Highlights and Areas of Inquiry in the HSR Special Issue: Social Care and the US Health Care Sector
这篇特刊导言总结了美国医疗体系中社会照护干预的研究进展，包括跨部门合作、食品与住房支持的效果评估，以及筛查社会风险的方法挑战，对关注健康公平和政策制定的研究者有参考价值。
Recognizing that social context influences health, over much of the last decade, federal, and state policies have encouraged healthcare institutions to identify patients without adequate access to basic material goods such as food, housing, and transportation and to develop partnerships with community-based organizations (CBOs) and social service organizations to support any identified service gaps [1]. As an example, 27 states and the District of Columbia have active or pending Medicaid Section 1115(a) waivers that include provisions related to health-related social need (HRSN), which the Centers for Medicare & Medicaid Services has described as “an individual's unmet, adverse social conditions that contribute to poor health.” [2] These waivers are based on research suggesting that interventions on adverse social determinants will likely improve health [3]. However, recent policy changes threaten the sustainability of these and other “social care” activities undertaken in the healthcare sector. The February 2025 Executive Order “Establishing the President's Make America Healthy Again Commission” does not mention social determinants of health. Instead, the order emphasizes individual accountability and health behaviors [4]. As this special issue was going to press in March 2025, the Centers for Medicare & Medicaid Services rescinded prior guidance about using Medicaid dollars to address Medicaid members' HRSN [5]. As these policy shifts play out, it is pertinent to remember that long-term health outcomes are shaped more by social factors than health behaviors [6]. However, the country is not aligned on how health policies should incorporate this fact. We believe disseminating what we have learned about social care—whether at patient, community, or policy levels—is key to strengthening alignment. At the very least, disseminating relevant research can improve awareness and provide insights in the face of great uncertainty. In this Special Issue, “Social Care and the US Health Care Sector,” we partner with the Social Interventions Research and Evaluation Network (SIREN) to elevate studies exploring how health care systems, CBOs, and social service agencies are aiming to improve health by assessing patient- and population-level HRSN and attending to identified needs. Below, we highlight ways that studies in the Special Issue build on the existing research in this field, and we highlight persistent gaps in knowledge. Several 2025 SIREN Special Issue articles describe community or policy-level interventions to improve social and medical care integration. California's Medi-Cal Whole Person Care (WPC) program was assessed through qualitative and mixed-method approaches by Chuang et al., and Safaeinili et al. [7, 8] The WPC initiative is a $3 billion social care intervention, offered through California Medicaid Section 1115(a) waiver demonstration. The WPC initiative pilots programs to deliver integrated care across multiple service sectors to address HRSN. Utilizing data from the WPC program, Chuang et al. identified 11 pilots that improved cross-sector collaboration and 15 that did not. Cross-sector collaboration was facilitated by collaborative governance and braided funding to support integrated care and meaningful use of data-sharing systems [7]. In a qualitative study of the WPC intervention, Safaeinili et al. assessed the sustainability of the 26 WPC pilots that integrated medical and social services for Medicaid enrollees. Twenty-two sustained the WPC model. Four factors facilitated program sustainability: adaptability and flexibility, consistent funding and reimbursement infrastructure, collaborative leadership with managed care plans, and contracts with partner organizations [8]. These studies contribute to understanding organizational factors that can promote the collaboration and sustainability of social care to improve members' health and well-being. Other articles in this special issue examine the social and health impacts of social services at a patient level. Thompson-Lastad et al. studied the effectiveness of Recipe4Health (R4H) in Alameda County, CA, a community-level social care intervention addressing food insecurity to improve mental health outcomes. As part of R4H, eligible participants received optional group medical visits from trained community health center (CHC) staff who provided nutrition and cooking education, movement and relaxation techniques, health coaching, and weekly home delivery of fresh fruits and vegetables. Eligible participants of R4H were racially and ethnically diverse children and adults who demonstrated either food insecurity or a nutrition-sensitive health condition. Data collected between 2020 and 2023 from four of the five participating CHCs suggested that for patients who only received food deliveries, food security improved for participants with baseline depression and anxiety. For patients who received both food deliveries and group medical visits, depression, anxiety, and food security significantly improved. Qualitative interviews also showed that group medical visits helped to reduce pervasive loneliness among participants. The subgroup analyses in this study are especially interesting since among those who received food deliveries and group medical visits, loneliness significantly improved only in those without baseline depression or anxiety. One takeaway is that subgroup analyses are critical for better appreciating for whom these programs work and under what conditions [9]. Chapman et al. evaluated Supportive Services for Veteran Families (SSVF) targeting short- and long-term housing outcomes among veterans struggling with housing instability. To discern SSVF's effectiveness, the research team examined clinical data from Veterans Affairs electronic health records (EHRs). They used natural language processing (NLP) to “assign” participants to SSVF or standard care in 36 simulated trials. In total, 238,059 patients were “enrolled” in one or more “trials.” The authors found that the risk of being unhoused was reduced in both intervention and control arms but was more pronounced among the intervention groups in the first 120 days. The authors note that significant limitations of this study include using data prone to error from NLP and selection bias in the simulated assignment to intervention and control groups [10]. One article also examines the impact of social care on health care costs and health care utilization [11]. Lovelace and colleagues used a pre-post design to examine changes in Medicaid and dually eligible participants of a supportive housing program who enrolled in Cultivating Health for Success (CHFS). CHFS is a joint coordination program between a payer organization and community-based housing partners that provides HUD vouchers, Medicaid premiums to cover other housing-related costs, a personalized health care plan, and residential support from community partners. The authors retrospectively analyzed over 6 months of data from 80 Medicaid recipients in an urban Pennsylvania county between 2018 and 2023 to determine health care costs and utilization rates, comparing 12 months before enrollment and 12 months after enrollment in CHFS. In the 12 months post-enrollment, Medicaid recipients demonstrated significantly lower utilization of emergency department services, reduced primary and specialist care use, and lower ED and total health care costs compared to the pre-housing period. The study highlights the vital contributions of CBOs, which, in this case, engaged unhoused participants. As efforts to address social determinants become increasingly integrated within health care settings, scientists have attempted to develop novel ways of screening for social risk, with varying degrees of success. Kim et al. utilized cTAKES, a form of NLP, to extract social needs information from EHRs since screening patients can be inefficient. While cTAKES was somewhat successful in extracting information related to housing insecurity, it was unsuccessful in extracting information related to food insecurity [12]. The study by Dankovchik et al. also centers on the opportunity to use EHR data to identify social risk referrals on a national network of community-based clinics [13]. This paper discusses how health care providers are increasingly called upon to report on social risk referrals. If a provider goes through with a referral, they will be incentivized through reimbursement rates. However, the paper emphasizes that the difficulty with providing these referrals is that if there are no standardized methods to document referrals in EHR, this may pose barriers to reporting. To investigate this topic further, researchers developed a study design that included smart phrases and procedure codes identified as potential indicators of a social risk referral. Furthermore, the report mentions that using the structured data fields currently in place with existing EHR data is labor-intensive and impractical for clinic use. This is the case for reporting and analyzing food insecurity, financial stress, and housing needs in social risk patterns. A significant barrier to reporting social risk referrals is that community-based clinics may have difficulty using smart phrases and procedure codes, as mentioned above, to monitor and report such occurrences effectively. More standardized methods for documenting these activities are recommended. Past research has suggested that the capacity to address social needs can vary by locality and the type of social care needed [14, 15]. This was supported by Brewster et al., who suggested that specific patient goals related to social needs (e.g., housing) were less likely to be resolved than others, like food insecurity [16]. Multiple studies explore identifying patients who may benefit from social care interventions. In one study comparing implementation supports for social risk screening in CHCs, Hessler-Jones and colleagues use a difference-in-difference design to evaluate two social risk-screening interventions (financial incentives versus practice facilitation) and one combined intervention in 32 intervention clinics and 32 control clinics. Their results show that average monthly social risk screenings increased only modestly post-intervention with both financial incentives and practice facilitation interventions, suggesting the need to consider additional investments for more significant screening gains [17]. Some of the answers to this conundrum of how to improve social risk screening come from the study by Olson et al., a mixed-methods, community-based participatory research study that assessed patient experiences with “Resourceful,” a community resource referral system. “Resourceful” relates to health systems' EHRs, which health care workers and CBOs use to track social care referrals. Data were drawn from online surveys with 62 patients and 14 health care workers and semi-structured interviews with 6 patients and 5 health care workers. The information elicited described the significance of “Resourceful” in facilitating trust, positively impacting patients' experiences of finding and connecting to help, improving communication between patients and health care workers, and accessing resources that addressed patients' HRSN. Interviews with patients suggested that high-quality, reliable relationships between workers and patients, an accessible “Resourceful” platform to identify resources, and a diverse set of reliable referral organizations were essential to reducing social care barriers. However, there was a substantial difference between health care workers' perceptions of the patient's experiences with the program and those expressed by patients themselves—health care workers consistently overestimated how beneficial “Resourceful” was for patients across multiple experiential dimensions compared to patients' perceptions [18]. While the studies in this special issue offer unique insights into the rapidly evolving field, there is much more to learn. None of the studies included in the Special Issue were randomized controlled trials, and only one used a high-quality, quasi-experimental design to assess the effects of the intervention [17]. Of the studies included, six were mixed methods, five included qualitative data only, and five included only quantitative data. Data included survey results, claims, clinic notes, and semi-structured interviews. The significant heterogeneity in methodological and measure approaches across these studies is consistent with findings from a previous review of interventions to address social determinants, which found that most social care interventions were tested via cross-sectional or quasi-experimental designs, and findings were often incomparable across studies [19]. This stems partly from the fact that many social care interventions address only one or a limited number of HRSN [20] although many individuals face a wide range of health and social needs [21]. These needs are not consistently accounted for when testing interventions, even in randomized controlled trials (e.g., trials of Housing First [22, 23]) [21]. Furthermore, relatively few studies in the special issue evaluated specific health outcomes. This again reflects trends in the broader literature, where studies predominantly focus on process measures and social risk outcomes rather than health or health inequality outcomes [24, 25]. To strengthen social care research and inform new investments in social care, future research will need to rigorously examine relevant health, health equity, and cost outcomes. HRSN-related interventions require significant monetary investment—11 of these studies occur in large healthcare systems, such as the Veterans Health Administration or Kaiser Permanente, or large initiatives and collaboratives focused on social drivers of health and HRSNs, such as WPC. Smaller health systems, independent practices, and community-based clinics not involved with state or federal efforts to promote social risk screening and interventions may not have the financial and personnel resources to identify and address patients' social needs. Research that supports developing, testing, and implementing well-designed interventions is needed to ensure that these kinds of healthcare providers can identify and meaningfully address HRSN in their patient populations. Additionally, the evidence in this special issue disproportionately reflects efforts in the western part of the United States—nearly half (7) of these studies take place in California exclusively, with at least two others set primarily in the Pacific Northwest. Only two studies use representative claims data to understand HRSN at the national level. Thus, the evidence generated may not generalize to other parts of the United States, where inequities in access to health and social care by race, ethnicity, sexual orientation, gender identity, and rurality are also potentially significant drivers of differences in health and health-related outcomes [26, 27]. While the health services research field has made significant strides in the breadth of evidence on social care, there is room to improve the quality and design of studies. Although progress may be slow due to possible federal disinvestments in HRSN, a clear focus on producing rigorous evidence will ultimately enable better, more informed policies in social determinants down the line. There is no funding to report. The authors declare no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.