A Case Study of Food Insecurity Screening and Referral: Implementation Barriers,Facilitators,and Lessons from a Healthcare–Food Bank Partnership

Abstract

Background:

Routine food insecurity screening in healthcare is recommended, yet the impact of community referrals remains uncertain. This study examined whether food-insecure patients identified through a free cancer screening program were successfully connected to resources via a local food bank partnership.

Methods:

Women screened for breast or cervical cancer through a state-funded grant were assessed for food insecurity using the Hunger Vital Signs tool. Patients identified as food insecure were offered referrals to food bank services during a follow up call. One to three weeks later, a brief telephone survey assessed barriers and facilitators to accessing food.

Results:

Of the 451 patients screened (average age 46 years, 54.3% Spanish-speaking), 92 (20%) were identified as food insecure and received follow-up calls, with 64 (69%) responding. Referrals accepted included government food assistance (23%), home delivery (6%), and both services (8%). Of the 24 patients referred to SNAP or home delivery services, 21 (87%) responded to follow-up calls, and 8 (33%) of these successfully received the referred services.

Conclusions:

Integrating food insecurity screening in health care settings is a promising approach to connect at risk patients with food. Future programs should consider barriers and facilitators to connecting with resources within their community.

Introduction

Social Determinants of Health (SDoH), the conditions in which people are born, grow, work, and age are recognized as primary drivers of health outcomes. ¹ Within the Healthy People 2030 framework, household food insecurity and hunger are highlighted as a key area for targeted SDoH interventions, given their significant impact on population health. ² Food insecurity is influenced by a variety of factors, including income, geography, race/ethnicity, and disability.^{3 -6} Food insecurity leads to inadequate or imbalanced nutrition because households experiencing food insecurity typically stretch their food budgets by purchasing cheaper, energy-dense foods, which usually have lower nutritional value. ⁷ Food insecurity is closely linked to an increased prevalence of diet-related chronic diseases and inadequate disease management, contributing to higher rates of type II diabetes, chronic kidney disease, and cardiovascular disease.^8,9

A strong link between food insecurity and health highlights the importance of screening and intervening for food insecurity in preventive health care settings. ¹⁰ Further, screening for SDoH is encouraged as routine practice in many health care settings. ¹¹ Identifying and acting on food insecurity within a healthcare setting allows healthcare providers to address these health issues and tailor effective treatment plans, reducing long-term consequences. ¹⁰ However, screening without providing any actionable response or resources could be ineffective and worsen patient treatment perceptions or outcomes. Therefore, healthcare organizations are actively seeking collaboration with community partners to address their patients’ needs for sufficient and nutritious food. Food banks and pantries, which aim to combat food insecurity, are prevalent in numerous communities, making them ideal partners for healthcare organizations to collaborate and promote healthy eating initiatives. ¹²

The aim of this study was to evaluate the referral rate of a pilot program developed through a collaboration between a free cancer screening initiative and a municipal food bank to connect individuals identified as food insecure with community-based food resources. Additionally, we sought to identify the barriers this population faces in accessing food bank programs.

Methods

Setting and Participants A cancer institute associated with a large medical center provides breast and cervical cancer screenings at no cost for uninsured women referred by community providers across over 40 counties in Texas through grants from the Cancer Prevention and Research Institute of Texas (CPRIT). ¹³ Eligibility for our study included all women who accessed breast or cervical cancer screening through the cancer center between January 9, 2023, through July 14, 2023. The breast and cervical cancer screening program is aimed at serving women and therefore men were not included by nature of the clinic population.

Data Collection and Measurements This is a program case study utilizing a questionnaire for food insecurity screening, food bank referrals, and qualitative data through interviews. This project was conducted as part of the Community Health Fellowship Program at an affiliated medical school, a program designed to equip medical students with hands-on experience in community-based participatory research. The Institutional Review Board at the affiliated institution determined this research to be nonregulated. The main outcome was measured as the percentage of individuals successfully connected to food bank programs.

During intake for breast or cervical cancer screenings, patients were screened for food insecurity using the 2-item Hunger Vital Signs™ tool, which includes the statements: “Within the past 12 months we worried whether our food would run out before we got money to buy more” and “Within the past 12 months the food we bought just didn’t last and we didn’t have money to get more.” The Hunger Vital Signs™ screener was selected for this project due to its validation as a brief, evidence-based tool that reliably identifies food insecurity in clinical settings using just 2 questions, making it feasible to administer during routine intake without overburdening staff or patients. ¹⁴ Respondents indicating “often true” or “sometimes true” to either statement were identified as food insecure.

Study personnel contacted food-insecure patients, offering referrals to the local food bank’s home food delivery services and Supplemental Nutrition Assistance Program (SNAP) application assistance. If interested, study personnel completed an online referral form to provide patient information to the food bank. The percentage of patients who accepted or declined services was recorded. Food bank navigators then assisted referred patients with SNAP applications or scheduling home deliveries of shelf-stable food boxes, delivered biweekly or monthly in certain ZIP codes under a pilot program.

To evaluate the referral rate, study personnel conducted follow-up phone calls with participants 1 to 3 weeks after the initial referral. During these calls, patients were asked about their experiences in accessing food services, any challenges they faced in receiving home deliveries, and difficulties encountered during the SNAP enrollment process. This follow-up assessment helped identify barriers to service utilization and informed potential improvements to the referral process.

Data Analysis

Descriptive analysis was performed in Microsoft Excel, calculating percentages of patients identified as food insecure, those who accepted or declined services, and those successfully connected to food. Patient feedback was documented and analyzed qualitatively to identify common themes.

Results

Four hundred and fifty-one patients (N = 451) were screened by the intake team, with 20% (92/451) found to be food insecure. The demographic characteristics of these women included a mean age of 46.2 years, 54.3% Spanish-speaking, and 83.5% residing in Tarrant County.

The referral intervention phone calls had a 64/92 response rate (69%). Among the 64 patients who responded, 15 were referred to SNAP assistance (23%), 4 to home food delivery services (6%), and 5 to both SNAP assistance and home food delivery (8%). Additionally, 22 received contact information for local pantries or the food bank (34%). Eighteen (28%) declined any services.

Of the 24 patients referred to SNAP or home delivery services, 21 responded to follow-up calls (87%). Eight patients reported receiving SNAP or home food delivery services upon referral. Among the 13 patients not connected with services, 9 stated that they did not receive communication within the time frame (69%), 3 were ineligible for SNAP due to citizenship status (23%), and 1 did not attend their appointment because they thought they did not qualify for SNAP (7%), see Figure 1.

Figure 1.

Flow diagram detailing inclusion and exclusion of participants.

Qualitative analysis from phone calls revealed financial constraints, exacerbated by unemployment, were major contributors to food insecurity for 20 women interviewed. One woman noted “money is tight” due to current unemployment while another woman mentioned challenges affording groceries because “inflation is crazy right now.” Citizenship status emerged as a barrier for 6 women who wished to receive food assistance but were ineligible for SNAP. Two patients mentioned they did not want to apply for SNAP because “other people need it more.”

Discussion: Barriers, Facilitators, and Lessons Learned

During this pilot initiative, we found that 20% of screened participants reported food insecurity which is higher than the regional average in North Texas (14.2%). ¹⁵ Although the program aimed to connect patients identified as food insecure to local food resources through a partnership between a cancer screening initiative and a municipal food bank, many patients referred for assistance were ultimately not connected to services.

Barriers

This case-study provides insight into the real-world implementation challenges and contextual factors that can influence the success of food insecurity screening and referral interventions. Barriers to connection included systemic issues such as patients’ citizenship status, difficulty making or receiving calls during work hours, lack of follow-up from the food bank, and a broader distrust of institutional or government-affiliated programs. These challenges reflect not only logistical obstacles but also the complex social dynamics that impact patient engagement.

Structural limitations within the program itself also likely contributed to low connection rates. For example, lack of a warm hand-off procedure, dedicated navigation personnel, and wraparound services that improve the success of clinical-community linkages, could have influenced low connection rates. Without these critical supports, patients may be left to navigate a complex referral process on their own, leading to disengagement and unmet needs. Additionally, the high rate of non-responders raises concerns about potential selection bias; future programs should assess differences between responders and non-responders to better understand this dynamic and improve generalizability.

In addition, communication difficulties, such as delays and miscommunication, likely reduced engagement. Long intervals between the initial Hunger Vital Signs™ screening and follow-up calls resulted in some patients forgetting the original interaction altogether. The food bank also had a waitlist for the home food delivery program, which delayed food distribution to most referred patients. An incomplete list of eligible ZIP codes led to some patients being incorrectly identified as ineligible for food delivery. Therefore, miscommunication about food bank programs possibly eroded patient trust, reducing participation.

Cultural perceptions were also a barrier. Some patients declined assistance out of concern that others might be more deserving, reflecting common social norms around food aid programs. ¹⁶ Stigma or fear of judgment may have also led to underreporting of food insecurity during screening, reducing the effectiveness of the intervention from the outset. ¹⁷

Staffing and funding constraints within the food bank posed additional barriers, limiting the organization’s ability to respond to referrals in a timely manner. During the follow-up phase, low survey response rates made it difficult to determine whether patients were successfully connected to services. In some cases, patients may have accessed food assistance after the 1-to-3-week follow-up window, making it challenging to accurately capture connection rates.

Collectively, these barriers underscore the need for comprehensive strategies that integrate clinical screening with immediate, patient-centered support and robust community partnerships. These findings are consistent with existing literature on food bank-healthcare partnerships. For example, Poulos et al ¹⁸ identified similar challenges, including inconsistent data collection, mismatched expectations between clients and food providers, and varying degrees of institutional support.

Facilitators

Despite these challenges, several facilitators strengthened the pilot initiative. First, the integration of interns fluent in Spanish significantly enhanced the program’s reach and inclusivity. Given that over half of screened patients were Spanish-speaking, language concordance likely improved comfort during screening, increased trust in the referral process, and minimized miscommunication, especially given that Spanish-speaking individuals are often underrepresented in medical research. ¹⁹

Second, strategic outreach timing emerged as an important facilitator. Making calls during evening hours, outside of typical workday constraints, improved response rates and enabled better engagement with patients who may otherwise have been unavailable. This approach reflects a patient-centered adaptation that can be scaled in future interventions.

Third, the use of an online referral form streamlined communication between clinical staff and the food bank. This efficient documentation process facilitated timely submission of referrals and improved data tracking across the 2 organizations. Although operational limitations on the food bank’s end delayed some follow-ups, the referral process itself represented a feasible and replicable model of electronic coordination.

Another strength of the program was its embedded feedback loop. By conducting follow-up interviews 1 to 3 weeks after referral, study personnel were able to gather patient-centered insights on barriers and facilitators to accessing services. This qualitative component, often missing from food insecurity interventions, enabled the team to capture the nuances behind service uptake, including issues of stigma, misinformation, and eligibility confusion. ¹² These insights are essential for refining future efforts and ensuring interventions are responsive to the actual needs of patients.

Finally, the integration of this food insecurity screening and referral process into an existing cancer screening program exemplifies a pragmatic and scalable model. Leveraging an already-established patient touchpoint allowed the program to efficiently identify unmet social needs during routine care, reducing the burden of additional clinic visits while enhancing the holistic care experience.

Lessons Learned

One of the key lessons from this pilot initiative is the importance of embedding food insecurity interventions directly into clinical workflows. Rather than relying on follow-up phone calls, which can be hindered by timing issues and staffing limitations, we recommend integrating screening and referral at the point of care. Patients should be screened at intake and asked if they would like assistance with applying for SNAP, information about local food pantries, or home delivery support. Following this discussion, intake personnel can complete the brief online referral forms with the food bank on the patient’s behalf. This streamlined approach reduces reliance on downstream navigation support, an element that was notably missing from our model, and may mitigate some of the drop-off seen in our follow-up phase.

Future programs should aim to collect more detailed participant data, including clinical and lifestyle factors such as family history of cancer, sexual history, and other relevant health behaviors. This program was not run in a traditional clinical setting where electronic medical record data was available, which limited access to such information. Collaborating more closely with referring clinics may help address this gap by allowing for the identification of potential confounding variables, including elevated cancer risk. This, in turn, could enhance the interpretation of screening outcomes and support the development of more tailored interventions.

Additionally, co-locating food assistance programs within clinical settings represents a promising model. Recent literature documents integrating food pantries in healthcare settings may be an effective intervention to enhance accessibility and tackle the specific challenges patients face. For example, a program in New York City provided food to over 2000 cancer patients across 11 clinics, which resulted in improvements in patient food security and wellbeing. ²⁰ These approaches foster holistic healthcare by conveniently addressing medical and social needs in the same location.

Although our program did not achieve high referral rates, it did establish a valuable collaboration between a cancer screening program and a regional food bank. This partnership laid the groundwork for future improvements and demonstrated the feasibility of cross-sector coordination to address food insecurity. Importantly, our experience highlighted both patient-and system-level barriers, including gaps in communication, citizenship-related ineligibility, and cultural attitudes toward food assistance. By identifying these obstacles and reflecting on the operational limitations of our model, we offer concrete recommendations for other programs seeking to improve clinical-community linkages.