Dr. Ana Poblacion and Richard Sheward discuss recent research related to measuring food insecurity, and how ongoing research supports better strategies to reduce food insecurity & its associated health risks.
This episode features development & validation of a scale measuring changing levels of food security in households with young children. The work was funded with a small USDA research grant and presented as part of marking 25 years of household food security measurement. Research presentations from all projects supported as part of this 25th anniversary are found at: https://sites.tufts.edu/foodsec25/
For the full Hunger Vital Sign explainer series, visit our website: https://www.vtfoodinhealth.net/hunger-vital-sign-toolkit
To learn more about the Hunger Vital Sign National Community of Practice, visit the Children's HealthWatch HVS special projects page: https://childrenshealthwatch.org/public-policy/hunger-vital-sign/
LABUN: Welcome to the Policy in Plainer English podcast, I’m your host, Helen Labun.
Our last series focused on explaining one of the most common tools for integrating food access and health care - the Hunger Vital Sign.
Hunger Vital Sign is a set of two questions used to identify whether patients are at risk for food insecurity. It’s built from decades of research by the U.S. Department of Agriculture, or USDA. This research provides a consistent definition for levels of food security. Here’s Rich Sheward from our original series:
The USDA household food security survey module, the HFSS or the USDA 18 item as it's often referred to, was developed in the 1990s. Essentially, it's considered the gold standard in assessing and determining the ranges of food insecurity from Food secure, high food security or marginal food security, to food insecurity, low food security, or, very low food security
LABUN: Hunger Vital Sign is also built from an extensive data set collected by an organization called Children’s HealthWatch, which has been studying how economic factors impact children’s health and development since 1998.
Over the years, children's health watch developed a research base, demonstrating that food insecurity in households with young children resulted in a myriad of negative health outcomes.
LABUN: Using these two sets of data, researchers could figure out the most efficient way to identify patients in a health care setting who were at risk for both food insecurity and related negative health impacts. If you want to hear details about that work and the use of social risk screening in health care, go to the Hunger Vital Sign explainer series linked in the show notes.
This work was a starting point for much of the ongoing research around how we identify food insecurity. This ongoing work is the focus of today’s episode. We’ve brought back our guest from Part 1 of the explainer series:
I'm Rich Scheward and I'm the Director of System Implementation Strategies at Children's Healthwatch. My former role in the organization, until recently, was Director of Innovative Partnerships.
LABUN: And our guest has brought a guest.
Hello, Helen, thank you for inviting me to share our latest research on social determinants of health with your audience. I am Dr. Ana Poblacion, research scientist and director of onsite operations with Children's Health Watch, headquartered at Boston Medical Center, and Assistant Professor of Pediatrics at the Boston University School of Medicine.
LABUN: And these two guests are connected to an international network of researchers focused on understanding, and addressing, food insecurity.
We lead, alongside the Food Research and Action Center (FRAC) a national community of practice focused on food insecurity screening and interventions that utilize the Hunger vital sign screening tool that we validated back in 2010.
LABUN: I know, that’s a national community of practice and I promised you international.
In 2014, the Children's HealthWatch newly validated Hunger Vital Sign was getting traction here in the U.S. and I was marveled with this, with its ample possibilities, and wondered if I could bring such a tool to Brazil.
LABUN: We’ll get to the Brazil part later.
In the Hunger Vital Sign explainer series we talked about how an active research community, continuously building on what has happened before, is an advantage for moving from theory to practice. This episode will give examples of why. Starting with research that was presented in the spring of 2022 as part of a conference marking 25 years of USDA food security measurement. The final research paper will appear in the Journal of the Academy of Nutrition and Dietetics in early 2023. Conference proceedings will be linked in the show notes.
Now, Step number one is identifying the research question. That’s the most obvious advantage of an active research community, they’re identifying gaps in current knowledge, filling those gaps, and sharing the results. And after 25 years of work, there are still plenty of questions remaining.
We asked ourselves, does an abbreviated food security scale for use in households with children capture severity of child and adult food security as effectively as the US household food security survey module?
LABUN: Haven’t we all, at some time in our lives, asked ourselves this question?
Well, the first step in understanding why this question stood out is to review what distinguished the 2-question Hunger Vital Sign risk screen from the underlying Household Food Security Survey from USDA. That long research questionnaire was designed to identify trends over time across the entire country and provides a foundation for much of what followed.
This research question was really a natural progression of the work that we've been doing since 1998 and really, looking at the food insecurity prevalence in the US over time. Thanks to the great work of the U S D A economic research service, as every year they come out with a new food insecurity in the US report, looking at food prevalence over time, we can clearly identify that food insecurity is a persistent major public health issue in the U.S.
LABUN: We know that in 1998, Children’s HealthWatch looked at that research and asked if there was a way to quickly identify who was at risk for food insecurity within the context of a health care response. A health care provider is not going to stand up an economic research project every time a patient walks through the door. To translate the underlying research into action, providers need a way to quickly know if the person sitting in front of them right now is at risk for food insecurity - simply knowing that the person is in a community trending that direction won’t point to an immediate next step for the individual. The Hunger Vital Sign provided this quick test - like other risk screens for any number of potential health conditions.
However, on the way to brevity, it dropped the function of describing levels of food insecurity, the risk screen simply offers a yes / no answer. This simplification left a gap in the health care toolkit -- somewhere between risk screen and economic research, providers needed a scale that could offer them more information about what was happening with at-risk patients.
Confirming the presence of food insecurity as well as determining the level of severity is important because then it allows us to truly shape an intervention that can appropriately address the level of severity along with other contextual information. That's really important to know. For example, if you're a clinician and you have a childless adult in your office with low food security, who is just laid off from their job, they're in the process of applying for a new job, maybe they haven't enrolled in SNAP, that individual will probably benefit from a lighter touch intervention compared to a patient where maybe they're struggling with chronic conditions like hypertension and type two diabetes with very level food security, and they have children in their household also experiencing food insecurity. So that household would likely benefit from a more intensive, let's say a six or 12 month medically tailored meal program.
So the Hunger Vital Sign, it doesn't really shed too much light on whether an intervention A or intervention B is the most appropriate for that patient and their household.
LABUN: An abbreviated scale for describing levels of food insecurity does exist in the form of a 6-question survey module based on the same USDA survey that underlies Hunger Vital Sign. But this scale had not been tested for households with young children. Households with young children often require slightly different measures. In these circumstances, a caregiver is answering on behalf of the entire household. It may be the first household experience with food insecurity, since the addition of a new child has a significant impact on food security risk and on eating patterns in general. And severe food insecurity usually affects the children in a household last. When food insecurity does affect children, the results can be particularly damaging.
Young children experience particular vulnerability to food insecurity as they are in a sensitive period of brain body growth. Also, they are physiologically more vulnerable than older children to adverse impacts asserted by chronic economic hardships, both directly on their development trajectory and indirectly via their parents' health and wellbeing.
LABUN: So, the researchers tested an abbreviated scale that would specifically recognize the factor of food insecurity for young children within a household.
This measure was validated using the 18 item household food security survey module, the gold standard scale food security measure in the United States. And, , this newly abbreviated scale is an eight item scale, in which six questions refer to the household and adult experience, , with footing security. And two questions refer to the experience of footing security among children.
And yes, we concluded that this novel abbreviated scale is highly sensitive, specific and valid for detecting levels of foody security among our sample of racially diverse households with young children. So we recommended this novel abbreviated scale as a standalone scale or a follow up scale after households with children screen positive for food insecurity, for example, using the Hunger vital sign
LABUN: That is an example of deepening the information available to health care providers about food insecurity among their patients. Researchers can also take a starting validated measure, like that original USDA food security survey, and the data sets connected to it, and bring the framework out to applications in new regions. Here is another benefit of the active research community.
And it’s also where Brazil comes in.
Seeing the broad use of the Hunger vital sign here in the United States made me think about taking this screening tool back to my home country, Brazil, where we already have a food security scale derived from the US gold standard Household Food Security survey module. So based on this, I replicated the validation steps conducted in the US over in Brazil.
LABUN: When researchers validate these surveys, they’re testing the relationship between answers to the survey and data that defines the condition they hope to measure - like food security. So bringing a Hunger Vital Sign screen into Brazil isn’t a matter of faithfully translating from English to Portuguese, it requires ensuring that those relationships between data sets translate. Including whether the underlying data itself is the right data to review. For example, Hunger Vital Sign is validated against a set of negative health outcomes related to lower quality diet, and we don’t take for granted that the most common clinical signs of these outcomes are the same in Brazil as they are in the U.S. Some of those original U.S. health factors were themselves based on health care risk assessment surveys that may not exist in other countries. Plus, clinical measurements often include patterns of use for certain health services - like emergency departments visits, or inpatient hospital stays. Those service structures change from country to country. You don’t hear U.S. researchers discuss implementation through a universal free health care system, for example.
In Brazil, while research built from the U.S. experience, the end result was neither the same full length scale nor the same two question screen.
So household food security survey module, which is the gold standard measure in the US, is the mother of all food security scales. And one of the, you know, children of the scale is the EBIA which is the food security scale that was translated and validated to Portuguese instead of the 18 item of the US scale in Portuguese, it actually has 14 items. These 14 items are used as the gold standard in Brazil. And when I went to validate the hunger vital sign to Portuguese, instead of the first two items, which is the hunger vital sign here in the U.S. actually we have questions two and four as our Tria.
LABUN: The Tria questions are “In the past 12 months, the food that (I/we) bought just didn’t last, and (I/we) didn’t have money to get more?” and “In the past 12 months, did (you or other adults in your house-hold) ever cut the size of your meals or skip meals because there wasn’t enough money for food?”
POBLACION: The two item Brazilian screen was named TRIA, which is short for triage. And this acronym actually has a powerful meaning in Portuguese. After completion of this research project, I was in discussions with the Ministry of Health to implement the TRIA in our universal free healthcare system. And, in the beginning of 2022, the Ministry of Health recommended the use of TRIA in all primary health visits. And now we are in the training phase of providers, community health workers and health center stuff.
LABUN: The example of Brazil, the example of the brief scale for measuring levels of food insecurity, and the original Hunger Vital Sign project all highlight bringing research into practice through offering ways that health care practitioners can connect the general insights gained from basic research to treatment of their particular patients. However, the benefit is greater than that, because each of these advances has a multiplier effect.
Let’s go back to the original example of the scale for food insecurity levels in households. An advantage of these abbreviated measures is that they can be in-the- minute and granular, as a complement to the USDA survey modules that illustrate historic trends nationally. If that’s true, then the new scale should reflect community changes during economic disruptions and also changes in response to the policies set up to address those disruptions. Researchers aren’t going to manufacture catastrophes to test this attribute of the scales, they need to rely on existing data at different points in time - data that we have in part because health care providers had put previous iterations of food insecurity measurement into regular practice.
To validate this novel abbreviated scale, we had two different protocols. To gather data for the 2008 Great Recession, we looked in our database for surveys of families who were interviewed twice. So they were repeaters in our data set before the 2008 Great Recession, and again, during and after the recession. So in this sense we had two interviews in different points in time, and we could compare the food security status of these households. For the pandemic data, in spring 2021, we looked in our database for surveys of families who were previously interviewed by Children's Health Watch before the pandemic started, and invited them to participate in a follow up interview between September, 2021 and March, 2022. So then we had the same protocol, but in different points in time and in different recession periods of the United States.
To be clear - collecting information on food insecurity during a crisis isn’t only an altruistic contribution to the future of academic knowledge. It informs immediate response as well. When we realize food insecurity risk has shot upwards across the country, people from all sectors mobilize to respond. If health care practices know - from a combination of clinical and social information - which of their patients are most vulnerable to adverse health impacts from reduced diet quality, then they can use their specialized knowledge to offer a targeted response to this high-risk, medically complex group.
Looking forward, the addition of these abbreviated scales for measuring food security levels could represent the beginning of a tipping point in the response to food insecurity risk identified in a health care setting. Rich provided the example of helping providers tailor their responses to different patients’ circumstances. But these scales are also what would be used to know if any response actually worked.
Currently we're only able to ascertain some level of risk using, for example, the accountable health communities screening tool and then ideally offer an intervention. This doesn't allow for the ability to measure the impact or the effect of that intervention. So let's say an accountable care organization might screen and refer, but this scale now allows the community based organization, the healthcare agency, whoever the intervention partner is to measure the level, the depth, the severity of food insecurity when that patient connects to the resource. And then do so again, as a follow up after that intervention concludes, to truly understand was there any change in that level, or severity, of food insecurity. Whereas today we're just able to understand, yes, there's some level of risk; yes, we need to connect this person, this household to a resource. There's no way to truly close that loop and create that feedback that's gonna change the outcome for the household. This scale brings us closer to being able to do that.
LABUN This feedback loop accelerates the movement from theoretical to applied. Especially in regions that may be under-represented in the research behind food as medicine interventions. There’s plenty of research to say that if we identify a group of patients at risk for food insecurity today, then something happens that improves their food access tomorrow, on average that group will have better health at some point in the future. There are also model projects from around the country that have published results on options for those “somethings” to improve food access and when that better health “some point in the future” might occur. But it’s easy to imagine that if you picked up one of those programs from central Los Angeles and dropped it into rural Vermont, it wouldn’t play out exactly the same way. It will take several iterations to adjust an intervention in its new context. Health care practices can speed up that improvement cycle by using the risk screen plus abbreviated scale for measuring food insecurity levels to test whether the core logic model is working - that the intervention is improving food security for patients facing barriers to food access. Because those measurements are validated for diverse patient groups all across the country, providers can compare the results to what other practices adapting similar interventions are seeing. This provides guideposts for tailoring a model intervention to a particular clinic’s circumstances to achieve similar positive health outcomes.
A great analogy is standardized parts in the auto industry where companies like Toyota took the traditional assembly line concept and expanded on it to utilize standardized parts across different vehicle models so they can swiftly and efficiently manufacture vehicles rather than rely on a process where each car is hand built by an individual craftsman where something goes wrong, then you're relying on that individual to fix the whole problem, the whole system. Really, at the end of the day, this is all about efficiency and precision, and we're getting a little bit closer and this scale is one step in that direction.
LABUN: The positive impacts of setting up the system once carry forward. Similar to how researchers build from the data sets of previous research, implementers can build from the systems they’ve set up to translate that research into practice - using past results from risk screening and food insecurity measurement to identify resource gaps to fill, measure progress on filling those gaps, and sharing the results.
Or at least, they should be able to.
This is where things get messy, and this is truly where the field of implementation science comes into play. The field is at a point where hunger vital sign screening has been shown to be both efficacious and effective in certain settings, but the setting or population of interest to, you know, you the listener or the researcher or the practitioner, let's say in a community health center in Vermont, is not adequately represented in the scientific literature. There's evidence that shows it could take in many cases as much as 17 years for research findings to be taken up into practice.
LABUN: Seventeen years! Hold on to that number in your mind.
There's evidence that shows it could take in many cases as much as 17 years for research findings to be taken up into practice. And so this has naturally led to a growing urgency in health services resource research to address the research to practice gap as it's called. We need to not just, , you know, utilize the latest research, but we need to put it into practice as quickly as possible.
LABUN: Seventeen years is possibly the entire lifecycle of Twitter. It’s the time it took Facebook to move from a term describing a physical directory distributed to Harvard students to the company now known as Meta, aspiring to rule the Metaverse. It’s older than Bitcoin. If health care research were cryptocurrency, you would have seen Household Food Security Survey Super Bowl ads while the Green Bay Packers played the Steelers more than a decade ago.
Will any of these references make sense to people listening to this podcast a few years from now? A few months from now? Possibly not and that’s the point. In many sectors we’ve seen the perils of a move fast and break things attitude. Here, we’re far from that danger - this is seventeen years to take something up in practice after it’s already been proven in research. We’re still seeing calls to prove that food quality is intrinsically connected to health quality, and I believe I missed the chapter in world history where we first decided it wasn’t connected. Innovation can move slowly, it can not move at all, and eventually you have to concede the point that it’s in danger of moving backwards.
Luckily, like we’ve said, as groups like the Hunger Vital Sign Community of Practice put each new piece into place, it has the potential for a multiplier effect throughout the system. And engaging practitioners as part of the research community doesn’t close the implementation gap by simply increasing awareness of research findings, it increases the speed of both developing new tools and adapting existing tools to new places. Having standardized ways to measure food insecurity supports a better response today while generating more data to use in future assessments, data points that can be compared over time, across communities, across disruptions in those communities; these scales build capacity for testing and refining specific interventions so they can work in new places; and mapping this iterative process adds to the evidence base that the *next* community might draw from to build their own food insecurity response. And that’s without even getting into how this work informs policy changes that themselves speed up the rate of progress.
If you want to learn about the Hunger Vital Sign community of practice and research in this field, be sure to check the show notes for links to more information and to review the tools and explainers shared in the original Hunger Vital Sign series.