn't have to
47:44manually type in all of
47:46the information about each patient,
47:49for the data collection purposes.
47:51A lot of folks who
47:52come into the health center
47:53to get reenrolled in Medicaid
47:54are not actually health center
47:55patients. Statistically, they will become
47:58health center patients, but at
47:59the time of Medicaid re
48:00enrollment,
48:02they were community members who
48:03were not in the electronic
48:04medical record, they had to
48:05manually enter that data. That
48:07was unpalatable,
48:08no data was collected.
48:10All right, so lessons learned.
48:12They're big ones.
48:14I'm going to keep going.
48:15Okay, great.
48:16So it's hard to do
48:18stuff
48:19in general, but when you
48:20have no money, it's extra
48:23And so safety net settings
48:24where I do all of
48:25my community partnerships
48:27and I think it's an
48:28ideal place to do a
48:29lot of this work because
48:31that's often where the most
48:32need is.
48:34But the health systems that
48:36serve those patients
48:37also are highly stretched, don't
48:40have a lot of money,
48:41don't have a lot of
48:42bandwidth. I always joke that
48:43everyone at the health center
48:44has like eighteen jobs and
48:45they're only paid for one.
48:47And so when we think
48:49about these highly
48:51controlled
48:52food is medicine, I'm screening
48:53you for your diabetes, I'm
48:54screening you for your resources,
48:56I'm making sure you're getting
48:57enough
48:58A1C follow-up so I can
48:59prove the point that it's
49:00doing something for your diabetes,
49:02that's never gonna fly in
49:03the safety net. And so
49:05how are we thinking about
49:05implementing food programs,
49:07in environments
49:09with limited resources?
49:12And then the second
49:13issue that I run into
49:15all the time is limitations
49:16of the electronic medical records.
49:18So much
49:19of all the work we
49:20do, right, especially in implementation
49:22science, clinical reminders, workflow,
49:24data, it's all in the
49:26EMR. All of it. And
49:27if your EMR sucks or
49:29if you're dealing with community
49:30and not patients,
49:31you can't use the EMR.
49:33It is useless, it is
49:34a useless tool. And in
49:36a lot of the work
49:37in terms of enrolling patients
49:39in food as medicine programs,
49:42There's so much alignment between
49:43the social system EMR and
49:45then the health system EMR,
49:47and how can we get
49:48them to talk? It's like,
49:49well, what if no one
49:50has an EMR? Can we
49:51just not do anything?
49:53And I think a lot
49:54of innovation in this space
49:55needs to happen in the
49:56non EMR
49:57implementation side.
49:59And then, this is true
50:00now more than ever, but
50:01in rapidly changing policy environments
50:03require agile systems. So we're
50:05funding food as medicine today,
50:07but we might not be
50:08funding it tomorrow
50:09because
50:10I don't know
50:11policy priorities. Right? We're in
50:13a government
50:14that's delegitimizing
50:15a lot of the health
50:16equity work I do, and
50:18so how do we continue
50:19doing the work when our
50:20funding streams are under attack?
50:23And how can we make
50:23sure that outside of the
50:25grant period,
50:26we can still stand up
50:28and sustain these food programs?
50:29Because you don't stop being
50:31hungry
50:32when the grant period ends.
50:34And I feel like a
50:35lot of the food is
50:36medicine space conversation is, well,
50:37after six months, you're on
50:38your own. And I don't
50:39think that's ethical. And then,
50:41ethical. And then finally,
50:43the narrow outcomes that we
50:45think about
50:46in evaluation of these food
50:47assistance programs are missing real
50:49benefits. And so
50:51very rarely are communities being
50:53asked what their food programs
50:54should look like, instead we're
50:55telling them what food they
50:56should be eating.
50:58And we're not asking communities,
51:00like, who should be targeted?
51:01Like, do they think that
51:02this is like a zero
51:03sum game the same way
51:04we do, where if we
51:05target diabetes, we're not targeting
51:07cardiovascular
51:08disease?
51:09And then my final question
51:10is, does food need to
51:12be medicine, or can it
51:13be more of a human
51:14right?
51:15And so where are we
51:17yeah, please, go.
51:18I
51:19just wanted to ask your
51:20thoughts about the substance law
51:22type question. You know, the
51:24health care,
51:27system is, like, twenty percent
51:28of GDP. Yes. So if
51:30you want money for something,
51:31you you Yep.
51:33Return. Yep. One percent of
51:35that for food, there would
51:36be fat. Yep. So but
51:38then there's the that comes
51:39like you're saying, you know,
51:40with serious limitations about
51:42how we conceptualize
51:44when and how food is
51:45delivered. So
51:46What are your thoughts about
51:47if we're looking for a
51:48solution, should we be looking
51:49within the health care system?
51:51Yeah. No. That's a really
51:53good question. It's actually my
51:54favorite thing to think about.
51:55So
51:58I
51:59I don't think it should
52:00come from the health care
52:01system. I think doing it
52:02from the health care system
52:02is the single most inefficient
52:04expensive way to do anything.
52:06We have so many layers
52:08of administrative
52:09garbage.
52:11And like I when I
52:12give similar, but not the
52:13same lectures to my medical
52:15students and the postdocs,
52:16I talk about the fact
52:17that, like, how many times
52:18have you received an email
52:20asking you to change the
52:21code on a patient you
52:22saw,
52:23that it's like heart failure,
52:25acute exacerbation,
52:27because you can get, like
52:28and so we have chosen
52:29as a society or no.
52:30No. We have chosen as
52:32a health care entity that
52:33we are going to spend
52:34money on those people to
52:36increase our revenue
52:38by upcoding
52:39for, you know, whatever
52:41whatever ICD-ten,
52:43gets us the most money.
52:44But we will not invest
52:45in, like, a social care
52:47workforce because, like, the return
52:48on investment is insufficient.
52:50And so I,
52:52you know, I think when
52:53you look at countries who
52:54don't face the same
52:56health equity issues we do,
52:58it's a robust social safety
53:00net not funded through health
53:01care, but funded through the
53:02government.
53:03And that is, like, the
53:04aspirational goal. But we as
53:06Americans, we're individuals. We don't
53:08want to rely on the
53:09government, we lift ourselves up
53:10from our bootstraps.
53:11And so it seems like
53:12a political infeasibility.
53:14So,
53:16I don't know. All right.
53:17Yeah. I feel like I'm
53:18out of time. Am I
53:19out of time?
53:24Fantastic. I think I have
53:25two more slides.
53:30I have no way to
53:31see. Okay. All right.
53:33Yes. So where am I
53:34going next?
53:35Meaningful outcomes for thinking about
53:37more than just food. Right?
53:38I'm interested in the entire
53:40ecosystem of social care delivery.
53:42This this has been a
53:44problem,
53:45especially in a large in
53:46in the FQHC
53:47immigrant, community I work with.
53:49My primary practice is in
53:50an FQHC.
53:51ICE sits across the street
53:53three out of five days
53:53of the week.
53:56So like we can't do
53:57anything in the health system
53:58because no one's coming anymore.
54:00That's an overstatement. And then
54:01to increase overall participation, right,
54:03if we're just targeting those
54:04with chronic illness, coming back
54:06to the hospital over and
54:07over again, we're missing the
54:08opportunity and window of prevention,
54:10which is so stupid, but
54:11that's what we're doing.
54:13So not that any of
54:14oh, whoops. Wrong way. Not
54:15that any of it matters
54:16much because we're about to
54:17lose, I think, everything,
54:19and that's it.
54:22This is my daughter helping
54:23me at the mobile market.
54:24This is, the COVID where
54:25you get handed your groceries
54:26and so get to pick
54:27it
54:27up. Alright. Questions?
54:30You have three minutes. Yes.
54:35Well, the elephant
54:38is, we sit in the
54:39health care system. Mhmm. Yeah.
54:41And not all of us
54:42can do what you're doing.
54:44Right.
54:45So from where we sit,
54:48what are your recommendations?
54:50I have two two recommendations,
54:51and they both involve,
54:54essentially advocacy,
54:55like the advocacy framework. So,
54:58one of the recommendations is,
54:59you
55:01know, as I have I
55:02give another talk that's
55:05as I have I give
55:07another talk that's exactly entitled,
55:07What Can You Do? Which
55:07is, you know, we well,
55:07second to nurses. Nurses are
55:08the most respected and trusted
55:10profession. Doctors are the second
55:12most respected and trusted profession,
55:14at least for now.
55:15And so
55:16lobbying and policy making, the
55:18reason I think all of
55:20this has ended up under
55:21the rubric of health care
55:23is because the business case
55:24is that it's making health
55:25more expensive.
55:26But I feel like if
55:27doctors were able to kind
55:28of lobby together instead of,
55:30like, the AMA lobbying for
55:34reimbursement policy, we're
55:36lobbying for
55:37increased social care funding,
55:40I think it would carry
55:41more weight. I feel like
55:42a lot of the community
55:43based organizations who are underfunded
55:45and doing all the work
55:47have a very limited voice
55:48in any rooms where policy
55:50is being made.
55:52So that's like big P
55:53policy. And then the little
55:54p policy is at the
55:56health system level. I mean,
55:58so much, I can't tell
55:59you how many times that
56:00health centers and health systems
56:02are like, well, we'll just
56:02set up our own food
56:03bank.
56:04No.
56:05No. There are food banks
56:07out there in the community
56:09that exist, that need our
56:10resources, that need our participation.
56:13You know, we have community
56:14benefit dollars that often go
56:15to feed us back
56:17business. And, like, can we
56:19make those actual community
56:21benefit dollars? I mean,
56:24I hold out hope that
56:25sometimes, yes.
56:28Yeah. Just kind of a
56:29follow-up to Amy's question. So
56:31the two solutions that you
56:32proposed are kind of related
56:33to advocacy and Mhmm. Structural
56:35administrative
56:36Mhmm.
56:37When you're sitting across Mhmm.
56:39From a patient Yes. Is
56:41working to form how how
56:42you take care of them?
56:43Because I think you've a
56:44very strong case that a
56:46lot of what we do
56:47in medical medicine is a
56:48Band Aid for
56:50Mhmm. Yeah. But the Band
56:51Aids are getting better and
56:52better.
56:53You know, Jill
56:54Okay. So when you got
56:56twenty minutes with the patient
56:57in your primary care clinic
56:58Yeah. How does this work
56:59inform how you take care
57:01of them? And is it
57:02about
57:03applying the Band Aids in
57:04the best way that you
57:05can and using the best
57:06Band Aids that there are?
57:07Or are you doing something
57:08different that you wouldn't otherwise
57:10have done based on your
57:11that your clinic?
57:15That's a good question. I
57:16don't know if I have
57:16a great answer for that.
57:17I mean, I I think
57:19I often find when I
57:20ask all of my patients
57:21about SNAP if they're enrolled,
57:23and I think I am
57:23uniquely qualified to counsel them
57:25on enrolling in it that
57:26I think the vast majority
57:27of primary care physicians are
57:29not. I'm also like, so
57:30Massachusetts is the site of,
57:33an eleven fifteen waiver demonstration,
57:36that is providing
57:37nutrition and housing supports for
57:39patients
57:40on Medicaid who meet certain
57:42criteria. I am intimately aware
57:44of all of those criteria.
57:45So, you know, my Band
57:46Aids are maybe better than
57:47the average Band Aids because
57:49I'm aware of, like, all
57:50of these
57:51kind of nuanced detailed programs
57:52that may or may not
57:53be available to individuals.
57:56But I think in the
57:57moment, I think that going
57:58back to the National Academy
57:59of Medicine five a is,
58:00like, actually asking the question,
58:02like, making sure there's room
58:03in the visit, for, like,
58:05actually my blood pressure is
58:06terrible because I'm not taking
58:07the medicines because I can't
58:08afford the co pay. I
58:09mean,
58:11like, it's so funny. When
58:12I was a when I
58:12was a resident in the
58:13hospital and you put all
58:14the patients on their home
58:16meds and their blood pressure
58:17bottoms out, I'm like, what
58:18the heck is happening? No
58:19one bothered to explain it
58:20to me. And now I
58:22know. They're not taking their
58:23medicines at home often because
58:25they can't afford them, or
58:26they're not picking them up
58:27because they can't get to
58:28the pharmacy.
58:29And I feel like just
58:30that piece of information would
58:32be so much
58:33kind of more instructive, like,
58:35in the clinical visit. Like,
58:36how are patients driving their
58:37priorities versus how am I
58:39driving my priorities to take
58:40care of them so that
58:41it can be aligned?
58:43I feel like a lot
58:44of a lot of doctors
58:45threw up their hands and
58:45were like, well, not my
58:46problem. Go see the social
58:47worker.
58:51Yeah. No. Thank you.