Produce Prescription Programs

Program Categories:

The term “Produce Prescription” appears in reference to several different program types.

Produce Prescriptions & Nutrition Incentives: The USDA’s Gus Schumacher Nutrition Incentive Program (GusNIP) includes two categories of initiatives to increase consumption of fruits & vegetables - nutrition incentives and produce prescriptions. Nutrition incentives are any programs that increase purchasing power to include more produce, while produce prescriptions achieve this goal with referral from a health care practice. The core Nutrition Incentive program at GusNIP builds from SNAP income eligibility & participation.

  • One challenge in reviewing the USDA-based literature on Produce Prescriptions is that the health care partnership as defined by GusNIP can mean many different things - were clinicians involved or other staff? Was this paired with clinical nutrition services? Was it paired with medication-based treatment for diet-related conditions? Did the participant even have a diet-related health condition / clinical risk indicators?

  • It isn’t necessary for a health care practice to be working within their capacity as a medical center to address food insecurity and potential negative health outcomes; in the FAHC context, we considered the non-medical role of health care practices primarily as partners in SNAP / 3SVT outreach, not through prescribing produce as a complement to SNAP / 3SVT. 


Produce Prescriptions & Food Prescriptions: A focus on produce targets the primary downfall in most American diets - not enough unprocessed vegetables & fruit. However, food prescriptions might also cover medically-indicated ingredients such as vegan protein alternatives, non-dairy milk alternatives, and gluten-free products. The key distinguishing factor is a prescription focused on whole ingredients, not prepared meals. See the Medically Tailored Meals section for examples of why meals might be a preferred option in some cases. 


Medically Tailored Groceries: A final step in the range of “produce prescription-like” interventions is medically tailored foods. Medically Tailored Groceries, similar to Medically Tailored Meals, indicate a higher level of clinical involvement and a more complete nutritional prescription than lower-intensity interventions like GusNIP Produce Prescriptions. Some Medically Tailored Grocery programs take the form of a meal kit, while others focus on staples. Some programs to support managing chronic conditions begin with Medically Tailored Meals then transition Medically Tailored Groceries as patients become more comfortable with a new dietary pattern and preparing new foods. 

A variation on Medically Tailored Groceries is using groceries as a supplement to meal programs that offer nutritional tailoring but provide less than 50% of daily needs, such as Meals on Wheels.

Information on Produce Prescriptions compared to Medically Tailored Groceries can be found in the 2022 Aspen Institute Food Is Medicine Research Action Plan.

[Back to Top]

Implementing Produce Prescriptions:

The Gretchen Swanson Center is funded by the USDA as the national technical assistance and evaluation (TAE) center for nutrition incentive and produce prescription programs. Their resource website is the Nutrition Incentive Hub. This hub also includes discussion groups for connecting around different topics. Subject matter expert contributors to the program include:  


Two key complementary organizations:


Some well known examples of national programs include: DC Greens, Produce Perks Midwest, Philadelphia Food Trust. In Vermont, multiple statewide initiatives support nutrition incentives, produce prescriptions and similar projects - in addition to local programs. For example: 

  • For several years Vermont has offered Crop Cash to enhance purchases at farmers’ markets. This is part of a New England consortium, organized by Farm Fresh Rhode Island.

  • Farm to Family - Coupons expanding the produce purchasing ability of WIC participants at participating farmers’ markets and farmstands.

  • The Vermont Foodbank offers a mobile food pantry, VeggieVanGo, which travels to health care providers, among other locations. This service combines fresh produce access, education, and a community gathering point. Like some of the retail interventions listed in the previous bullet point, the Foodbank also offers resources to help food shelves display and promote fresh produce, including local produce, linked here.

  • CSA & Health Care - See following section.

In early review of produce prescription capacity in Vermont, FAHC identified technology as a limiting factor in clinical integration. There are low-tech options for offering a starting program, as the examples above show. There are also low-tech ways to establish a “closed loop”, in which community partners & health care providers share basic information about food program participation, as the first step in clinical integration (see, for example, this 2022 presentation). The task becomes more complicated as programs advance and look for options to support elements such as:

  • Patient choice in food access locations - including combining local shopping and traditional retail stores. 

  • Patients with individualized prescriptions - including not just different types of foods but also different amounts of daily requirements. 

  • Programs with multiple funding streams - especially funding with different reporting requirements, or that offer some services eligible for coverage in health insurance plans.

  • Program evaluation systems that can compare patient cohorts with different characteristics (including those that didn’t participate), different “doses” of food treatment, and changes over time. See also next bullet point.

  • Ease of integration with electronic health records - including food insecurity screening results, diet quality diagnostic results, SDOH coding, and complementary medical services. 

  • Ease of integration across geographies to allow all patients within a plan or health service area comparable access to a food benefit. 


One example of an organization that has focused on platforms that can help produce prescriptions and medically tailored groceries grow as integrated health services is Fresh Connect, in Boston. The strongest business case for Fresh Connect implementation requires either a research project or a food program with a high level of clinical integration that engages patients across a full course of treatment (vs. a one-time connection to resources). FAHC had no participating partners with these advanced needs that might serve as an anchor institution to introduce the platform. At the same time, the other options available did not fully satisfy partners’ needs. Limitations included:

  • Lack of flexibility in locations where food subsidies could be used - which in turn exacerbates existing transportation barriers. 

  • Challenges in tailoring the food covered to individuals’ needs, particularly when going beyond produce. 

  • Concern about tying health outcome-focused initiatives to financial outcome-focused programs like SNAP, as the two have different starting criteria for referral and (in Vermont) enrollment data is not shared between health care providers and economic services.  

  • Challenges integrating different funding sources into a single coherent program. 

  • Low capacity for tracking patient participation and outcomes to evaluate program impact on clinical risk factors and/or existing diet-related health conditions. (Note: This is a different task from evaluating impact on food security levels, which connect to long term population health trends).  


As described in the Next Steps section, FAHC choose to focus on more easily resolved issues in the near term rather than directly address anticipated barriers to future growth.

  • Bi-State Primary Care Association investigated possible pathways to address these limitations from an advocacy perspective. A particular concern is that if health care payer coverage of food prescriptions emerges ahead of funding to set up local infrastructure, then the largest food retail companies will control the space and rural communities will lose the health benefits of community-anchored care. This outcome is especially problematic for sustainable dietary change, which is closely linked to social connections, family, and culture. The linked issue brief provides additional details.

[Back to Top]

Community Supported Agriculture and Health Care

In Vermont, Community Supported Agriculture (CSA) programs have been one of the leading produce prescription formats. CSA for health care programs provide a subscription to weekly food shares for households with a goal of preventing diet-related health conditions and/or in response to an existing condition or clinical indicator of risk. Referrals come from a health care provider’s office. Programs last approximately 12 - 16 weeks, primarily in summer, although some programs offer extended seasons.


In 2021, Vermont’s Farm-to-Plate Network, with support from FAHC, launched the CSA and Health Care Community of Practice. The Vermont Farmers Food Center (VFFC) serves as the convener, with support from Bi-State Primary Care Association staff. 


This Community of Practice emerged as part of a Farm-to-Plate organizational restructuring, replacing the Food and Health cross cutting team. This shift was designed to bring focus to a very broad topic area (food and health) in a way that aligned with the local food systems interests of Vermont Farm-to-Plate. 


The Community of Practice explains the choice of the new approach in this way: 


The goals of the CSAs have a unique role to play in the intersection of local food and health in Vermont. This model is one of the most mature farm + health care practice connections in our state, this style of program is familiar to both local food organizations and health care professionals. Plus, we have a good mix of well-established CSA and health care programs and newer programs getting started. There seem to be many opportunities for peer-connection. These programs also have many points in common with evidence-based health programs used outside of Vermont, creating a context for learning from national and regional groups. With our collective experience in CSA-Health Care connections, we can pinpoint specific gaps and areas for development to explore. This development can be helped with input from people who have particular expertise (clinical, reimbursement, logistics, etc.), who may not have capacity to meet with multiple individual programs but could lend perspective in a group setting.   


We believe that a peer-based Community of Practice is a useful way to approach the big topic of food and health, which can be overwhelming without additional focus. 


Examples of Vermont CSA and Health Care programs include: 


Many CSA programs in Vermont also offer supported shares that are closer to a nutrition incentive structure, for example NOFA Farm Share, NOFA Senior Share, and Intervale Supported Shares. Care coordinators can include as options for community resources, information on the community resource referral process is available here.

[Back to Top]

Measuring Impact

One focus of the Vermont Farm-to-Plate Network has been data collection to demonstrate progress towards Network goals & calculate potential impact of different investments in local food systems. The CSA Community of Practice had a similar interest, carried over from the Food & Health cross cutting team. One particular concern for community CSA programs has been an inability to secure capacity-building grants that would allow them to increase the reach and impact of their programs. They find that fundraising revolves around seasonal contributions to cover the cost of shares. This leaves minimal margin for program growth or deepening their health care partnerships. FAHC’s strategic planning grants included evaluation of existing capacity, which echoed the Farm-to-Plate interest in evaluation. Therefore, FAHC reviewed the earlier work on CSA and health care, along with national models of produce prescriptions, to develop a starting point to address the data question.


We divided the challenge into three general categories:

  • Evaluation tools for individual CSA & Health Care programs

  • Cost-benefit evaluation of CSA programs for possible health care payer investment 

  • Collective impact measurements for Vermont CSA & Health Care programs as a sector

Evaluation tools for individual CSA & Health Care programs

To address the first question, FAHC collected the most commonly used evaluation tools onto our Program Evaluation page. We also offered a presentation on the 10 most frequently misunderstood concepts. These are resources for CSA programs and their health care partners. While much of the work on individual program evaluation rests with those programs, some network-based opportunities to improve current systems include:

  • General education in the health care and food & health sectors around Hunger Vital Sign and food insecurity screening implementation (see separate plan)

  • Opportunity for subject matter experts to answer points of confusion around setting common goals in food and health care (to be continued in 2023, see Next Steps). 

  • Opportunity for peer-based exchanges on best practices (the Community of Practice). 

  • Reducing the organizational costs of program evaluation - may be a technology solution (see West Virginia Farmacy Program for an example) or a workforce solution (such as offering experiential learning opportunities to students training for relevant careers).

Cost-benefit evaluation of CSA programs for possible health care payer investment 

FAHC also reviewed the question of health care payer investments. We began with the regulatory structure for health care coverage. Meals have a much greater precedent for coverage than produce prescriptions (see Meals Funding##). The Center for Health Law and Policy Innovation has been investigating options for produce coverage:

The most likely avenue in Vermont’s current structure is found in our 1115 waiver - however, at the time of this writing, Vermont awaited clarification on how the structures might be implemented in the absence of risk-bearing managed care entities. 

We also reviewed the underlying research into the costs and benefits to health care from Produce Prescriptions, led largely by the GusNIP TAE in their evaluator capacity. Here we saw many limitations to the evidence base, including common weaknesses:

  • Studies do not control for broader health care context - such as patient engagement in primary care services, relevant pharmaceutical prescriptions, relevant health services accessed such as RD counseling. 

  • Studies do not offer control groups or compare to cohorts participating in other interventions with similar health goals. A corollary to not tracking eligible non-participants is that we do not know if a common characteristic affects both enrollment and health outcomes (for example, if patients accepting farmers’ market vouchers disproportionately represent those with reliable transportation, kitchen equipment, and cooking knowledge). 

  • Studies do not control for how health care provider referral is made - for example by diagnosis (if any), food insecurity screening systems (if any), or other means.

  • Studies do not identify patients’ health goals when entering the program, primary care providers’ engagement in supporting those goals, or whether those individual goals are advanced (vs. measuring general biomarkers). 

  • Data sets of biomarkers do not necessarily control for frequency of lab tests. 

From these reviews, we determined that the expansion of some “produce prescription” interventions as a covered part of Vermont health care was possible. However, it did not appear imminent at scale. The CSA format has a lower likelihood of qualifying for payer coverage due to constraints in food selection, year round availability, ability to modify to individual needs / enrollment periods, and geographical coverage. We additionally determined that it was beyond the scope of FAHC to address current gaps in the published evidence base.

Collective impact measurements for Vermont CSA & Health Care programs as a sector

 

The final question was of options for measuring collective impact of the diverse CSA programs found in Vermont. The Community of Practice set basic parameters for this task: 

  • Builds from existing commonalities among programs, it isn’t pushing anyone into a framework that doesn’t apply to them or interfere with reporting to current funders.  

  • Information collected provides direct benefits to programs; in other words, it isn’t a pure research project. 

  • Participants do not disclose medical information to non-medical partners as part of this effort. 

  • Minimal additional burden on program staff, volunteers, and participants – preference for using data already collected elsewhere; replacing or streamlining current questions; voluntary participation (for example, focus group for people interested in a particular topic).

The attached slide deck reviews the results of this analysis. Key conclusions:

  • Previous work brought the group close to setting common measurements for impact on the local food economy, and VFFC could lead program managers through resolving final questions.

  • Before specific biomarkers can be incorporated into evaluation, the relationships between individual local food programs & local health care providers needed to be strengthened across several dimensions (see Next Steps). 

  • There was no consensus on measures for dietary change, and setting a common survey tool for all programs could create unnecessary administrative burden for programs that partner with nutrition educators, dietitians, grantmakers, and others with their own dietary change measurement systems. We want to encourage these partnerships, not add unnecessary complications. 

  • A core common health impact question for CSA programs is whether participants can sustain improved diet quality after the end of the CSA season. This question remains true even for CSA programs that run in multiple seasons, as the content of the share changes outside of the summer season. 

 

CSA and Health Care program leaders already knew from previous evaluations that sustaining dietary change after the summer season caused anxiety for their participants. They also knew, year over year, that the same participants returned to the program and reported that they struggled to continue a healthy diet in between CSA seasons. 

 

While the short duration of a CSA share highlights the question of sustainable dietary change, most food-based health interventions assume a time limit. Ongoing support for food access and basic nutrition is the responsibility of the USDA; health care funded food access programs that last indefinitely (without a clear medical need) risk shifting USDA costs onto the already-costly health care sector. 

 

CSA programs may offer some advantages for understanding and resolving the sustainability challenge:

  • The time period is clearly defined, known in advance, and the same for all participants - so there can be collective planning for the end of the season. 

  • This established schedule also means there are established planning and preparation periods for the program in the “off” season, which can be used to address the sustainable diet question. 

  • CSA boxes, as they reflect the local harvest, often introduce participants to new foods and new preparation techniques. In this way they inherently expand skills and palates, benefits that don’t end with the end of the season.

  • CSA programs are networked within their broader community (it is the “C” in “CSA” after all) and networked statewide through programs like Farm to Plate; this introduces a range of possible partners for assistance in sustaining dietary change. 

  • CSA program participants build their own networks as part of the program, participating in pickups, meeting farmers, taking part in educational activities and events, etc. These social networks can also help with sustaining change. 

 

Measuring CSA & Health Care Programs’ collective impact on sustained improvement in dietary patterns requires a combination of strategic planning (described in Next Steps) and data coordination. The literature on behavior change offers many options for choosing a common statewide question to assess CSA participants’ self-perceived ability to sustain dietary change. Examples:

[Back to Top]

Next Steps:

The CSA and Health Care Community of Practice will remain a part of the Vermont Farm to Plate Network, with sign-ups available here. The first full year of meetings identified many potential topics for discussion. These linked meeting notes from the first session of the Community of Practice shows an original brainstorm for topics, which evolved over the course of the year. One concern from the previous Food & Health Crosscutting Teams was the ability to access subject matter expert knowledge on health care practice operations and health care funding.

  • The Bi-State Primary Care Association grant proposal for Network Development starting in 2023, if funded, would include support for staff to assist with this work, and also add Vermont Farm to Plate to the consortium of organizational partners.

  • One item that did not extend past the original phase of FAHC was the monthly newsletter, podcast, and posted updates about food access and health care topics. Individual organizations will continue to circulate funding opportunities, however policy tracking and issues analyses shared with anyone interested in the topic is no longer a background activity. These activities supported the CSA & Health Care Community of Practice by offering general interest education so that the Community of Practice could remain focused on their program format. Vermont Farm to Plate, which manages the Network structure, will assist VFFC in navigating any adjustments that might be necessary.

Some of the topics brought up related to food program-health care provider relationships around Produce Prescriptions appear in separate FAHC plans:

  • Hunger Vital Sign Screening - HVS came up in the community of practice as an indicator that had, over the years, migrated from the original purpose (helping health care practices identify patients who might be interested in the food program) to become a data collection point for the community program partner after referral. A food insecurity risk indicator is less helpful after someone has already enrolled in a food assistance program, a more helpful tool would be the abbreviated food security scale - allowing programs to set a pre-intervention baseline. Informational resources are available to assist any community food program partners interested in adopting the abbreviated food security scale.

  • Closed Loop Referrals - Individual CSA & health care programs can set up systems for reporting program participation back from the CSA to the health care provider, allowing the provider to track patient health goals and relevant clinical indicators. Other produce prescription formats can be more challenging in this regard, as there is not the simple information collection point of who picks up their share each week. Bi-State Primary Care Association is evaluating these closed loop systems as part of the Healthy Rural Hometown Initiative grant. Bi-State is investigating the potential for bringing in a shared basic IT platform, such as has been referenced in West Virginia. In the case of West Virginia, the Farmacy WV platform provides information on individuals’ participation in the program, supports project evaluation tools / surveys, and facilitates HIPAA-compliant communications with food program participants.

The Vermont Farmers Food Center (VFFC) will use mini-grants from FAHC and the Vermont Sustainable Jobs Fund to run a ‘proof of concept’ project in 2023 to test patient-centered design techniques for identifying a strategy to support year-round dietary improvement. This approach will allow VFFC to understand how CSA participants think through their dietary plans. Instead of working from generic barriers - “transportation”, for example - VFFC and their program participants can work together from specific details of how food planning occurs in participants’ everyday lives. For example, “transportation” as a barrier might be replaced by learning many participants have the resource of friends who can offer rides but availability is sporadic. They might go a step further to say the sporadic availability means they wish to build skills in bulk buying, safe food storage, and batch preparing meals to accommodate inconsistent shopping schedules. A CSA program might respond by teaching these food management skills during the CSA season and partnering with community health workers to extend skills building through the year. They might also work with participants to identify what fresh ingredients run out quickly in between major shopping trips and build on their broader food system connections to make those items more widely available.

Patient-centered design, or Patient Co-Design, is an approach health care that begins with how individuals experience the health care system. Some of the approaches resemble consumer testing in other industries - for example, usability testing for computer systems can help finalize patient-centered or provider-led design in telehealth. Other approaches borrow from motivational interviewing, some engage mind mapping techniques. Examples of organizations performing this work include:

Aspects of this design approach already appear informally in programs today. For example, when VFFC wanted to improve how they helped CSA participants learn about federal nutrition program opportunities such as Crop Cash (increasing purchasing power at Farmers’ Markets), they engaged Crop Cash Ambassadors. The Ambassadors were CSA participants who had also had a good experience with Crop Cash. Ambassadors could use their experiences to anticipate what might be preventing others from signing up while also talking them through what they might gain from enrolling. Examples of ideas from the Ambassadors included that 1:1 conversations worked better than letters or posters that were ignored, and that these conversations worked even better with a connected activity - for example setting up a mini-demonstration at the CSA Share pick up with cooking, chances to exchange ingredients with each other & with samples from the farmers market, and farmers market volunteers attending to talk about Crop Cash and assure participants they would be there to help make using Crop Cash easy when they came to the market.

The goal of a patient co-design proof of concept with VFFC and consultant Mandy Dols is to experiment with formalizing informal input processes, directing that input towards the question of sustaining dietary change, and observing if the process could be scaled up to other CSA & health care programs. The first phase of this work is to run detailed interviews with a small group of VFFC participants to begin mapping how they think about their food plan after the end of the CSA season - this component is described in the original project proposal. VFFC can then identify small projects to trial in summer 2023 in response. If all looks promising, VFFC and other interested CSA & Health Care programs can begin to build a strategy for expanding into a full pilot and, from there, consider a larger collaborative project.

[Back to Top]