While efforts to assess and address social determinants of health (SDOH) in primary care are inherently limited in their ability to address larger systemic challenges, approaches to respond to individual level patient social needs show real potential for improving health and reducing disparities by integrating clinical care with social services. Our work at Lincoln Community Health Center (LCHC) is part of a larger nationwide movement toward re-imagining the role and responsibility of health systems with respect to the communities that they serve. The long-term success of efforts to assess and address SDOH in primary care will involve health systems as catalysts for community engagement and cross-sector collaboration with real potential to achieve better population health and health equity.
Assessing and addressing social and non-medical drivers of health requires an upfront investment in clinic infrastructure, workforce development, and collaborative community partnerships.
Implementing a protocol to assess and address patient social needs requires a multidisciplinary team-based approach. Our experience highlights the important role of social workers and behavioral health professionals in clinical settings.
A patient-centered model for responding to SDOH risk factors in routine outpatient clinical encounters must include a participatory shared decision-making process to develop a tailored non-medical care plan.
Capturing and disseminating SDOH data can be leveraged to identify gaps in community resources and inform population health management strategies and investments.
It is well established that social determinants of health (SDOH), the broad set of economic and social systems that shape the conditions of daily life, impact health outcomes and contribute to health disparities. For example, there is a strong relationship between income and life expectancy. Material deprivation and other environmental factors that are faced disproportionately by low-income households are associated with greater risk of chronic stressors and health problems.1 A recent study found that that the wealthiest Americans live almost 15 years longer than those at the bottom of the income distribution.2 There is a growing body of literature demonstrating that initiatives to mitigate SDOH risk factors by responding to individual level social needs, such as housing,3,4 or food insecurity,5 result in downstream health benefits and savings in health care expenditures. Given this evidence base, health systems have been increasingly called upon to identify and, in partnership with social service agencies and community-based organizations, to address health-related social and resource needs in order to manage and improve the health of individuals and populations. Doing so is imperative for the success of efforts to reform health care delivery toward value and effective population health management. However, care redesign that incorporates the assessment of and response to social needs in primary care is an implementation challenge, and best practices have not yet been identified.
In the present report, we describe our experience at Lincoln Community Health Center in developing and adopting a protocol for screening for SDOH and responding to identified risk factors in a diverse, low-income, and medically vulnerable population. LCHC is a FQHC based in Durham, North Carolina, that serves approximately 34,000 unique patients each year. We believe that our experience provides useful information for policy makers, health systems, and clinicians who are interested in implementing clinical workflows for screening and responding to patients’ social needs.
As health care delivery reform efforts focus on high-value approaches to managing the health of defined populations, there is a need for practical methods to acquire SDOH data that are relevant and actionable in order to predict risk and tailor care to an individual or community. Our goal was to implement a comprehensive, yet easily administered, SDOH assessment tool to improve risk prediction and a patient-centered clinical workflow to provide tailored and effective referrals to community resources and social services.
Information on family health history, personal health habits, and screening for mental health concerns are all collected to inform care planning. However, there is a growing recognition of the importance of measuring non-medical needs as part of clinical encounters to predict and mitigate potential social and economic barriers to treatment adherence and to inform social service integration, especially for high-risk patient populations. The Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences (PRAPARE), developed in collaboration by the National Association of Community Health Centers, Inc. (NACHC), the Association of Asian Pacific Community Health Organizations, the Oregon Primary Care Association, and the Institute for Alternative Futures, has emerged as a standardized measure of SDOH that has been widely adopted at FQHCs, including LCHC. PRAPARE consists of core measures to be included in all assessments as well as optional measures based on the needs and priorities of individual communities.
Designing Clinical Workflows
Our implementation of PRAPARE consisted of two distinct workflows: (1) assessing needs with use of a screening tool and (2) responding to needs via internal resources and/or through linkages to community-based organizations and social service agencies. Both workflows required investment in staff training and time as well as in the infrastructure needed to collect, document, analyze, and act on data. In addition, external referrals required an environmental scan of appropriate local, state, and federal resources to serve the needs of the patient population.
At LCHC, the primary pathway for PRAPARE delivery exists within the Adult Medicine, Family Medicine, and Pediatric clinics. Adult patients are referred by a primary care provider to behavioral health staff for evaluation and assistance when social risk factors, poorly controlled chronic conditions, high-risk behaviors, or emotional distress are present.
Once a referral is made, the PRAPARE assessment is conducted during or after a clinical encounter by a social worker case manager on the behavioral health team. Administration takes approximately 5 to 10 minutes. We believe that it is better to engage with patients in person to explain why assessing non-medical needs allows the clinic to better serve and support them as opposed to offering the screening tool for self-administration during check-in. The behavioral health team already had received training in patient-centered communication as well as in assessing and addressing non-medical patient needs. Given these competencies, they were uniquely equipped to administer the tool and then refer those with identified needs to appropriate resources within LCHC or to outside community agencies and services (Table I). The use of a patent-centered communication style allowed for the integration of patient values, preferences, and needs into a tailored plan of community referrals and follow-up as needed.
Incorporating Shared Decision-Making Techniques
Our experience suggests that promoting health-system-wide awareness of the tool and the team-based approach for assessing and addressing SDOH has facilitated acceptance and adoption. PRAPARE is synergistic with the patient-centered medical home model, which emphasizes a multidisciplinary approach to providing comprehensive and participatory “whole person” care that attends to both medical and non-medical health needs. We have recognized that conversations with patients about their social needs can be challenging. These conversations are facilitated when initiated by behavioral health professionals whose training has prepared them to inquire about personal issues with sensitivity and tact. For example, the use of motivational interviewing, reflective listening, and empathic communication techniques are important competencies for engaging with patients around non-medical needs.6,7,8 In addition, when needs and risks are identified, it is advantageous to respond to these issues promptly. A behavioral health professional has the knowledge and skills to respond in a manner that prioritizes the patient’s needs and goals. We found that this type of engagement around non-medical needs requires training that recognizes the sensitive nature of these conversations while emphasizing the use of shared decision-making between clinical team member and patient. Initial and ongoing trainings, as well as institutional support, have been effective elements of our implementation strategy at LCHC for gaining broad staff and clinician support.
Integration of SDOH Data into the EHR
A key goal of our project was the integration of SDOH data into patient EHRs. Integration makes information readily identifiable, actionable, and trackable by all members of the care team, as is the case with a laboratory test or vital sign. This consideration is all the more relevant given recent progress toward including SDOH measures in the next stage of meaningful use and ICD-10 clinical codes, with corresponding CPT codes, that can be used to incentivize screening and addressing social determinant risk factors.9,10,11 In addition to identifying and tracking patient needs and risk factors through a customized EHR flowsheet, it is equally important to identify and follow up on those resources and services to which patients are referred. To this end, we have also designed a structured EHR template for systematic collection of referral data for future reporting (Fig. 1). These referral data can facilitate case management and assist population health planning by describing factors that impede or facilitate connection to community resources, by identifying resources that are most impactful to the health and well-being of our patients, and by identifying gaps in the community.
Our team included six social workers on the behavioral health integration team, who were responsible for leading the effort to identify patients with complex non-medical needs; a family medicine physician and chief medical officer, who were responsible for championing the clinic-wide implementation project; a nursing clinical informatics specialist to oversee EHR integration and report SDOH data; and an implementation and organizational development specialist to lead a quality-improvement evaluation and overall project management.
Approximately 1,700 LCHC patients were screened for SDOH risk factors, with 1,222 patients receiving referrals for LCHC or community services. Several useful and surprising trends regarding SDOH risk factors were identified. We found that the most commonly reported affordability challenge that patients faced was related to medicine and medical care. Given that this patient population is served by an FQHC, which offers medical services at fees that are adjusted to the income status of recipients, this finding was a surprise. Additional qualitative data are being collected to understand the driver of this affordability challenge (Fig. 2). Food insecurity is a prevalent SDOH risk factor for our patients, with almost 30% of patients having material insecurity indicating challenges with affording food within the past year (Fig. 2). We also found housing insecurity among the patients who were screened. However, in contrast to patients who have challenges with affording food or clothing, those who face chronic challenges with housing affordability have fewer resources at their disposal (Fig. 2). Given the impact of housing instability on health outcomes, health systems must partner with policy-makers to alleviate this challenge. In addition, >60% of LCHC patients were employed, disabled, or a student. PRAPARE response structures allowed for greater awareness of patients who were not engaged with the formal labor market but were still involved with work (e.g., as a caregiver). Special outreach was provided to unemployed patients, especially for the 13% of respondents who were currently seeking work. Social isolation was prevalent, with approximately 25% of patients indicating that they see or talk to someone with whom they feel close only 1-2 times per week or less. Given the risk that social isolation poses to health,12 we have expanded upon our community resource directory to connect patients with support groups, social and recreational activities, and civic organizations to improve social support (Fig. 2). Of particular note is that a proportion of patients who were screened did not receive any community referrals either because they did not indicate any SDOH risk factors or because suitable resources were not available. This finding may have been related to patients being unwilling or uncomfortable with sharing social and economic information. The most commonly referred community/LCHC resources were related to medicine and medical care access and food insecurity (Fig. 3).
Overall, our experience suggests that a multisectoral approach through community support and linkage is critical in order to successfully address the social needs of patients. We found that our experience has resulted in further strengthening of existing community linkages while also forging new ones. From a clinical workflow perspective, we believe that the integration with our Epic EHR system decreased the burden of clinic documentation by behavioral health staff administering PRAPARE. Such integration may require substantial programming and customization. Our experience underscores the importance of a designated clinical informatics champion as part of the implementation team.
To fully evaluate efforts to respond to identified social risk factors, it is necessary for health systems, community-based organizations, and external partners to communicate and coordinate effectively in order to ensure that patients can access the resources to which they are referred. However, “closing the loop” on these community-based referrals is challenging and requires novel approaches in order to support patients across this medical and non-medical care continuum. Current investments at the state level in technology to support care coordination could accelerate these efforts. There is evidence that SDOH screening and referral processes such as PRAPARE have led to increased uptake of community resources.13 However, there is a need to determine best practices to maximize the likelihood of a successful referral and to understand the contextual factors that impact patients’ ability to access and benefit from community referrals. We intend to collect data from patients to refine the PRAPARE clinical approach and identify barriers to and facilitators for accessing resources in the community.
While our experience suggests that social workers and other behavioral health team members are well equipped to lead PRAPARE implementation, there remains a need to further involve primary care medical providers in effective shared treatment planning. There is also an opportunity to evaluate how other members of the health care team could support these efforts, including a role for community health workers and other clinical support staff.14 We are planning focus groups with clinicians and frontline staff to explore strategies to increase medical providers’ awareness of the SDOH risks and challenges faced by their patients, engage them in reinforcing referrals, and assist them in implementing more informed and effective shared treatment planning.
Although our protocol was implemented in a low-resource patient population, a proportion of our patients did not indicate that they had an SDOH risk factor. We believe that the method of screening will impact the validity of the results. For example, patients may be more comfortable indicating that they have unmet social needs on a form as opposed to answering questions in person. We intend to conduct structured interviews on the acceptability of the process in order to better understand patient preferences for engagement to address non-medical needs in a clinical setting.
Another key consideration is the financial sustainability of implementing PRAPARE. There are few data on costs associated with novel practice patterns for assessing and responding to social needs. We intend to evaluate the costs to LCHC associated with PRAPARE implementation. The main drivers of costs to the clinic are associated with startup (e.g., training, EHR integration, and community resource directory development) and the ongoing effort of trained personnel to assess and respond to social needs. There is a need for further research to understand the impact that identifying and addressing SDOH has on health and health care costs and utilization. Specifically, we are interested in whether PRAPARE lowers future emergency department and acute hospital or inpatient utilization. Equipped with evidence of a beneficial effect on health and cost, we anticipate that investments in support of this activity will continue.
The implementation effort and quality-improvement study were supported by the Blue Cross and Blue Shield of North Carolina Foundation. The Duke University Health System Institutional Review Board reviewed and exempted this project.