Food Insecurity Screening of Hospitalized Patients: A Descriptive Analysis

The Hunger Vital Sign,¹⁰  a validated 2-question screening tool for FI, was added to the intake questionnaire at a quaternary-care children’s hospital in August 2020. The intake questionnaire is embedded within the electronic medical record (EMR) and is completed by bedside nursing with the patient’s family/guardian at admission. The Hunger Vital Sign consists of the following 2 statements:

1. “Within the past 12 months, we worried whether our food would run out before we got money to buy more”; and

2. “Within the past 12 months, the food we bought just didn’t last and we didn’t have money to get more.”¹⁰ 

Responses to each statement were: “never true,” “sometimes true,” “often true,” and “don’t know/refused.” A positive screen for FI was defined as a response of often true or sometimes true to 1 or both statements. A negative screen for FI was defined as a response of never true. A screen that was not completed with the provided answer choices or completed as don’t know/refused was considered as an incomplete screen. A screen that was completed with an answer of don’t know/refused to 1 question and never true to the other question was also considered an incomplete screen. The screening was not a mandatory function in the EMR.

A positive screen for FI triggered a social work (SW) consult via the EMR. SW would perform an assessment specific to FI (Fig 1). Interventions for FI included providing the family with meal tickets to obtain food from the hospital cafeteria during the hospitalization, providing local food bank information for the patient’s county of residence, and providing information on enrollment to the Supplemental Nutrition Assistance Program (SNAP) and/or the Women, Infant, and Children (WIC) Program if the family was not already enrolled.

Demographics were collected on admitted patients hospitalized from August 2020 to April 2021.

Further chart review was performed on patients with a positive screen for FI. Statistical analyses included 1-way analysis of variance in comparison of continuous data. For categorical data, Fisher’s exact testing was used as the low incidence in FI observed as associated with very low expected values. To assist with the calculation of significance values, Monte Carlo simulation (n = 50 000) was used. All statistical analyses were performed with SAS 9.4 (Cary, NC). This study was approved as exempt by the institutional review board.

There were 7751 admissions during the study period of August 2020 to April 2021, of which 4777 (61.6%) had an FI screen completed. There were 2974 (38.4%) admissions with a screen considered incomplete (1348 [17.4%] with a response of don’t know/refused and 1626 [21.0%] with no standard response to the screen documented). Among those with a completed screen, 233 patients (4.9%) were positive for household FI (Fig 2). Patients with household FI had an age range from 15 days to 22 years (mean age 8.7 years, interquartile range [IQR] 2–14 years) and were admitted to 17 unique service lines. Mean age was similar in patients with household FI and patients who had a negative screen for FI (mean age 8.3 years, IQR 2–14). Patients with an incomplete screen were younger with mean age of 7.7 years (IQR 1–13 years, P <.001). Patients with household FI were more likely to be Black (P <.001) and have Medicaid (P <.001). Additional demographics can be found in Table 1. Of the positively screened FI patients, at least 1 previous medical diagnosis and at least 1 admission at our institution within the previous year was documented for 166 (74%) and 85 (36%) patients, respectively. SW documentation specific to FI was recorded for 125 of the 233 (54%) FI patients, of which 39 (31%) were not enrolled in WIC or SNAP.

This initiative highlights acute hospitalization as an opportunity to screen for FI using a multidisciplinary approach. Our findings are in line with previously described racial and insurance disparities related to FI.^5,6,9,11,12  The high proportion of existing medical diagnoses and previous hospitalization to our institution among patients with FI underscores how often health care providers may interact with these patients. This furthers the importance of screening for FI and other SDOH in all health care settings. Similar practices could be implemented at other children’s hospitals to identify FI and connect families to local and federal resources such as WIC and SNAP.

The proportion of patients with FI at our hospital was lower than expected on the basis of initial FI rates in our community before the COVID-19 pandemic. At the time of implementation, we anticipated a higher observed rate on the basis of reports of increasing FI rates.⁷  This finding could be explained by the limitation that the screening tool was not a mandatory function of the admission intake questionnaire. Therefore, we may have missed patients living in food-insecure households in which the screen was not completed. Additionally, the screening was performed either verbally or written, which could lead to underreporting given the potentially sensitive nature of these questions and social desirability. Patient profile questionnaires are to be conducted in the preferred language of the patient; however, this study is limited in that we did not assess the language spoken, which may have led to underreporting of FI in non-English speaking populations. The pandemic also brought additional challenges with providing education to the bedside providers who performed the screening because the typical in-person education sessions had been postponed. This may have led to an answer response of don’t know/refused to be used in situations where the bedside provider did not know the answer rather than the family reporting this answer, such as if a family was not at bedside. Though patients with an incomplete FI screen had a significantly lower mean age of 7.7 years as opposed to 8.7 years for patients who were FI and 8.3 years for patients who were not FI, it is unclear if this has clinical significance. Further investigation could be done to better characterize why this difference was seen, which could also help inform how best to address incomplete screens.

While FI and poverty are associated, they are not synonymous.⁸  Literature has endorsed the relationship between racism and FI given the connection between racism on differences in socioeconomic status (SES) and in turn, FI.⁸  In our population, patients with FI were more likely to be Black, and this association is consistent with previous research.¹²  Although poverty and low SES are associated with FI, there are racial disparities in SES, and structural racism can contribute to these disparities.^8,13  One example of this may include hesitation to obtain government assistance because of concern for racial stereotypes.⁸  Beyond economic disadvantage, social disadvantage among people of color should also be considered. In a study among African-American households in South Carolina, severity of household FI was associated with lifetime racial discrimination.¹⁴ 

We also hypothesize a potential relationship between FI and poor health care outcomes not addressed in this study. Appropriate nutrition is considered critical for physical health and neurodevelopment, especially from the time of conception through the first 2 years of life,¹⁵  and FI can be associated with decreased quantity, as well as quality of available food. FI can also be a form of toxic stress.^1,16  Households experiencing FI may face difficult choices regarding how money is spent.¹⁷  This exposure to FI and thus toxic stress affects the neuroendocrine-immune network, which could account for differences in health outcomes.¹⁶  Although FI remains a complex issue and root causes should be addressed, screening and identifying FI can allow the medical team to intervene and mitigate the effects of FI. Future areas of study include evaluating FI as a potential high-yield screening tool for assessing risk of worse health care outcomes.

In response to initial findings and feedback, quality improvement methods have been implemented to standardize and improve the screening and resource provision process. This has included improving the education to providers regarding FI, screening flow, and resources. Future steps include follow-up with the family to determine if resources were obtained and used. If successful, there is opportunity to expand inpatient screening to include other SDOH to assess for risk of disparate hospital outcomes and connect patients to needed resources with the ultimate goal of mitigating the effects of SDOH, including FI on child health.