Skip to main content
NCBI home page
Journal of Primary Care & Community Health logoLink to Journal of Primary Care & Community Health
. 2026 Jul 28;17:21501319261472633. doi: 10.1177/21501319261472633

Coordinating Multisector Efforts to Address Social Needs in a Rural Midwestern Primary Care Setting

Samuel T Savitz 1,✉, Margaret M Paul 1, Gretchen E Mauerman 2, Amy E Glasgow 1, Sarah B Harper 3, Jessica J McCoy 1, Erin C Westfall 2
PMCID: PMC13420049  PMID: 42521199

Abstract

Introduction/Objectives

The rise in screening for social drivers of health (SDoH) in primary care settings has led to questions about effective intervention design. A large primary care practice at Mayo Clinic developed a model centered on a full-time Health Equity Coordinator (HEC) to facilitate clinic-community partnerships in 2022. The goal of this analysis was to assess the model on process outcomes, resolution of SDoH needs, and changes in healthcare utilization.

Methods

Using a cohort study design, we identified primary care patients with unmet SDoH needs at the intervention clinic and two comparator clinics from March 2022 to December 2023. We assessed process outcomes descriptively and resolution of SDoH needs and healthcare utilization using multivariable regression.

Results

The HEC contacted 348 patients and conducted 748 meetings with community-based organization staff. While there were decreases for SDoH needs and missed or late cancelled appointments in intervention group, there were no significant difference in resolution of SDoH and healthcare utilization relative to comparator groups.

Conclusions

The HEC model for multi-stakeholder coordination showed promise in referring patients to services and engaging key stakeholders in a primary care setting. Longer term follow up may be needed to identify changes in SDoH needs and healthcare utilization.

Keywords: social drivers of health, rural healthcare, primary care, community partnerships, social needs screening

Introduction

Health systems are increasingly required to have mechanisms to identify social drivers of health (SDoH).1,2 These requirements are in response to the recognized impact of unmet SDoH needs on health outcomes, healthcare access, and healthcare utilization.3-8 Health systems have the potential to screen and refer patients to social services organizations in the community with the goal of decreasing unmet SDoH and improving outcomes. 9

Health systems have initiated programs to coordinate across multiple stakeholder groups in the community to make referrals and address SDoH.10-12 Multi-sector coordination was also a critical component of the Centers for Disease Control and Prevention (CDC)’s Improving Social Determinants of Health—Getting Further Faster (GFF) initiative. In this initiative, CDC partnered with the Association of State and Territorial Health Officials (ASTHO) and the National Association of County and City Health Officials (NACCHO) to identify 42 community partnerships that address SDoH through multistakeholder coordination. The awardees included health systems, public health departments, and community-based organizations. 13 The Centers for Medicare and Medicaid (CMS) Accountable Health Communities Model was another effort to address SDoH needs through an assistance track offering navigation services and an alignment track offering both navigation and coordination across stakeholders. However, the resolution of SDoH was similar for both tracks compared to the control group with lack of capacity at community-based organizations identified as the greatest barrier. 14

In this context of these prior efforts to address SDoH needs through multi-stakeholder collaboration, Mayo Clinic Southwest Minnesota (SWMN) developed a Health Equity Coordinator (HEC) role to guide daily planning and operations of a multisector social needs screening and referral program at a large rural primary care practice that serves 16,000 patients annually. The Mayo Clinic SWMN Primary Care Subcommittee on Social Drivers of Health is responsible for identifying and addressing patients’ SDoH that lead to adverse health outcomes and inequities throughout the Mayo Clinic SWMN. The overarching goal of the committee is to develop the infrastructure, systems, and workflows required for healthcare teams to achieve this goal consistently and effectively. The Subcommittee’s prior experience with SDoH screening and referral throughout Mayo Clinic SWMN indicated that the work involved in building and maintaining strong relationships with social service agencies and coordination to connect patients to services was too onerous to build into existing staff roles. This is also reflected in literature, 15 which suggests that hiring a full-time staff member to manage SDoH screening interventions is an effective implementation strategy for these programs. 16 The SDoH Subcommittee developed the HEC role through the work to transition grant-funded health equity efforts into sustainable clinical operations.

The HEC pilot aimed to build infrastructure enabling the clinical practice to engage with patients, providers, and local organizations with the goal of identifying patients with SDoH and facilitating effective referrals to community partner organizations. The aim of this manuscript was to evaluate the HEC pilot in terms of process outcomes, resolution of SDoH, and healthcare utilization. We hypothesized that the HEC pilot would lead to improvement in the resolution of SDoH. We further hypothesized that by addressing SDoH, the intervention could lead to reductions in missed or late cancelled visits and emergency department (ED) visits because unmet SDoH may affect the ability to attend outpatient visits,17-22 which has been linked to increased healthcare utilization.23-25 Additionally, SDoH have been linked to worse health outcomes3,26,27 that could also drive increased healthcare utilization.28-31 The results have implications for the design of programs to conduct multistakeholder coordination to address SDoH needs.

Methods

Study Design

We conducted a retrospective cohort study to analyze the impact of the HEC intervention on unmet SDoH needs and healthcare utilization outcomes. We compared patients who met with the HEC to patients who did not meet with the HEC at the same clinic as well as patients being seen at similar clinics. This approach allowed us to assess changes in outcomes over time and whether the change was greater than the change that occurred for similar patients who were not exposed to the intervention.

Setting

The Eastridge primary care clinic piloted the HEC. This clinic is located Mankato, MN and serves approximately 16,000 patients annually. We pulled data for Eastridge Family Medicine clinic, as well as the comparator clinics in Mankato (Eastridge Family Medicine) and St. Peter (St. Peter Family Medicine) Minnesota. All of the study clinics are within 20 miles of each other and are part of the greater Mankato area. All three clinics serve a large farming community and a substantial East African immigrant and refugee community. Mankato is a regional hub hospital and clinic for the SWMN region of Mayo Clinic, which includes three critical-access hospitals and clinics and one mid-size hospital and clinic. Six additional clinics are in the SWMN region, serving communities with populations ranging from 3,000 to 11,000 people. Many of the counties served by the SWMN region have a higher-than-average rate of poverty and people with no high school diploma or GED. 32 In the 2022 SWMN region community health needs assessments key stakeholders including community members, community-based organizations (CBOs), and healthcare entities identified SDoH needs and health disparities as the top community health needs to address. 32

Data Collection

The primary data source for this evaluation comes from the Mayo Clinic Electronic Health Record (EHR). In addition, we used information from a comprehensive database that was maintained by the HEC to document core components of their activities. The database included information on the primary care provider, the reason for the visit, SDoH needs, a description of the HEC interaction with patients, and the amount of time spent meeting with each patient. A structured Institutional Review Board (IRB) decision tool classified this project as quality improvement and therefore the Mayo Clinic IRB does not require IRB review. We report the analysis in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. 33

Participants

The analysis sample for the SDoH and healthcare utilization outcomes included adult patients (age ≥18) who were a primary care patient at the Eastridge Family Medicine, Eastridge Family Medicine, or St. Peter Family medicine clinics between March 1, 2022 to December 31, 2023 (see Appendix Figure 1). Patients in the intervention group had to meet with the HEC between March 1, 2022-December 31, 2023 and be primary care patient at Eastridge Family Medicine. Patients in the comparison groups had to have a primary care clinic visit during the period of March 1, 2022-December 31, 2023. The patients in the comparison group also had to have an SDoH need identified through the SDoH screening within the past 30 days prior to their appointment. The SDoH screening uses a standard set of screening questions available through the Epic EHR platform and Mayo Clinic assigns it to patients annually. During the intervention period, we focused on the screening questions for financial resource strain, 34 transportation needs, 35 food insecurity, 36 and housing insecurity 37 using questions that were adapted from prior research. The index date was the date of first contact with the HEC for the intervention group and the date of the primary care appointment for the comparator groups.

Intervention

The HEC was introduced to the Eastridge clinic in March of 2022. During the initial year of the program, the HEC established the necessary infrastructure for the clinic to effectively engage with patients, providers, the local health department, and CBOs, with the aim of enhancing capacity to meet patient needs. Patients entered the program through referrals to the HEC from patients’ care team and from the HEC reaching out to patients who had screened positive for unmet SDoH needs. The interaction between the HEC and the patient typically occurred at family medicine provider visits. When a patient had an SDoH identified on the screening questionnaire or if a concern came up in conversation with a provider, the provider would invite the HEC to join the visit if the patient provided permission. This interaction could occur at any point during the provider visit (before, during, or, most commonly, after). If a provider had a concern about a patient outside of an appointment, the provider could refer the patient to the HEC for outreach and resource referrals through phone calls or patient portal messages. In addition to working with patients and providers, the HEC also met regularly with community organizations to maintain awareness of community resources and capacity to address SDoH.

The HEC position was novel for the health system. The job description allowed for flexibility to work within the clinic alongside clinicians and patients, as well as in the community to build and sustain a network of relationships and resources to address patients’ identified social needs. To maximize impact, the HEC had to be a trusted member of the community, have experience working with community programs and health organizations, and be sensitive to the unique needs of low-income and culturally diverse families. The HEC met the qualifications for the role by having lived experience with SDoH concerns and a professional background in public health, chronic disease prevention, and community engagement. The HEC had autonomy to fluidly move into and out of the clinic and engage community partners whenever they needed it most. The job description did not specify the associated tasks a priori, which facilitated innovation and iterative development and implementation of screening and intervention workflows and community capacity building. The HEC position was 1.0 FTE that was spread across meeting with patients, meeting with community organizations, giving presentations to internal and external stakeholders, staff education, maintaining a produce assistance program, technical assistance, workflow build, and administrative tasks.

Measures

The first set of outcomes were process outcomes on the reach of the program and the work necessary to serve patients. This set of outcomes included the number of patients served, the number of meetings with CBO staff, and the number of hours interacting with different stakeholders. The second set of outcomes were SDoH outcomes that measured the change in SDoH status before and after the index date. We focused on four domains collected from the SDoH screening over the entire study period: financial resource strain, transportation needs, food insecurity, and housing insecurity. We compared the responses before meeting with the HEC or before the index primary care appointment for the comparator groups to the next value during the study period. The baseline screening value could take place up to one year before the index date. We excluded financial resource strain as an outcome due to insufficient variation. We excluded patients without an initial or follow-up response to the SDoH screener from this analysis. The third set of outcomes were changes in healthcare utilization outcomes after the intervention began. We focused on ED visits and missed or late cancelled primary care visits. Visits were considered late cancelled if they were cancelled by the patient less than 48 hours before the scheduled appointment time. We evaluated these outcomes for the two years before the index date through one year following the index date in six-month intervals. For ED visits, we operationalized the outcome as a binary variable for whether there were any visits since the percentage with more than one ED visit in each period was small. For missed or late cancelled appointments we analyzed the count of missed or late cancelled visits along with the total number of scheduled visits in the period.

We also collected other patient characteristics to characterize the population and adjust for baseline differences in the regression models. First, we identified patient demographics including age, gender, race, ethnicity, material status, and education level. Second, we collected patient interaction with the medical system including whether the patient used the patient portal in the last three months and insurance status. Third, we collected medical conditions including the total sum of Elixhauser medical comorbidities and history of alcohol use disorder. The Elixhauser comorbidity index includes 30 medical conditions identified through ICD-10 codes. 38 Fourth, we pulled the Area-Deprivation Index, which is a measure of socioeconomic status derived from publicly available census data. 39 We included all variables in the regression models except for education level due to model convergence issues for this variable.

Analysis

We analyzed the reach of the program and the time spent engaging with stakeholders descriptively. We compared the change in SDoH screening results using logistic regression using four comparisons: 1) pre-post intervention group, 2) pre-post for intervention group compared to Eastridge comparator group, 3) pre-post for intervention group compared to Eastridge and St. Peter comparator group, and 4) pre-post for all Eastridge patients (intervention and comparator) compared to Eastridge and St. Peter comparator group. The first comparison shows the difference without accounting for changes that may occur without the intervention. The second and third comparisons account for expected change by using two alternative comparison groups. Finally, the fourth comparison conceptualizes the intervention as encompassing all patients at Eastridge and evaluating the impact on the entire clinic. We assessed utilization of ED visits and missed or late cancelled primary care appointments using regression analysis using the same comparison as described above. For comparisons 2-4, we created dummy variables for each period with the reference period being 0-6 months pre-intervention for ED visits and 6-12 months pre-intervention for missed or late cancelled appointments. We used 6-12 months pre-intervention for the missed or late cancelled outcome because one of the inclusion criteria was having attended a primary care appointment for the comparison groups. Time zero was defined as the index date of either the meeting with the HEC or the first primary care appointment of the period. We then interacted these variables with an indicator for the intervention group. We focused on the interaction term between the indicator variable for the time period and the intervention or comparator group with the interaction for the post-period indicators as the estimated effect of the intervention. We also included the main effects for these variables in addition to the interaction terms. We clustered standard errors at the level of the primary care provider for SDoH models and the patient for the healthcare utilization models since these models included multiple observations per patient. We report the results as odds ratios (ORs) with 95% confidence intervals (CI) and consider results statistically significant at p<0.05. We conducted multiple imputation using the Markov Chain Monte Carlo (MCMC) approach with 20 imputations to address missing data. We extracted the data using SAS version 9.4 (SAS Institute, Cary, NC USA) and analyzed it using Stata version 18 (StataCorp, College Station, TX, USA).

Results

Process Outcomes

Between March 2022 and December 2023, the program served 409 patients, requiring a total of 315.5 contact hours to facilitate patient engagement and referrals to appropriate resources. These contact hours comprised both in-person visits and telephone interactions. The adult patient cohort ranged in age from 18 to 92 years old, with a minority of patients under the age of 18 (n=18). In these cases, parental/guardian engagement was determined on a case-by-case basis.

During the same period, the HEC conducted 788 meetings with CBO staff, amounting to 1,088 contact hours. Beyond these direct engagements, the HEC delivered over 20 presentations to various Mayo Clinic SWMN committees, CBOs, and other community stakeholders, aimed at disseminating program progress and garnering support.

In addition to time spent actively engaging patients, colleagues, and community-based collaborators, the role entailed substantial administrative duties to ensure system functionality across platforms including the Mayo Clinic’s EHR (Epic), and EHR-integrated community resource referrals platform (Findhelp), and various internal committees and workgroups, totaling over 1,200 hours.

Participant Characteristics

There were 326 patients in the intervention group in the analysis sample (Table 1). There were 755 patients in the Eastridge comparison group and 572 in the Eastridge and St. Peter comparison group. Patients in the intervention group (46.9%) were more likely to have Medicaid insurance than Eastridge comparison group (29.0%) or the Eastridge or St. Peter comparison group (26.2%). Patients in the intervention group also had greater medical complexity (mean Elixhauser: 3.61) compared to the Eastridge comparison group (2.00) or the Eastridge or St. Peter comparison group (2.04). Patients in the intervention group were also more likely to report transportation needs and food insecurity.

Table 1.

Patient Characteristics in Intervention Period

Variable Level Eastridge comparison group Intervention group Northridge and St. Peter comparison group
N = 755 N = 326 N = 572
Transportation Needs Yes 98 (13.0%) 75 (23.0%) 63 (11.0%)
No 650 (86.1%) 191 (58.6%) 504 (88.1%)
Missing 7 (0.9%) 60 (18.4%) 5 (0.9%)
Housing Needs Yes 243 (32.2%) 76 (23.3%) 177 (30.9%)
No 503 (66.6%) 184 (56.4%) 394 (68.9%)
Missing 9 (1.2%) 66 (20.2%) 1 (0.2%)
Food Insecurity Yes 306 (40.5%) 156 (47.9%) 207 (36.2%)
No 443 (58.7%) 111 (34.0%) 359 (62.8%)
Missing 6 (0.8%) 59 (18.1%) 6 (1.0%)
Financial Needs Yes 545 (72.2%) 170 (52.1%) 435 (76.0%)
No 202 (26.8%) 89 (27.3%) 136 (23.8%)
Missing 8 (1.1%) 67 (20.6%) 1 (0.2%)
Gender Female 448 (59.3%) 210 (64.4%) 368 (64.3%)
Male 307 (40.7%) 115 (35.3%) 204 (35.7%)
Missing 0 (0%) 1 (0.3%) 0 (0%)
Race Black 38 (5.0%) 42 (12.9%) 22 (3.8%)
White 682 (90.3%) 269 (82.5%) 530 (92.7%)
Other 25 (3.3%) 8 (2.5%) 13 (2.3%)
Missing 10 (1.3%) 7 (2.1%) 7 (1.2%)
Marital Status Married 109 (14.4%) 64 (19.6%) 76 (13.3%)
Divorced/Separated/Widowed 240 (31.8%) 66 (20.2%) 208 (36.4%)
Single 394 (52.2%) 193 (59.2%) 281 (49.1%)
Missing 12 (1.6%) 3 (0.9%) 7 (1.2%)
Education Some high school, did not graduate 197 (26.1%) 77 (23.6%) 158 (27.6%)
High School graduate/GED 71 (9.4%) 45 (13.8%) 45 (7.9%)
Some college 145 (19.2%) 63 (19.3%) 116 (20.3%)
Undergraduate or graduate degree 312 (41.3%) 81 (24.8%) 231 (40.4%)
Missing 30 (4.0%) 60 (18.4%) 22 (3.8%)
Ethnicity Non-Hispanic 701 (92.8%) 294 (90.2%) 529 (92.5%)
Hispanic 44 (5.8%) 28 (8.6%) 36 (6.3%)
Missing 10 (1.3%) 4 (1.2%) 7 (1.2%)
Primary Language English 749 (99.2%) 315 (96.6%) 568 (99.3%)
Non-English 6 (0.8%) 10 (3.1%) 4 (0.7%)
Missing 0 (0%) 1 (0.3%) 0 (0%)
Insurance Status Medicare 117 (15.5%) 85 (26.1%) 97 (17.0%)
Medicaid 219 (29.0%) 153 (46.9%) 150 (26.2%)
Private 390 (51.7%) 73 (22.4%) 305 (53.3%)
Other 29 (3.8%) 15 (4.6%) 20 (3.5%)
RUCA Urban 616 (81.6%) 215 (66.0%) 347 (60.7%)
Rural 139 (18.4%) 108 (33.1%) 225 (39.3%)
Missing 0 (0%) 3 (0.9%) 0 (0%)
ADI Q1 8 (1.1%) 5 (1.5%) 5 (0.9%)
Q2 95 (12.6%) 34 (10.4%) 58 (10.1%)
Q3 141 (18.7%) 44 (13.5%) 120 (21.0%)
Q4 192 (25.4%) 105 (32.2%) 199 (34.8%)
Q5 196 (26.0%) 123 (37.7%) 90 (15.7%)
Missing 123 (16.3%) 15 (4.6%) 100 (17.5%)
Alcohol Use Disorder ​ 52 (6.9%) 31 (9.5%) 39 (6.8%)
Age In Years 40.85 (16.27) 44.23 (18.56) 42.65 (16.70)
Elixhauser Sum Sum of Conditions 2.00 (2.14) 3.61 (3.29) 2.04 (2.04)
Distance to Clinic In Miles 19.09 (85.34) 14.65 (44.52) 21.61 (109.96)
Prior Cancelled Visits (0-6 Months Pre) Count 0.22 (0.67) 0.63 (1.02) 0.25 (0.65)
Prior Cancelled Visits (7-12 Months Pre) Count 0.27 (0.69) 0.44 (0.84) 0.21 (0.60)
Prior ED Visits (0-6 Months Pre) Any 93 (15.6%) 90 (37.7%) 62 (12.9%)
Prior ED Visits (7-12 Months Pre) Any 84 (14.6%) 63 (29.0%) 55 (11.7%)
Open in a new tab

Note: ADI stands for Area Deprivation Index; RUCA stands for Rural-Urban Commuting Areas.

SDoH Outcomes

The intervention was only associated with resolution of SDoH needs in the analysis focusing on the pre-post without a comparison group (Table 2). For that analysis, the ORs for housing (OR: 0.59 [0.37, 0.94]) had lower odds in the follow-up period. While the point estimates for transportation needs and food insecurity were below one, they were not statistically significant. In contrast, the ORs for the comparison with the Eastridge comparison group and Eastridge and St. Peter comparison group were all non-significant.

Table 2.

Regression Output for Unmet Needs

Period Eastridge FM contacted Eastridge FM contacted vs. not contacted Eastridge FM contacted vs. Northridge and St. Peter FM Eastridge FM vs. Eastridge and St. Peter FM
Odds Ratios (95% CI) Odds Ratios (95% CI) Odds Ratios (95% CI) Odds Ratios (95% CI)
Food Insecurity 0.78 (0.53, 1.14) 1.68 (0.92, 3.08) 1.51 (0.85, 2.68) 1.22 (0.82, 1.81)
Transportation 0.69 (0.44, 1.07) 1.22 (0.53, 2.83) 1.31 (0.69, 2.51) 0.94 (0.52, 1.70)
Housing 0.59* (0.37, 0.94) 0.94 (0.45, 1.95) 1.01 (0.51, 2.00) 0.77 (0.42, 1.40)
Open in a new tab

Note. The models controlled for gender, race, ethnicity, age, marital status, insurance, portal use, distance to clinic, alcohol use disorder, ADI, RUCA, Elixhauser comorbidity index, and social drivers of health (housing instability, food insecurity, financial insecurity, and transportation needs).

*: p<0.05.

Healthcare Utilization Outcomes

The results for ED visits and missed or late cancelled appointments were not statistically significant. Further, the point estimates for the interaction terms 0-6 months post and 6-12 months post were close to one for all analyses (Table 3).

Table 3.

Regression Output for Utilization Outcomes

​ Period Eastridge FM contacted Eastridge FM contacted vs. not contacted Eastridge FM contacted vs. Eastridge and St. Peter FM Eastridge FM vs. Eastridge and St. Peter FM
Odds Ratios (95% CI) Odds Ratios (95% CI) Odds Ratios (95% CI) Odds Ratios (95% CI)
ED Visits 18-24 Months Pre 0.66 (0.44, 1.01) 0.78 (0.46, 1.35) 0.68 (0.39, 1.20) 0.82 (0.50, 1.34)
12-18 Months Pre 0.87 (0.59, 1.28) 0.90 (0.54, 1.50) 0.78 (0.45, 1.35) 0.87 (0.57, 1.33)
6-12 Months Pre 0.72 (0.50, 1.03) 0.78 (0.49, 1.25) 0.81 (0.48, 1.37) 0.95 (0.58, 1.55)
0-6 Months Pre Reference Reference Reference Reference
0-6 Months Post 1.03 (0.73, 1.45) 0.84 (0.54, 1.31) 0.94 (0.57, 1.54) 1.11 (0.73, 1.70)
6-12 Months Post 0.80 (0.56, 1.15) 1.03 (0.64, 1.65) 1.15 (0.67, 1.97) 1.22 (0.73, 2.05)
Missed or Late Cancelled Appointments 18-24 Months Pre 1.57* (1.13, 2.17) 2.53 (1.55, 4.10) 1.71 (1.07, 2.72) 1.08 (0.71, 1.63)
12-18 Months Pre 1.21 (0.84, 1.75) 1.67 (1.01, 2.76) 1.36 (0.81, 2.28) 1.07 (0.69, 1.67)
6-12 Months Pre Reference Reference Reference Reference
0-6 Months Pre Excluded Excluded Excluded Excluded
0-6 Months Post 1.04 (0.76, 1.43) 1.07 (0.71, 1.62) 1.00 (0.63, 1.60) 0.98 (0.65, 1.47)
6-12 Months Post 1.11 (0.77, 1.63) 1.31 (0.81, 2.12) 0.99 (0.59, 1.65) 0.88 (0.57, 1.35)
Open in a new tab

Note. The models controlled for gender, race, ethnicity, age, marital status, insurance, portal use, distance to clinic, alcohol use disorder, ADI, RUCA, Elixhauser comorbidity index, and social drivers of health (housing instability, food insecurity, financial insecurity, and transportation needs).

*: p<0.05.

Discussion

In this evaluation of the HEC role in a large primary care clinic in Southwest Minnesota, we found that the role facilitated the development of necessary infrastructure to support healthcare teams and CBOs and led to referrals for patients interacting with the HEC. We observed evidence for a reduction in housing needs SDoH in the pre-post analysis for just HEC patients. However, we did not find strong evidence for reductions in SDoH, ED visits, or missed or late cancelled appointments relative to the comparison groups.

To our knowledge, this is the first analysis of such an intervention led by a full-time clinic staff member with a flexible job description designed to adapt over time to meet evolving program needs. The process outcomes indicated that the HEC successfully reached over 300 patients and engaged with key stakeholders to facilitate referrals to community-based organizations. This work contributes to the literature on efforts to coordinate across key stakeholder groups to address unmet SDoH needs. This model serves as another example of a program to engage stakeholders and help refer patients to services to address SDoH.10-12

There are several potential explanations for the limited evidence for improvement in the SDoH and healthcare utilization outcomes. First, these outcomes may be difficult to change given the follow-up period. While there has been prior evidence linking SDoH and healthcare utilization,28-31 referral programs designed to address SDoH often fail to significantly improve SDoH and healthcare utilization during analysis follow-up periods.12,14,40,41 This is not necessarily indicative that the programs have not succeeded, but that they may need more time to significantly improve these outcomes which are affected by many factors including health status. Second, there may have been underlying differences between patients in the intervention group and the comparison groups that biased the results. The intervention group was identified as those who either received a referral or were contacted by the HEC to receive help and then accepted help. In contrast, the comparison groups included individuals who had a primary care appointment and also had an SDoH that was identified through the screening questionnaire in the last 30 days. This selection process appeared to lead to greater medical complexity in the intervention group, as measured by the Elixhauser comorbidities. While we controlled for the Elixhauser comorbidities, the differences may also point to greater unmeasured medical complexity, which may have made it more difficult to show a difference for ED visits. In addition, the people receiving referrals or accepting offers for assistance may have more complex SDoH needs, which could not be assessed given that the SDoH screening categorizes needs as either being present or not being present. Additionally, some of the patients in the comparison groups may have already been linked to resources in both comparison groups but especially patients in the Eastridge comparison group who were not referred to the HEC despite being eligible. Third, the sample size in this pilot evaluation may have been insufficient to identify small to medium effect sizes. The confidence intervals encompassed large reductions as well as large increases and the lack of precision would have required large effects to be statistically significant. However, many of the point estimates relative to the comparison group were close to one or greater than one. The only exceptions were that some comparisons showed that resolution of housing needs had point estimates consistent with a reduction.

This position was initially based in a single Family Medicine department at the Eastridge Clinic, but over one year expanded to support six different clinical sites, with a planned expansion to three additional sites in 2026. Assessing the capacity of community-based organizations, the HEC position, as well as the tools and technology and reimbursement mechanisms needed to support multiple sites across regions will be essential when determining the scalability and sustainability of the role. Identifying patient panel size that a HEC can support, while ensuring ongoing practice and CBO engagement and capacity building will require a leadership-level commitment to retaining flexibility within the job description. The flexibility, unique to the health system and traditional clinical roles, was crucial to core HEC tasks, such as monitoring capacity of community resources, providing situational awareness to providers and care teams, and maintaining CBO involvement in accepting referrals. Within the system, implementation of the HEC role revealed that although the intention of the program was to connect established patients with community resources, most of the effort required for effective implementation centered on engaging providers and staff and coordinating with CBOs rather than direct patient interactions. The flexible job description created by clinical practice leadership was pivotal to the success of both the HEC role and the broader program. Its adaptability allowed the HEC’s responsibilities to evolve with program needs and practice priorities, which contributed substantially to overall program effectiveness. There is broad support for sustaining this role within the health system, and future research will help determine its scalability across clinics, a feature that may make roles like the HEC feasible in a variety of clinical contexts.

This analysis includes limitations. First, this was an observational analysis and as such there may have been unmeasured confounding. We addressed this concern by adjusting for observable patient characteristics and identifying a comparison group, but there may have been other differences unmeasured as noted above. Second, our analysis on ED visits was limited to Mayo Clinic EHR data and therefore did not account for patients who may have used healthcare services at another non-Mayo Clinic location. Third, the program was focused on primary care patients receiving care at Mayo Clinic and did not include individuals who do not access health care services and who may have worse access to care.

Conclusions

While the HEC position was not associated with clear evidence of improvement in resolution of SDoH needs and healthcare utilization, it strengthened coordination and communication across key stakeholders and enhanced the capacity of primary care teams to address patients with unmet SDoH needs. These findings suggest that embedding a dedicated HEC should be considered as a potential model for health systems looking for an approach to support patients with unmet SDoH needs.

Appendix.

Appendix Figure 1.

Appendix Figure 1.

Open in a new tab

Sample Flow Diagram

Footnotes

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 Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery at Mayo Clinic, Rochester, Minnesota.

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

ORCID iDs

Samuel T. Savitz https://orcid.org/0000-0003-3190-7740

Margaret M. Paul https://orcid.org/0000-0003-3281-6234

Erin C. Westfall https://orcid.org/0000-0002-8563-340X

Ethical Considerations

This project was classified as quality improvement work through a structured Institutional Review Board decision tool and was therefore not reviewed by an Institutional Review Board.

Data Availability Statement

The data for this study has not been made publicly available because of privacy protections for medical records data. However, the data is available from the corresponding author upon reasonable request from individuals with human subjects training.*

References

  • 1.Centers for Medicare & Medicaid Services (CMS), HHS . Medicare Program; Hospital Inpatient Prospective Payment Systems for Acute Care Hospitals and the Long-Term Care Hospital Prospective Payment System and Policy Changes and Fiscal Year 2024 Rates; Quality Programs and Medicare Promoting Interoperability Program Requirements for Eligible Hospitals and Critical Access Hospitals; Rural Emergency Hospital and Physician-Owned Hospital Requirements; and Provider and Supplier Disclosure of Ownership; and Medicare Disproportionate Share Hospital (DSH) Payments: Counting Certain Days Associated With Section 1115 Demonstrations in the Medicaid Fraction. Fed Regist. 2023;88(165):58640. [Google Scholar]
  • 2.The Joint Commission . New Requirements to Reduce Health Care Disparities. R3 Report Requirement, Rationale, Reference. 2022(36):1-8. https://digitalassets.jointcommission.org/api/public/content/2db60f969e1e4d33a99f997629b60a45?v=20f7c288 [Google Scholar]
  • 3.Walker RJ, Smalls BL, Campbell JA, Strom Williams JL, Egede LE. Impact of social determinants of health on outcomes for type 2 diabetes: a systematic review. Endocrine. 2014;47(1):29-48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Lim A, Benjasirisan C, Liu X, et al. Social determinants of health and emergency department visits among older adults with multimorbidity: insight from 2010 to 2018 National Health Interview Survey. BMC Public Health. 2024;24(1):1153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ardekani A, Fereidooni R, Heydari ST, Ghahramani S, Shahabi S, Bagheri Lankarani K. The association of patient‐reported social determinants of health and hospitalization rate: A scoping review. Health Science Reports. 2023;6(2):e1124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Okoro CA, Strine TW, Young SL, Balluz LS, Mokdad AH. Access to health care among older adults and receipt of preventive services. Results from the Behavioral Risk Factor Surveillance System, 2002. Prev Med. 2005;40(3):337-343. doi: 10.1016/j.ypmed.2004.06.009. [DOI] [PubMed] [Google Scholar]
  • 7.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. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Fiori KP, Heller CG, Rehm CD, et al. Unmet Social Needs and No-Show Visits in Primary Care in a US Northeastern Urban Health System, 2018-2019. Am J Public Health. 2020;110(S2):S242-s250. doi: 10.2105/ajph.2020.305717. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Gurewich D, Garg A, Kressin NR. Addressing social determinants of health within healthcare delivery systems: a framework to ground and inform health outcomes. Journal of general internal medicine. 2020;35(5):1571-1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Shirey MR, White-Williams C. Leading Multisector Partnerships to Address Social Determinants of Health and Advance Health Equity for an Underresourced Heart Failure Patient Population. Nurs Adm Q. 2022;46(3):197-207. doi: 10.1097/naq.0000000000000533. [DOI] [PubMed] [Google Scholar]
  • 11.Shaikh U, Gosdin M, Helmke E. Multisector collaborations at children’s hospitals to address social drivers of health. Int J Qual Health Care. 2025;37(3):mzaf067. doi: 10.1093/intqhc/mzaf067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Yan AF, Chen Z, Wang Y, et al. Effectiveness of Social Needs Screening and Interventions in Clinical Settings on Utilization, Cost, and Clinical Outcomes: A Systematic Review. Health Equity. 2022;6(1):454-475. doi: 10.1089/heq.2022.0010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Centers for Disease Control (CDC) . Improving Social Determinants of Health—Getting Further Faster. Centers for Disease Control; 2023. https://www.cdc.gov/health-equity-chronic-disease/media/pdfs/2024/04/GFF-eval-brief-508.pdf [Google Scholar]
  • 14.Renaud J, McClellan SR, DePriest K, et al. Addressing health-related social needs via community resources: lessons from accountable health communities: study examines how the accountable health communities model addresses health-related social needs via community resources. Health Affairs. 2023;42(6):832-840. [DOI] [PubMed] [Google Scholar]
  • 15.Marchis EH, Aceves BA, Brown EM, Loomba V, Molina MF, Gottlieb LM. Assessing Implementation of Social Screening Within US Health Care Settings: A Systematic Scoping Review. J Am Board Fam Med. 2023;36(4):626-649. doi: 10.3122/jabfm.2022.220401R1. [DOI] [PubMed] [Google Scholar]
  • 16.Berry C, Paul M, Massar R, Marcello RK, Krauskopf M. Social Needs Screening and Referral Program at a Large US Public Hospital System, 2017. Am J Public Health. 2020;110(S2):S211-S214. doi: 10.2105/AJPH.2020.305642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Cosgrove MP. Defaulters in general practice: reasons for default and patterns of attendance. Br J Gen Pract. 1990;40(331):50-52. [PMC free article] [PubMed] [Google Scholar]
  • 18.Bean AG, Talaga J. Appointment breaking: causes and solutions. J Health Care Mark. 1992;12(4):14-25. [PubMed] [Google Scholar]
  • 19.Boshers EB, Cooley ME, Stahnke B. Examining no-show rates in a community health centre in the United States. Health Soc Care Community. 2022;30(5):e2041-e2049. doi: 10.1111/hsc.13638. [DOI] [PubMed] [Google Scholar]
  • 20.Samuels RC, Ward VL, Melvin P, et al. Missed Appointments: Factors Contributing to High No-Show Rates in an Urban Pediatrics Primary Care Clinic. Clin Pediatr (Phila). 2015;54(10):976-982. doi: 10.1177/0009922815570613. [DOI] [PubMed] [Google Scholar]
  • 21.Kaplan-Lewis E, Percac-Lima S. No-show to primary care appointments: why patients do not come. J Prim Care Community Health. 2013;4(4):251-255. doi: 10.1177/2150131913498513. [DOI] [PubMed] [Google Scholar]
  • 22.Chapman KA, Machado SS, van der Merwe K, Bryson A, Smith D. Exploring Primary Care Non-Attendance: A Study of Low-Income Patients. J Prim Care Community Health. 2022;13:21501319221082352. doi: 10.1177/21501319221082352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Wallace R, Hughes-Cromwick P, Mull H. Cost-effectiveness of access to nonemergency medical transportation: comparison of transportation and health care costs and benefits. Transportation research record. 2006;1956(1):86-93. [Google Scholar]
  • 24.Rust G, Ye J, Baltrus P, Daniels E, Adesunloye B, Fryer GE. Practical barriers to timely primary care access: impact on adult use of emergency department services. Archives of internal medicine. 2008;168(15):1705-1710. [DOI] [PubMed] [Google Scholar]
  • 25.Nguyen DL, DeJesus RS. Increased frequency of no-shows in residents’ primary care clinic is associated with more visits to the emergency department. Journal of primary care & community health. 2010;1(1):8-11. [DOI] [PubMed] [Google Scholar]
  • 26.Khariton Y, Nassif ME, Thomas L, et al. Health status disparities by sex, race/ethnicity, and socioeconomic status in outpatients with heart failure. JACC: Heart Failure. 2018;6(6):465-473. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Metlock FE, Hinneh T, Benjasirisan C, et al. Impact of social determinants of health on hypertension outcomes: a systematic review. Hypertension. 2024;81(8):1675-1700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Hatef E, Ma X, Rouhizadeh M, Singh G, Weiner JP, Kharrazi H. Assessing the impact of social needs and social determinants of health on health care utilization: using patient-and community-level data. Population health management. 2021;24(2):222-230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Krause TM, Schaefer C, Highfield L. The association of social determinants of health with health outcomes. Am J Manag Care. 2021;27(3):e89-e96. [DOI] [PubMed] [Google Scholar]
  • 30.Baker MC, Alberti PM, Tsao T-Y, Fluegge K, Howland RE, Haberman M. Social Determinants Matter For Hospital Readmission Policy: Insights From New York City: Study examines social determinants and hospital readmissions. Health Affairs. 2021;40(4):645-654. [DOI] [PubMed] [Google Scholar]
  • 31.McCarthy ML, Zheng Z, Wilder ME, Elmi A, Li Y, Zeger SL. The influence of social determinants of health on emergency departments visits in a Medicaid sample. Annals of emergency medicine. 2021;77(5):511-522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Mayo Clinic Health System . Community Health Needs Assessment. Mayo Clinic Health System. https://www.mayoclinichealthsystem.org/about-us/community-health-needs-assessments. Accessed October 15, 2022. [Google Scholar]
  • 33.Von EE, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. The Lancet. 2007;370(9596):1453-1457. [DOI] [PubMed] [Google Scholar]
  • 34.Puterman E, Adler N, Matthews KA, Epel E. Financial Strain and Impaired Fasting Glucose: The Moderating Role of Physical Activity in the Coronary Artery Risk Development in Young Adults Study. Biopsychosocial Science and Medicine. 2012;74(2):187-192. doi: 10.1097/PSY.0b013e3182448d74. [DOI] [PubMed] [Google Scholar]
  • 35.National Association of Community Health Centers, Association of Asian Pacific Community Health Organizations . Oregon Primary Care Association, Institute for Alternative Futures. The Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences (PRAPARE). https://prapare.org/the-prapare-screening-tool/. Accessed May 5, 2026. [Google Scholar]
  • 36.Hager ER, Quigg AM, Black MM, et al. Development and validity of a 2-item screen to identify families at risk for food insecurity. Pediatrics. 2010;126(1):e26-e32. [DOI] [PubMed] [Google Scholar]
  • 37.Sandel M, Sheward R, Ettinger de Cuba S, et al. Unstable Housing and Caregiver and Child Health in Renter Families. Pediatrics. 2018;141(2):e20172199. doi: 10.1542/peds.2017-2199. [DOI] [PubMed] [Google Scholar]
  • 38.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130-1139. doi: 10.1097/01.mlr.0000182534.19832.83. [DOI] [PubMed] [Google Scholar]
  • 39.Kind AJH, Buckingham WR. Making Neighborhood-Disadvantage Metrics Accessible - The Neighborhood Atlas. N Engl J Med. 2018;378(26):2456-2458. doi: 10.1056/NEJMp1802313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kulie P, Steinmetz E, Johnson S, McCarthy ML. A health-related social needs referral program for Medicaid beneficiaries treated in an emergency department. The American Journal of Emergency Medicine. 2021;47:119-124. [DOI] [PubMed] [Google Scholar]
  • 41.McQueen A, von Nordheim D, Caburnay C, et al. A Randomized Controlled Trial Testing the Effects of a Social Needs Navigation Intervention on Health Outcomes and Healthcare Utilization among Medicaid Members with Type 2 Diabetes. Int J Environ Res Public Health. 2024;21(7):936. doi: 10.3390/ijerph21070936. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data for this study has not been made publicly available because of privacy protections for medical records data. However, the data is available from the corresponding author upon reasonable request from individuals with human subjects training.*

Articles from Journal of Primary Care & Community Health are provided here courtesy of SAGE Publications

RESOURCES