Telehealth for Primary and Preventive Care Among Food-Insecure Individuals

ABSTRACT

Objectives: Using data from the 2020-2022 Medical Expenditure Panel Survey, we examined the relationship among food insecurity, access to a usual source of care, and telehealth utilization across 4 types of office-based and outpatient visits: general checkup, diagnosis or treatment, psychotherapy or mental health counseling, and follow-up or postoperative care.

Study Design: Retrospective analysis of nationally representative data.

Methods: The study employed logistic regression models for access to care, Poisson models for annual visit counts, and 2-stage Heckman selection models for telehealth utilization and associated expenditures.

Results: Food-insecure individuals had a 7.2% lower rate of annual visits (in-person and telehealth combined) than their food-secure counterparts. Food insecurity was associated with a 1.7–percentage point increase in the share of telehealth visits. Among individuals with a usual source of care, food insecurity was linked to higher telehealth use as travel time increased: 2.6 percentage points higher with 15 to 30 minutes of travel time, and 4 percentage points higher for travel times exceeding 30 minutes. Additionally, each 1–percentage point increase in telehealth share corresponded to a $117.64 reduction in health care expenditures per visit.

Conclusions: These findings highlight significant disparities in health care utilization based on food security status in the US. Even after accounting for geographic access, food insecurity remains strongly associated with reduced health care use. It is important to test whether integrating food insecurity screening in health care settings and developing hybrid telehealth models (eg, mobile clinics) may help close gaps in access and improve outcomes for food-insecure populations.

Am J Manag Care. 2026;32(1):In Press

Takeaway Points

• Food-insecure individuals have fewer total annual visits (in-person and via telehealth) across 4 types of office-based and outpatient visits: general checkup, diagnosis or treatment, psychotherapy or mental health counseling, and follow-up or postoperative care.
• Food-insecure individuals are more likely to use telehealth for these 4 categories of care.
• A higher share of telehealth for these 4 categories of care is associated with a significantly lower health care expenditure per visit.

Chronic diseases account for nearly 90% of the $4.5 trillion annual health care spending in the US.^1,2 Adults in food-insecure households are at greater risk for conditions such as diabetes, hypertension, and mental illness,^3-12 and they tend to use more costly services such as emergency and inpatient care.^3,13-18 It is unclear whether these higher costs are solely due to greater disease burden or also reflect barriers to primary and preventive care. If access is a key factor, telehealth, using technologies such as videoconferencing and wireless communication to deliver clinical care, may improve access for food-insecure populations and reduce health care costs.¹⁹

Multiple studies have examined telehealth utilization for disease management^20-23 and the relationship between food insecurity and health care utilization separately. However, their intersection remains largely unexplored. Food insecurity is often linked to transportation challenges and financial constraints,^24-27 which can limit access to in-person care.^28-30 Telehealth presents a potential solution by providing a more accessible alternative, reducing the burden of travel and associated costs. However, disparities in digital access, technological literacy, and insurance coverage may prevent food-insecure individuals from fully benefiting from telehealth services.³¹ Addressing these gaps through targeted policies can help maximize telehealth’s role in improving health care outcomes for food-insecure populations.

This study had 2 objectives: (1) to explore the relationships among food insecurity, access to primary and preventive care, and health care utilization across 4 types of office-based and outpatient visits: general checkup, diagnosis or treatment, psychotherapy or mental health counseling, and follow-up or postoperative care, and (2) to examine the demand for telehealth for primary and preventive care based on individuals’ food insecurity status and their travel time to usual care settings, including doctors’ offices, clinics, and health centers.

METHODS

Data Source

This study used data from the Medical Expenditure Panel Survey (MEPS), a longitudinal survey of health care use, expenditures, payment sources, and medical conditions among noninstitutionalized US residents. The study sample comprised adults 18 years and older from the MEPS Household Survey with available measures of food insecurity. Publicly available deidentified MEPS data were merged with harmonized sociodemographic data from IPUMS.³²

The first set of analyses, which explored differences among food-insecure and food-secure individuals in their access to care and total health care utilization, used unbalanced panel data from survey years 2020 to 2022. The second set of analyses, which examined telehealth utilization and related expenditures, used pooled data from survey years 2021 and 2022. Questions related to telehealth use were first introduced in the second half of the 2020 survey year, with considerable missing observations until November 2020.

Although MEPS typically includes 2 overlapping panels, 3 were used during the COVID-19 pandemic to offset lower response rates. This study’s sample included data from 21,876 adults in 2020, 22,781 adults in 2021, and 18,101 adults in 2022, totaling 62,758 observations from 2020 to 2022. Due to missing data on key variables, the analytic sample for panel regressions was reduced to approximately 54,000 observations. Table 1 details variable-level missingness.

Measures

Food insecurity was a key predictor in this study. Because MEPS collects food insecurity data only for the family reference person and does not include a person identifier in public-use files, the reference person’s food security status was assigned to all family members. Following standard practice, individuals were initially categorized into 4 levels—high, marginal, low, or very low food security—based on a 10-item questionnaire. The measure was then dichotomized so that individuals with marginal, low, or very low food security were coded as food insecure (coded 1) and those with high food security as food secure (coded 0).^3,17 eAppendix Table 1 reports category-specific sample sizes by year (eAppendix available at ajmc.com).

Access to primary and preventive care was measured by an indicator, no usual care, coded as 1 if an individual lacked a usual source of care (such as a doctor’s office or clinic) and 0 otherwise. Travel time to usual care, asked only of those with a usual source, was categorized as less than 15 minutes, 15 to 30 minutes, and more than 30 minutes (the latter combining longer times due to sparse responses). Insurance type was grouped into 3 categories: uninsured, public only, and any private coverage.

Annual visits is a count data measure reflecting the total number of office-based and outpatient visits, either in person or via telehealth, for the 4 visit types in a year. Although there are other reasons for visiting office-based and outpatient care settings, such as immunizations, accident- or injury-related emergencies, well-child exams, and pregnancy-related care, the sample was restricted to those who used the 4 types of visits, as telehealth was most commonly used for those categories of care.

The second analysis focused on telehealth utilization and health care expenditures, using data from 2021 and 2022, as telehealth measures were not collected until late 2020. To include individuals with and without office-based or outpatient visits, data were pooled across both years. The pooled sample included 15,176 individuals: 6101 with no visits and 9075 with at least 1 visit. eAppendix Table 2 reports average telehealth shares by visit type.

Telehealth utilization was measured as the weighted share of telehealth visits among all office-based and outpatient visits, adjusted for visit type. It was calculated by weighting the telehealth share within each of the 4 visit types by the total number of visits per type and dividing by the sum of weights. On average, 7.5% of such visits were via telehealth in 2021 and 2022. Expenditures were calculated by averaging spending per visit within each type and setting, then summing across all categories. The mean expenditure per visit was $374.79.

All regression models included covariates for chronic conditions, functional or activity limitations, sex, and marital status. Race/ethnicity was categorized as White non-Hispanic, Black non-Hispanic, Hispanic, or other. Age, education (in years), and poverty level were included as continuous variables. Summary statistics are provided in Table 1.

Statistical Analysis

The first set of analyses explored the interrelationships among food insecurity, access to health care, and demand for primary and preventive care. First, unbalanced panel data from 2020 to 2022 were used to estimate the difference in the likelihood of not having a usual place of care due to food insecurity using a logistic regression model. Second, multinomial logistic regression models were used to estimate the relationship between food insecurity and travel time to usual place of care. Third, Poisson regression models were used to estimate the association between food insecurity and total number of office-based and outpatient visits in a year.

Both fixed and random effects models were estimated for each outcome. Fixed effects control for unobserved, time-invariant individual heterogeneity but exclude individuals with no variation in the dependent variable or data from only 1 year in nonlinear models with unbalanced panels. Random effects use more data and are more efficient but may be inconsistent if unobserved heterogeneity exists. Hausman tests rejected the consistency of random effects in all cases, indicating unobserved factors must be accounted for.

Heckman selection models were used in the second set of analyses to account for individuals with no office-based or outpatient visits using pooled data from 2021 and 2022. For these individuals, no telehealth utilization was observed for any of the 4 visit types. The Heckman selection model consists of 2 estimation stages. First, the likelihood of having at least 1 office-based or outpatient visit for any of the 4 categories of care was estimated. Conditional on this likelihood, the share of telehealth use (or health care expenditure) was estimated in the second stage. Insurance status was included as a predictor only in the first-stage model because it was associated with use of primary and preventive care, but not with share of telehealth care. All other independent variables were included in both stages. Although insurance coverage is known to affect health care utilization, its differential impact on telehealth vs in-person care utilization remains unclear. In the current sample, this analysis did not find a statistically significant relationship between telehealth utilization and insurance coverage. Additionally, the analysis of the 2021-2022 period leveraged pandemic-induced policies that expanded telehealth coverage by requiring all health insurers to reimburse telehealth services at the same rate as in-person services. Accordingly, insurance coverage was used as an instrument in the first stage to predict any office-based or outpatient care utilization across the 4 visit types. The second stage then examined the share of telehealth utilization among individuals with nonzero health care use.

RESULTS

Logistic regression models with bootstrap SEs were estimated to examine the relationship between food insecurity and the likelihood of not having a usual place of care. Estimated ORs from both fixed and random effects models are shown in Table 2 for select independent variables. Results were interpreted from the fixed effects model, while acknowledging the substantial reduction in sample size. The odds of not having a usual place of care among food-insecure individuals were 1.12 times those among food-secure individuals; however, this estimate is not statistically significant. In contrast, the random effects model yielded a substantially higher OR with a smaller SE, and it is statistically significant at the 99% confidence level. In both models, having insurance coverage, whether public or private, was associated with significantly lower odds of not having a usual place of care compared with being uninsured. In the fixed effects model, prior health status was not statistically significantly associated with the likelihood of lacking a usual place of care.

eAppendix Table 3 presents the relationship between travel time to individuals’ usual place of care and food insecurity status, using fixed effects multinomial logistic regression models. In the first model, individuals without a usual place of care were excluded, as travel time data were not available for them. In the second model, “no usual place of care” was included as an additional category of the dependent variable. Taken together, both models suggest that food-insecure individuals were more likely than their food-secure counterparts to experience longer travel times when accessing their usual source of care.

Finally, the unbalanced panel data were used to estimate Poisson regression models to examine the number of annual visits to office-based and outpatient care settings for the 4 visit types. Incidence rate ratios (IRRs) are reported in Table 2. As with the previous models, the Hausman specification test rejected the null hypothesis, indicating that fixed effects estimated are preferred. The fixed effects model, estimated using data from 15,103 individuals, shows that both food insecurity and lack of a usual place of care were significantly associated with fewer health care visits. Specifically, food-insecure individuals had a 7.2% lower rate of annual visits (IRR, 0.93), a result that was statistically significant at the 99% confidence level. Similarly, individuals without a usual place of care had a 12.7% lower rate of annual visits (IRR, 0.88), significant at the 95% confidence level. In a second model, travel time to the usual place of care was included as an additional explanatory variable. However, travel time was not found to be statistically significantly associated with the number of visits.

Next, the relationship between food insecurity and demand for telehealth services for primary and preventive care was examined. Two sets of models were estimated. In the first set, the dependent variable was the share of telehealth use, calculated as the weighted proportion of visits conducted via telehealth out of all visits for general checkups, diagnoses or treatments, psychotherapy or mental health counseling, and follow-up or postoperative care. In the second set of models, the dependent variable was the mean health care expenditure per visit for the same 4 care categories. To address the issue of individuals with no observed health care visits, and therefore no telehealth use or associated expenditures, Heckman selection models were employed in both analyses. This 2-stage modeling approach accounts for potential selection bias due to zero utilization in the observed 2-year period.

The relationship between food insecurity and travel time to the usual place of care is illustrated in Figure 1. The figure displays the differences in the predicted probabilities across travel time categories by food insecurity status. Food-insecure individuals were 2.4 percentage points less likely than their food-secure counterparts to reside within 15 minutes of their usual source of care; on the other hand, they were 1.7 percentage points more likely to live at least 30 minutes away. Both differences are statistically significant at the 90% confidence level. These findings from the pooled data align with the trends observed in the unbalanced panel data.

To examine telehealth utilization, 3 sequential models were estimated, each introducing an additional explanatory variable. Across all models, food insecurity status remained the primary independent variable of interest. Model 1 estimated telehealth utilization without controlling for whether an individual had a usual place of care. Model 2 added an indicator for having a usual place of care, and model 3 further included travel time to the usual place of care, retaining the indicator variable for individuals without a usual source of care. Table 3 presents the estimated coefficients of key independent variables from the second-stage regression. Corresponding first-stage regression results are shown in eAppendix Table 4. In the pooled data, 5192 of 13,420 individuals did not report a single office-based or outpatient visit for the 4 care types. As a result, the second stage of the Heckman model was estimated using data from the remaining 8228 individuals who had at least 1 relevant visit. The number of observations decreased in subsequent models due to missing data on the usual place of care and travel time variables. Inverse Mills ratios are reported in Table 3. They are not statistically significant in any model, suggesting that the null hypothesis, which is that the errors in the selection and outcome models are uncorrelated, cannot be rejected.

Food insecurity was found to be a statistically significant predictor of greater telehealth utilization. In the third model, which includes all covariates, food insecurity was associated with a 1.7–percentage point increase in the share of telehealth visits. Additionally, in the third model, lacking a usual place of care was linked to an approximately 7–percentage point decrease in the share of telehealth visits, suggesting that having an established source of care may facilitate greater engagement with telehealth services. This analysis also examined the interaction between food insecurity and travel time to the usual place of care in relation to telehealth utilization. The results are presented in Figure 2. Among individuals with a usual place of care, there was a monotonic increase in the share of telehealth visits among food-insecure individuals relative to food-secure individuals as travel time increased. Specifically, food-insecure individuals with a travel time of 15 to 30 minutes had a 2.6–percentage point higher share of telehealth visits. This difference increased to 4 percentage points for those with travel times exceeding 30 minutes. Both differences are statistically significant at the 90% confidence level, highlighting that longer travel times may amplify reliance on telehealth among food-insecure individuals.

To analyze mean health care expenditures per visit, the share of telehealth visits was included as a key explanatory variable, while retaining all other covariates from previous models. Model 1 found that a 1–percentage point increase in the share of telehealth visits was associated with a $117.64 decrease in health care expenditures per visit, a result that is statistically significant at the 95% confidence level. This substantial reduction in mean health care expenditures associated with increased telehealth utilization is interpreted as follows: Outpatient expenditures are generally higher than office-based expenditures because they include both provider payments and facility charges. Although the public-use MEPS files do not disaggregate office-based costs into provider and facility components, the raw data suggest that expenditures are more comparable across the 2 settings when care is delivered via telehealth. This may be due to the reduced incidence of facility charges in telehealth-related expenditures. In model 2, the indicator for not having a usual place of care was not significantly associated with health care expenditures. Additionally, although not shown in Table 3, there was no evidence of a significant relationship between travel time to usual care and health care spending.

DISCUSSION

These findings highlight important disparities in primary and preventive health care utilization based on food security status and access to a usual source of care. Individuals experiencing food insecurity or lacking a usual place of care consistently reported fewer office-based and outpatient visits, even after controlling for individual-level time-invariant factors. Although there was a weak, yet positive, association between longer travel time to care settings and food insecurity, it did not appear to independently affect the number of annual visits. This suggests that broader structural barriers—such as food insecurity, continuity of care, and the interplay between food insecurity and geographic accessibility—are more critical determinants of health care utilization than geographic accessibility alone. Telehealth technology could be one means to strengthen connections to regular care providers and play a key role in improving access to essential health services. This is supported by these study results. The finding that a larger share of telehealth utilization was associated with a decrease in health care expenditure is notable. Future research will explore the relationship between telehealth utilization and total cost of care.

The strong association between having a usual source of care and both increased health care utilization and telehealth adoption reinforces the importance of promoting continuous, patient-centered care relationships. Telehealth, although not a complete substitute for in-person care, presents a valuable tool for enhancing health care access—especially for individuals facing socioeconomic and transportation barriers. Tailoring telehealth strategies to better serve food-insecure populations, such as offering mobile health units or hybrid models of care, could further amplify their impact.³³ Continued investment in telehealth infrastructure, including broadband expansion in rural and low-income areas and policies that support reimbursement parity and provider training, will be critical for sustaining telehealth access.

Limitations

These results have important limitations. First, unmeasured confounding may bias findings despite adjusting for observed covariates, so associations between food insecurity and telehealth use are not causal. Second, food insecurity is episodic, and a single assessment may misclassify some individuals; assigning the reference person’s status to all family members may also introduce measurement error. Third, having a usual place of care as a proxy for access may not fully capture preventive service use or health care engagement relevant to telehealth utilization.

CONCLUSIONS

The temporary pandemic-era expansions in telehealth coverage demonstrated the potential to reduce barriers for underserved populations, including those experiencing food insecurity. Making such policies permanent, such as enacting reimbursement parity for telehealth services, could help sustain telehealth use among low-income and rural populations. Additionally, given the intersection of food insecurity and health care access, integrated approaches that link Medicaid beneficiaries to food assistance programs (such as through Medicaid managed care organizations or accountable care organizations) may improve both health and nonhealth outcomes. Policy makers should also consider supporting state-level Medicaid waivers and pilot programs that incorporate telehealth into care models for vulnerable populations, including those who are dually eligible. As federal and state agencies revisit telehealth regulations post pandemic, cross-sector coordination will be essential to ensure that access to care is not only preserved but also expanded for those most in need.

Author Affiliation: School of Economic Sciences, Washington State University, Pullman, WA.

Source of Funding: This work was supported by the US Department of Agriculture National Institute of Food and Agriculture, Hatch/Multistate Project 0844: “Household Financial and Health Decision-Making Under Economic Uncertainties.”

Author Disclosures: The author reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design; acquisition of data; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript for important intellectual content; and statistical analysis.

Address Correspondence to: Bidisha Mandal, PhD, Washington State University, 101 Hulbert Hall, PO Box 646210, Pullman, WA 99164-6210. Email: bmandal@wsu.edu.

REFERENCES

1. Benavidez GA, Zahnd WE, Hung P, Eberth JM. Chronic disease prevalence in the US: sociodemographic and geographic variations by zip code tabulation area. Prev Chronic Dis. 2024;21:E14. doi:10.5888/pcd21.230267

2. Fast facts: health and economic costs of chronic conditions. CDC. July 12, 2024. Accessed February 10, 2025. https://www.cdc.gov/chronic-disease/data-research/facts-stats

3. Dean EB, French MT, Mortensen K. Food insecurity, health care utilization, and health care expenditures. Health Serv Res. 2020;55(suppl 2):883-893. doi:10.1111/1475-6773.13283

4. Ejiohuo O, Onyeaka H, Unegbu KC, et al. Nourishing the mind: how food security influences mental wellbeing. Nutrients. 2024;16(4):501. doi:10.3390/nu16040501

5. Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health Aff (Millwood). 2015;34(11):1830-1839. doi:10.1377/hlthaff.2015.0645

6. Laraia BA. Food insecurity and chronic disease. Adv Nutr. 2013;4(2):203-212. doi:10.3945/an.112.003277

7. Leung CW, Kullgren JT, Malani PN, et al. Food insecurity is associated with multiple chronic conditions and physical health status among older US adults. Prev Med Rep. 2020;20:101211. doi:10.1016/j.pmedr.2020.101211

8. Nagata JM, Palar K, Gooding HC, Garber AK, Bibbins-Domingo K, Weiser SD. Food insecurity and chronic disease in US young adults: findings from the national longitudinal study of adolescent to adult health. J Gen Intern Med. 2019;34(12):2756-2762. doi:10.1007/s11606-019-05317-8

9. Seligman HK, Laraia BA, Kushel MB. Food insecurity is associated with chronic disease among low-income NHANES participants. J Nutr. 2010;140(2):304-310. doi:10.3945/jn.109.112573

10. Stuff JE, Casey PH, Szeto KL, et al. Household food insecurity is associated with adult health status. J Nutr. 2004;134(9):2330-2335. doi:10.1093/jn/134.9.2330

11. Gregory CA, Reed-Jones M. Predicted prevalence of five chronic diseases increased as household food security worsened. US Department of Agriculture Economic Research Service. January 8, 2024. Accessed February 10, 2025. https://www.ers.usda.gov/data-products/charts-of-note/chart-detail?chartId=108211

12. Ward BW, Schiller JS, Goodman RA. Multiple chronic conditions among US adults: a 2012 update. Prev Chronic Dis. 2014;11:E62. doi:10.5888/pcd11.130389

13. Berkowitz SA, Seligman HK, Meigs JB, Basu S. Food insecurity, healthcare utilization, and high cost: a longitudinal cohort study. Am J Manag Care. 2018;24(9):399-404.

14. Ghani F, Wang H, Manning SE, Sambamoorthi U. Interactive association of chronic illness and food insecurity with emergency department utilization among school-age children in the United States: a cross-sectional study. Health Sci Rep. 2023;6(2):e1123. doi:10.1002/hsr2.1123

15. Gupta N, Sheng Z. A population-based study of the association between food insecurity and potentially avoidable hospitalization among persons with diabetes using linked survey and administrative data. Int J Popul Data Sci. 2019;4(1):1102. doi:10.23889/ijpds.v4i1.1102

16. Peltz A, Garg A. Food insecurity and health care use. Pediatrics. 2019;144(4):e20190347. doi:10.1542/peds.2019-0347

17. Tarasuk V, Cheng J, de Oliveira C, Dachner N, Gundersen C, Kurdyak P. Association between household food insecurity and annual health care costs. CMAJ. 2015;187(14):E429-E436. doi:10.1503/cmaj.150234

18. Thorpe KE, Allen L, Joski P. The role of chronic disease, obesity, and improved treatment and detection in accounting for the rise in healthcare spending between 1987 and 2011. Appl Health Econ Health Policy. 2015;13(4):381-387. doi:10.1007/s40258-015-0164-7

19. What is telehealth? HHS. Updated March 27, 2020. Accessed March 2, 2025. https://www.hhs.gov/hipaa/for-professionals/faq/3015/what-is-telehealth/index.html

20. Baker LC, Johnson SJ, Macaulay D, Birnbaum H. Integrated telehealth and care management program for Medicare beneficiaries with chronic disease linked to savings. Health Aff (Millwood). 2011;30(9):1689-1697. doi:10.1377/hlthaff.2011.0216

21. Omboni S, Ballatore T, Rizzi F, Tomassini F, Panzeri E, Campolo L. Telehealth at scale can improve chronic disease management in the community during a pandemic: an experience at the time of COVID-19. PLoS One. 2021;16(9):e0258015. doi:10.1371/journal.pone.0258015

22. Polisena J, Coyle D, Coyle K, McGill S. Home telehealth for chronic disease management: a systematic review and an analysis of economic evaluations. Int J Technol Assess Health Care. 2009;25(3):339-349. doi:10.1017/S0266462309990201

23. Williams C, Shang D. Telehealth for chronic disease management among vulnerable populations. J Racial Ethn Health Disparities. 2024;11(2):1089-1096. doi:10.1007/s40615-023-01588-4

24. Baek D. The effect of public transportation accessibility on food insecurity. East Econ J. 2016;42:104-134. doi:10.1057/eej.2014.62

25. Buzza C, Ono SS, Turvey C, et al. Distance is relative: unpacking a principal barrier in rural healthcare. J Gen Intern Med. 2011;26(suppl 2):648-654. doi:10.1007/s11606-011-1762-1

26. Mui Y, Headrick G, Raja S, Palmer A, Ehsani J, Porter KP. Acquisition, mobility and food insecurity: integrated food systems opportunities across urbanicity levels highlighted by COVID-19. Public Health Nutr. 2022;25(1):114-118. doi:10.1017/S1368980021002755

27. Rajasooriar D, Soma T. Food access, mobility, and transportation: a survey and key informant interviews of users of non-profit food hubs in the City of Vancouver before and during the COVID-19 crisis. BMC Public Health. 2022;22(1):6. doi:10.1186/s12889-021-12434-9

28. Arcury TA, Preisser JS, Gesler WM, Powers JM. Access to transportation and health care utilization in a rural region. J Rural Health. 2005;21(1):31-38. doi:10.1111/j.1748-0361.2005.tb00059.x

29. Syed ST, Gerber BS, Sharp LK. Traveling towards disease: transportation barriers to health care access. J Community Health. 2013;38(5):976-993. doi:10.1007/s10900-013-9681-1

30. Wolfe MK, McDonald NC, Holmes GM. Transportation barriers to health care in the United States: findings from the National Health Interview Survey, 1997-2017. Am J Public Health. 2020;110(6):815-822. doi:10.2105/AJPH.2020.305579

31. Phan M, Hickey E, Beck AF, Klein MD, Real FJ. A “smart” way of addressing food insecurity in the digital age. Pediatrics. 2018;142(1):e20181336. doi:10.1542/peds.2018-1336

32. Blewett LA, Drew JA, Backman D, et al. Medical Expenditure Panel Survey (MEPS): version 2.4. IPUMS Center for Data Integration. 2024. doi:10.18128/D071.V2.4

33. Nguyen CJ, Gold R, Mohammed A, et al. Food insecurity screening in primary care: patterns during the COVID-19 pandemic by encounter modality. Am J Prev Med. 2023;65(3):467-475. doi:10.1016/j.amepre.2023.03.014