A pilot food prescription program promotes produce intake and decreases food insecurity

Abstract

Food insecurity is associated with limited food resources that may lead to poor nutritional intake and diet-related chronic disease. Food prescription programs offer an avenue for facilitating access to fresh and healthy nonperishable food while reducing food insecurity. The purpose of this pilot study is to examine the feasibility, perceptions, and impact of a collaborative food prescription program in an area with a high rate of food insecurity. The study was a single group pre–post evaluation design. Participants were recruited from two school-based clinics and one Federally Qualified Health Center in north Pasadena, an area with a high rate of food insecurity in Harris County, TX. Adult, food insecure participants were screened at health clinics for eligibility. Participants received nutrition education materials and 30 pounds of a variety of fresh produce plus four healthy, nonperishable food items every 2 weeks for up to 12 visits at a local food pantry. Surveys and tracking tools monitored food insecurity, program dosage, reach, fidelity, acceptability, and program costs. Surveys and key informant interviews assessed perceptions of health care providers, implementation staff, and participants. Participants (n = 172) in the program reported a 94.1% decrease in the prevalence of food insecurity (p < .01) at the end of the program. An average of 29.2 pounds of fruits and vegetables were distributed per family per distribution, and 99% of participants reported eating “all” or “most” of the food provided. Program costs were $12.20 per participant per redemption. Interviews revealed that providers and participants felt the program was well received and highly needed. This pilot study demonstrates the framework and feasibility of a collaborative clinic-based food prescription program to address food insecurity. Future research should examine the sustained impact of such programs on behavioral and health outcomes.

Implications

Practice: Systematic screening for food insecurity is underway nationwide across many health care institutions. Healthy food prescription programs may be a feasible and effective way to address food insecurity and improve healthy eating behaviors in low-income communities.

Policy: With the institutionalization of food insecurity screening in health care settings, sustainable, healthy food access solutions for those found to be food insecure are concurrently needed to not only decrease food insecurity but also improve health outcomes in diet-related illnesses.

Research: Future research should examine the sustained impact of food prescription programs on long-term behavioral and health outcomes.

INTRODUCTION

According to the U.S. Department of Agriculture, household food insecurity is defined as the inability “… at some time during the year, to provide adequate food for one or more household members due to lack of resources” [1]. At least 15.6 million American households were food insecure at some point in 2017. On average, 16% of households with children experienced food insecurity [1].

Paradoxically, the same individuals that report food insecurity tend to suffer disproportionately from diseases typically associated with overnutrition, including diabetes, hypertension, coronary heart disease, congestive heart failure, and obesity [2]. Lack of access to healthy foods such as fruits and vegetables among people on a limited income has been associated with an increased consumption of energy-dense, nutrient-deficient foods, thus contributing to the obesity–hunger paradox [3,4]. In recent years, strategies to improve produce intake using incentive or matching programs as part of existing nutrition assistance programs have expanded, with evidence suggesting their effectiveness [5,6].

Given the demonstrated linkages between health outcomes and food insecurity, health care systems are increasingly implementing food insecurity screening as part of standard care [7]. However, evidence suggests that health practitioners are hesitant to implement food insecurity screening without an adequate way to provide assistance for their patients who screen positive [8,9]. In order to address this gap, clinic-linked resources and food prescription programs have gained popularity as potential ways to improve outcomes for patients identified by routine health care screenings. However, research on the feasibility and impact of food prescription programs remains nascent. The purpose of this study was to demonstrate the framework and examine the feasibility, acceptability, cost, and preliminary impact of a collaborative pilot food prescription program in Harris County, TX.

METHODS

North Pasadena, TX, where this program was implemented, reportedly has the highest prevalence of food insecurity (19%) and childhood (65%) and adult obesity (66%) in Harris county [10,11]. This population is 75% Hispanic, 40% of residents over age 25 do not have a high school diploma, and 26% of the total population (36% of children) are living in poverty [11]. In response to these public health issues, the Harris County BUILD Health Partnership (Harris County BUILD) was formed. The goal of the Harris County BUILD project was to implement a healthy and sustainable food system in north Pasadena using a collective impact framework. The framework concurrently implemented multiple strategies to improve healthy food production, distribution, and consumption in north Pasadena (redacted). Three core partners led the initiative—(redacted) (local health department), (redacted) (local community-based organization and fiscal agent), and (redacted) (local hospital). Further details of Harris County BUILD are described elsewhere [12]. As part of this initiative, a 9 month pilot food prescription program was implemented to mitigate food insecurity across three clinics serving low-income adults in targeted north Pasadena zip codes.

Study design, study sample, and inclusion/exclusion criteria

A one-group, pre–post mixed-methods evaluation study was conducted over 9 months (September 2016–May 2017). This study was reviewed and approved by the Quality Improvement Assessment Board of (redacted).

The food prescription program was implemented across three clinics serving north Pasadena residents: two school-based clinics and one federally qualified health center (FQHC). Adult patients and parents of pediatric patients were screened for eligibility during routine clinic visits at participating sites. Patients were eligible if they were (a) 18 years of age or older, (b) categorized as food insecure based on a two-question clinic screener, and (c) resided in one of three targeted zip codes. A two-question validated rapid screener was used to assess food insecurity among the clinic patients during their clinic visit [13,14]. Of the 549 patients screened across three clinics, a total of 331 patients were eligible to receive the program (60.3%). Of those eligible, 242 opted to participate in the food prescription program (73.1% participation rate; Fig. 1).

Fig 1

Food prescription flow chart.

Food prescription program description

Eligible participants were screened by designated clinic staff, who subsequently described the Food Rx program verbally and using visual aids to the participants who screened as being food insecure and invited them to participate in the program. The medical provider issued a “Food Rx” card to interested participants detailing their food prescription for them to bring to their first visit to the food pantry; the prescription was eligible for redemption every 2 weeks for up to 6 months for a total of 12 redemptions. Upon arrival at the food pantry for redemption, the participants underwent an orientation walk-through of the pantry, where volunteers described the food prescription components, reinforced with nudges, labeling, and messaging around the pantry. Food prescriptions were redeemed at a local food pantry that was open every Thursday and two Saturdays a month. Nutrition guidelines were designed to encourage healthy eating with an emphasis on fresh produce [10,15,16]. A “client choice model” was used, where participants could choose two or more varieties of both fruits and vegetables up to 30 pounds as well as four nonperishable “Food Rx-friendly” items identified with nutritional callout labeling that reinforced basic nutrition messages by food group about various topics, including healthfulness of the food and ease of preparation. Nutrition education booklets in English and Spanish provided easy to read, tailored information on general nutrition, healthy recipes, easy food storage, and basic food safety.

Staff trainings and ongoing support for participating health care providers, food bank, and food pantry workers and volunteers was provided, outlining the study project and individual roles. Ongoing support with food pantry visits, weekly team meetings, and electronic communication was provided. Representative clinic staff received monthly conference calling to discuss implementation challenges and successes across sites. Data collector training and staff interview trainings were conducted.

Measures

Demographics

The self-report screener included questions on gender, age, race/ethnicity, language spoken at home, types of assistance received, and number of children and adults in the household.

Food insecurity

Self-report surveys at the 3rd, 6th, 9th, and 12th redemptions used the same two-item questions administered on the screener to track change in the prevalence of food insecurity [13,14].

Dosage, reach, fidelity, and participant acceptability

Trained volunteer pantry staff tracked attendance and program dosage of individual participants (pounds and variety of produce and types of other BUILD-friendly items chosen) and number of nutrition education booklets distributed. Trained bilingual data collectors administered self-report surveys to participants at their 3rd, 6th, 9th, and 12th redemptions that measured perceived helpfulness of program components in improving participant dietary intake, usage of foods provided, and estimated weekly savings on groceries. The survey included nine questions. For example, one item asked about ease of pantry redemption, “Is it easy to come to the pantry every two weeks?” (response options: yes/no), and “If not, why was it hard for you to come to the pantry?” (response options: transportation, limited time, days pantry is open, pantry hours, and other). Five items asked about participants’ experience with the food provided, “Which of the following best describes your overall experience using the: fruits, vegetables, protein, dairy and grains that were provided by the Food Rx at your last visit?” Response options for each food group included: “I ate all, I ate most, I ate half, I ate less than half, I did not eat, or I did not receive.” If participants reported eating less than half of any of the food groups, further feedback as to why was probed with response options: “The food was spoiled, my family did not like the food available, there was too much food for my family to consume, I did not know how to prepare the food, or other.” Participants were asked to rate the overall success of the program in helping them eat more fruits, vegetables, lean protein, low-fat dairy, and whole grains on a 1–5 scale for each food group with 1 as “not helpful at all” and 5 as “very helpful.” Participants were asked about the average amount of monies they spent on groceries before and after the program (dollars per week) and whether they felt the program was saving them money weekly. Exposure to nutrition education materials was asked “Did you receive the booklets” (yes/no) and “Did you use the booklets” (yes/no). Finally, there was one question regarding program acceptability, “If you were to seek help again, would you come back to our program?” with scaled responses from 1 (no, definitely not) to 4 (yes, definitely).

Program costs

Program costs per participant per redemption were computed using data from implementation partners, including cost of personnel and staffing, food, delivery/inventory, material and supplies including shelf labeling, and other marketing materials.

Clinic provider acceptability and satisfaction. An online 14-item survey measured perceived benefits of program implementation as it related to patient care and food security and perceived program effectiveness, satisfaction, barriers, and successes. The survey was administered to clinic providers at the end of program implementation; responses were on a 10-point Likert scale, with 6 open-ended questions.

Implementation staff and participant interviews

Key informant block-style interviews using standardized questions were conducted by trained bilingual interviewers with implementation staff including the food bank, food pantry, health care implementation partner, and program participants to better understand the implementation, satisfaction, and perceived effectiveness of the program. Interviews lasted about 1 hr and were conducted in-person or by phone. Staff interviews focused on program implementation process and feasibility, success of partnerships, organizational commitment to program sustainability once the funding ends, and perceived program impact on participants’ food insecurity. Participant interviews focused on enrollment experience at the clinic, food received, food pantry experience, and overall rating of the effectiveness and satisfaction with the program’s ability to meet their family’s food needs.

Data analysis

All quantitative data were analyzed using Stata software (Version 15.0, College Station, TX). Summary statistics including means, standard deviations, and frequencies as well as program costs per participant per redemption were computed. Unpaired t-test was used to determine change in prevalence of food security among participants over the implementation. Adept Word Management translated qualitative key informant interviews. Thematic results were analyzed using NVivo software (NVivo 11.0, Melbourne, Australia). The qualitative analytical approach was deductive and concept driven; two researchers independently coded interviews and discussed to consensus.

RESULTS

Sample demographics are summarized in Table 1. The majority of participants were Hispanic (79.7%), female (79.1%), Spanish speaking at home (58.9%), with a mean age of 47 years. The average household size was 4.6 people, with 1.9 children per household. Overall, 24.2% of participants across programs reported receiving additional food assistance, with <10% each receiving Supplemental Nutrition Assistance Program (SNAP) or Medicare/Medicaid assistance and <4% receiving the Supplemental Nutrition Program for Women, Infants and Children (WIC) assistance.

Table 1

Demographics of total enrolled sample

	All (n = 242) n (%)	Clinic 1 (n = 145)	Clinic 2 (n = 81)	Clinic 3 (n = 16)
Race
African Americana	6 (3.5)	6 (4.1)	0 (0.0)	0 (0.0)
Hispanic	137 (79.7)	113 (77.9)	16 (84.2)	8 (100.0)
Female	136 (79.1)	111 (76.6)	17 (89.5)	8 (100.0)
Age (mean in years [SD])	47.3 (13.6)	49.2 (13.5)	35.9 (9.0)	36.0 (4.6)
Language spoken at home
Spanish	133 (58.9)	77 (53.1)	41 (62.1)	15 (100.0)
English	95 (41.1)	68 (46.9)	25 (37.9)	0 (0.0)
Household (mean [SD])
Adults in household	2.7 (1.4)	2.6 (1.3)	3.0 (1.4)	2.6 (2.1)
Children in householdb	1.9 (1.5)	1.5 (1.4)	2.9 (1.2)	3.3 (1.0)
Assistance enrollment
WIC	8 (3.3)	2 (1.4)	4 (4.9)	2 (12.5)
Medicare/Medicaid	17 (7.0)	10 (6.9)	4 (4.9)	3 (18.8)
SNAP	23 (9.5)	17 (11.7)	3 (3.7)	3 (18.8)
No assistance	59 (24.2)	47 (32.4)	11 (13.6)	1 (6.3)

SD standard deviation; SNAP Supplemental Nutrition Assistance Program; WIC Supplemental Nutrition Program for Women, Infants and Children.

^aIncludes “Black” and “African American” classifications.

^bAges 17 and younger.

Dosage, reach, fidelity, and acceptability

Of the 242 patients who opted into the food prescription program and were provided with a prescription, a total of 172 (71.1%) redeemed their prescription at the food pantry at least once. One of the four recruiting health care clinics was dropped in analysis due to low eligibility among screened patients resulting in an inability to recruit participants. Redemption was defined by the number of times the participant visited the food pantry and picked up their food share. Participants were included in redemption rate calculations if they redeemed at least one of the 12 visits. Overall, the average redemption rate across the 12 visits by clinic varied from 35.4% to 39.4%. Average redemption rates declined from 71.1% (redeemed at least once) to 18.2% (redeemed 12 times) over the 6 month program period (Fig. 2a). On average, participants redeemed 6.5 times of the available 12 redemptions. When examining redemption by frequency of visits, 15.7% redeemed only once, 25.0% redeemed 2–4 times, 19.2% redeemed 5–8 times, and 40.1% redeemed 9–12 times. Across those who redeemed their prescription, participants obtained an average of 29.2 pounds of produce per redemption with an average of 10.1 different varieties of produce.

Fig 2

Overall dosage, fidelity, and acceptability of food prescription.

Participant surveys were administered at the 3rd (n = 106, 85.5% response), 6th (n = 80, 84.2% response), 9th (n = 65, 92.9% response), and 12th (n = 42, 95.5% response) redemptions; detailed results report program processes, perceived helpfulness, food insecurity metrics, and reported cost savings. The surveys demonstrated that 99% of participants reportedly ate all or most of the fruits and vegetables provided. Perceived helpfulness of fruits and vegetables in improving their dietary behaviors was high overall at 94.4% and 90.6%, respectively. Perceived helpfulness of other food items provided in improving their dietary behaviors was slightly lower, with lean proteins at 85.2%, whole grains at 82.1%, and low-fat dairy at 73.4%. (Fig. 2c).

Three-fourths of participants reported receiving the nutrition education booklets; 65.7% reported using the booklets. Utility was evident, with one participant noting she learned “…to eat more fruits, vegetables, or to look at the calories in the label, to buy healthier things.”

Food insecurity

Self-reported food insecurity decreased significantly over the course of the program with an average 5.9% of participants screening positively for food insecurity at the program end (94.1% decrease in food insecurity from baseline; p < .01). Food insecurity decreased from 100% at baseline to 10.2% at visit 3 and declined even further to 5.9% by visit 12 (Fig. 2b). Interviews concurred, with one participant stating, “I was struggling a little bit, but then after the pantry everything—well, it was a lot easier; the benefit of the pantry was very good.”

Program costs

The cost of ~$12 per family per redemption included cost of produce plus costs of educational materials and personnel to implement the program. A portion of the food cost was offset by donated produce to the centralized food bank serving the food pantry, reducing the amount of food that needed to be purchased. Furthermore, because the pantry was manned by volunteers, the implementation costs at the pantry site were minimal and personnel costs were primarily related to the food bank personnel costs. Participants reported an average $57 savings per week on their grocery bills, with one participant noting, “The truth is that it does help a lot to the house economy...so I do feel that it has been a huge help.”

Clinic provider acceptability and satisfaction

Overall, clinic providers (n = 3) answered 9.3 on a 10-point Likert scale that because of the food prescription program, “I am more aware of food insecurity in the community I serve” indicating high awareness of food insecurity as a result of the program. When asked whether, as a result of the food prescription program, “Their health center has stronger ties with community programs that promote healthy eating and/or healthy weight,” the providers’ average score was 6.3 on a 10-point scale, indicating a moderate impact of the program on community ties. When clinic providers were asked “How would you rate the overall effectiveness of the Food Prescription (Rx) program” and “How would you rate your overall satisfaction with the Food Prescription (Rx) program,” the mean scores across both items were 9.0 out of 10.0, indicating a high perceived effectiveness and satisfaction of the program. One clinician noted, “I see financial benefits. I think it saves them on groceries…I also think it exposes them to things…and since we give them to them, they’re going to cook it and eat it.”

Implementation staff and participant interviews

Key informant interviews with implementation partners (food bank [n = 1], food pantry [n = 2], and health care implementation partner [n = 1]) highlighted the importance of training and ongoing support for programming, and strong communication between implementation partners was identified as important predictors of implementation success. One interviewee noted, “Overall, I think the trainings, particularly with the clinics, went extremely well because we had to rely on them a lot.” Another highlighted the importance of trust and communication between organizational partners, noting, “Once you build a relationship with them, it’s all about community trust, and if you do not close that loop of communication, you’re going to break that trust.”

Program participants’ interviews (n = 4) concurred with quantitative findings including high-perceived financial, dietary, and health benefits of participating in the program for their families. One parent commented, “...well, I still spend between 60 and 100 dollars, but what helped me was that before I bought very little amounts of fruits and vegetables, so I felt like a little but more comfortable in the sense that I knew that my children were eating more vegetables even though I spent the same.”

DISCUSSION

Our pilot food prescription program implementation and evaluation demonstrated successful feasibility, acceptability, and a 94% decrease in prevalence of food insecurity among food prescription participants for the 6 month duration of the program. Program participants reflected the source clinic population—more than three-fourths were Hispanic/Latino, one quarter to one half had a language preference other than English, and two-thirds were below the poverty line or economically disadvantaged per clinic or attached school demographics [17–19]. Other pilot studies of food prescription programs have successfully partnered with health clinic populations [20–23]. Freedman et al., in recruiting low-income diabetic patients from an FQHC in rural South Carolina, examined visits to an onsite farmer’s market and financial incentive influence on fruit and vegetable intake [20]. A dose–response relationship between the intervention and fruit and vegetable intake was found. Goddu et al. implemented a food prescription program for diabetic patients across six health centers in Chicago. Participating patients redeemed their prescription at pharmacy sites, which were equipped with fresh produce and healthy food offerings, and at local farmers markets [22]. The implementation lessons learned included involving the entire clinic staff in providing prescriptions, incorporating appropriate screening measures, importance of convenience of location of redemption sites, partner collaboration, and having a coordinating team to achieve program implementation success. Similarly, Trapl et al., in a mixed-methods evaluation of a produce prescription program for pregnant women, showed that providers reported high feasibility of implementing the program in a clinic setting and that it created opportunities for them to talk about diet with participants, greater awareness about farmers markets, and new shopping habits [22]. Also, Bryce et al. determined the impact of a farmers market food prescription program, using a one-group pre–post evaluation design, on low-income diabetics receiving care at an FQHC [24]. The program allotted up to $40 ($10 per week for up to 4 weeks) for purchase of produce from an FQHC co-located farmers’ market. Results demonstrated significant decreases in hemoglobin A1C after participation in the 3 month program.

Our study differed from these programs as it targeted food insecurity as the main outcome, an important social determinant of health, and linked to chronic disease including obesity and diabetes [2,24–26]. However, a limitation was lack of data on dietary intake of study participants. Also, while the other food prescription programs were primarily monetary incentive strategies [20–23], the Harris County BUILD food prescriptions were for 30 pounds of various healthy foods. Furthermore, our study captured program costs, which is important for future scalability and sustainability of program components.

While our process evaluation data suggest that participants who did attend the Food Rx distributions found the various foods and nutrition education materials provided to be highly acceptable and effective in improving their eating behaviors, there were challenges with clinic and participant retention in our study. One of the four participating clinics did not adequately screen their patients and, hence, was dropped from the study. Furthermore, the redemption rate among the participants dropped from 70% to 18% over the 6 month redemption period, with an average redemption rate of 39%. However, it is worth noting that once a participant redeemed, a significant proportion (more than 40%) continued on and redeemed 9–12 times. Redemption rates in other food prescription studies ranged from 56% to 100%, with proximity to the fruit and vegetable provisions cited as an important factor among those who redeemed the prescription [22,23]. In one study, which had the food market on site, the redemption rates were higher as compared to those seen in our study, which could be related to convenience of the location of the food market [20]. However, in our study, given the sampling from the three clinics for the study, there was one local food pantry as the redemption site. For clinics that may have space constraints, such a model of having a food pantry in the area may provide a more feasible solution. To address the low redemption rate in our study, project staff conducted reminder phone calls. Some of the most common reasons for lack of attendance cited by participants on these calls were that they thought the Food Rx was a one-time redemption and the food pantry redemption times did not align with their needs, which prompted redemption day modifications. The qualitative evaluation highlighted the need for bolstering attrition-reduction strategies, including early heightened contact and reinforcement with participants, ongoing communication between key partners, and data tracking with real-time feedback loops between partners to support program participation and retention in future studies. These results helped identify challenges in engaging participants and identified the need for clear and reinforced program messaging with participants; it also underscored the importance of incorporating qualitative assessments to help contextualize quantitative findings.

Our study demonstrated the feasibility of collaboration and communication between the clinics and the food pantry, where the participants redeemed their prescriptions. Tracking was systematic and robust, but there was a lack of clarity with detailing and reinforcing program messaging to participants and an inadequate feedback loop between the pantry and the clinic, which should inform future studies. Our study operationalized linkages between clinics, the food bank community service provider, and the food pantry, while serving those high-risk individuals most in need of assistance. This provides a translational framework for food prescription program implementation among communities experiencing high levels of both food insecurity and health disparities.

Limitations

The small sample size, lack of a control group and randomization to reduce bias, and lack of validated measurement of dietary intake make it challenging to attribute changes seen to the intervention. The results underscore the need for evaluation using larger sample sizes with sufficient statistical power and robust study designs. Further, the pilot study did not include an outcome evaluation to assess participants’ food insecurity status beyond the conclusion of the intervention, including dietary intake, home nutrition environmental outcomes, and cooking behaviors, which should be considered in future studies. Although the food prescription program continued in north Pasadena through clinics and the local food pantry after the pilot study ended, ways to maintain the healthy eating habits established and healthy food access should be explored in collaboration with other food assistance programs, such as SNAP and WIC.

Translational implications

As Glasgow and colleagues opine, dissemination and implementation science ensures that health research becomes the accepted standard in other settings with varying populations [27]. Our pilot study provides an important first step trialing a program design for implementation in other health care settings with food vulnerable populations. As next steps, it would be important to identify and understand predictors of implementation success across clinic settings using validated frameworks such as Consolidated Frameworks for Implementation Research (CFIR) [28]. Using hybrid study designs and conducting effectiveness-implementation trials could inform successful dissemination of the Food Prescription program across clinics serving diverse populations similar to that seen in our study [29]. Health care dissemination of innovative ideas can be slow [30]. Keys to improving dissemination success include identification of vulnerable populations and fostering strong relationships with health care partners. Concerted effort was spent cultivating mutual relationships with implementation partners that proved invaluable; the same dedication will be paramount to expanding program success across organizations and locales.

The findings of this study may also be particularly applicable to minority or immigrant populations. Participants in the current study were primarily Hispanic, and few (less than one-fourth) reported receiving additional food assistance. If these are potentially immigrant families, recent data suggest that Mexican immigrant families are less likely to participate in SNAP, although they have higher food insecurity compared to nonimmigrant families [31]. Thus, programs such as the food prescription program that screen and provide for food insecure families as part of routine health care, outside of federal or state assistance programs, may help fill the resource gap among marginalized populations.

Qualitative feedback from implementation staff and participants added rich detail to both successes and challenges that need to be considered in future applications, specifically related to strategies to optimize participant retention and implementation and support staff communication. Great attention should be paid to setting up these strategies and feedback loops early in implementation to ensure ongoing program success.

CONCLUSION

This pilot study demonstrates the feasibility of a coordinated, clinic-community-based food prescription program to reduce food insecurity among underserved populations. While there were participant retention challenges, program acceptability was high among those who did participate. As health care institutions nationwide screen for food insecurity, evidence-based strategies and frameworks are needed to address the issue. While the preliminary findings demonstrate feasibility of implementation, acceptability, and satisfaction among providers, and program participants, further research using a stringent study design with a comparison group, larger sample size, and a statistically powered outcome is warranted.

Acknowledgements

The authors would like to acknowledge the Harris County Build Partnership and thank all the organizations involved with this project, including technical assistance provider Wholesome Wave, AmeriCorps VISTA, Pasadena Community Ministries, participating Memorial Hermann Clinics in Pasadena, the Pasadena Health Center and the Houston Food Bank. The authors specifically would like to acknowledge AmeriCorps VISTA David Kronenberger for his work with the Pasadena Community Ministries food pantry, on behalf of the Houston Food Bank. TheHarris County Build Partnership was the result of a call for proposals issued from the BUILD Health Challenge, a national awards program supporting bold, upstream, integrated, local, and data-driven community health interventions in low income, urban neighborhoods. The overarching strategic goal of Harris County BUILD was to launch a new food system in north Pasadena that is healthy, sustainable, affordable, accessible, and community supported. Core to the success of Harris County BUILD was the use of a collective impact framework with various community service providers concurrently implementing the programmatic components and a Backbone Committee overseeing and leading these efforts.

Funding

The Harris County BUILD project was funded by the BUILD Health Challenge, a national awards program supporting “bold, upstream, integrated, local, and data-driven” (BUILD) community health interventions in low-income, urban neighborhoods founded by The Advisory Board Company, the de Beaumont Foundation, the Colorado Health Foundation, The Kresge Foundation, and the Robert Wood Johnson Foundation. The University of Texas MD Anderson Cancer Center’s match was funded through the Office of Health Policy and the Cancer Prevention & Control Platform within the Moon Shots Program at MD Anderson. The Cancer Prevention & Control Platform is supported by generous philanthropic contributions from the Lyda Hill Foundation. Dr. M. Raber is supported by the National Cancer Institute of the National Institutes of Health, Award Number R25CA057730 (PI: Shine Chang, PhD).

Compliance with Ethical Standards

Conflict of Interest: The Harris County BUILD Health Partnership’s hospital partner was The University of Texas MD Anderson Cancer Center (MD Anderson). As a condition of the BUILD Health Challenge award, the hospital/health care partner was required to provide a match through financial and in-kind support. MD Anderson provided the match through in-kind support (program management) as well as direct financing of the evaluation activities. MD Anderson and UTHealth entered into a services agreement to organize and lead the evaluation. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Ethical Approval: 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. This article does not contain any studies with animals performed by any of the authors.

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