Introduction

Access to health insurance is critical to receiving preventive care and an improved quality of life [1]. The number of uninsured people in the United States (U.S.) has historically been high but rose to more than 30 million in 2020 as the COVID-19 pandemic caused an economic recession and massive unemployment rates [2]. This global crisis has consequently amplified the health inequities and poor outcomes already burdening disadvantaged patient populations with chronic cardiovascular and psychiatric diseases [3, 4]. To meet the rising rates of these illnesses, medical student-run free clinics (SRFC) have become an increasingly more prevalent method of providing primary care services to low-income and medically underserved patients across the U.S. [5] In addition to providing medical care, these free clinics further offer a unique opportunity to address the underlying social risk factors that perpetuate long-term, inequitable health outcomes [6].

The World Health Organization describes the social determinants of health (SDH) as the conditions in which people are born, grow, work, live, and age [7]. However, while it is recognized that modifiable SDH factors are responsible for up to 60% of preventable deaths, it is not yet standard of care to routinely screen and address these at preventive health appointments [6, 8–10]. In 2019, our team implemented a social intervention program to better identify and manage the adverse SDH factors providing disproportionally high barriers to care for patients seen at our SRFCs. This innovative organization, called Care Connections, mobilizes medical students as social workers to routinely screen and assist patients using social need questionnaires based on literature proven social intervention programs [11, 12]. Our method specifically capitalizes on the developing clinical acumen of medical students to alleviate overburdened caregivers and the healthcare system, and also holistically manage individual social risks that worsen chronic health conditions [4, 13].

Integration of these services into SRFCs is critical given the increasing volume of patients precipitated by financial and medical limitations imposed by the COVID-19 pandemic. In addition, our model facilitates early real-world clinical and community outreach opportunities for medical students to connect with patients at the highest risk of COVID related complications [14]. Our objective was to perform a retrospective review to identify the social needs of our SRFC patients and improve the administration of community health referrals within our clinics. We hypothesized that patients with more chronic health problems and financial stressors would have the highest requests for resources. Community related health interventions have been shown to improve several outcomes in underinsured populations with comorbid chronic health conditions [1]. However, it is crucial to better understand how routine SDH screening tools can be implemented to efficiently guide more individualized, comprehensive care to the marginalized and vulnerable patient populations seen in free community health clinics.

Methods

Setting

We performed a retrospective review of the socioeconomic screening questionnaires distributed at the University of Arizona College of Medicine–Tucson’s (UACOM-T) SRFCs from August 1^st, 2019 to April 1^st, 2022. Our institution serves patients with limited or no access to health care through the medical student-run and directed Commitment to Underserved People (CUP) Program [15, 16]. First implemented as an elective seminar in 1978 [15], this umbrella initiative has grown to encompass several primary care based clinics including those devoted to mental health, refugees, LGBT+ patients, integrative medicine, family care, pediatrics, and women’s health [16]. These privately funded SRFCs are directed by first-year medical students under the supervision of faculty, typically operate on weekday evenings or weekend mornings, and can either be performed in person at the medical campus or via telehealth. Located near the Mexico-Arizona border, our clinics serve a large, diverse population of medically underserved patients.

In the standard workflow, patients are greeted by clinic intake personnel who gather demographic information and distributes the social screening form (Table 1). The patient is then delegated to a clinical care team, which consists of a preclinical (first or second-year) medical student paired with a clinical (third or fourth-year) medical student. Prior to clinical evaluation, the patients who opted to fill out the social screening tool are interviewed by the first-year representative from our social intervention group, Care Connections. Interview and SDH screening questions (Table 1) are based on the open-source models published by Health Leads [12] and the Social Intervention Research and Evaluation Network’s Gravity Project [11]. Care Connections then debrief with the care team after their medical evaluation and presentation to the attending physician. A comprehensive list of community resources is subsequently assembled for the patient according to their unmet needs. Notes and screening tool results are transcribed into the free cloud-based Electronic Health Record (EHR) platform Practice Fusion (Version 3.7.1.192.0.4608, Practice Fusion, Inc, San Francisco, CA).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Socioeconomic screening questions employed at clinics.

https://doi.org/10.1371/journal.pone.0291682.t001

Results

Our query identified 273 total patients who were evaluated for social needs using our screening questionnaire. Nine of these patients were excluded due to not meeting our inclusion criteria. The mean age of patients was 39.9 (0.4–90), and there were 177 females (67.0%) and 87 males. All participants were uninsured. Eighty-eight patients (33.3%) identified themselves as Hispanic/Latino and 176 patients identified as Non-Hispanic/Latino (Table 2). Of the included 264 patients, 63% requested a community referral to receive resources for financial support, which was also the most frequently reported unmet social need (62%). The least requested unmet need was related to legal resources (5.7%) (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics patients who filled out the socioeconomic screening questionnaire.

https://doi.org/10.1371/journal.pone.0291682.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Summary of responses from patients who filled out the socioeconomic screening questionnaire.

https://doi.org/10.1371/journal.pone.0291682.t003

Univariate analyses (Tables 4 and 5) revealed that more patients with a history of illicit drug use (27.27% vs 13.40%, p = 0.022), musculoskeletal (MSK) conditions (61.82% vs 42.11%, p = 0.010), psychiatric illness (63.64% vs 26.32%, p<0.001), psychiatric medications (30.91% vs 12.44%, p = 0.002), or a higher mean number of comorbidities (4 vs 3, p<0.001) had a significant association with food insecurity compared those without. Patients were significantly more likely to report unmet housing needs if they had a history of alcohol use (48.65% vs 30.40%, p = 0.037), tobacco use (45.95% vs 24.23%, p = 0.009), speaking English (72.97% vs 48.02%, p = 0.007), psychiatric illness (62.16% vs 29.52%, p<0.001), or a higher mean number of comorbidities (5 vs 3, p<0.001). A history of psychiatric illness (58.54% vs 29.60%, p<0.001), or a higher mean number of comorbidities (4 vs 3, p = 0.010) were significantly associated with unmet utility needs. Patients who reported needing financial resources were significantly associated with illicit drug use (21.47% vs 7.92%, p = 0.003) and a higher mean number of comorbidities (>3 vs <3, p = 0.028).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Summary of patient characteristics by various unmet social needs.

https://doi.org/10.1371/journal.pone.0291682.t004

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Summary of patient comorbidities and medication uses by various unmet social needs.

https://doi.org/10.1371/journal.pone.0291682.t005

Transportation insecurity was significantly associated with English-speaking patients (67.44% vs 48.42%, p = 0.030). The need for childcare assistance was significantly more likely to be reported by younger patients (mean age: 32.0 vs 39.5, p = 0.022), patients taking fewer mean medications (0.5 vs 1.0, p = 0.041), and patients with a lower mean number of comorbidities (2.5 vs 3.0, p = 0.044). Employment instability was significantly associated with alcohol use (47.37% vs 28.99%, p = 0.011) and psychiatric illness (49.12% vs 29.95%, p = 0.011). This was also significantly associated with patients who did not have Diabetes Mellitus (DM) (24.64% vs 10.53%, p = 0.028), or who did not take a DM (14.49% vs 1.75%, p = 0.005) or dyslipidemia medication (6.76% vs 0%, p = 0.045). Unmet education needs were significantly more likely to be reported by younger patients (30 vs 40, p = 0.012), patients with psychiatric illness (50.94% vs 29.86%, p = 0.006), and patients without dyslipidemia (19.43% vs 7.55%, p = 0.042). Patients without DM were significantly more likely to request legal resources (22.89% vs 0%, p = 0.047).

Unmet companionship needs were significantly associated with psychiatric illness (69.23% vs 28.00%, p<0.001), taking a psychiatric medication (28.21% vs 14.22%, p = 0.036), and a higher mean number of comorbidities (4 vs 3, p = 0.006). Health literacy needs were significantly associated with dyslipidemia (35.48% vs 14.59%, p = 0.009), gastrointestinal disease (51.61% vs 28.33%, p = 0.012), female (83.87% vs 64.81%, p = 0.041), Hispanic/Latino (54.84% vs 30.47%, p = 0.014), Spanish-speaking (45.16% vs 24.89%, p = 0.030), and older patients (mean age: 47 vs 38, p = 0.012). Patients with a higher mean number of comorbidities (3.0 vs 2.5, p = 0.006) were significantly associated with requesting a community referral to address an unmet social need of any kind.

Univariate analyses were also performed to identify significant associations between lab values (Table 6) with social needs or receiving resources. No significant associations were found for patients who reported unmet employment, education, companionship, childcare assistance, or legal needs. There were significantly lower mean high-density lipoprotein (HDL) cholesterol levels in those who reported food insecurity compared to those who did not (31 vs 48, p = 0.006). There were significantly lower mean hemoglobin A1c levels in those who reported unmet financial needs compared to those who did not (5.9 vs 6.3, p = 0.039). There was also significantly lower mean systolic blood pressure (SBP) in those who reported unmet transportation needs compared to those who did not (119 vs 128, p = 0.033). Finally, there were significantly higher mean total cholesterol levels in those who reported health literacy needs compared to those who did not (224 vs 197, p = 0.037).

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Objective measurements of the patients who filled out the socioeconomic screening questionnaire.

https://doi.org/10.1371/journal.pone.0291682.t006

Discussion

We present an innovative social intervention model in which medical students were mobilized to assess SDH factors afflicting underserved patients in several free community health clinics. Using a validated socioeconomic screening questionnaire, we identified the unmet needs of patients seen in our SRFCs and found that 63% of patients requested assistance finding community resources. Overall, the most reported SDH factor was unmet financial needs (62%) which were significantly associated with a higher number of comorbidities. Our results support utilizing poverty as an efficient, highly sensitive screening topic to identify at-risk patients in need of social resources [17].

Individual-targeted assessments of detrimental SDH factors are effective and promising methods of mitigating key drivers of health inequities [8]. A 2019 systematic review and meta-analysis demonstrated that the health behavior benefits of social interventions were significantly related to improvements in overall well-being, hemoglobin A1c levels, and cessation of alcohol/tobacco misuse (p<0.05) [18]. Absence of these interventions has also been identified as a critical contributor in primary nonadherence to chronic disease medication [19]. However, screening by primary care providers is often limited by their short appointment times, disproportionately high workloads, and management of related administrative responsibilities [13]. Mobilizing medical students as social workers lessens these burdens through the efficient allocation of bureaucratic duties within busy clinics [8, 13, 20]. This included directly assisting patients with complex insurance application forms and integrating community resources to empower future self-sufficiency outside of the clinic [4, 13].

Patients with Type 2 DM who are connected to community social resources report improved self-care, quality of life, and improved glycemic control [1]. Gorrindo et al. demonstrated that appointing even a preclinical first-year medical student to educate SRFC patients about their health and other non-medical needs resulted in significantly improved hemoglobin A1c levels, (9.6 to 7.9, p<0.0001) [21]. Similar to our model, Dadlani et al. also implemented a system to screen for detrimental SDH factors at SRFCs. Their most requested resources were reported as utilities (18%) and mental health (18%), which was slightly higher than our clinics (16% and 15%, respectively) [22]. In contrast, our study identified financial resources as the most common unmet need (62%) and focused on associations between social needs and medical conditions, whereas Dadlani et al. utilized patient zip codes to identify geographic gaps in the distribution of resources [22].

Our results indicate a significant association between reporting unmet social needs and the presence of more cardiovascular risk factors. Notable associations included childcare assistance with the number of comorbidities or medications, food security with the number of comorbidities, housing instability with tobacco use, transportation with SBP, and financial needs with hemoglobin A1c levels. One likely possibility is that patients who are chronically sick and financially disadvantaged are exhausted from the overwhelming wait times in the current healthcare system [23]. Progressively worsening chronic health conditions result in additional appointments that require more time and resources than underserved patients are equipped to meet, leading to avoidance of seeking social services [8, 23]. The significant association between employment stability and DM, taking a DM medication, or taking a dyslipidemia medication may further imply that these patients are forced to work in order to afford the financial burden associated with their chronic diseases [24].

However, there were predictably higher significant mean total cholesterol and lower mean HDL levels seen in patients who reported unmet health literacy and food security needs, respectively. These exceptions could be due to the immediate necessity of food or difficulty interpreting the screening tool at the appointment [3, 8]. We demonstrated significantly more requests for community resources after a simple conversation during a time when patients would be waiting regardless. Given the disparity between cardiovascular disease and the active medication list of our cohort (especially antiplatelet therapy), our results strongly suggest that medical student-facilitated social intervention may contribute to lowering the morbidity and mortality rates associated with cardiovascular disease [25]. A higher number of comorbidities and a history of psychiatric illness were the two most common medical problems significantly associated with several unmet social needs and may be an area of interest for future models.

It is well documented that patients with more chronic health conditions suffer from a higher socioeconomic burden [1]. Psychosocial factors similar to those in our study are responsible for exacerbating the baseline risk factors that cause cardiac events and long-term disability [26]. Neglected complications of chronic, untreated peripheral vascular involvement can also lead to life-altering foot ulcers, especially in minority and socially disadvantaged patient populations [27]. The resulting impact on ambulation and finances can be emotionally overwhelming, and even devastating to patients who inevitably require amputations [24]. Strategies aimed at preventing these outcomes in future, larger social interventions should reflect the needs of the targeted community through a more expansive assessment of well-known risk factors such as tobacco use, family medical history, and metabolic syndrome [26]. Preemptive identification and appropriate management of physical, mental, and social care is therefore critical to improving disproportionate disparities in long-term health outcomes [1, 26, 27].

Conclusions

In summary, we identified a high prevalence of synergistic cardiovascular and psychiatric comorbidities in underserved patients that was complicated by several underlying unmet social needs. We found that 63% of participants requested assistance finding community resources. While the most reported SDH factor was unmet financial needs (62%), a higher number of comorbidities and history of psychiatric illness were the two most common medical problems significantly associated with several unmet social needs. Future research can focus on larger-scale, prospective patient health outcomes based on the success of community referrals aimed at addressing the unmet psychosocial needs of underserved patients.

Supporting information