“The coupons and stuff just made it possible”: economic constraints and patient experiences of a produce prescription program

Patients with hypertension who also experience food insecurity experience economic barriers to participating in a produce prescription program even after vouchers for produce were provided.

Abstract

Although produce prescription (PRx) programs have been shown to improve fruit and vegetable (FV) consumption, few studies have examined how economic constraints influence participant experience. We conducted a qualitative study of patient experience of a 3-month PRx program for hypertension (PRxHTN) including 3 safety-net clinics and 20 farmers’ markets (FMs). We interviewed 23 PRxHTN participants using semistructured guides to understand their program experiences. Interviews were audio-taped, transcribed, and analyzed to identify a priori and emergent themes. PRxHTN participants completing qualitative interviews were mostly middle-aged (mean: 62 years) African American (100%) women (78%). Economic hardship as a barrier to maximum program participation and sustainability was a main theme identified, with three subthemes: (i) transportation issues shaped shopping and eating patterns and limited participant ability to access FMs to utilize PRxHTN vouchers; (ii) limited and unstable income shaped participant shopping and eating behavior before, during, and after PRxHTN; and (iii) participants emphasized individual-level influences like personal or perceived motivations for program participation, despite significant structural constraints, such as economic hardship, shaping their program engagement. Future PRx programs should bolster economic and institutional supports beyond FM vouchers such as transportation assistance, partnering with local food banks and expansion to local grocery stores offering year-round FV access to support sustained behavior change. Additionally, structural competency tools for providers may be warranted to reorient focus on structural influences on program engagement and away from potentially stigmatizing individual-level explanations for program success. These efforts have potential to enhance the translation of PRx programs to the needs of economically vulnerable patients who struggle to manage chronic illness and access basic nutrition.

Implications.

Practice: Training clinical staff in how to assess patients’ level of structural vulnerability and needs for extra-clinic support may improve long-term lifestyle management of chronic disease.

Policy: Clinic-based prescription programs referring patients with chronic disease to community-based services, such as farmers’ markets, must take into consideration structural constraints limiting patients’ ability to engage such programs, including lack of transportation and the need for other support services such as Supplemental Nutrition Assistance Program.

Research: Future research should be aimed at understanding the impact of addressing structural needs of lower-income patients on management of chronic disease.

INTRODUCTION

Chronic diseases are among the most common and costly, yet preventable and manageable, health problems in the USA [1]. Cardiovascular disease (CVD) is the leading cause of death in the USA, accounting for approximately 31% of deaths each year [2]. CVD disproportionately affects racial minority groups and adults with lower socioeconomic status. Compared with all other subpopulations, African Americans have the highest prevalence and poorest control of high blood pressure (BP). Almost half of African Americans have CVD and they are twice as likely as whites to die from these diseases [2]. In addition to racial disparity, adults with low socioeconomic status have 20–40% worse BP control compared to adults with higher socioeconomic status, even in countries with universal health insurance coverage [3]. Poor diet is a main risk factor for CVD and is exacerbated in low-income and racial and ethnic minority communities. Residents of these neighborhoods face challenges accessing high quality, healthy food due to residential segregation, uneven economic development, and endemic poverty. Research demonstrates how “food deserts” negatively influence diet [4] and can leave individuals at high risk for CVD more vulnerable due to economic and geographic structures.

Efforts to address healthy food access increasingly utilize environmental interventions. For example, programs linking individuals with farmers’ markets (FMs) to bolster fruit and vegetable (FV) consumption while treating chronic disease are increasingly employed public health interventions with a growing evidence base for effectiveness [3]. Awareness of structural barriers as issues to address in promoting access to fresh produce has increased interest in FMs [4]. How structural barriers to healthy eating shape FM use, however, has been relatively neglected in research. In a systematic review of factors affecting FM use, only 59% of the studies reported the economic status of participants and no studies with participants from multiple income levels stratified results by income [5]. Other factors indicative of participant structural position, such as race or ethnicity, were reported in only 41% of the studies [5].

Interventions that use “prescriptions” for produce and link individuals to local FMs are one way to address structural influences on diet. These programs integrate clinic-level behavioral interventions with economic and community supports to promote FV consumption through FM shopping [6–9]. Socioeconomic factors are known to shape program effects in important ways. For example, increased FV consumption was less likely to be reported by participants living in financially strained households as indicated by household crowding, running out of food in the last year, and limited access to a car [10]. Lack of transportation has also been shown to be a barrier to FM use [11–13]. Financial strain and isolation within food deserts restricts the potential of FM use as an environmental intervention to address structural influences on health. These interventions will only be effective in addressing health disparities if FMs are broadly accessible to residents in need [14].

While food environments are not consistently associated with dietary behavior and related health outcomes [15–17], several studies have shown improvements in FV consumption with produce prescription (PRx) programs leveraging local farmers markets [6–9]. Yet few studies directly examine how economic constraints may influence program participant experiences [10]. Of prior studies that focus on socioeconomic factors shaping participant engagement with programs linking individuals to local FMs, none focus on how economic hardship shaped participant experiences of such interventions.

Awareness and documentation of the contexts of diverse, low-resource clinical and community settings, particularly social and environmental influences, is needed to better understand their impact on health outcomes and anticipate how best to translate the program to other settings [18]. This research has potential to aid in making sense of mixed results regarding the influence of improvements in food environments on dietary behavior change.

We draw on a qualitative process evaluation examining how participants in one produce prescription program experienced it. We focus on one theme in our data: how economic constraints shaped participant program engagement and sustainability of behavior change related to FV consumption. We attend to participant experiences of the program in clinics, markets, family homes, and neighborhoods. Our approach is informed by the RE-AIM framework with a focus on program adoption, implementation, and maintenance [19]. We aim to understand how the program was implemented and experienced in “real world” settings versus highly controlled contexts characteristic of efficacy research [19] Thus, we add a critical holistic, patient-centered perspective that provides decision makers with information necessary to more effectively implement and sustain future PRx programs in routine practice and community settings [18–20].

METHODS

Intervention: produce prescription program for hypertension

Members of the Health Improvement Partnership of Cuyahoga County (HIP-Cuyahoga), a county-wide health improvement consortium, developed and implemented the produce prescription for hypertension program (PRxHTN). PRxHTN represented a partnership between 3 safety net clinics (those that serve at least 30% patients on Medicaid or uninsured) and 20 FM in Cuyahoga County, Ohio. Clinics served neighborhoods that had roughly 75% of residents living in U.S. Department of Agriculture (USDA)-recognized food deserts, low median household income ($27,349), more than 30% adults living in 100% poverty, greater than 25% of adults over age 25 without high school education, and majority African American resident (97%) populations [21]. The goal of PRxHTN was to increase access to and consumption of fresh FV among participating patients.

Details on the implementation of PRxHTN can be found elsewhere [22] and are described here briefly. The program was based on the theory of implementation intentions and repeated behaviors that focuses on developing specific plans to achieve a decided goal (i.e., the when, where, and how) [23]. Trained healthcare providers (pharmacists, medical assistants, or nurse care coordinators) at each clinic enrolled adult patients who screened positive for food insecurity based on a two-item validated questionnaire [24] and had a diagnosis of hypertension. Enrolled patients met with the provider monthly for 3 months during the FM season (July through December 2015) to check their BP, receive tailored counseling on ways to improve their diet toward better BP control, and were prescribed fresh FV in the form of free produce vouchers ($40 per month for 3 months) to be redeemed locally at participating FM.

Qualitative process evaluation

To better understand patient experiences of PRxHTN, we utilized a qualitative process evaluation approach [25]. We aimed to understand: (i) what aspects of the program worked well and what needed improvement, and (ii) how patient participants interpreted and engaged with the program. In-person, semistructured interviews with participants were conducted to assess participants’ beliefs about food, healthy eating, and FMs; experiences of the program at clinics; and experiences of the program at FMs. Following an iterative research approach, additional questions about food histories, eating patterns, and grocery shopping routines were added to the interview guide after preliminary analyses (see Table 1: Sample Interview Questions). Interviews were approximately 45 min in length and were conducted from March to August 2016 (3–8 months postintervention) in community settings (e.g., health centers, libraries, patient homes) most convenient to the interviewee.

Table 1.

Sample patient participant interview questions

Concept area	Example questions
1. Food Acquisition	How do you typically do your grocery shopping?
2. Program Satisfaction	Can you tell me about the clinic visit when you initially found out about the program, walking me through what happened from the beginning to the end of the visit? What did you think about the materials that were included in your packet each month? How did you use them?
3. Farmers’ Market Experience	Can you walk me through your first visit to a farmers’ market during the produce prescription program, step by step from beginning to end?
	How was shopping at farmers’ markets different from your typical grocery shopping?
	Have you continued to visit farmers’ markets since the program ended? Why or why not?

Sample and recruitment

A total of 23 patients that participated in PRxHTN (6–8 per clinic) were recruited for interviews via a mail-in response card sent to all participants who consented to follow-up contact (210 of 224 participants). Of the 80 patients who returned response cards, interviewees were selected to achieve variation in clinic, age, gender, and economic position. Participating patients were compensated with $15 worth of FM vouchers. Participant characteristics are shown in Table 2. These characteristics of our patient interview sample largely mirrored the sociodemographic characteristics of the overall group of PRxHTN participants, with the exception that interviewees were more highly educated than the overall sample.

Table 2.

PRxHTN patient characteristics

Characteristic	Overall sample	Interviewee sample
Patients Enrolled (n)	224	23
African American (%)	97	100
Age (mean years, SD)	62 (11.2)	62 (9.2)
High School Education or Below (%)	62	36
Female (%)	72	78
Received SNAP (%)	48	43

SNAP Supplemental Nutrition Assistance Program.

This research was approved by the [MetroHealth Medical System] Institutional Review Board. All names used are pseudonyms to protect participant confidentiality.

Data analyses

All interviews were audio-recorded, transcribed, and analyzed using NVIVO qualitative data analysis software. We conducted thematic analyses of a priori themes related to the theory of implementation intentions and repeated behaviors and program processes (i.e., barriers and facilitators to implementation) [23], and emergent themes related to participant experiences [26]. We used a coding and inter-rater reliability process informed by procedures outlined by Hruschka et al. [27]. Two investigators initially developed the codebook after preliminary review of interview transcripts. The full research team then met to discuss the coding scheme, refining code definitions and collapsing and separating codes. The two investigators then independently coded 10% of the interviews, including interviews with participants at each clinic site; we achieved an inter-rater reliability score of 80% [28]. The research team met to discuss remaining questions regarding code definitions and discrepancies before finalizing the codebook. One investigator with graduate-level training in qualitative research methods coded the remaining transcripts.

Data were analyzed for differences by key selection factors (i.e., gender, age, and clinic site). Findings related to economic hardship were consistent across groups. One outlier case, a participant significantly younger than others, was compared to the rest of the sample and no differences were found.

RESULTS

Three subthemes related to patient economic hardships and influence on program engagement emerged from the qualitative analyses. First, many participants lacked transportation, shaping shopping and eating patterns and limiting their ability to access FMs to utilize PRxHTN vouchers. Second, participants often lived on limited and unstable incomes that shaped their shopping and eating behavior before, during, and after PRxHTN participation. Third, economic hardships were at times minimized by participants who focused on individual-level explanations of health behavior change.

Access to transportation

Many participants reported limited access to reliable and affordable transportation, shaping food shopping patterns and engagement with PRxHTN. About half of the participants we interviewed had no car or unreliable vehicles (n = 11). While some took buses to get to markets and did not report problems using public transportation, others, like Larry (African American man, age 63), experienced challenges:

It all keeps coming back to my car. Let me just say this. I’m 63 years old. When I was in high school, I didn’t have to walk … So I’m not used to catching buses or walking. I didn’t ever have to walk when I was a kid and wait ‘til I get an old man and I gotta start walking? So this is all new to me, so that’s why I keep bringing back about my car. I’m talking about the car more than any of the produce and the vegetables.

Participants who similarly struggled with lack of transportation often reported that these issues preoccupied their PRxHTN program experience.

Even participants with cars had difficulty accessing FMs due to limited economic resources. Several participants described their cars as unreliable or entirely unusable during their participation in PRxHTN (n = 4). Those with access to a car altered shopping patterns to conserve gas money. Cynthia (African American woman, age 57) noted that driving to multiple FMs or FMs farther from home was an economic burden: “That burns a lot of gas, and back then, last year, about that time, gas was high.” Donna (African American woman, age 61) used her car to get to multiple FMs for preferred produce, but noted challenges related to public transit:

There was something else I wanted that they usually didn’t have and I had to go to the other [FM] in order to get it. [I: But you got it?] Yes. I did, but you have to travel, and that’s where transportation came in. If you’re on a bus, then you don’t have time to go to another market … Some of the markets did not always have a wide variety of what you might’ve wanted … And by them not being open on the same day, if you didn’t have a car, that would be an inconvenience.

Some clinics attempted to address transportation by coordinating free transportation from the clinics to FMs or establishing a FM at the clinic itself. There was also one mobile FM truck that increased access to patients living along that route.

Despite these efforts, transportation remained a barrier to market access for many. Participants with limited personal transportation sought other means to get to FMs. They drew on their social networks, getting rides from children, friends, neighbors, other kin, and a health aide (n = 8). Leveraging social networks often had the additional effect of exposing drivers to FMs. Brenda (African American woman, age 64) rode to markets with her godsister who lived in the same apartment building, and they shopped together: “We’d go to the farmers’ market, park the car and go in and walk around, look at the vegetables and the fruits, and we’d pick out stuff. She bought stuff too.” In this way, the limitation of lack of transportation had the positive effect of expanding FM exposure to members of participant social networks.

Participants also got rides to markets from what James (African American man, age 58) describes as “travel associates”: people who accept cash in exchange for transportation. These individuals typically waited while participants shopped and did not attend markets with them. This enabled access to FMs when other rides or public transit were not accessible, but with an added financial cost. Finally, some participants used local transportation services through healthcare or social services organizations to access FMs. However, these services had limited pick-up times, making it difficult for participants to coordinate them with market schedules.

When participants were able to access transportation to travel to markets, many, like Tina (African American woman, age 61) reported positive effects of being able to “get out” to FMs:

I would be working my schedule around [FM schedules]. My car is 16 years old and I just got it a year and a half [ago]. I hadn’t driven in 20 years, and to be able to finally get out and get around to do something that was gonna benefit me health wise, and to be able to go to these places without having to get on public transportation and things of that sort, it gave me something to do, you know?

Nonetheless, limited transportation may have both restricted the number of FMs that participants could visit and their access to diverse and desired FV.

Limited and unstable income

Patients commonly reported limited incomes and significant economic insecurity, relying on a patchwork of social services and community resources to meet their basic needs. These economic constraints shaped program participation. PRxHTN vouchers were understood by participants as fitting within this patchwork of often unstable and temporary resources. Like many participants, Tina, whose enjoyment of visiting markets we described above, supplemented her modest and uncertain monthly Supplemental Nutrition Assistance Program (SNAP) allotment, which participants referred to as “food stamps,” with PRxHTN vouchers and visits to a local food bank:

Interviewer:Did you say you just get $15 in food stamps?

Tina:I had $16, but they took a dollar away.

Interviewer:For the month?

Tina:Um hmm. And I got an interview this coming Friday. They might take the whole thing away. When they have the little neighborhood food bank produce things going around the different areas I go to those.

Interviewer:You go to those. So you are a healthy eater.

Tina:I try to be.

Interviewer:You look for opportunities.

Tina:Yeah. I mean you gotta do what you gotta do.

Participants who did not receive SNAP also struggled to obtain basic nutrition. When asked what would support continued FM shopping, Louise (African American woman, age 62) explained, “The [PRxHTN] tokens, ‘cause they really came in handy. I don’t receive food stamps. I don’t receive WIC, none of that, so them tokens be a big help.”

Limited and unstable incomes also influenced participants to shop in bulk or purchase only sale items. As Tina notes: “Some people go shopping and go buy up a whole bunch of stuff, and it lasts a whole month. I only get 15 bucks [in SNAP benefits], so I basically set my shopping schedule around what’s on sale.” PRxHTN helped participants supplement their diets that were largely comprised of free or low-cost items—nonperishables such as canned goods—with fresh produce. Rhonda (African American woman, age 69) explained that PRxHTN, “enabled me to be able to get extra fruits and vegetables, ‘cause when they cut back on your food stamp voucher, all you can do is get the basic-basics.” Fresh produce was seen by some patients as a luxury afforded only by PRxHTN participation. Lisa (African American woman, age 33) explained her motivation for program participation was to afford these typically out of reach foods: “It was like getting healthier food for little or nothing. [For] No more than some time, and I’m willing to dedicate time for that type of food.”

Co-occurring illnesses in addition to hypertension, such as diabetes and chronic pain, exacerbated financial hardship. Some participants faced challenges with affording both food and medicine that shaped grocery shopping patterns. Beverly (African American woman, age 69) limited her food shopping each month to allot funds for future medications. “I had the money to go grocery shopping,” she explains, “but I have to sparingly spend it ‘cause the next month I might need more money for medicine.” Participants who relied on disability or workman’s compensation benefits struggled to afford healthy food. PRxHTN helped these participants manage healthy eating on minimal and sometimes uncertain benefit incomes. Larry, whose struggles using public transportation we described above, used PRxHTN vouchers to fill in gaps in his workman’s compensation income: “It helped financially, because my benefits was suspended for a while. I had to get them reinstated. So money was low, and you have to eat every day, not just sometimes.” The produce vouchers eased these economic burdens during PRxHTN, but provided only temporary relief due to the program’s time-limited nature.

PRxHTN vouchers gave participants economic support to better follow recommendations to increase FV intake, but because of lack of money to supplement vouchers, most participants limited their FV shopping to the amount that could be purchased with program coupons. Larry (African American man, age 63) puts it bluntly: “The fact of the matter is, the [PRxHTN] coins could only buy so much, so I wasn’t really trying to stay in the Pyramid [following program education]. I was just trying to get what the coins could buy.” The inability of the majority of participants to supplement program vouchers with personal funds constrained their integration of FM shopping and FV consumption into long-term routines.

Finally, one participant, Mary (African American woman, age 54), cited lack of basic food preparation tools during PRxHTN participation as something restricting her ability to use the recipe book provided as part of program:

Interviewer:Do you remember using any of those recipes?

Mary:No, ‘cause I didn’t have a stove there at that time where I was [living] at.

Interviewer:So how did you cook the fruits and vegetables that you would get?

Mary:Microwave, and we had like an electric skillet and a crock pot.

Despite increased FV consumption and FM shopping during the program, most patients, even those highly invested in the program like Tina, Rhonda, and James, were unable to sustain these changes due to their day-to-day struggles related to modest incomes. Like other participants, James (African American man, age 58) expressed appreciation for program education, but placed it within his economic reality:

When [FV] are in season was very helpful to learn. Eating better than I do when I’m able to get groceries at the grocery store. I’m not able to get a lot of these items. I’m not able to afford them, and these are the items that I need to be concentrating more on in regards to my health and well-being.

Motivation and deservingness

Despite awareness of economic constraints on FV consumption and FM shopping among participants, they tended to focus on individual motivation for healthy eating. At times, interviewees downplayed structural influences on health behavior and emphasized personal choice, self-control, and discipline.

Although they clearly described the significant economic constraints limiting their food and shopping choices, some participants discussed their ability to apply program information primarily in relation to individual motivation. Often, these narratives were set in the context of participants’ attempts to manage chronic illnesses such as hypertension, diabetes, and chronic pain. Patricia (African American woman, age 60), for example, attributed her health improvement after a long struggle with multiple chronic illnesses to her individual motivation and self-control:

I had never ever been sick a day in my life and I developed a brain tumor, and after my brain tumor, it was like all hell broke loose and I became diabetic. I gained all this weight. My blood pressure was out of control and it was like I was down for the count and it’s like, I don’t want to be this person. I just don’t. It’s not me. I was really depressed because I couldn’t do the things that I wanted to do. I didn’t even look the same . . . And until I decided I have to take control of all of this stuff, because when you have this ongoing illness, like the chronic illness or the acute illness, doctors know what to do if you slash your arm open. They sew it up, or they know what to do about all that stuff, but this internal stuff, it’s like a crapshoot.

Rhonda (African American woman, age 69), who struggled to access healthy food with limited SNAP benefits, similarly emphasized individual motivation, with a focus on self-discipline. She described her improved BP: “After I started disciplining myself, my [blood pressure] came right on down and now it’s under 150.” She continued to incorporate FV into her diet after PRxHTN, attributing her sustained behavior change to “persistence.”

These perspectives regarding program success focused on individual-level patient motivation, self-control, and proactive behavior existed alongside acknowledgment of structural constraints such as limited income and transportation. At times the focus on individual-level factors influencing behavior change shaped how participants were viewed by peers. Some patients distinguished between participants perceived as “serious” about the program versus those who were not and were simply seeking “free” food. Patricia, who emphasized self-control of her health, observed, “You always have some people that are just there for the freebies, but there’s people in there that are really serious.”

These perspectives focused on individual-level influences on health behavior change were expressed by a subgroup of participants despite their parallel understandings of economic limitations on patients’ behavior change. These narratives imply that patients who utilize the program for basic food access without additional health-related goals are less deserving of program participation. These beliefs suggest a hierarchy in which patients viewed as embracing self-control are granted deserving patient status and those motivated by basic needs are labeled less deserving.

DISCUSSION

Our findings highlight the significant economic barriers hindering full program engagement and sustainability of healthy behavior changes among PRxHTN participants. While PRxHTN was designed to improve access to and consumption of FV by promoting FM use with produce vouchers, limited access to affordable and reliable transportation shaped participant shopping and eating patterns before, during, and after PRxHTN in ways that challenged improved FV consumption. Participants also reported low and unstable incomes and relied on food assistance programs to meet basic nutritional needs. Due to economic constraints, many participants struggled to integrate FV consumption and FM shopping into their shopping and eating patterns during the program, and were often unable to do so after the program ended. Despite the clear significance of economic hardship in shaping participant experiences of PRxHTN, some emphasized individual-level influences on behavior change, obscuring the structural factors that significantly shaped patient program engagement.

Findings from this study are consistent with previous research underscoring that economic contexts shape produce prescription program experience [9]. Our qualitative data adds depth and nuance to the existing literature, and novel themes emerged such as the idea of program-related deservingness in the context of produce prescription programs. Our attention to program experiences across multiple sites—participant homes, neighborhoods, clinics, and FMs—deepens our understanding of how individuals experience produce prescription programs in their everyday lives in and outside of formal healthcare settings. Finally, our sample of low-income African American adults living in food deserts and accessing healthcare at safety net clinics provides information on how this structurally vulnerable population experiences produce prescription interventions.

Our findings have implications for the implementation and sustainability of future intervention efforts. Lack of reliable and affordable transportation restricted the ability of participants to simply get to FMs, limiting their engagement with the program on a basic level. This finding is consistent with research showing that proximity to a FM and access to transportation facilitate produce prescription programs’ effectiveness [5,10,29]. PRxHTN and similar programs have been designed to leverage and highlight existing supports for healthy eating [6,8,9,30]. Our data elucidate specific approaches to further integrate existing resources consistent with the needs of underserved patients to enhance program experience and impact. Future efforts should address transportation needs by partnering with clinics, health insurance programs, or social services to develop programs such as free transportation to FMs, mobile FMs, and clinic-based FMs that some of the safety-net clinics collaborating on PRxHTN provided. Patient social networks and informal economies, such as “travel associates” common in impoverished areas [31], could also be formally integrated into programs to support participants without access to a car. Additionally, future programs should consider developing integrated partnerships with local grocery, convenience, or corner stores. Such partnerships, built alongside FM collaborations, would likely facilitate greater access to diverse FV in food shopping venues closer to home, more familiar to patients, and available year-round, supporting greater sustainability of healthy eating.

Our data also have implications for broader policy such as federal food assistance programs. The limited incomes and economic insecurity reported by PRxHTN participants, who were recruited from impoverished, food desert areas, required them to rely on a patchwork of food assistance programs to meet basic nutritional needs. While all participants qualified as food insecure to be included in the program, many received only small amounts of SNAP support or none at all, highlighting the structural limitation of current food assistance programs in reaching individuals in need. This finding is consistent with population level data where less than half of eligible older adults (over 60 years) receive SNAP benefits nationally [32]. Programs like PRxHTN are valuable in helping to fill this gap, but these benefits are restricted by their time-limited nature, and produce prescription programs are not designed to replace essential food assistance programs. However, clinics already screening patients for food insecurity to enroll them in produce prescription programs could serve a dual role to increase SNAP enrollment by connecting food insecure patients to Jobs and Family Services administrators to access these benefits. Healthcare systems promoting SNAP enrollment among eligible nonuser older adults has been shown to result in substantial cost savings within healthcare systems due to averted nursing home admissions and less costly stays [33].

Additionally, our findings underscore the need to view participants in interventions like PRxHTN in the context of their overall lives. Lack of basic resources shaped participants’ eating and shopping, and their lives more generally. Limited transportation isolated them within impoverished communities and low and unstable incomes forced some to make choices between healthy food and medications. These circumstances highlight the importance of addressing other critical needs in individuals’ lives prior to or in parallel to a focus on eating healthy, such as cost-related underuse of medication [34]. Participants faced competing priorities that strained their meager economic resources. Future efforts to support the health of this population should understand diet within participants’ lives holistically, attending to how additional economic supports might contribute to improved health in other ways besides facilitating FM shopping and FV consumption. For example, healthcare systems may identify high risk populations by first assessing for both food insecurity and cost-related medication underuse, and subsequently connecting identified patients to local, state, or private benefit programs that may help reduce the burden of competing choices [34].

Finally, our finding that participants both acknowledge economic barriers to program participation and understood participant engagement in terms of individual-level factors has important implications for framing future programs. Programs like PRxHTN that target socioeconomically marginalized populations exist in spatial and ideological contexts of inequality. Researchers are beginning to examine the race and class politics of these contexts. For example, research has identified distinctions between “cash-paying” and “voucher-using” FM patrons, with “voucher-using” individuals stigmatized by fellow FM shoppers as less invested in the local food system and only seeking “free” produce [14]. These beliefs are consistent with narratives of deservingness that emerged from our data and should be considered in developing similar programs.

The narratives of motivation and deservingness that emerged in this work may be rooted in broader cultural beliefs about “appropriate” patients as rational, proactive, and self-controlling. This has been referred to as Cultural Health Capital, which privileges patients who fit a model of rational autonomous personhood [35]. Yet this model obscures structural contexts, such as poverty, that significantly shaped most PRxHTN participants’ engagement with the program and even factored into their eligibility for the program. These issues highlight central questions: What populations should produce prescription programs engage? What is “appropriate” program engagement? Is motivation to access basic nutrition enough? What can programs like PRxHTN do to better recognize and address the pressing, multifaceted economic constraints faced by participants?

Enhancing the “structural competency” of produce prescription programs has potential to address these issues. Structural competency promotes, “a shift in medical education away from pedagogic approaches to stigma and inequalities that emphasize cross-cultural understandings of individual patients, toward attention to forces that influence health outcomes at levels above individual interactions” [26, p.126]. Individual choice and culture clearly influence health behaviors, yet attention to these factors may overlook their interaction with structural contexts [36]. Although PRxHTN specifically aimed to serve an underserved, food insecure patient population dealing with chronic disease, individual-level explanations of program engagement among some participants remained.

PRxHTN as currently designed already focuses on several components of structural competency, including efforts to “[c]reate alliances between doctors and other professionals who serve the same vulnerable patients to better address the multiple and entangled structural forces that affect patients’ health” and “[b]e creative in addressing extraclinical structural problems” [26, p.683]. Our suggestions for additional partnerships with transportation resources and local retailers are consistent with these goals. Additionally, integrating specific structural competency education for clinic providers, market managers, and patients could reinforce the existing program orientation toward structural barriers and prevent potentially stigmatizing individual-level explanations for program deservingness [37].

Despite the critical insights generated from this study, a few limitations deserve mention. Our participant recruitment strategy that relied on patient initiative to return response cards likely generated an interviewee pool that was invested in the program and may have had relatively greater resources (e.g., stable housing and phone service). This sample provides understanding of how the participants with greater program investment and possibly greater resources experienced PRxHTN. Since the subsample of interview participants had higher levels of education than the overall sample, these data likely capture only the tip of the iceberg of economic strain among program participants. We may have captured less information on how the most socioeconomically marginalized participants experienced the program. However, additional interviews with participants of lower socioeconomic status are likely to only heighten the importance of the economic factors we identified.

Translational implications

Researchers and service providers are increasingly investigating and attempting to respond to the structural needs of socioeconomically marginalized patients to support healthy eating and to prevent and manage chronic illnesses. Our data point to the need for future in-depth studies focusing on how economic contexts of everyday life shape healthy eating and engagement in programs like PRxHTN in order to maximize their effectiveness.

Qualitative process evaluation methods [38] designed to include key dimensions of program adoption, implementation, and maintenance have potential to provide deeper contextual understanding, informing efforts to bolster the structural competency of future programs, thereby potentially improving their effectiveness and sustainability.

Future implementation research that tests the translation of PRx interventions targeting diverse populations in various clinical and community settings is needed. Studies should examine the effectiveness of PRx programs that integrate transportation support, community–clinic collaborations, and structural vulnerability measures in various contexts of care, from urban safety-net clinics to rural specialty clinics, with attention to adaptations necessary to meet the needs of economically vulnerable patients in these diverse care settings. Additionally, our research reveals the limited maintenance of dietary behavior change due to economic barriers. Future research should integrate a sustainability phase to examine the long-term course of PRx interventions targeting socioeconomically vulnerable populations (e.g., prospective trials to test the long-term course of interventions) [18].

These efforts have potential to enhance the external validity [20] and effectiveness of PRx programs that target environmental barriers to improved nutrition, helping better meet the needs of the most vulnerable patients who struggle to manage chronic illness and access basic nutrition.

Compliance with Ethical Standards

Conflicts of Interest: All authors declare that they have no conflicts of interest.

Human Rights: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study protocol was approved by the Institutional Review Board at MetroHealth Medical System.

Informed Consent: Informed consent was obtained from all individual participants included in the study.