Addressing Social Needs in Health Care Settings: Evidence, Challenges, and Opportunities for Public Health

Abstract

There has been an explosion of interest in addressing social needs in healthcare settings. Some efforts, such as screening patients for social needs and connecting them to needed social services, are already in widespread practice. These and other major investments from the healthcare sector hint at the potential for new multi-sector collaborations to address social determinants of health and individual social needs. This paper discusses the rapidly growing body of research describing the links between social needs and health and the impact of social needs interventions on health improvement, utilization and costs. We also identify gaps in the knowledge base and implementation challenges to be overcome. We conclude that complementary partnerships among the healthcare, public health and social services sectors can build upon current momentum to strengthen social safety net policies, modernize social services, and reshape resource allocation to address social determinants of health.

Introduction

Efforts to address social determinants of health (SDOH), long a priority for public health professionals (55), have been re-energized by recent attention and investment from the healthcare sector, spurred by a range of opportunities, initiatives and incentives, including the Affordable Care Act (102). This paper reviews the rich and rapidly growing body of research driving the efforts to address SDOH in clinical and community settings, and identifies key gaps and promising future directions for new scientific approaches to addressing social needs in health care.

The World Health Organization’s (WHO) oft-cited definition of SDOH includes political, social and economic forces that affect health by shaping the conditions in which people live (112). These determinants are usually viewed as system- and policy-level influences that affect everyone in a society and can either drive or reduce inequities in housing, education, jobs, pay and other social institutions and opportunities. At a person-level, the downstream consequences of SDOH for people who have been systematically disadvantaged include unmet social needs, such as unstable, unaffordable and/or low-quality housing, food insecurity, unemployment, lack of quality, affordable child care, or needing utility payment assistance.

The WHO’s landmark “Closing the Gap” report (2008)(112) outlined three broad actions to address social determinants of health and achieve health equity: (1) improve daily living conditions; (2) tackle inequitable distribution of power, money and resources; and (3) measure and understand SDOH and assess the impact of action. The report’s conceptual framework (97) casts healthcare systems as an intermediary determinant, and suggests that through intersectoral actions like food and transportation assistance, healthcare can help address inequities in “material circumstances” such as housing, financial means to buy food and clothing, and neighborhood conditions and safety.

The healthcare sector’s recent involvement in SDOH has followed the WHO conceptual model and recommended actions fairly closely: assuming responsibility for measuring the problem (e.g., screening patients for social needs), assessing the impact of action (e.g., intervening on social needs and tracking outcomes), and focusing on deprivation in areas of material circumstances (as opposed to more upstream determinants like laws and policies). Castrucci and Auerbach (2019) correctly argue that these person-level experiences such as food and housing insecurity are really individual social needs, not SDOH (25). Although sometimes still referred to as SDOH by healthcare systems and in research (4), the term “social needs” is increasingly being used as shorthand for unmet material needs experienced by individuals.

The momentum to address social needs in health care settings is now visible in several routine practices. Recent national surveys of Medicaid managed care plans, hospitals and health systems show that in particular, two individual-level approaches – screening for social needs and making referrals to social services – are already in widespread practice (6, 67). However, not all healthcare organizations have a formal process for screening and referral (67) and rates of uptake appear to be lower among healthcare providers in low-resource settings (82). The growing interest and investment in screening and referral for social needs has already spawned dozens of new technology platforms designed to facilitate the process of addressing social needs (23).

Many larger health systems have also invested in community-level and structural solutions. According to one study of 57 health systems, $1.6 billion was spent from 2017-2019 on housing interventions alone, including construction of affordable housing for homeless patients and those with high use of healthcare, helping employees purchase homes in neighborhoods targeted for revitalization, and eviction prevention programs (51). Although these tended to be pilot projects undertaken by larger health systems in partnership with other community organizations (and generally not represented in the scientific research literature), they are nonetheless major investments that hint at the potential for multi-sector collaborations to address these needs.

At present, however, such partnerships may be the exception: only 30% of hospitals and health systems in a national survey reported having fully-functional formal partnerships with community-based social needs providers, and 70% did not have dedicated funds to address social needs for all their target populations (67). Moreover, many health systems likely lack community-level social needs data to inform investments of this scope. Studies show that SDOH information is the least developed component of the Community Health Needs Assessments that are required of all 501(c)(3) organizations that operate hospitals (45, 83, 101).

The scientific literature on healthcare efforts to address SDOH is overwhelmingly skewed towards research on individual social needs, which is the focus of this paper. With 12 new studies of social needs interventions published every month (32), special supplements of leading journals dedicated to social needs research (40, 110), and a growing number of evidence reviews on the subject (9, 42, 70, 96, 104), it is important that public health professionals understand this research and practice, its complementarity with public health efforts to address SDOH, and the opportunities it provides for partnership to improve population health and reduce disparities.

Drawing on this growing body of research, we describe the social needs experiences of diverse populations, the effects of social needs on health, and the impact of interventions on social needs and health improvement, utilization and costs. We conclude by identifying gaps in the knowledge base and implementation challenges to be overcome, and we suggest future directions for a science of social needs.

Assessing social needs

Although there is strong interest in community-level indicators, social needs are most commonly measured among individuals, using self-report. Healthcare organizations have a long history of screening for specific concerns (e.g., interpersonal safety) in specific clinical populations (e.g., pregnant women, seniors, pediatric patients) or settings (e.g., emergency departments) (5). Today’s social needs screeners are more multidimensional and include a range of screening tools (39, 49, 77, 78) that vary in the total number of questions asked (e.g., 2-23), the time interval assessed (e.g., needs experienced in the past 12 months, current needs, anticipated need in the next month), whether needs are assessed for the respondent only or all members of the household, and the different sets of needs assessed.

In 2015, the Institute of Medicine recommended 11 key measures including race/ethnicity, education, financial strain, stress, depression, interpersonal safety, tobacco use, physical activity, and social connections (2, 54). A Centers for Medicare and Medicaid Services (CMS) measure covers five domains – housing, food, transportation, utilities and safety/abuse – using 10 items (15). The American Academy of Family Physicians (AAFP) added to the CMS measure questions assessing education, employment status, financial insecurity and childcare needs (78). The Health Leads screening tool also assesses whether any of their social needs were urgent, and whether they needed help reading hospital materials (49). The PRAPARE tool adds items assessing incarceration history in the past year, stress level, and frequency of feeling socially connected (77).

Across these and other social needs screening tools, the most commonly included items assess needs related to food, utilities, housing, transportation, and personal safety. Within these multi-dimensional domains, screening tends to focus on aspects of the need that align with priorities of the healthcare organization or its perceived locus of influence. For example, screening items are more likely to assess housing stability than housing quality, even though both can have a great impact on people’s lives and health. Most screening tools do not include an open-ended question about other social needs patients may want to report that were not already addressed, which could improve the patient-centeredness of screening tools and elicit additional needs.

In addition to assessing social needs, the AAFP and Health Leads screeners also assess whether a respondent wants assistance addressing a social need. Some research suggests many may not (28). For example, a multi-site survey of 969 adult patients and caregivers of pediatric patients across nine states found that screening for social needs in healthcare settings was widely viewed as appropriate (29), but qualitative interviews with a subset of respondents revealed that not all wanted help from healthcare teams addressing social needs (22). Interviews showed that patients wanted their healthcare providers to be aware of their social needs, but did not expect them to resolve the needs; some even felt it was outside the scope of medicine to do so. In related findings, only 40-60% of people who report an unmet need agree to participate in navigation or other social needs programs (11, 39, 82, 92).

Some social needs screeners include items that identify non-modifiable factors that could influence the types of social need assistance available to a person. For example, incarceration history or veteran status are not social needs, but could affect a person’s eligibility for services to address social needs (3, 77). Use of such items and treatment of the resulting data should be carefully considered to avoid unintended consequences. For example, a history of incarceration might bias the treatment a patient receives, and remain in their electronic medical record long-term.

Although most social needs screening tools in widespread use were designed for adult patients, many screeners have also been developed for and/or administered to parents of pediatric patients (74, 81). A recent review identified 11 different social needs measures for pediatric patients, most of which addressed social needs in the broader family context (96).

There is considerable variability in the wording, response format and time frame assessed in different social needs screening items, as well as in the procedures for administering them. These differences across studies make it difficult to directly compare results (50). In some screening tools, a single item assesses multiple needs (e.g., “How likely is it that you will have enough money for food, rent and clothing in the next 6 months?”), making it difficult to determine exactly which needs are unmet. Other questions require respondents to connect in causal fashion a social need to a health outcome (e.g., “Within the past 12 months, I couldn’t afford to eat balanced or healthy meals”) (94). These examples also illustrate the variability in time frame considered for each social need exposure; some are prospective while others are retrospective, and the relevant time period can range from a day or week to a year (15).

There is no consensus on how or how often patients should be screened for social needs. This includes determining which patients should be screened, in which settings, and by whom. Currently, social needs screening is administered in diverse healthcare contexts (e.g., community health centers, emergency departments, inpatient services), with different populations (e.g., pediatrics, those with chronic disease, low-income patients), using different modalities (e.g., verbal, paper, tablet computer), and implemented by a range of interviewers (e.g., doctors, nurses, social workers, community health workers, volunteers), and during different points in a clinical encounter (e.g., waiting room, exam room, post-visit). Some have cautioned that selective screening – assessing social needs only in certain patient populations, based on perceptions or data on place of residence, race or ethnicity, or perceived education – could erode patient trust and exacerbate stigma, discrimination, and health disparities (36). Screening all patients for social needs would reduce this risk (71).

Prevalence of social needs

Unmet social needs are widespread in marginalized populations. Studies assessing the prevalence of social needs have relied mostly on self-report from screening questions, though some have constructed social needs indicators based on administrative data (58) or tracked requests to community helplines (30; N. Verdecias, R. Garg, J. Steensma, A. McQueen, R. Greer, M.W. Kreuter, manuscript in review). Studies frequently report the prevalence of individual social needs as well as the total and/or mean number social needs experienced, including the percent with 0, 1, or 2+ social needs. Table 1 reports findings from 15 studies to illustrate the prevalence of selected social needs in different populations and healthcare settings.

Table 1.

Sample of studies that report prevalence of social needs.

Study information			Social needs prevalence			Individual social needs
	N	Study population	Social needs assessed	Total needs (mean)	% with 0/1/2+ needs	Housing	Utilities	Food	Jobs	Transportation	Health care access	Child care
Clinical setting
Berkowitz, 2016	416*	Urban primary care	8	2.6	NA	21	36	40	20	15	47	ND
Bisgaier, 2011	1,506	Urban emergency dept	5	NA	52/17/31	18	ND	23	19	ND	26	ND
de la Vega, 2019	1,696	Urban adult primary care	8	NA	74/NA/NA	8	6	11	12	7	11	3
Garg, 2015	336	Mothers at urban CHCs#	6	NA	10/22/68	43	9	20	57	ND	ND	29
Gottlieb, 2016	1,809	Parents at safety net hospitals	14	2.7	17/NA/NA	29	41	41	31	ND	21	ND
Hassan, 2015	401	Urban young adult clinic	9	NA	24/28/47	34	ND	29	10	ND	37	ND
Heller, 2020	24,633	Primary care patients	10	NA	80/10/10	5	3	6	ND	5	4	3
Page-Reeves, 2016	3,048	Primary care patients	11	NA	54/17/29	11	22	12	15	9	ND	3
Polk, 2020	10,916	Patients at pediatric practices	11	3.2	NA	22	25	30	22	3	15	37
Community or non-clinical setting
Emechebe, 2019	13,708*	MCO helpline callers#	NA	NA	NA	6	20	15	ND	17	7	ND
Kreuter, 2020 (Study 1)	1,898	2-1-1 callers	7	2.4	5/16/79	17	ND	15	ND	ND	ND	ND
Kreuter, 2020 (Study 3)	10,267	Medicaid enrollees	10	1.1	45/28/27	3	12	6	ND	9	ND	43
Kreuter, 2020 (Study 4)	1,370	2-1-1 callers	10	2.5	12/20/68	10	48	15	ND	21	ND	50
Schickedanz, 2019	3,721	Potential high use patients	15	NA	47/NA/NA	6	19	29	5	17	17	2
Thompson, 2019	1,214	Medicaid enrollees	7	1.3	32/32/36	6	ND	10	ND	ND	ND	ND

^*N only includes participants who reported at least one social need to reflect how prevalence data were reported in the study.

^#CHC = Community Health Center; MCO = Managed Care Organizations.

NA = data not available (i.e., not reported)

ND = no data collected

Studies find that participants frequently have 2 or more social needs, even though most assess a relatively small number of needs (38, 41, 62, 64, 86). Studies that screen a broader patient population tend to find fewer social needs per person (50). In Table 1, findings are grouped into two types of social need studies. Studies in clinical settings assess social needs during healthcare visits. This screening usually occurs in person but sometimes through other modalities such as online surveys (48). Studies in a community or non-clinical setting assesses social needs either independently from healthcare, among individuals who are seeking assistance through social service helplines, or through phone outreach to members of a healthcare plan or system.

Social needs related to housing, food, childcare and general financial strain are among the most common social needs experienced. In Table 1, 5-43% of participants screened in clinical settings had housing stability needs. Other housing related needs, including low-quality housing or limited space in the home, are often more prevalent than housing instability in studies that assess both (50, 64, 92, 105). However, many studies report housing stability alone and do not assess quality separately (11, 16, 27, 37, 38, 79, 80, 88, 95, 99).

Food insecurity, operationally defined in various studies as eating less, skipping meals, not having enough food for one’s family, or running out of food before having money or food stamps to buy more, is one of the most highly prevalent needs (11, 27, 41, 48, 79, 88, 95). In Table 1, 6-41% of participants screened in clinical settings had food-related needs.

Among adults with young children, the need to find quality and/or affordable childcare is highly prevalent (37, 38, 62, 86), although it often appears less so in general population screening when the denominator includes those without young children (27, 50, 80, 92). In Table 1, 29-50% of adults with young children reported childcare needs while only 2-3% reported childcare needs in general population studies that also included those without young children.

General financial strain is one of the most commonly identified social needs across studies in many different settings, although screening questions for financial strain vary widely, making comparisons of prevalence across groups difficult. For example, Kreuter et al found that not having enough money for unexpected expenses was the most prevalent social need (47-89%) in four independent studies (62). Other studies also found that financial strain was the most common social need, either in the context of paying for health care (11, 16, 99) or general employment and income concerns (27, 37, 38, 58).

These rates of housing, food, childcare and financial needs among low-income samples are generally higher than population-wide estimates from public health surveillance efforts. In the U.S., slightly more than 10% of households experience food insecurity at some time each year (106). Around 7% of U.S. households are cost burdened and around 0.2% of Americans are homeless (108).

The most prevalent needs can vary widely by study setting. Among helpline callers, for example, social needs related to utility bill payment and transportation are more common than they are among populations screened in healthcare settings (30, 62, 64). This may be because screening in healthcare settings often assesses transportation needs in the specific context of healthcare access, and because those who arrive for a healthcare visit have overcome, at least temporarily, whatever transportation need they may have.

Studies in clinical and community settings have identified several common correlates of having more unmet social needs, including lower income, less education, and unemployment (11, 47, 62, 64, 95, 105). For other demographic characteristics, however, findings are mixed. Some studies have found that older participants report more social needs (62), while other studies have found that younger participants have more social needs (87). There have also been conflicting findings about the association between gender and social needs (11, 50, 62).

Although national surveillance data suggest that members of racial and ethnic minority groups are at increased risk for food insecurity and housing instability (107, 108, 109), studies of social needs have not found a consistent association. One study among helpline callers found that White participants were more likely to have social needs (62), while other studies have found no association between race and social needs (47, 62, 95) or that Black or Hispanic participants are more likely to have social needs (11, 50, 62, 87). The conflicting findings may be explained, at least in part, by differences in study samples and/or measures.

The link between social needs and health

Both in the U.S. and globally, broad social determinants such as income and education are associated with health outcomes including chronic disease and mortality (1, 20, 21, 68, 69, 98). There is also a growing body of research linking individual social needs to a range of health outcomes from behaviors to mortality to health care utilization and costs (76). For example, material need has been associated with lesser access to and use of treatment and prevention services, later diagnosis and resolution, and greater hospitalization, length of stay, readmission, complications, and mortality (60). These associations may be mediated by lack of health insurance (111). Because much of this research is based upon cross-sectional data, causality often cannot be established and could be bi-directional or reversed (72). For example, longitudinal studies have shown basic needs predict depressive symptoms (18), but also vice versa (52).

Many studies have examined the relationship between a single social need and a particular set of health outcomes. Most commonly, this research has focused on social needs related to food and housing. Food insecurity is associated with negative health consequences across the life span including obesity, stunting, wasting, and cardiometabolic disease (24). It has been associated with sleep problems (44), inflammation (43), poor diabetes control among diabetic patients (12), and poor health among children (7, 89). Among the general U.S. population, adults reporting food insecurity have higher subsequent health care expenditures (10), whereas low-income adults who participate in the Supplemental Nutrition Assistance Program have lower health care expenditures compared to those who do not (13).

Housing insecurity and homelessness are also associated with negative health outcomes among adults (66, 100) and children (26). Low-income adults that receive rent assistance have three-fold lower odds of rating their health as “fair” or “poor” compared to peers on a waitlist for rest assistance (59). Among Canadian adults with HIV, having at least one unmet need (for food, clothing, or housing) was associated with lower physical and mental health quality of life (95).

There is strong evidence for a dose-response relationship between social needs and health. Studies show that increased social needs are associated with worse physical, mental and self-rated health, more chronic conditions, depressive symptoms, and higher perceived stress (11, 16, 62, 64, 95, 105). Among children, higher levels of unmet needs are associated with lower levels of child wellness (34).

Social needs are also associated with a range of health-related behaviors including smoking, illicit drug use, eating fewer fruits and vegetables, getting less exercise, getting less sleep, and not seeking preventive healthcare (16, 58, 62, 64, 84, 105). For many of these behaviors, the association with social needs also follows a dose-response gradient.

Although most research linking social needs with health behaviors and health outcomes has been cross-sectional, longitudinal studies have also established that social needs predict negative health outcomes. In community-dwelling adults, higher levels of unmet social needs were associated with increases in depressive symptoms (18), increases in problems with physical functioning (90) and higher mortality (17).

This accumulation of evidence has led the National Academies to conclude, “The consistent and compelling evidence on how social determinants shape health has led to a growing recognition throughout the heath care sector that improving health and reducing health disparities is likely to depend—at least in part—on improving social conditions and decreasing social vulnerability” (76; p. 27).

More evidence is needed about the mechanisms through which unmet needs affect health. Several conceptual models and frameworks have been proposed to explain the indirect effects of addressing social needs on health outcomes: through reduced stress and competing demands for resources and improved adherence to medication and physician visits (46, 75), and through improved health behaviors, physiologic functioning, and psychosocial factors (8, 9, 21, 72, 98).

Social needs interventions

Some social needs interventions focus on addressing a particular need, while others are increasingly addressing a range of needs (9, 35, 42). Among interventions addressing multiple social needs identified through screening, the simplest approaches compile and distribute social needs resource guides, often with little follow up or evaluation. More comprehensive “linkage” interventions involve systematic approaches to screening for social needs followed by social prescriptions or referrals for specific resources available from independent or partner community organizations (e.g., food bank, YMCA) or a co-located service provider (e.g., medical-legal partnerships).

Given eligibility requirements and limited resources at many social service provider agencies, however, there is no guarantee that linking patients with community programs will resolve the patient’s social needs. Although some patients may be able to resolve their unmet need independently using a resource information booklet or following a verbal or written referral for an assistance program, others will be more successful with personal help from someone like a case manager or social needs navigator (72) who can advise them on the documentation needed to meet eligibility criteria, help them connect with the agency, and assist them with applications.

Social needs navigators are often affiliated with a healthcare organization or community agency (e.g., social workers, nurses, community health workers, volunteers) and provide ongoing support and follow up in person and/or by phone. In many healthcare organizations, patients with high acuity or utilization are offered case management or navigation services to address health and social needs (i.e., hotspotting (33)), but it is less common for navigation programs to be made available universally to a patient or member population, without regard to individual differences in health risk profiles.

Several reviews of the literature have identified and summarized interventions designed to address social needs. Most studies report the prevalence of needs and what percent of needs were resolved through referrals over a particular time period (19, 38); fewer report health outcomes or cost savings. Results have varied based on which needs were addressed. In one review, intervention studies addressing housing needs found effects on health outcomes, costs, or both, whereas studies of nutrition, income, or care coordination supports were more sparse and had mixed effects (104). Several interventions have focused on housing and childhood asthma triggers; some reported reductions in the use of urgent health care services and increases in symptom-free days and quality of life (9). Incorporating social work interventions within primary care has shown promising effects on subjective health measures, self-management of chronic conditions, and reduced psychosocial morbidity and barriers to care (70).

There is also evidence for benefits of social needs navigator programs that address multiple needs. For example, one study determined that having a university-affiliated navigator for 3 months led to greater improvement in parents’ evaluation of overall child health than a printed community resource guide (41). Another found that having a community health worker provide navigation for 6 months was more effective and cost-effective in reducing hospitalization and costs compared with standard care (56, 57). A comparative-effectiveness study showed that delivering navigation via a clinic setting vs. a community setting did not differentially affect psychosocial outcomes (85). An evaluation of Health Leads programs across three sites found that patients who screened positive for unmet social needs and agreed to participate in navigation showed improvements in blood pressure and cholesterol (but not HbA1c) over time compared to those who declined participation (11). Providing a medical-legal intervention to parents with newborns increased access to supports (e.g., food, utility assistance, public assistance), increased rates of immunization and preventive care, and decreased emergency department visits (93).

Although individual studies provide promising results for navigation interventions, reviews of the literature identify multiple areas for methodological improvement to strengthen our knowledge of what works (9, 42, 104). Most trials to-date have used social needs screening tools with unknown reliability and validity and focused on process outcomes. Few involved large samples or included cost-effectiveness analysis, some lacked appropriate comparison groups to provide rigorous results that are generalizable to broad patient populations, and long-term follow-up varied. Additionally, many trials involve pre-post designs among high-risk or high-utilizing patients; thus results may be confounded by regression to the mean, as seems to be the case in hotspotting interventions (33). It is also important to evaluate whether interventions to reduce healthcare utilization (less ED use) are beneficial and not potentially harmful (e.g., having unintended consequences such as reducing preventive care) (14, 92).

In addition to the need for more rigorous studies evaluating whether social needs interventions are effective, there is an equally important need for research seeking to explain how these interventions influence health and social needs outcomes. Fichtenberg and colleagues (2019) suggest four potential pathways: increasing patient access to resources, reducing patient stress, helping healthcare providers give better care, and reducing provider burnout (32). The latter two of these are provider-related, and have received less attention in research conducted to date. It’s possible that knowing a patient’s social needs could alter in a quality-enhancing way a provider’s approach to medication management, behavior change recommendations, or other patient interactions, or that knowing your healthcare system has an infrastructure and resources to assist with social needs reduces providers’ stress and worry when treating patients with unmet social needs. Studies exploring these and other mechanisms of effect will advance the science and practice of addressing social needs.

Challenges and next steps

A key assumption underlying the strategy to address social needs through screening and referral interventions is that social service agencies have sufficient capacity to address the demand in their communities. A few longitudinal studies have tracked the outcomes of social needs referrals and found that overall, only about one-third of those receiving a referral end up getting assistance from the referral agency (19, 91). But the capacity of social service agencies can vary widely by community and type of social need (86), with community capacity higher in urban versus rural areas, for example, or higher for responding to food-related needs compared to housing needs (63). Screening and referral interventions may not be suitable for social needs that cannot be reliably addressed due to limited capacity of the local social service system.

Our current knowledge of how a person’s social needs change over time is limited. In both research and practice to date, screening or assessment of social needs is at best sporadic, and when it is assessed at multiple time points there are often long gaps between assessments. This is not a problem if social needs are relatively stable, but there is no evidence supporting that. Rather, studies examining the financial needs of low-income individuals suggest that they fluctuate frequently and often dramatically (73, 103). Thus changing needs over time cannot be resolved by a one-time intervention and may perhaps be more effectively addressed by longer-term interventions such as ongoing navigation. Understanding the patterns, timing and sequences in which some social needs rise and others fall could also help social needs interventions shift from being reactive to proactive, and could even help identify issues that require addressing underlying causes a population level. Longitudinal research measuring dynamic needs and their effects on each other and on health is needed.

Current screening and referral interventions tend to treat each social need a person reports as independent. In other words, if a person has food and transportation needs, they would likely receive one referral for food assistance and another for transportation services. While having the advantage of being simple and straightforward, this approach ignores possible links among different social needs and may use limited community resources inefficiently. As an example, needing utility payment assistance and money for daily necessities like food, clothing and shelter are highly correlated across multiple social needs studies (r_s = .49 to .71) (62). When planning social needs interventions, it’s possible that addressing one need – say transportation, by providing a needed car repair or bus pass – would leave enough money in the person’s budget to help meet food needs or increase their ability to get to food stores. Future work should identify and test promising alternative strategies that solve individual needs while preserving limited community resources.

It is also plausible that certain social needs may cluster in certain population sub-groups (61). For example, a proof-of-concept analysis showed that compared to others in a low-income sample, women with children were more likely to report not having enough space in their home and needing help with utility bills, men under 50 were more likely to report being physically threatened and needing a place to stay, and adults 50 and older in fair or poor health were more likely to need food and transportation (62). Hudson and colleagues argue that Black men may experience distinct stressors over the life course and may benefit from social needs interventions that address their particular social, legal, or economic needs (e.g., getting disability benefits or clearing criminal and credit histories) (53). If certain combinations of social needs occur more commonly in distinct population sub-groups, healthcare and other organizations could develop integrated packages of interventions supported by coalitions of community partners to increase effectiveness. We found no published studies comparing effects of one-size-fits-all social needs interventions with interventions that are highly targeted for specific population sub-groups with similar social needs experiences.

Understanding the complex and dynamic relationships among social needs could also make screening for social needs more efficient. For example, compared to those who do not express transportation needs when calling a community helpline, individuals who do express transportation needs have three times greater odds of also having food needs and twice the odds of needing health insurance or a regular doctor, even though they had not sought help for food or healthcare needs from the helpline (N. Verdecias, R. Garg, J. Steensma, A. McQueen, R. Greer, M.W. Kreuter, manuscript in review). Identifying such interdependencies could lead to adaptive social needs assessment tools that are at once shorter and more informative.

All of these considerations – community capacity, temporal patterns and clusters of social needs, common social needs experiences of different population sub-groups, and smart social needs screening tools that are adaptive and conditional – have practical implications for delivering social needs interventions. They help answer key questions for advancing the field, including which social needs to target and when, which individuals might benefit most from which interventions, and which strategies can identify and address social needs most efficiently, cost-effectively, and sustainably (32).

At the same time, even if social needs interventions in healthcare evolve and are increasingly guided by a robust science of social needs, they should be understood in a broader context and optimized through complementary partnerships and policies. Not all people have health insurance or access to healthcare. Integration of healthcare and social services provides a useful example of the range of collaborative possibilities. On a continuum of degrees of integration, screening and referral interventions that send patients from a health plan to certain social service agencies would rank low compared to community-wide collaborations that reach and benefit all people, not just health plan members (31). In fact, different health plans that implement similar social needs referral programs may ultimately compete with each other for the same social services for their own members.

Addressing people’s social needs in a way that works and lasts will be difficult and likely expensive. Results may be modest and the time horizon for seeing health benefits long, or perhaps never if underlying causes are not addressed upstream. There is a real risk that the healthcare sector could lose interest along the way, decide that the investment is not worth it, or conclude that others are better suited to address social needs. That would be a tremendous missed opportunity. Sustaining the attention, interest and investment of the health care sector must be a high priority. Public health professionals and organizations are needed to help build, lead, or participate with healthcare organizations in cross-sector and multi-level community efforts to improve population health and reduce disparities by addressing SDOH and social needs.

Efforts at all levels of the continuum will work better if there are investments in and modernization of the social service sector, including social safety net policies and resource allocation to sustain them (40). As health care interventions to address social needs continue to be developed, refined, and tested, it is crucial for public health professionals to also strive to shape the upstream SDOH that drive both health disparities and the inequitable distribution of unmet social needs (65).