Hospitalized and Hungry: A Mixed Methods Study Assessing Immigrant Caregiver Perspectives on an Inpatient Food Insecurity Intervention

Abstract

Inpatient food insecurity (FI), or caregiver inability to obtain adequate food for themselves during child hospitalization, negatively affects caregiver participation in care. Using mixed methods, we assessed inpatient FI prevalence, factors associated with inpatient FI, and perspectives on an inpatient FI intervention among immigrant caregivers (ICs) at a children's hospital from 2021–2022. We performed a sub-analysis of data from a larger FI intervention study, which provided meal trays and food bank public benefit navigator referrals for caregivers screening positive for household or inpatient FI. Logistic regression assessed factors associated with inpatient FI among ICs. We interviewed ICs enrolled in the intervention and identified themes. Of 369 ICs, 56% reported inpatient FI. Low income, poor caregiver health, and household FI were associated with inpatient FI in regression analysis. Nine qualitative interviews revealed positive reception to the intervention. Immigrant caregivers noted that it facilitated participation in care and alleviated financial burden.

Children in immigrant families (i.e., children with at least one immigrant parent) represent a quarter of United States (U.S.) children.¹ These children have over twice the prevalence of household food insecurity (FI) compared with children with U.S.-born parents, which negatively affects health and development.^2–4 Public benefit use reduces household FI disparities among children in immigrant families.⁵ However, language discordance, discrimination, interpersonal racism, and immigration-related fears create barriers for immigrant families to obtain benefits to address FI, such as the Supplemental Nutrition Assistance Program (SNAP) and Special Supplemental Nutrition Program for Women, Infants, and Children (WIC).^6–8

Inpatient FI, or caregiver inability to obtain adequate food for themselves during a child's hospitalization, impedes caregiver ability to participate in their child's care.⁹ Little is known, however, about the prevalence of inpatient FI among immigrant caregivers (ICs), or caregivers born outside of the U.S. Children in immigrant families face more negative outcomes during hospitalization than those from non-immigrant families. For example, caregiver preference for languages other than English is associated with prolonged length of stay, discharge delays, and transfer to intensive care.^10–14 Additionally, some immigrant populations, such as children of undocumented immigrants, face additional stressors during hospitalization.^15,16 For example, one survey found that about a quarter of undocumented caregivers of hospitalized children feared that their child's hospitalization could affect their immigration status.¹⁵

In light of the many barriers to receiving high-quality care that immigrant families face, our primary aims were to assess factors associated with inpatient FI among ICs and elicit perspectives on an inpatient FI screening and intervention bundle.

Methods

Study design/population

We performed a sequential, explanatory, mixed-methods study of English- and Spanish-speaking ICs whose children were admitted to medical acute care services (e.g., general pediatrics and subspecialty services) and surgical acute care services at a large quaternary children's hospital in Houston, TX.

Quantitative design/analysis

For the quantitative portion, we performed a sub-analysis using data from a larger FI intervention study from March 2021–March 2022 which used random sampling to distribute a 136-item survey to assess demographics, public benefit use, household FI, and inpatient FI among all families of hospitalized children (Appendix Figure 1, available from authors on request).¹⁷ This larger study assessed household FI using the validated United States Department of Agriculture (USDA) FI survey.¹⁸ It assessed inpatient FI using a modified version of the validated USDA household FI survey, which has been published in previous work.⁹ The larger study surveyed caregivers on the second day or later of their child's hospitalization and provided interventions if a caregiver screened positive for either household FI and/or inpatient FI. The intervention was funded by philanthropic donations and included a minimum of two free meal trays each day of their child's hospitalization after their enrollment and active referral to a public benefit navigation service through the local food bank. The public benefit navigator service provided resources for applying for public benefits, with a focus on SNAP. Other local resources, such as rent/utility assistance, were provided as well. The larger study then contacted families two weeks after discharge to follow up on their satisfaction with the intervention. Prior to the onset of the larger study, food options for families included: bringing food from home, purchasing a meal tray, purchasing food from our hospital cafeteria, ordering delivery of food from outside restaurants, or going to nearby restaurants. For patients determined to have a need by a medical staff member or for patients who declined to participate in the study, social workers assessed whether caregivers qualified for the caregiver meal tray program outside of the study. For our sub-analysis, we analyzed the subset of data from ICs, or caregivers who self-reported being born outside of U.S. We compared the prevalence of inpatient FI in ICs with the prevalence in non-immigrant caregivers of hospitalized children using chi-squared analysis. We then performed a univariate analysis comparing demographics, household FI, employment, distance of home from hospital, transportation, child/caregiver insurance, perceived health, methods of obtaining food, and public benefit use in ICs with and without inpatient FI using chi-squared testing or Fisher's exact testing. Using stepwise logistic regression, we generated adjusted odds ratios (aOR) with 95% confidence intervals (CI) to assess factors associated with inpatient FI among ICs. Variables with p-values of <.2 in univariate analysis were considered as candidate predictor variables for the regression model. We then iteratively examined the statistical significance of each independent variable in the regression model from the set of candidate predictor variables by entering and removing predictors. Significant variables were chosen for the final model.

We also reported descriptive follow-up survey data on satisfaction with the intervention, as well as follow-up data from the local, food bank navigator, closed-loop referral system concerning those who connected with the navigator and applied for SNAP.

Qualitative

To better understand the quantitative survey findings about the feasibility and acceptability of our inpatient FI intervention, we used a generalized, descriptive qualitative approach¹⁹ to explore IC perspectives. Authors used an iterative process to develop a semi-structured interview guide, incorporating content expertise as well as adaptations from two frameworks published in the literature: the "Getting to Equity" framework and the "Immigrant Health Service Utilization" framework (Appendix Figure 2, available from authors on request).^20,21 Authors piloted the interview guide with 10 families (5 in English, 5 in Spanish) prior to initiating recruitment to assess understandability and cultural appropriateness, and minor adjustments were made based on pilot feedback. We used purposive sampling to identify ICs enrolled in the intervention and recruited participants from November 2021–February 2022. Either the bilingual principal investigator (PI) or a bilingual collaborator (under the supervision of the PI, a trained qualitative interviewer) performed semi-structured individual interviews in the participants' language of choice (English or Spanish). We asked probing questions to explore emerging concepts. The interview, conducted in a private setting, lasted approximately 15–20 minutes. We provided a $20 gift card for participation. We recorded interviews using an audio recorder and transcribed them verbatim in either English or Spanish. Professional translation services translated Spanish transcriptions into English. After removing personal identifiers, three authors trained in qualitative research methodology (MM, KG, and RA) coded interviews independently. We coded data using Microsoft Word and then extracted and organized coded quotations into a Microsoft Excel document. We maintained a codebook and reviewed it as coding progressed. Data collection and analysis occurred concurrently. To ensure credibility of the data, peer debriefing with members of the study team occurred frequently. Analytic memos, reflexive notes, and bracketing in field notes were included in the analysis. A detailed research record provided an audit trail of the study. We used an inductive thematic analysis approach to identify, define, and refine themes. Recruitment continued until our coders judged that we achieved thematic saturation.

Results

Quantitative

Of 1,000 caregivers surveyed, 369 (37%) were born outside of the U.S. and were categorized as ICs. The prevalence of inpatient FI was higher among ICs compared with non-immigrant caregivers (56.1% vs. 39.9%; p=<.01).

Of ICs, 162 (44%) did not have inpatient FI and 207 (56%) had inpatient FI. Demographics comparing ICs with and without inpatient FI are found in Table 1. Compared with ICs without inpatient FI, more ICs with inpatient FI identified as Hispanic, preferred languages other than English, were single, had lower incomes, and had lower education levels. Table 2 compares social and health factors between ICs with and without inpatient FI. More ICs with inpatient FI experienced household FI, were unemployed, had limited transportation (e.g., only had transportation to get home from the hospital), were uninsured, had perceived poor health, and shared food from their child's food tray compared with those without inpatient FI. Additionally, more ICs with inpatient FI were using WIC (66.7% vs 55.7%; p=.02) and food pantries (34.8% vs 17.3%; p<.01) compared with ICs without inpatient FI. Free/reduced-price school lunches and SNAP use were similar across groups.

In regression analysis, factors associated with inpatient FI among ICs included annual income <$10,000 (aOR 9.96; CI 3.49–28.45; ref: >$40,000), annual income $10,000–19,999 (aOR 3.54; CI 1.28–9.74; ref:>$40,000), annual income $20,000–29,999 (aOR 5.75; CI 1.88–17.59; ref: >$40,000), perceived caregiver "fair" health (aOR 4.50; CI 1.55–13.02; ref: "excellent" health), and household FI (aOR 34.49; CI 16.39–72.55; ref: no household FI) (Table 3).

Of the 248 ICs who enrolled in the intervention, 177 (71.4%) completed the follow-up survey after discharge (Table 4). Of those who completed the follow-up survey, the median comfort level with being asked about food during hospitalization was nine out of 10 (on a scale of 1–10, with 10 being the most comfortable). One hundred and forty-six (146, 82.5%) of the follow-up survey participants reported receiving a caregiver meal tray. Of those who received the meal tray, median satisfaction with the tray was 10 (on a scale of 1–10, with 10 being the most satisfied). Of 43 participants who reported connecting with the food bank at the time of the follow-up survey, the median food bank satisfaction score was 10 out of 10 (on a scale of 1–10, with 10 being the most satisfied). From food bank follow-up data, the navigator successfully connected with 141 (57%) participants. Of the 141 participants, 44 (31%) applied for SNAP.

Qualitative

Nine ICs receiving the intervention, all of whom were mothers, enrolled in the qualitative portion of the study (1 English-speaking and 8 Spanish-speaking). Five were from Mexico, three from Honduras, and one from El Salvador. Eight themes emerged from the interviews, and representative exemplar quotations are found in Box 1.

Theme 1: Having a hospitalized child is costly, and this cost is compounded by the inability of ICs to work or earn wages during the hospitalization

Participants described several costs during the hospitalization (in addition to the high cost of food). These costs included parking/transportation, childcare for their other children at home if

Table 1.

DEMOGRAPHICS OF IMMIGRANT CAREGIVERS WITH AND WITHOUT INPATIENT FOOD INSECURITY (FI)

primary caregiver is at the hospital, and the opportunity cost of lost wages with many reporting working in jobs that did not pay them (or fired them) when they took days off to be with their sick child. These costs made it particularly burdensome to spend extra money on food. One participant recounted that her husband was working less hours because he had to drop off his children at school while the hospitalized child's mother stayed at the bedside: "So, instead of working eight hours, he's working five. So, then he's going to have, well, less money for this week."

Theme 2: Leaving the child's bedside to obtain food or coordinating with family to bring in food can be logistically complicated during a hospitalization

Participants described that it was often cheaper to bring in food from home, but that they would prefer not to leave their child's bedside. Some coordinated with family members to bring in food, but this was difficult if family members were caring for other children at home: "It gets complicated because they're taking care of my other girl. And well, they have their own things, right? They can't be available exactly at the hours I have to eat. So, I have to figure it out for myself."

Theme 3: Immigrant caregivers feel grateful for inpatient FI screening during their child's hospitalization

The screening made them feel that their health care providers cared about their wellbeing: "I think it's very good that they're interested in the health of my child and also that they take an interest in us, the family members."

Theme 4: Immigrant caregivers appreciate the free meal tray, feel that it is accessible, and feel that it facilitates participation in their child's care

The meal tray allowed caregivers not to worry about where to find food, was easy to order, and allowed the caregiver to remain at the bedside. One caregiver expressed that she "would be weak,

Table 2.

UNIVARIATE ANALYSIS COMPARING TRANSPORTATION, HEALTH CARE FACTORS, AND PUBLIC BENEFIT USE AMONG IMMIGRANT CAREGIVERS WITH AND WITHOUT INPATIENT FOOD INSECURITY (FI)

perhaps dehydrated" without the meal tray, stating, "They have given me everything here, though. They have given me proper, healthy nutrition."

Theme 5: Immigrant caregivers suggest possibly adding more food on the meal tray and a variety of menu options

Most caregivers were very grateful for their meal tray, but when prompted for areas of improvement, they suggested additional food and additional options.

Theme 6: Immigration-related fear is a barrier to obtaining public benefits for ICs, and ICs recommend education and reassurance to quell immigration-related fear

Many ICs are reluctant to enroll in public benefits because they fear that it will negatively affect their or their child's immigration status if they apply for citizenship, or their ability to stay in the U.S. To maximize enrollment in public benefits, ICs suggested providing more education and reassurance to families about how the enrollment will not affect their (or their child's) immigration status "because a lot of people coming from other countries are not very informed."

Theme 7: Immigrant caregivers are grateful for the food bank public benefit navigation service referral

Immigrant caregivers were appreciative that their health care providers were thinking about their long-term health and providing them with resources to address hunger at home after discharge. One participant noted, "It feels nice you know

Table 3.

LOGISTIC REGRESSION: FACTORS ASSOCIATED WITH INPATIENT FOOD INSECURITY AMONG IMMIGRANT CAREGIVERS^a

that you guys […] actually […] want to take care of us while we're here, but you know after [discharge] […] to see if we if we need the help […] it's very generous."

Theme 8: Immigrant caregivers suggest screening for other health-related social needs and referrals during their child's hospitalization

Examples of social needs were transportation, employment, housing, legal matters, mental health support, and obtaining a driver's license.

Discussion

Our study found 1) a high prevalence of inpatient FI among ICs and 2) positive receptiveness to a hospital-based intervention to address inpatient and household FI among ICs.

It is known that immigrant families experience a higher prevalence of household FI than non-immigrant families.^2,3,5,22 However, to our knowledge, this is the first study

Table 4.

POST-DISCHARGE SURVEY RESULTS AND DATA FROM LOCAL FOOD BANK FOR IMMIGRANT CAREGIVERS WHO ENROLLED IN THE INPATIENT FOOD INSECURITY INTERVENTION

that quantifies the prevalence of inpatient FI among immigrant caregivers. The high prevalence of inpatient FI among ICs may lead to worse disparities in health outcomes for children in immigrant families, as inpatient FI makes it difficult for caregivers to participate in their child's care.⁹ Inpatient FI among ICs is likely multifactorial. Our regression analysis showed that lower income levels and household FI were associated with inpatient FI. Participants in qualitative interviews felt that accessible food options were expensive, with one even expressing that they could not buy a plate of food because of hospital parking costs. Children in immigrant families are more likely to live in poverty than those in non-immigrant families, and thus may not be able to afford food delivery, purchasing a tray at the hospital, or purchasing food at a restaurant nearby.²³ Furthermore, many of our participants confided that they were unable to work and earn income during their child's hospitalization. Many immigrants have temporary employment in industries with inflexible leave policies, such as construction, food services, or manufacturing.²⁴ This places them at risk for losing their job or missing wages while their child is hospitalized, which would compound their financial

Box 1. REPRESENTATIVE QUOTATIONS FROM IMMIGRANT CAREGIVERS IN THE INPATIENT FOOD INSECURITY INTERVENTION

strain.^24,25 Moreover, many of our participants reported that family members were unable to bring them food due to their inability to miss work and/or need to care for the hospitalized child's siblings at home. Health care providers should be mindful of the financial hardships experienced by many ICs. Caregivers also suggested expanding health-related social needs screening to include transportation, employment, housing, legal matters, and mental health support, which are opportunities to better meet the needs of children in immigrant families.

Based on our study results, interventions for children in immigrant families should focus on interventions to address IC needs in the hospital in addition to providing resources for home. Specific interventions will vary by hospital. In one prior study, meal vouchers were effective in reducing FI on pediatric inpatient units.²⁶ In our setting, complimentary caregiver meal trays delivered to the bedside were well received as they were accessible and provided "proper, healthy nutrition," facilitating caregiver availability and participation in care. Our screening and intervention, which was performed universally (administered to both immigrants and non-immigrants) made ICs feel that our health care team not only cared about the needs of their child, but also their family. Screening for FI should be conducted in a respectful, non-judgmental, and trauma-informed manner to demonstrate provider approachability and increase caregiver comfort.²⁷ When counseling families, staff should provide resources in the caregiver's preferred language, keeping in mind that immigrant caregivers may have varying levels of education and literacy.^28,29 Interventions such as ours that support ICs while they are in the hospital have the potential to improve equity in caregiver engagement at the bedside and strengthen the IC-health care team relationship.

Although the vast majority of ICs with inpatient FI also experienced household FI, our data suggest that public benefits may be underutilized (with less than half of participants with inpatient FI using SNAP and food pantries, about half using free/reduced priced school lunches, and about two-thirds of families with children under five using WIC). Our study found that inpatient referral of immigrant families to a public benefit navigator was feasible and well received. Our food bank successfully connected with over half of our ICs enrolled in the intervention and assisted about a third of those with enrolling in SNAP. Consistent with what our participants shared in qualitative interviews, public benefits are often underused in immigrant families because of immigration-related fear, stigma, and confusion with enrollment processes^5,8,15 Our participants additionally expressed interest in immigrant-specific public benefits counseling, which is similar to prior literature on the inpatient setting showing that ICs desire provider guidance on benefits for which their children may qualify.¹⁶ Because our setting has a high proportion of children in immigrant families, the local food bank has experience counseling ICs and addressing immigration-related fears. Inpatient providers who do not have access to public benefit navigators may familiarize themselves with basic eligibility criteria for food-related public benefits for children in immigrant families by using resources such as the American Academy of Pediatrics (AAP) Immigrant Child Health toolkit.³⁰

Some hospital settings have co-located benefits offices.³¹ Although our public benefits navigator through the food bank reached over half of families and about a third of those applied to SNAP, it is possible that additional types of active interventions such as an in-person navigator during the hospitalization or collocation of benefits offices in the hospital would have had better enrollment rates and the ability to more effectively counsel ICs. Additionally, screening for health-related social needs, including FI, at the time of admission would better identify those in need of assistance with benefits applications as well as hospital resources (such as meal trays) to prevent inpatient FI. This intervention could be provided for those who screen positive for household FI, given the strong association of inpatient and household FI in our regression analysis.

Our study has limitations. We conducted our study at a single institution and limited recruitment to English- and Spanish-speaking patients, so our study may not be generalizable to all immigrant populations. Additionally, we did not assess documentation status, and experiences can vary drastically across various statuses. Additionally, we only surveyed one caregiver per child and did not assess the country of origin of all caregivers in the family—thus, we may not have captured all children in immigrant families from our larger dataset. Furthermore, although we recruited a high proportion of ICs, representative of our population, we may have been underpowered to capture significant associations for some of the factors associated with inpatient FI. Our qualitative sample assessing IC experience with our intervention was also small and may not represent the views of all ICs of hospitalized children. Furthermore, our study specifically compared ICs with and without inpatient FI, but it is worth noting that about 25% of ICs without inpatient FI had household FI. Our study was unable to assess whether this 25% had already received assistance from social work or other protective mechanisms against inpatient FI in this portion of the sample. Moreover, our survey data suggest underutilization of food-related public benefits, however we were not able to determine participant eligibility for public benefits. Finally, we did not assess the long-term impact of our intervention on families, and this is a future direction of the work.

Conclusions

As more and more hospitals begin to screen for health-related social needs such as household FI, hospitals should also consider caregivers' needs in the hospital such as inpatient FI. Immigrant caregivers of hospitalized children have high levels of inpatient FI, with several barriers to accessing food and likely underutilization of food-related public benefits such as SNAP and WIC. Our hospital-based intervention to address inpatient and household FI was well-received by immigrant families, has the potential to improve caregiver ability to participate in their child's hospitalization, and may improve enrollment in food-related public benefits such as SNAP for children in immigrant families experiencing FI.