Cross-Sectional Analysis of Insurance Claims Shows Health Care Utilization and Costs Associated With Health-Related Social Needs (HRSNs) in an Urban Population

Abstract

Introduction:

Unmet health-related social needs (HRSNs) contribute to higher health care spending. Health systems and governments are increasingly addressing these needs to improve outcomes and reduce costs.

Objective:

This study examined differences in health care costs and utilization among patients who screened positive versus negative for HRSNs within a large urban health system.

Methods:

This retrospective, cross-sectional study included adult Medicaid patients (ages 21-65) who completed a 10-item HRSN screening between April 2018 and December 2019. Each patient’s first screening was linked to insurance claims for the following 12 months. Cost and utilization outcomes were analyzed using linear regressions adjusted for sociodemographic covariates.

Results:

Of 4432 patients, 1194 (26.9%) screened positive for at least 1 HRSN. Their mean annual health care cost was $8169.72, compared with $4393.22 among those without HRSNs, a difference of $3776.50 (P < .005). Patients with HRSNs also had higher utilization: 3.65 more specialty visits, 1.08 more mental health visits, 0.86 more emergency visits, and 0.40 additional inpatient days annually (P < .005).

Conclusions:

Unmet HRSNs were linked to significantly higher costs and utilization, highlighting the potential of addressing social needs to improve outcomes and reduce health care spending.

Keywords

cost utilization health equity health-related social needs social determinants of health Medicaid

Introduction

Social determinants of health (SDOH) are the “conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life” as elucidated by the World Health Organization (WHO).¹ According to the County Health Rankings (CHR) model, approximately 80% of addressable health outcomes are associated with SDOH.² In other words, while clinical care contributes only 20% to health outcomes such as length of life and quality of life, the other 80% is contributed by socioeconomic factors (education, employment, income, family and social support, and community safety), health behaviors (tobacco use, diet and exercise, alcohol and drug use, and sexual activity), and physical environment (air and water quality, housing, and transit).² In recent years, there has been a growing body of evidence indicating that unmet socioeconomic needs worsen health outcomes.³ This is especially relevant in Medicaid populations due to their complex clinical, behavioral health, and social needs.⁴ Thus, in 2019, the National Academies of Sciences, Engineering, and Medicine (the National Academies) released recommendations for the Centers for Medicare & Medicaid Services to address social needs to improve health outcomes, reduce health disparities, and lower costs.⁵

Although the United States spends nearly twice as much on health care as any other high-income country, Americans continue to suffer worse health outcomes.⁶ U.S. national health expenditures have increased significantly in recent years, reaching $5.3 trillion in 2024.^7,8 Additionally, that same year, Medicaid spending grew by 6.6% and Medicare by 7.8%.⁷ Health-related social needs (HRSNs), the social risks an individual or household requests assistance with, account for 40% to 50% of the cost structure associated with Medicaid and Medicare programs.^9-11 Investments in targeting HRSNs with the goal of improving health outcomes and lowering costs have come into greater focus and favor. However, the extent to which HRSNs drive costs as well as utilization, especially in urban Medicaid populations, needs to be better understood. Furthermore, although insurance claims data based on encounters within health systems are systematically tracked in the electronic health record (EHR), HRSNs data are frequently incompletely documented or not documented at all.⁹ It is also uncommon for a health system to have access to all the Medicaid spending for a particular individual regardless of where the care is delivered, which makes studying the connection between HRSNs and health care spending especially challenging.

Some studies have measured the reduction in costs and utilization for patients with HRSNs after interventions have taken place, with the most notable being the Accountable Health Communities (AHC) Model.^12,13 This study found that for Medicaid and Medicare beneficiaries connected to community resources for their HRSNs, there was a reduction in total health care expenditures, inpatient stays, and ED visits.¹³ A more recent retrospective analysis of the AHC Model analyzed short-term acute care utilization and the resolution of individual HRSNs.¹⁴ In contrast, the present study uniquely quantifies both the overall health care cost burden and utilization across multiple care settings, including outpatient visits, associated with unmet HRSNs, providing population-level baseline data. Few studies have characterized health care costs associated with HRSNs across a population at baseline. One study with Medicaid patients from Washington, D.C., found that mean future health care spending based off claims data was significantly higher in the upper 2 social risk classes than in the lowest class (defined by the number of reported social adversities and employment status), which were determined using their own HRSN survey.¹⁵ Another study found that among Medicaid patients, those with social needs (determined using a validated screening tool) had significantly higher costs (determined using an engineered cost accounting approach) than those without any social needs.¹⁶ Furthermore, many studies have explored the relationship between health care utilization and HRSNs, often using their own surveys either to measure utilization or HRSNs.^17-20 For instance, a study of largely Medicaid insured patients found those having HRSNs had an increased likelihood of visiting the emergency department and an increased number of primary care visits, but fewer outpatient mental health visits.¹⁷

Although prior studies have demonstrated associations between social needs and health care utilization or costs, important gaps remain. Many analyses focus on post-intervention outcomes, rely on self-reported data, or lack access to complete insurance claims across multiple care settings.^14,15,17-21 Furthermore, few studies have quantified baseline differences in both health care costs and detailed utilization patterns using validated HRSNs screening data linked to comprehensive Medicaid claims. Establishing these baseline cost and utilization benchmarks is particularly important as Medicaid programs increasingly invest in large-scale social care initiatives. To address this gap, this study links a validated, system-wide HRSNs screening tool embedded in the EHR to 12 months of complete Medicaid claims data with the objective of examining differences in health care costs and utilization among adults who screened positive versus negative for unmet HRSNs within a large, urban safety-net health system. By characterizing population-level cost and utilization differences prior to intervention, this study provides actionable evidence for researchers, health system leaders, and policy makers seeking to design, target, and evaluate social care investments in Medicaid populations.

Methods

This retrospective cross-sectional study analyzed the electronic health record (EHR) and insurance claims of patients from a large urban health system in the Bronx, New York, in which 80% of the patient population is insured through Medicare or Medicaid.²² The health system, which has historically prioritized social medicine and targeting health disparities, has introduced standardized, systematic, and scalable HRSN assessments in clinical settings.²³

A 10-item HRSNs screening tool was adapted from a widely used, validated instrument, the Health Leads screening toolkit.²⁴ In 2018, this screening tool was integrated into the health system’s EHR and made accessible to all clinical practices. The clinical practices implemented the screener at their discretion, and it was designed to be self-administered by patients during clinical visits. The screener was made available in the 9 most common languages in the patient population and addressed housing insecurity and quality, food insecurity, utilities, health transportation, medications, child or elderly care, legal services, and family stress and safety. Microsoft SQL Server, version 18, was used to query HRSNs from the Epic Electronic Health Record Data Warehouse. Analysis of the HRSNs data was approved by the Albert Einstein College of Medicine Institutional Review Board (IRB).

Claims data were accessed through the Medicaid Data Warehouse by the Population Health Analytics team under a data use agreement with the New York State Department of Health (approved by the Albert Einstein College of Medicine IRB). This access enables the data team to review Medicaid claims for specific populations and perform analyses on aggregated data, ensuring that no protected health information is extracted or transmitted. Adult patients (aged 21-65) screened for HRSNs between April 10, 2018 and December 8, 2019 and enrolled with a major Medicaid managed care plan were included in the analysis. We restricted the study population to adults aged 21 to 65 to focus on a Medicaid-only population and avoid heterogeneity introduced by pediatric Medicaid policies and Medicare-Medicaid dual eligibility among adults aged ≥65, which differ in coverage, utilization patterns, and reimbursement structures. The study size was determined by the number of adult patients who met inclusion criteria, defined as those with continuous Medicaid enrollment and at least 1 completed HRSN screen during the study period. No a priori sample size calculation was conducted, as all eligible patients within the available EHR and claims data were included in the analysis. Patients with missing or invalid claims data were excluded from the analysis.

For patients screened multiple times, only the patient’s first HRSNs screen during this period was used to represent their baseline HRSNs status. The results of the first HRSNs screener administered per unique patient were then interrelated with claims in the 12 months following the date of screening for that patient. This ensured that the claims data was complete, which would be less likely if the data were extracted less than 1 year after the service was provided. Cost and utilization data extracted from the claims were then categorized based on setting (inpatient, outpatient, or emergency department) and outpatient services were further classified by type (PCP medical, medical specialty, and mental health).

The primary outcome in the analysis was annual health care cost with utilization as a secondary outcome, which were both determined using insurance claims from the 12-month period following their first HRSNs screen. The average annual cost per patient was calculated from the paid claims that represent the expenditures submitted to the payer. Health care utilization was measured by assessing the number of visits to outpatient services, number of visits to emergency departments (ED), and number of medical inpatient days on average annually per patient. Outpatient services included primary care physician (PCP) visits for medical needs, medical specialty visits, and Article 28 specialty mental health visits.

The primary predictor for this analysis was HRSNs status, which was defined as whether the patient self-reported at least 1 HRSN on the 10-item HRSNs screening tool. Patients were categorized as either positive or negative for HRSNs. For the patients included, sociodemographic characteristics were also collected and incorporated as covariates in the analyses. The covariates were age (categorized: 21-29, 30-44, and 45-65) and sex (categorized: male or female).

For the descriptive analysis, cost data were reported as unadjusted mean values, calculated as the mean cost for each subgroup defined by sex, age group, and HRSNs status. These values were computed by grouping the raw data and calculating simple averages, without any regression adjustment.

A 2-step analytic approach was then used for both cost and utilization. A simple linear regression was applied to assess whether there was a significant difference in the mean annual cost per patient based on HRSNs status.^17,25,26 The association between annual cost per patient and HRSNs status was also analyzed with multiple linear regression, adjusted for age and sex.²⁶ To assess utilization, PCP medical, medical specialty visits, mental health visits, emergency department visits, and medical inpatient days were analyzed separately. For each form of utilization, a simple linear regression was fit to determine whether there was a significant difference in the mean number of visits or days per patient by HRSNs status. Multiple linear regression was also performed for each type of utilization to analyze the association between the number of visits or days per patient and HRSNs status, adjusted for age and sex.

Adjusted mean outcomes were reported by subgroup, including sex (male or female), age group, and HRSNs status. These values were obtained by running linear regression models with the relevant outcome variable as the dependent variable, which was annual cost, visits, or days. The models included HRSNs status (0 for Negative HRSNs and 1 for Positive HRSNs) as the primary predictor and sex (coded as 0 for Male and 1 for Female) and age group (21-29, 30-44, and 45-65) as the covariates. The reported mean values reflect the predicted mean outcomes for HRSNs status after adjusting for the covariates. P values and adjusted means were also reported separately for members with Negative HRSNs and Positive HRSNs.

All P values less than .05 were considered to be statistically significant. Statistical analyses were performed using Python 3.9 on Google Colab.

Several measures were implemented to reduce potential sources of bias. Selection bias was minimized by including all adult patients with continuous Medicaid enrollment and at least 1 completed HRSN screen during the study period. Information bias was reduced through the use of a standardized, validated screening tool, and routinely audited EHR and claims data. Misclassification bias was limited by using only the first completed HRSN screen per patient to define baseline status. Confounding was addressed by adjusting for age and sex in multivariable models and restricting analyses to a 12-month claims window to ensure complete follow-up.

We used the Supplemental STROBE cross sectional checklist when writing our report.²⁷

Results

Four thousand four hundred sixty-two adult patients (aged 21-65) were screened for HRSNs between April 10, 2018 and December 8, 2019, enrolled with a major Medicaid managed care plan, and had claims from the 12-month period following their first HRSNs screen. Thirty patients were excluded due to either missing or negative (likely due to data entry errors or refunds) values in the claims data. There were 4432 unique patients included in the final analysis. Three thousand ninety-four (69.81%) of the patients were female and 1338 (30.19%) were male. One thousand and fifty (23.69%) patients were aged 21 to 29 years old, 1533 (34.59%) patients were aged 30 to 44 years old, and 1849 (41.72%) were aged 45 to 65 years old. The average annual health care cost of the study population was $5410.62 (95% CI: $4851.55, $5969.70; Table 1). The costs for males were greater than for females at $6089.01 (95% CI: $4923.96, $7254.05) with females spending $5117.25 (95% CI: $4494.56, $5739.95) on average annually (Table 1). Spending also increased with age with patients aged 21 to 29 years old spending $2556.26 (95% CI: $2115.36, $2997.16), patients aged 30 to 44 years old spending $3979.71 (95% CI: $3268.88, $4690.55), and patients aged 45 to 65 spending $8217.90 (95% CI: $7052.39, $9383.42; Table 1). Males also had more medical specialty, mental health, and ED visits and medical inpatient hospitalization days per patient than females on average (Tables 2 and 3). Outpatient visits, ED visits, and inpatient days increased with age as well (Tables 2 and 3).

Table 1.

Demographic Characteristics of Patients Screened for HRSNs.

	Total		Negative HRSNs		Positive HRSNs
Sociodemographic characteristics	No. (%)	Average annual cost (95% CI)	No. (%)	Average annual cost (95% CI)	No. (%)	Average annual cost (95% CI)
Total	4432 (100.00)	$5410.62 ($4851.55, $5969.70)	3238 (73.06)	$4393.22 ($3859.66, $4926.78)	1194 (26.94)	$8169.72 ($6691.70, $9647.73)
Sex
Male	1338 (30.19)	$6089.01 ($4923.96, $7254.05)	999 (30.85)	$4437.73 ($3543.79, $5331.67)	339 (28.39)	$10 955.17 ($7219.42, $14 690.91)
Female	3094 (69.81)	$5117.25 ($4494.56, $5739.95)	2239 (69.15)	$4373.36 ($3712.38, $5034.34)	855 (71.61)	$7065.31 ($5628.24, $8502.39)
Age, in years
21-29	1050 (23.69)	$2556.26 ($2115.36, $2997.16)	787 (24.31)	$2305.87 ($1831.83, $2779.91)	263 (22.03)	$3305.51 ($2262.37, $4348.64)
30-44	1533 (34.59)	$3979.71 ($3268.88, $4690.55)	1110 (34.28)	$3052.74 ($2434.33, $3671.15)	423 (35.43)	$6412.20 ($4423.72, $8400.68)
45-65	1849 (41.72)	$8217.90 ($7052.39, $9383.42)	1341 (41.41)	$6727.79 ($5589.42, $7866.17)	508 (42.55)	$12 151.44 ($9175.90 $15 126.99)

Abbreviations: CI, confidence interval; HRSNs, health-related social needs.

Table 2.

Predicted Outpatient Provider Visits Per Patient on Average Annually.

	PCP medical visits				Medical specialty visits				Mental health visits
Sociodemographic characteristics	Total	Negative HRSN	Positive HRSN	P value	Total	Negative HRSN	Positive HRSN	P value	Total	Negative HRSN	Positive HRSN	P value
Single linear regression
Total	4.96	4.12	5.37	.0042	5.48	4.39	8.10	.0029	1.86	1.49	2.55	.0041
Multiple linear regression
Total	4.42	4.01	5.20	.0023	5.49	4.39	8.04	.0021	1.89	1.47	2.55	.003
Sex
Male	4.61	3.99	5.19	.00012	5.83	4.27	7.92	.0016	1.97	1.44	2.44	.00047
Female	4.20	4.10	5.27	.00098	5.07	4.51	8.16	.001	1.77	1.52	2.59	.00059
Age, in years
21-29	3.99	3.84	5.12	.0071	4.21	3.89	6.93	.0068	1.49	1.30	2.16	.0089
30-44	4.36	4.04	5.23	.002	5.40	4.49	8.08	.0023	1.91	1.47	2.58	.0028
45-65	4.68	4.16	5.29	.0036	6.12	4.60	8.20	.0035	2.21	1.52	2.64	.0042

Abbreviations: HRSNs, health-related social needs; PCP, primary care provider.

Table 3.

Predicted ER Visits and Inpatient Days Per Patient on Average Annually.

	ER visits				Inpatient days
Sociodemographic characteristics	Total	Negative HRSNs	Positive HRSNs	P value	Total	Negative HRSNs	Positive HRSNs	P value
Single linear regression
Total	2.07	1.66	2.43	.0038	0.89	0.73	1.12	.0052
Multiple linear regression
Total	2.08	1.73	2.59	.0026	0.90	0.72	1.13	.0034
Sex
Male	2.17	1.71	2.55	.00027	0.96	0.71	1.09	.00037
Female	1.95	1.75	2.62	.00033	825	0.73	1.15	.00056
Age, in years
21-29	1.82	1.62	2.50	.0075	0.76	0.70	1.02	.0073
30-44	2.04	1.74	2.58	.0024	0.87	0.72	1.12	.032
45-65	2.33	1.77	2.63	.0039	1.01	0.74	1.17	.0048

Abbreviations: ER, emergency room; HRSNs, health-related social needs.

Of these 4432 patients, 1194 (26.94%) self-reported at least 1 HRSN and 3238 (73.06%) self-reported no HRSNs. The average annual cost per patient was $8169.72 (95% CI: $6691.70, $9647.73) for those with HRSNs and $4393.22 (95% CI: $3859.66, $4926.78) for those without HRSNs (Table 1). Those with HRSNs had $3505.98 (P < .005) more health care costs each year (Table 4). After adjusting for sociodemographic factors, it cost $3776.50 (P < .005) more for those with HRSNs (Table 4). This difference between HRSNs status groups increased with age: $3639.12 (P < .001) for ages 21 to 29 years old, $3703.85 (P < .05) for ages 30 to 44 years old, and $3725.13 (P < .005) for ages 45 to 65 years old (Table 4). Females also had a much greater difference at $4122.11 (P < .001) compared to males with a difference of $3671.84 (P < .005; Table 4).

Table 4.

Predicted Average Annual Cost Per Patient.

Simple linear regression
	Total	Negative HRSNs	Positive HRSNs	P value
Total	$5248.30	$4173.56	$7679.53	.0031
Multiple linear regression
	Total	Negative HRSNs	Positive HRSNs	P value
Total	$5410.62	$4393.22	$8169.72	.0021
Sex
Male	$6089.01	$4102.57	$7774.41	.0013
Female	$5117.25	$5044.62	$9166.72	.00074
Age, in years
21-29	$2556.26	$1644.74	$5283.86	.00081
30-44	$3979.71	$2957.71	$6661.56	.017
45-65	$8217.90	$7194.45	$10 919.58	.0034

Abbreviation: HRSNs, health-related social needs.

Patients who screened positive for at least 1 HRSN had 3.72 (P < .005) more medical specialty visits, 1.06 (P < .005) more mental health visits, 0.78 (P < .005) more ED visits, and 0.39 (P < .01) more medical inpatient hospitalization days per patient on average annually than patients who screened negative for any HRSNs (Tables 2 and 3). After adjusting for sociodemographic covariates, those with HRSNs had 3.65 (P < .005) more medical specialty visits, 1.08 (P < .005) more mental health visits, 0.86 (P < .005) more ED visits, and 0.40 (P < .005) more medical inpatient hospitalization days per patient on average per year (Tables 2 and 3). The differences in medical specialty visits, mental health visits, ED visits, medical inpatient hospitalization days between HRSN status groups (P < .05) were similar across the sociodemographic groups (sex and age).

Discussion

This study provides one of the few population-level analyses linking validated HRSN screening data to comprehensive Medicaid claims to quantify baseline differences in health care costs and utilization. By establishing these cost and utilization benchmarks prior to widespread social care interventions, this study’s findings offer critical context for evaluating the impact of emerging Medicaid social care policies and investments. For instance, the present study found that HRSNs were associated with higher annual health care costs and utilization in a Medicaid population. Patients who screened positive for at least 1 HRSN had almost double the cost compared to patients who screened negative for any HRSNs. This study also revealed that patients with unmet HRSNs had increased PCP medical, medical specialty outpatient, mental health outpatient, and emergency department visits as well as medical inpatient days. Furthermore, those with unmet HRSNs had nearly twice the medical specialty outpatient visits per patient than those with no HRSNs. These findings have several relevant implications. This study suggests that unmet HRSNs are associated with significantly higher health care costs. An open question is whether meeting these needs would ultimately reduce expenditures, particularly in Medicaid populations, as those with HRSNs likely share a host of underlying factors that may play a confounding role in increasing costs. Important data will be gathered to answer this question with the implementation of the New York Medicaid 1115 Waiver, a $7 billion statewide investment over a 3-year period through March 2027, for Social Care Networks to provide HRSNs interventions to Medicaid members.²⁸ Moreover, incorporating screening for and addressing HRSNs in clinical settings could identify high-risk patients earlier and allow for targeted social interventions that may prevent avoidable health care utilization.¹³ However, it is important to note that increased PCP visits for those with HRSNs is not an issue that necessitates resolution, but rather an appropriate and desirable finding for a vulnerable population that may benefit from this more frequent level of care. The data from this study also support recent policy initiatives, such as those by the National Academies of Sciences, Engineering, and Medicine (NASEM), that encourage integrating social care into health care delivery to improve outcomes and reduce costs.⁵

The findings from this study are consistent with previous studies showing that unmet HRSNs are associated with higher health care expenditures.^15,16,29,30 For instance, a recent cross-sectional study in a large integrated health system found higher costs among patients with severe HRSNs. In contrast, this study evaluates a Medicaid-only population and examines both total costs and utilization across care settings.³⁰ The present study also aligns with earlier studies showing that HRSNs are associated with increased health care utilization patterns.^14,17-21 However, many prior studies have focused on intervention effects, self-reported data, or limited subsets of services, whereas the current analysis leverages comprehensive Medicaid claims to capture system-wide utilization and costs.^14,15,17-21 Moreover, 1 study in Oregon found that social risks were associated with higher odds of inpatient admissions, emergency department visits, and mental health visits during a 1-year period as well.¹⁸ However, another study in California found that those having HRSNs had an increased likelihood of visiting the emergency department and an increased number of primary care visits, but a decreased number of outpatient mental health visits.¹⁷ It’s possible that because these studies were conducted in different U.S. states, patients, particularly those with HRSNs, may differ in how they prioritize mental health and in their access to related resources since mental health is often underfunded,³¹ resulting in the differing change in mental health visits compared to those with no HRSNs. These variations underscore the importance of context-specific, claims-based analyses such as the present study to inform local policy and program design.

The current study is one of few to combine insurance claims data with a validated screening tool to analyze the association between HRSNs and both health care costs as well as specific health care utilization patterns across multiple care settings. This study also provides costs associated with HRSNs across a population at baseline, unlike other studies.¹² As such, the findings serve as a reference point for health systems and policy makers seeking to evaluate the effectiveness of emerging social care investments over time. The findings focus on a Medicaid population, providing valuable insight into an especially vulnerable population that has in recent years required increased investment to improve health outcomes, reduce health disparities, and lower costs.⁷ Although there is growing evidence suggesting that supportive interventions targeting HRSNs can impact both health outcomes and health care costs,³² there is still a need to elucidate what the most appropriate and efficient interventions would be.

The overall increased health care expenditures and utilization among patients with unmet HRSNs compared to those with no HRSNs are associated with poorer health outcomes, as other studies have shown, for instance, that patients with transportation needs were 84% more likely to have an alcohol/drug use disorder diagnosis.^17,33 The results from this study suggest that addressing HRSNs might lead to better health outcomes and optimized spending and utilization.^32,34 According to the Office of Health Policy April 2022 report, there is strong evidence to support that housing and nutrition interventions can both improve health outcomes and reduce costs.³⁵ Supporting social and economic mobility has also been associated with better health outcomes.³⁵ Specifically, addressing transportation needs by providing non-emergency medical transportation has proven to be cost-effective by promoting greater use of preventive and outpatient care while reducing reliance on more expensive health care services.^35,36 In example, 2 studies estimated that offering non-emergency medical transportation to patients with chronic conditions could lower health care utilization and costs by enhancing disease management, particularly for populations such as those undergoing dialysis and receiving diabetes wound care.³⁷ Furthermore, in a randomized controlled trial conducted in 5 major U.S. cities, low-income participants who were given vouchers to relocate from high-poverty to low-poverty neighborhoods experienced reduced rates of obesity and diabetes.³⁸ These findings highlight the importance of integrating social care with medical care to address HRSNs as a strategy for reducing health disparities and avoidable health care costs.

The current study has several limitations. The study population is limited to adult patients (aged 21-65) enrolled in a specific major Medicaid managed care plan serving individuals in the greater New York City metropolitan area, therefore results may not be generalizable. Specifically, restricting the study population to adults aged 21 to 65 may introduce selection bias and limit generalizability to pediatric populations and older adults, but this restriction was necessary to reduce heterogeneity in insurance structure and utilization patterns and to improve internal validity. While the Medicaid population has greater enrollment of females, this study population had an even higher percentage of female participants. This study also limited its covariates to age and sex. The HRSNs data were collected at the discretion of each clinical practice and thus, subject to selection bias. Due to the cross-sectional design of the study, any causal relationships could not be identified and it is unknown when the patients in this study first experienced or for how long they have been experiencing the unmet HRSNs identified by the screening tool. Since the cost and utilization data were collected from insurance claims a year after they first completed the screening, it is uncertain if the patients still had unmet HRSNs during this period or were exposed to interventions after screening. The authors were also unable to control for disease acuity. Lastly, the current study does not identify specific types of or the number of unmet HRSNs and their relation to costs or utilization.

Future research should explore the differential impact by type and number of HRSNs, incorporate clinical comorbidities and disease severity, and examine longitudinal changes in costs and utilization following targeted social care interventions. As Medicaid programs increasingly invest in social care infrastructure, linking screening data to long-term outcomes will be essential for evaluating effectiveness and equity.

Conclusions

This study found that patients who screened positive for unmet HRSNs incurred nearly twice the care costs and had more outpatient visits, emergency department visits, and medical inpatient hospitalization days compared to those who screened negative. These findings reinforce the understanding that unmet HRSNs are associated with increased health care utilization and expenditures, suggesting that addressing HRSNs may contribute to improved health outcomes and more efficient use of health care resources. Health care systems may benefit from investing in community partnerships, care coordination, and social service referrals to better manage patients with unmet HRSNs. To illustrate this point, the current study’s health system has recently invested in a workforce of community health workers (CHWs). However, further research is needed to determine how best to allocate resources like CHWs to target specific HRSNs. Since unmet HRSNs often reflect broader structural inequities, addressing them is essential not only for cost and care efficiency, but also for reducing health disparities.

Supplemental Material

sj-docx-1-jpc-10.1177_21501319261420562 – Supplemental material for Cross-Sectional Analysis of Insurance Claims Shows Health Care Utilization and Costs Associated With Health-Related Social Needs (HRSNs) in an Urban Population

Supplemental material, sj-docx-1-jpc-10.1177_21501319261420562 for Cross-Sectional Analysis of Insurance Claims Shows Health Care Utilization and Costs Associated With Health-Related Social Needs (HRSNs) in an Urban Population by Shimrani L. Banik, Samantha Levano, Kevin P. Fiori, Allison B. Stark, Scott Wetzler and Howard L. Forman in Journal of Primary Care & Community Health

Footnotes

Acknowledgements

The authors thank the data analytics team who helped with this study, including Christopher Viviano and Erjon Brucaj.

Author’s Note

Scott Wetzler is now affiliated with Montefiore Health System, Bronx, NY, USA.

ORCID iDs

Shimrani L. Banik

Samantha Levano

Kevin P. Fiori

Allison B. Stark

Scott Wetzler

Howard L. Forman

Ethical Considerations

This study received ethical approval from the Albert Einstein College of Medicine IRB (approval #2017-8421) on August 16, 2022. This study obtained a waiver of informed consent since this was a retrospective analysis of de-identified patient data.

Consent to Participate

This study obtained a waiver of informed consent.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Albert Einstein College of Medicine Research Fellowship Program.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The de-identified HRSN screener datasets used and/or analyzed during the current study are available from the corresponding author* on reasonable request. However, de-identified medical claims data cannot be shared with third parties due to restrictions outlined in our data use agreement with the New York State Department of Health.

Supplemental Material

Supplemental material for this article is available online.

References

World Health Organization. Social determinants of health. World Health Organization. 2025. Accessed March 1, 2025. https://www.who.int/health-topics/social-determinants-of-health#tab=tab_1

Hood

Gennuso

Swain

Catlin

BB.

County health rankings: relationships between determinant factors and health outcomes. Am J Prev Med. 2016;50(2):129-135. doi:10.1016/j.amepre.2015.08.024

World Health Organization. Closing the gap in a generation: health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. World Health Organization; 2008.

Spencer

Freda

McGinnis

Gottlieb

Measuring Social Determinants of Health Among Medicaid Beneficiaries: Early State Lessons. Center for Health Care Strategies; 2016;

National Academies of Sciences E, Medicine. Integrating Social Care into the Delivery of Health Care: Moving Upstream to Improve the Nation’s Health. The National Academies Press; 2019:194.

Munira

Gunja

EDG

Reginald

Williams II. U.S. Health Care From a Global Perspective, 2022: Accelerating Spending, Worsening Outcomes. Commonwealth Fund; 2023. doi:10.26099/8ejy-yc74

Centers for Medicare and Medicaid Services. National Health Expenditure (NHE) Fact Sheet. 2024. Accessed February 1, 2026. https://www.cms.gov/data-research/statistics-trends-and-reports/national-health-expenditure-data/nhe-fact-sheet

Mohan

Gaskin

DJ.

Social determinants of health and US health care expenditures by insurer. JAMA Network Open. 2024;7(10):e2440467. doi:10.1001/jamanetworkopen.2024.40467

Goldberg

Nash

DB.

For profit, but socially determined: the rise of the SDOH industry. Popul Health Manag. 2021;25(3):392-398. doi:10.1089/pop.2021.0231

10.

McGinnis

Williams-Russo

Knickman

JR.

The case for more active policy attention to health promotion. Health Aff. 2002;21(2):78-93. doi:10.1377/hlthaff.21.2.78

11.

Houlihan

Leffler

Assessing and addressing social determinants of health: a key competency for succeeding in value-based care. Prim Care. 2019;46(4):561-574. doi:10.1016/j.pop.2019.07.013

12.

Taylor

Tan

Coyle

, et al. Leveraging the social determinants of health: what works? PLoS ONE. 2016;11(8):e0160217. doi:10.1371/journal.pone.0160217

13.

RTI International. Accountable Health Communities (AHC) Model Evaluation Third Evaluation Report. 2024. Accessed October 28, 2025. https://www.cms.gov/priorities/innovation/data-and-reports/2024/ahc-3rd-eval-report

14.

Sidebottom

Martins

Vacquier

Dechaine

Behrens

Health-related social needs and health care utilization in the accountable health communities model. JAMA Netw Open. 2025;8(12):e2548036. doi:10.1001/jamanetworkopen.2025.48036

15.

McCarthy

Elmi

Wilder

Zheng

Zeger

SL.

Social determinants of health influence future health care costs in the medicaid cohort of the district of Columbia study. Milbank Q. 2022;100(3):761-784. doi:10.1111/1468-0009.12582

16.

Fiori

Levano

Colman

, et al. Signals in health inequity: examining social needs and costs in a large health system. J Ambul Care Manage. 2025;48(1):39-51. doi:10.1097/jac.0000000000000515

17.

Jones

Roth

Vartanian

KB.

Health and health care use strongly associated with cumulative burden of social determinants of health. Popul Health Manag. 2021;25(2):218-226. doi:10.1089/pop.2021.0255

18.

Tucher

Steele

Uratsu

McCloskey

Grant

RW.

Social risks and health care use in medically complex patients. JAMA Netw Open. 2024;7(9):e2435199. doi:10.1001/jamanetworkopen.2024.35199

19.

Rhee

Marottoli

Cooney

Jr Fortinsky

RH.

Associations of social and behavioral determinants of health index with self-rated health, functional limitations, and health services use in older adults. J Am Geriatr Soc. 2020;68(8):1731-1738. doi:10.1111/jgs.16429

20.

Wray

Tang

López

Hoggatt

Keyhani

Association of social determinants of health and their cumulative impact on hospitalization among a national sample of community-dwelling US adults. J Gen Intern Med. 2022;37(8):1935-1942. doi:10.1007/s11606-021-07067-y

21.

Abbott

Taylor

Vargas-Torres

, et al. Are unmet health related social needs associated with emergency department utilization among Medicare beneficiaries? BMC Health Serv Res. 2025;25(1):477. doi:10.1186/s12913-025-12554-7

22.

Chase

Montefiore Medical Center: Integrated Care Delivery for Vulnerable Populations. The Commonwealth Fund; 2010:53.

23.

Fiori

Rehm

Sanderson

, et al. Integrating social needs screening and community health workers in primary care: the community linkage to care program. Clin Pediatr. 2020;59(6):547-556. doi:10.1177/0009922820908589

24.

Fiori

Heller

Flattau

, et al. Scaling-up social needs screening in practice: a retrospective, cross-sectional analysis of data from electronic health records from Bronx county, New York, USA. BMJ Open. 2021;11(9):e053633. doi:10.1136/bmjopen-2021-053633

25.

Frees

Jie

Rosenberg

MA.

Predicting the frequency and amount of health care expenditures. North Am Actuar J. 2011;15(3):377-392. doi:10.1080/10920277.2011.10597626

26.

Baihaqi

Dianingrum

Ramadhan

KAN

Hariguna

Linear regression method to model and forecast the number of patient visits in the hospital. Paper presented at: 2018 3rd International Conference on Information Technology, Information System and Electrical Engineering (ICITISEE); November 13-14, 2018; Yogyakarta, Indonesia. IEEE; 2018: 247-252.

27.

von Elm

Altman

Egger

Pocock

Gotzsche

Vandenbroucke

. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-349.

28.

Centers for Medicare and Medicaid Services. New York Medicaid 1115 Waiver Approval Letter. 2024. Accessed October 28, 2025. https://www.medicaid.gov/sites/default/files/2024-01/ny-medicaid-rdsgn-team-appvl-01092024.pdf

29.

Falconi

Johnson

Chi

Stephenson

Marc Overhage

Agrawal

Health related social needs and whole person health: Relationship between unmet social needs, health outcomes, and healthcare spending among commercially insured adults. Prev Med Rep. 2023;36:102491. doi:10.1016/j.pmedr.2023.102491

30.

Silverman

ARC

Chung

Gould

, et al. Health-related social needs and total healthcare cost: a cross-sectional study in a large integrated health system. J Gen Intern Med. 2025. Published online June 25, 2025. doi:10.1007/s11606-025-09647-8

31.

Mahomed

Addressing the problem of severe underinvestment in mental health and well-being from a human rights perspective. Health Hum Rights. 2020;22(1):35-49.

32.

Chaiyachati

Grande

Aysola

Health systems tackling social determinants of health: promises, pitfalls, and opportunities of current policies. Am J Manag Care. 2016;22(11):e393-e394.

33.

Heller

Rehm

Parsons

Chambers

Hollingsworth

Fiori

KP.

The association between social needs and chronic conditions in a large, urban primary care population. Prev Med. 2021;153:106752. doi:10.1016/j.ypmed.2021.106752

34.

Honeycutt

Khavjou

Tayebali

Dempsey

Glasgow

Hacker

Cost-effectiveness of social determinants of health interventions: evaluating multisector community partnerships’ efforts. Am J Prev Med. 2024;67(6):916-923. doi:10.1016/j.amepre.2024.07.016

35.

Whitman

De Lew

Chappel

Aysola

Zuckerman

Sommers

BD.

Addressing Social Determinants of Health: Examples of Successful Evidence-Based Strategies and Current Federal Efforts. vol. 1. ASPE; 2022:1-30.

36.

Tsega

Lewis

McCarthy

Shah

Coutts

Review of Evidence for Health-Related Social Needs Interventions. The Commonwealth Fund; 2019.

37.

McCarthy

Lewis

Horstman

Bryan

Shah

Guide to Evidence for Health-Related Social Needs Interventions: 2022 Update. The Commonwealth Fund; 2022.

38.

Ludwig

Sanbonmatsu

Gennetian

, et al. Neighborhoods, obesity, and diabetes — a randomized social experiment. N Engl J Med. 2011;365(16):1509-1519. doi:10.1056/NEJMsa1103216

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB