Abstract
Background: Initiatives to identify and intervene on patients' socioeconomic needs in the context of health care delivery are expanding. Little information has been compiled across studies on health care providers' knowledge, attitudes, beliefs, and behaviors (KABB) regarding socioeconomic risk screening and referral interventions.
Methods: We conducted a systematic scoping review of providers' KABB related to health care–based socioeconomic risk screening and referral interventions using several search engines. Included studies assessed health care providers' KABB about screening and interventions conducted in clinical settings.
Results: Of 14,757 studies evaluated, 53 were eligible for inclusion. Study designs were heterogeneous. Outcome measures included attitudes and beliefs (n = 42), provider behaviors (n = 35), and provider knowledge (n = 26). The majority of providers expressed positive attitudes toward addressing patients' socioeconomic risks. Participants endorsed concerns regarding insufficient knowledge and resources, time and workflow disruption, and potential negative impacts of screening and referral programs on relationships. Exposure to screening and referral programs led to increases in providers' positive attitudes, socioeconomic risk screening rates, and reported knowledge about intervention options.
Conclusions: Participation in screening and referral programs seems to influence providers' perception of implementation barriers. Future research should explore providers' concerns about addressing identified risks.
- Attitude of Health Personnel
- Outcome Measures
- Referral and Consultation
- Scoping Review
- Social Determinants of Health
- Social Support
- Socioeconomic Factors
Introduction
Strong and consistent evidence has demonstrated that social determinants of health (SDoH)—including socioeconomic factors such as food availability, transportation, income, and housing—are associated with health and well-being.1⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓–12 These associations have spurred innovative clinical practices and payment models that incentivize intervening on social adversity to improve health outcomes, reduce health spending, and achieve health equity.13,14 As described in a recent National Academies of Sciences, Engineering, and Medicine report, these health care–based “social care” practices are diverse and include initiatives that involve detecting patients' social needs and intervening through referrals to community and government-based programs.14 Evidence is growing that these socioeconomic risk screening and referral programs (screening and referral programs) can contribute to improved health and reduced health care costs.15⇓⇓⇓–19
As the effectiveness evidence evolves, a crop of new studies has emerged exploring health care providers' perceptions of both the need for and capacity to implement screening and referral programs. Understanding the provider impacts of these activities is foundational to program implementation and sustainability and should be weighed alongside other elements of the Quadruple Aim.20 We conducted a scoping review to summarize this new body of research on health care providers' knowledge, attitudes and beliefs, and behaviors (KABB) about screening and referral programs.
Methods
This review was guided by existing methodological frameworks and reporting guidelines for scoping reviews.21⇓–23 This approach encourages researchers to broadly summarize an emerging field and identify gaps in the literature. Given the breadth of study types and in accordance with scoping review methodology, we did not assess the quality of included studies.21,22
Search Strategy
We employed a 4-step search strategy for identifying relevant studies. First, we conducted a preliminary search of PubMed to identify key articles on our topic and begin the process of term harvesting. From 20 key articles, we extracted text words and Medical Subject Headings (MeSH terms) from titles and abstracts to build a list of keywords and controlled vocabulary terms to create our search strategy. Next, we worked collaboratively with a medical librarian (JBW) to design and refine our search strategy. Potential search terms were tested by examining the unique results for each term to determine relevance to socioeconomic risks and inclusion in the search. Some concepts that were excluded in this way were violence and non-US-based studies, while concepts like literacy and social isolation were included. Third, we searched the evidence library of the University of California, San Francisco Social Interventions Research and Evaluation Network (SIREN), a curated online database of social interventions research. Finally, we searched the reference lists of included articles to identify additional studies and developed search alerts for academic databases to capture similar articles.
The search strategy was developed in PubMed and adapted to other databases using controlled vocabulary (eg, MeSH, Emtree, and thesaurus terms) where available. A second librarian completed a peer review of the final search strategy using the Peer Review of Electronic Search Strategies (PRESS) guidelines.24 Scoping reviews are not included in the Prospero database, and therefore this review was not registered.
The final search strategy combined 4 concepts: SDoH, health care providers, interventions, and attitudes. Boolean logic was applied by combining similar keywords and controlled vocabulary by using OR and by using AND between each concept. For example, (“social determinants” OR “food insecurity”) AND (“physician” OR “social worker”). To ensure that our search covered the range of interventions related to SDoH, we included broad search terms. No date or language limits were used in the final search. The original database search was conducted in PubMed, Embase, Web of Science, and PsycINFO on January 9, 2018 and then updated on January 29, 2019. The SIREN evidence library was searched on January 29, 2019 and again June 15, 2019 using the preset filter “provider outcome.” These additional searches captured relevant articles published through June 15, 2019. The complete search strategy for all databases can be found in Appendix 1.
Study Selection
Two reviewers (MY, AQR) independently screened a random sample of 200 studies and collaboratively reviewed screening decisions to ensure inter-rater reliability. Reviewers then divided and screened studies based on title and abstract to determine if they met the inclusion criteria for full-text review. Articles that were classified as “maybe” relevant at this stage were then double-screened by the entire review team (MY, AQR, HW, JH). Final screening was completed by the same 4 reviewers. Full texts of each article were reviewed for inclusion by at least 2 authors, and discrepancies were resolved by discussion between the review team. Eligible studies described health care providers' skills, knowledge, attitudes, or behaviors around identifying or addressing patients' socioeconomic risk factors (eg, food insecurity, unstable housing, transportation, etc.) in clinical health care delivery systems. Studies evaluating screening and/or intervention activities that did not take place in a clinical setting were excluded. Studies also were excluded if they were not available in English, primary research, peer reviewed, or United States based. Studies focused on adverse childhood events, intimate partner violence, and interpersonal violence without referencing other socioeconomic risk factors were excluded.
Data Extraction and Synthesis
A standardized data extraction form was developed to catalog information on (1) study setting, (2) study type and methodology, (3) characteristics of the intervention (eg, intervention type, duration, and outcome measures), (4) study population (provider type), (5) socioeconomic risk factor focus, and (6) provider outcomes. Three reviewers (MY, AQR, JH) completed all data extraction activities. Studies fell into 2 groups: those assessing providers' KABB in the context of a screening and referral program and those outside the context of a specific program. Provider outcomes were subdivided into 3 categories: attitudes and beliefs, knowledge, and behaviors.
Results
The initial literature search yielded 19,417 articles. The search update in January 2019 contributed 2,103 additional studies. After excluding duplicates, 14,742 studies were screened for inclusion based on title and abstract. A total of 15 additional studies were identified through searches of the SIREN database. The full texts of 385 studies were assessed for eligibility, and 332 studies were eliminated based on previously established exclusion criteria. Fifty-three studies were included in the final analysis as indicated in the PRISMA chart (Figure 1). Thirty-six of the 53 studies took place in the context of a screening and referral program. These studies are referred to as intervention-related studies.25⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓–60 Thirteen of the intervention-related studies included substantial provider-focused education and training components.30⇓–32,36⇓–38,42⇓⇓⇓⇓–47,57 Across the studies, providers' KABB were assessed using a diverse range of tools, including surveys, interviews, focus groups, and medical record reviews. Intervention study designs included 3 randomized controlled trials (RCTs),25⇓–27 7 quasi-experimental designs with comparison groups,28⇓⇓⇓–32,43,44 and 26 quasi-experimental designs without control groups (including studies with no preintervention data, as well as studies with both preintervention and postintervention data).33⇓⇓⇓⇓⇓⇓⇓⇓–42,45⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓–60 Seventeen studies were conducted outside the context of a specific screening and referral program and instead presented results from surveys more generally exploring providers' attitudes and behaviors about addressing patients' social needs in clinical settings. We refer to these as nonintervention studies.61⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓–77 Descriptions of all included studies are available in Appendix Table 1.
The heterogeneity of methods, interventions, settings, targeted populations, and socioeconomic domains across the included studies limited study comparisons (see Table 1). The majority of studies explored providers' KABB related to multidomain social risk screening and referral initiatives, though some focused specifically on 1 social need, eg, food security. Thirty studies (56%) assessed KABB of physicians who had completed medical school and residency training; 21 (40%) included residents; 16 (30%) included registered nurses and/or nurse practitioners; 6 (11%) included medical students. Sample sizes ranged widely across included studies (n = 6 to n = 1298).
Both intervention and nonintervention studies assessed a range of outcomes regarding providers' KABB about integrating screening and referral programs into clinical care. Forty-nine percent included outcomes related to providers' knowledge (n = 26). Attitudes and beliefs were the most commonly reported outcomes (79%, n = 42), including measures related to the acceptability of specific interventions and perceived challenges to implementing screening and referral programs in health care delivery contexts. Providers' behaviors were described in 66% (n = 35). The range of outcomes included in the full sample of included articles is listed in Table 2.
Provider Knowledge
Multiple intervention-related and nonintervention studies assessed provider knowledge. Topics such as knowledge about the prevalence of social needs within the served population, impacts of social needs on health, and resources to address social needs were included.25,27,30,32,35,36,38,41⇓⇓⇓⇓⇓–47,55⇓–57,59,63,64,68⇓⇓⇓–72,75 Provider level of knowledge around these topics varied. The most consistent gap in knowledge was about tools and resources needed to address socioeconomic needs once identified.43,55,56,59,64,69 One study showed that nonphysician respondents (n = 104) were more likely than physician respondents (n = 154) to be aware of relevant socioeconomic resources (64% vs 45%).75
Impact of Interventions on Provider Knowledge
The majority of studies examining the impacts of provider education and training about screening and referral programs reported improvements in provider knowledge about socioeconomic needs, including increases in knowledge about identifying and intervening on these needs in clinical care settings.27,30,43⇓⇓–46
Provider Attitudes and Beliefs
Nine nonintervention studies assessed providers' global attitudes about identifying and intervening on patients' social needs in health care contexts.61⇓–63,67⇓⇓–70,72,73 Three additional studies assessed providers' attitudes before implementation of a program/intervention.30,45,46 Across these 12 studies, the majority of respondents reported believing that patients' socioeconomic needs have significant impacts on health. Respondents endorsed “the importance of addressing social needs”61 and that knowledge about SDoH was crucial to developing effective programs for vulnerable populations.62 The majority of providers surveyed in the nonintervention studies were willing to conduct screening and reported feeling that screening for socioeconomic risks in health care settings was acceptable.62,67,69,70,72,73,75,76 Studies also reported a range of provider comfort and confidence with the practice of screening for socioeconomic risks.39,43,64,67,68
Beyond screening, the majority of respondents agreed that intervening on socioeconomic risks is also within their individual and/or collective scope of practice.43,45,59,61⇓–63,67,70,73,75 Three studies found significant positive associations between a clinical setting's ability to address patients' socioeconomic risks and providers' job satisfaction and perceptions of health care quality.61,65,77 In 12 of 15 studies assessing postintervention-only attitudes after program exposure, the majority of providers endorsed positive attitudes about the specific intervention in which they participated or screening and referral programs generally.33,35,37,38,48⇓⇓⇓⇓–53,56,58 Measures reflecting attitudes included topics such as acceptability of screening, process measures (eg, compatibility with clinic flow, screening ease), effect of the program on patient care, and provider satisfaction with the program. Providers in 2 post-only studies primarily expressed hesitation with utility to patient care.34,54 In 5 of 15 post-only studies, providers endorsed an increase in their positive attitudes toward screening and referral programs.37,50,51,53,56
Ten studies compared pre vs post or comparison vs control group provider attitudes.25,26,30⇓–32,39,43,45,46,55 In all of these studies, the majority of providers expressed positive attitudes toward their programs and in 8 studies providers' attitudes about their comfort with screening, and perceived barriers to screening and referral programs became more positive following program implementation.30⇓–32,39,43,45,46,55
Twenty-two of the 53 studies highlighted providers' concerns about the potential negative consequences of program implementation. Providers' concerns included discomfort with socioeconomic risk screening43,61,64,70,72; insufficient time and workflow disruption32,34,35,39,43,56,59,60,62,63,66,69,70,72,75; patient discomfort and negative impact on the patient-provider relationship33,34,53,55,56,58,60,70,72,75; and insufficient knowledge and resources to adequately address screening results.34,35,43,52⇓–54,56,59,61⇓⇓–64,66,69,70,72,75,76
Provider concerns about screening and referral programs arose primarily in either nonintervention studies or the preprogram implementation period. When evaluated postprogram implementation, concerns about the negative consequences in each of these areas generally abated. Concerns generally fell into 4 categories:
Discomfort with screening: Four provider education and training studies examining the impacts of provider training about SDoH generally and/or socioeconomic risk screening, specifically, demonstrated improvements in provider comfort with screening in at least 1 SDoH domain after participating in a screening and referral program.32,43,45,46 One of these studies involved providing training on food insecurity to faculty, residents, and medical students and showed significantly increased self-reported knowledge of food insecurity, resources, and willingness to discuss social needs with patients.45 Two additional studies used videos and facilitated discussions to model screening; training reduced residents' discomfort with screening.32,43
Concerns about time and workflow: After participating in screening and referral programs, providers often characterized their experiences with screening as “not burdensome,” “quick,” and “time and workflow were not barriers to screening.”26,32,33,39,51⇓–53,56,58,60 In 1 study of 45 pediatric residents, 75% of the residents in the experimental group reported that screening forms did not slow down patient visits.26 In 3 intervention studies, providers were initially concerned about workflow but were surprised to find that the program had limited impact on their work or was otherwise beneficial, reporting that the time burden was “less than anticipated” or that screening was “worth the time.”26,39,56 In 5 studies, providers reported that time remained a concern after program exposure.43,48,54,56,60
Concerns about patient-provider relationships: Six studies highlighted preprogram implementation provider concerns that screening activities might negatively impact patient-provider relationships.34,53,55,56,70,72 In 5 studies assessing postimplementation attitudes, providers instead indicated that screening enhanced their relationships with patients or did not have a negative impact.33,52⇓⇓–55
Addressing socioeconomic risks: Concerns about the availability of resources to address identified risks revolved around providers' (1) desire for more knowledge/confidence on addressing socioeconomic needs; (2) uncertainty regarding the effectiveness of resource lists and referral networks in their interventions; and (3) interest in more systems-level logistic support and material resources. Eight studies documented provider concerns postprogram implementation about the ability to provide adequate resources to address identified socioeconomic needs.34,35,43,52⇓–54,56,59 Three studies suggested that providers' confidence around addressing patients' needs increased following program implementation.39,43,46
Attitudes and beliefs about screening and referral programs differed across groups. Several studies found that individual-level provider characteristics such as provider specialty, under-represented in medicine (URM) identity, and gender were associated with provider attitudes. One large study (n = 240 faculty physicians) found that compared with nonminority physicians and specialists, minority physicians and primary care providers (PCPs) were more likely to think that the benefits of collecting patients' socioeconomic risk information outweigh negative consequences (51% nonminority vs 66% minority physicians; 54% specialists vs 78% PCPs.).62 In the same study, all physician groups identified lack of clinic and health system infrastructure to address social needs as the primary barrier to program adoption. The second-most cited concern among nonminority physicians and specialists was uncertainty around what to do with social risk data once collected. Among minority physicians and PCPs, the second-most cited concern was liability related to not addressing an identified risk.62 In the same study, female physicians were more likely than other groups to think it was appropriate to include socioeconomic data in electronic health records.62
In 1 study with pediatric emergency medicine providers (n = 114), a majority of respondents rated screening as important; a higher percentage of nursing staff (58.1%) than physicians (28.2%) preferred screening to be conducted by physicians.67 In 1 nonintervention study (n = 258 providers), time constraints were cited as barriers by 70% of physicians compared with 47% of nonphysician providers.75 In another study of clinical faculty providers (n = 240), 54% of physicians agreed that PCPs should be primarily responsible for managing socioeconomic risk factors, but when asked whether they should be solely responsible for addressing needs, significantly more specialists than PCPs agreed.62 In 4 studies, providers reported wanting ancillary staff to provide support with screening and referrals.25,51,56,69 In 1 large study (n = 258), 94% of clinicians thought social workers should conduct screening.75
Provider Behaviors
Across nonintervention studies, providers consistently reported screening for a wide range of socioeconomic risks. Reported screening frequency, tools, and approaches varied across these studies, however. Some studies reported that providers screen a pre-established target population (eg, patients with diabetes) while others reported that providers are prompted to screen in response to patient factors (such as comorbid conditions) identified in the context of clinical encounters.69,70 Several studies highlighted associations between provider characteristics, practice settings, and provider screening and referral behaviors. Two nonintervention studies found that providers practicing in clinical settings with staff whose role is to connect patients to community resources screened more frequently and referred to more community resources than providers with no or limited on-site resources.74,76 One large study of pediatricians (n = 602) found that those identifying as female, URM, practicing as generalists, or in rural settings were more likely to routinely conduct screening.76 Another study of family practice pediatric physicians and nurse practitioners (n = 186) found a negative association with lack of provider time and monitoring for family nutrition, specifically.70
Impact of Screening and Referral Programs on Provider Behaviors
All studies assessing screening rates before and after implementation of a screening and referral program showed statistically significant increases in screening behaviors in at least 1 SDoH domain.26,27,29,32,35,39,40,45,46,56 Few studies assessed long-term impacts on screening behavior. In 1 study the “median survival time of increased screening was 8.1 month”29; in another, the screening model continued to be in use 2 years following program implementation, but screening rates were not assessed at follow-up.56
Discussion
In this scoping review, we found a diverse group of studies exploring providers' KABB related to socioeconomic risk screening and referral programs in clinical settings. Thirty-six of these studies took place in the context of specific screening and referral initiatives. Across studies, a majority of clinicians and other clinical staff generally expressed positive attitudes about addressing patients' socioeconomic needs in health care settings. Three studies in this review described significant associations between the clinical setting's capacity to address patients' socioeconomic needs and providers' sense of professional satisfaction.61,65,77
In studies conducted outside the context of a screening and referral program (n = 17), providers expressed more concerns about the potential negative consequences of program adoption. In contrast, providers exposed to these programs more consistently reported positive impacts on patient-provider relationships, increased comfort with screening and referral practices, and fewer concerns about clinic workflow disruption. However, even in the context of program exposure, concerns persisted about capacity to adequately address patients' socioeconomic needs. These findings suggest that program exposure may decrease some, but not all, provider-related barriers to implementation.
Providers' persistent concerns about capacity to adequately address patients' socioeconomic needs should not be overlooked. Existing evidence on the impacts of screening and referral programs is mixed,78 though a growing number of studies suggest referrals programs may help to reduce patients' socioeconomic needs and affect health.17,79⇓⇓⇓–83 No studies in this review, however, described a program in which providers track patient outcomes postreferral, which may contribute to providers' persistent concerns about the adequacy of their referrals. In parallel, providers' focus on intervention effectiveness should be considered alongside studies in this field that have reported patients' perspectives on screening and referral programs. These studies suggest that patients value the patient-provider rapport building that develops from discussions about socioeconomic needs regardless of the social services referrals provided.27,84⇓–86 This inconsistency between provider and patient expectations warrants more attention in provider training and education. If providers are more aware of patients' expectations, they may better understand the value of these programs beyond referrals to social services.
Findings from studies undertaken in the context of screening and referral programs strongly suggested that provider education and training initiatives can impact provider behaviors and attitudes; such programs consistently reported post-training increases in socioeconomic risk screening and referral rates. However, most of these studies assessed these behavior changes over relatively short periods (max 12 to 18 months). Characterizing core training components and assessing more longitudinal outcomes are important areas for future research. This research would be facilitated by consensus on tools that gauge providers' KABB in this area.
Interestingly, 9 of the 13 studies that included provider education and training focused on physicians or physician trainees.30⇓–32,36,43⇓⇓⇓–47 Improving physician education on these topics is a growing focus in undergraduate and postgraduate medical education.87,88 Given the multidisciplinary workforce involved in many screening and referral programs, however, more attention should focus on meeting the training needs—and assessing the impacts of training—of a more diverse workforce.14,89
Limitations
There are 4 important limitations to consider in interpreting our findings. First, we narrowly focused on provider outcomes associated with socioeconomic risk screening and referral programs. This excludes health care activities that might address other social and structural determinants of health and equity, including racism. Future reviews will need to explore the impacts of programs aimed at addressing these other drivers on patients, population health, and clinical settings. Second, study abstracts that did not refer to provider outcomes were excluded. This may have limited our ability to identify relevant studies. Third, the review reflects a wide variety of different studies, many of which assessed provider attitudes via surveys, which are likely influenced by multiple forms of bias. Qualitative work and chart-based studies could more comprehensively explore changes in KABB. Finally, study heterogeneity—including different types of socioeconomic risk screening conducted—and lack of standardized outcome measures limited our ability to draw comparisons across studies. In addition, while a critical appraisal of the studies was not included in this review, many included studies were of nonexperimental designs.
Conclusions
In this systematic scoping review, we identified a wide range of studies that describe providers' KABB related to health care–based socioeconomic risk screening and referral programs. Study contexts influenced providers' concerns about program adoption: providers not participating in these programs were more likely than providers participating in such programs to report significant feasibility and implementation barriers. Rigorously conducted studies involving both clinicians and ancillary staff members involved in screening and referral initiatives will help to identify both implementation barriers and strategies to overcome them. Future studies also should clarify training required to ensure professional competence related to social and medical care integration.
Acknowledgments
The authors gratefully acknowledge the contribution of Dylnne Gonzalez in editing the manuscript.
Appendix 1. Search Strategy
Final search strategy table, January 9, 2018
Updated search January 29, 2019
Appendix 2. Characteristics of Included Studies
Notes
This article was externally peer reviewed.
To see this article online, please go to: http://jabfm.org/content/33/5/820.full.
Financial disclosures: AQR has no financial disclosures; HEW has no financial disclosures; JBW has no financial disclosures; MY has no financial disclosures; JMH has no financial disclosures; LMG has grants from Commonwealth Fund, Robert Wood Johnson Foundation, Episcopal Health Foundation, AHRQ, Kaiser Permanente, and NIMHD. LMG also has current contracts with Mathematica, RTI, and NORC, and previously with the Institute for Healthcare Improvement.
Conflict of interest: Support for this research was provided by Kaiser Foundation Health Plan Inc. The study sponsor did not play a role in study design; collection, analysis, and interpretation of data; writing the report; and the decision to submit the report for publication.
- Received for publication January 28, 2021.
- Revision received March 31, 2021.
- Accepted for publication April 14, 2021.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵