Social Needs Screening and Referral Program at a Large US Public Hospital System - NowPow

NowPow In the News

July 14th, 2020 | American Journal of Public Health (AJPH)

Many health care providers and systems are developing and implementing processes to screen patients for social determinants of health and to refer patients to appropriate nonclinical and community-based resources. The largest public health care system in the United States, New York City Health + Hospitals, piloted such a program in 2017. A qualitative evaluation yielded insights into the implementation and feasibility of such screening and referral programs in health care systems serving low-income, minority, immigrant, and underserved populations.

It is widely recognized that social determinants of health (SDOH), or the conditions in which people are born, grow, live, work, and age, affect health and well-being.1,2 A subset of social needs, including housing, food, and safety, are associated with health care utilization and health outcomes, particularly among low-income populations.3 The push toward value-based care in the United States at both state and federal levels has sparked a growing impetus among physician groups and hospital systems to identify these needs in the clinical setting and refer patients to appropriate social services.4,5

We report on lessons learned from the implementation of such a SDOH screening and referral program at New York City Health + Hospitals (NYC H+H), the largest public health care system in the United States.

NYC H+H provides essential inpatient, outpatient, and home-based services to more than one million New Yorkers every year in more than 70 locations, including 11 acute care hospitals, across the city’s five boroughs.6 In 2017, NYC H+H piloted the SDOH screening and referral programs in three of its ambulatory care clinics: an adult outpatient clinic and a pediatric clinic in Manhattan and an adult outpatient clinic in the South Bronx.

The intervention was designed to serve patients of NYC H+H’s ambulatory care clinics. Most of NYC H+H’s patients are low income: nearly one third (30%) are uninsured, and nearly half (49%) are Medicaid or Medicare beneficiaries.6 Additionally, nearly 90% of patients are people of color, many of whom are recent immigrants.

The overarching goal of this initiative was to mitigate the impact of social risk factors on the health of patients receiving care at NYC H+H primary care clinics by identifying and addressing a predefined set of SDOH-related needs and, ultimately, to inform expansion of SDOH screening and referral across the entire NYC H+H system. Although studies have assessed SDOH screening and referral interventions in diverse settings both nationally and internationally, few studies have assessed such interventions within safety net (Appendix A provides the Institute of Medicine definition of “safety net” [available as a supplement to the online version of this article at]) or public hospital systems. The findings we present can inform the development and implementation of SDOH screening and referral programs in primary care clinics at other similar hospitals and health systems across the United States.

Primary care providers and staff at each of the three pilot clinics worked with NYC H+H Office of Population Health staff (Appendix B provides a list of the individuals involved in screening tool development [available as a supplement to the online version of this article at]) to develop patient-administered (self-report) screening tools based on the Center for Medicare and Medicaid Innovation Accountable Health Communities screening tool.4 The screening tools were required to include eight critical domains (food insecurity, health insurance coverage, housing concerns, public income benefits, household interpersonal violence, adult education and literacy, daycare, and general and immigration legal problems) for which government or community-based resources were available to address patients’ needs. Clinics were encouraged to retain language from the validated screening tool but were able to edit specific questions and modify wording based on feedback from patients and frontline staff. Finally, each clinic defined its target populations (e.g., only the ≥ 90% of patients with scheduled appointments), developed workflows, and built on existing local referral networks to allow optimal implementation of the program. The screening tool was available in English and Spanish, and translation services were available for other languages.

At the request of the NYC H+H Office of Population Health, an evaluation team conducted a qualitative process evaluation of the pilot SDOH screening and referral program, interviewing 28 key informants (hereafter referred to as “respondents”) from the three pilot clinics, including leadership personnel, frontline staff, volunteers, and primary care providers, as well as central H+H leadership (Appendix C provides detailed study methods; Appendix D: interview protocols; Appendix E: a list of respondents [available as a supplement to the online version of this article at]). No patients or caregivers of pediatric patients were interviewed.

The evaluation also identified key barriers, facilitators, and other lessons learned from implementation as told by the leadership, staff and volunteers most closely involved in the program’s implementation (see the box on page S212). Using these data, we also identified a number of best practices for the expansion and sustainability of the SDOH screening program at all NYC H+H facilities (see the box on this page).

Challenge/Facilitator Qualitative Evaluation Findings
Challenge: Integration of screen into workflow A key lesson that emerged from early implementation was the importance of integrating all aspects of the screening process into existing workflows to avoid the perception that it was something “special” or “additional” (e.g., registration clerks at one site initially gave the screening tool to patients as a separate document when they checked in for their appointments, which resulted in a high rate of noncompletion because handing the tool to patients was an extra step for clerks, and some patients perceived the tool as optional). The registration clerks then integrated the screening tool into the standard packet of forms and screens given to all patients. This simple tweak “normalized” the screening tool for both clerks and patients, and program staff reported a much higher screening rate as a result.
Challenge: Burden on staff One of the biggest issues all three sites faced was how to screen and refer patients without placing undue burden on existing staff. Staffing challenges existed throughout the workflow, from the initial point of screening, to reviewing the screen with patients, to making and receiving referrals. Clinics addressed some staffing challenges (e.g., one site switched the role of screening administrator from nurses to trained, clinic-based volunteers after nurses indicated that the task was too burdensome given their existing responsibilities).
Challenge: Patient-related factors Low literacy, limited English proficiency, concerns about immigration status, and screening fatigue posed challenges to screening. Even with substantial refining of the language on the screening tools, respondents at all three sites identified patients’ language and literacy levels as significant and continuing barriers to completion of the screening tool.
Challenge: Tracking capability Respondents recognized the value of tracking screening and referrals to assess progress and gaps but acknowledged that consistent tracking was time consuming and thus challenging. One site tracked data ad hoc using an electronic database, but respondents identified this as a strain on time. Toward the end of the pilot period, all sites began researching the use of technology to streamline data tracking and facilitate other aspects of the SDOH screening process, including making referrals to the social work department through the electronic health record system. In addition, sites had access to NowPow,7 a social services referral software system, to help identify resources for patients who screened positive for various issues. NowPow can also track whether patients receive the services to which they are referred (closed-loop referrals); however, that functionality had not yet been rolled out to the sites at the time of the pilot program.
Facilitator: History of community-based partnerships and colocated services All clinics participating in this pilot had a history of referring patients to various internal and external services, albeit unsystematically, before this initiative. Perhaps as a result, respondents did not note significant gaps or barriers in identifying appropriate services to which to refer patients to address their social needs. Each site also benefited from the existence of a number of social services and services colocated within their facilities, including legal services, home health care, and enrollment in the federal Supplemental Nutrition Assistance Program (SNAP). Leadership at one site described challenges in finding referral resources outside the local hospital area, as many of their patients do not live near the hospital and therefore are outside the catchment area for many of the hospital’s local community-based organization partners.
Best Practices Identified From the Pilot of the Social Determinants of Health Screening and Referral Program
Best Practice Related Findings
Ensure the screening tool is appropriate in terms of literacy, culture, and language Concerns about literacy and patients’ understanding of the screening tool were paramount at all sites. Modifications to the Center for Medicare and Medicaid Innovation–based screening tool included adding items, eliminating items, and rewording items. Despite operating within the same health care system in the same city, each clinic ended up with its own unique tool adapted to its specific workflows and population served.
Integrate screening into regular clinic workflows The site that included the screening tool in the packet of materials all patients receive on registration illustrated the value of making social needs screening a part of regular clinic registration procedures and not something extra or optional for staff and patients alike. We recommend including the screening tool or integrating it into the regular set of forms that patients receive at clinic visit registration.
Make screening universal Providing screening tools only to certain subsets of the clinic population, such as patients with scheduled appointments or specific patient panels within the clinic, was burdensome for registration staff and subsequently led to lower completion rates. All three sites initially decided to screen only patients with scheduled appointments, but one of the adult clinics eventually transitioned to screening every patient at every visit, which enhanced screening rates and reduced the burden on registration clerks and patients.
Establish a solid referral base Universal screening only provides value when systems can address patients’ identified needs through internal resources or referrals to community resources. The sites in our pilot benefited from established relationships with community partners and colocated services.
Maximize the use of technology All three sites used an electronic referral system, but none were yet using such systems to screen patients. Respondents were clear that technology both reduced staff burden and increased screening rates. Integrating the screening tool into the existing electronic health record using referral software such as NowPow or a combination of the two could enhance the ease and universality of the screening and referral process.
Implement a data-tracking system Sites should be able to track patients throughout the screening process, from initial screen to referral and ultimately receipt of services. Such a system facilitates program monitoring and identification of gaps in the workflow and provides evidence of the value of the program. NYC New York City Health + Hospitals has integrated the screening tool into its electronic health record, allowing more sophisticated data tracking, and is currently pursuing the use of NowPow to track referrals.
Staff the program appropriately Optimal staffing includes a dedicated program coordinator to implement and monitor the program. Respondents were somewhat divided on whether using volunteer staff was feasible or sustainable, with frontline staff in particular strongly favoring paid staff over volunteers. If volunteers are the most feasible option, it is critical that they be committed to the program’s success, thoroughly trained on clinic operations and workflows, and well known to staff and providers. Staff coverage during all clinic hours is also important to ensure that all patients are screened.

Respondents generally supported the program, but frontline clinical staff at two clinics felt that ensuring proper completion of screens or reviewing them in a meaningful way with patients was time consuming on top of their other responsibilities. Relatedly, an unintended adverse consequence arose among primary care providers: although the program was deliberately designed for nonclinicians to have responsibility for linking patients to appropriate resources, some primary care providers felt that even simply reviewing the screens with patients brought up issues that warranted extended discussion with their patients but that such conversations were challenging to fit in during standard 15-minute patient visits.

Respondents widely reported that the SDOH screening and referral program had great value for their patients, and they wanted it to continue. However, some respondents expressed concerns about both the sustainability and expansion of the program without additional resources. Frontline clinical staff in particular thought that relying on volunteers would not be sustainable in the long term and suggested that investing in paid staff would enable both sustainability and expansion.

The NYC H+H Office of Population Health was already pursuing expansion of the program during the pilot phase and used the data from this evaluation to guide and support those efforts to the extent possible given resource constraints. A screening tool based on feedback from the pilot has been integrated into the system’s electronic health record, which was recently standardized and unified across all facilities in the system, and results of social needs screenings will be available on patient charts (Appendix F provides the standardized screening tool [available as a supplement to the online version of this article at]).

Because of resource constraints, hiring additional staff to manage the program was not feasible, so NYC H+H, in partnership with a local nonprofit organization, started the Health Advocate Volunteer program to provide trained volunteers to participating clinics. The Health Advocate program addresses the main concern with the use of volunteers, lack of reliability, and inadequate coverage. In addition, NowPow,7 an electronic closed-loop referral system has been enabled across all of H+H so that primary care providers can identify whether patients were able to connect with and receive services from community organizations and clinic-based services (e.g., social work) to which they were referred. NYC H+H is exploring integration of these data into the electronic health record. Finally, to further its mission of improving patients’ health and enabling them to live their healthiest lives, NYC H+H will use results from SDOH screens to target intensive care management services to patients with multiple health conditions and high health care utilization who also have significant unmet social needs.

Across the country, public health care systems are the primary sources of health care to low-income, minority, immigrant, and underserved populations that have many unmet social needs. Although further research and evaluation are needed to assess the impact of SDOH screening on staff and especially patients, this evaluation provides valuable insights into the implementation and feasibility of such screening and referral programs for vulnerable populations.


This research was supported in part by the National Center for Advancing Translational Sciences (National Institutes of Health and New York University Clinical and Translational Science Awards program grant UL1TR001445).

The authors would like to acknowledge the assistance of Kayla Fennelly in interviewing and coding and thank the key informants for sharing their insights on the implementation of the program.


The authors have no conflicts of interest to report.


Institutional review board approval was not needed for this quality improvement study because the project did not constitute human participant research


1. Braveman P, Egerter S, Williams DRThe social determinants of health: coming of ageAnnu Rev Public Health2011;32:381398CrossrefMedlineGoogle Scholar
2. Marmot M, Bell RFair society, healthy livesPublic Health2012;126(suppl 1):S4S10CrossrefMedlineGoogle Scholar
3. Berkowitz SA, Hulberg AC, Hong Cet al. Addressing basic resource needs to improve primary care quality: a community collaboration programmeBMJ Qual Saf2016;25(3):164172CrossrefMedlineGoogle Scholar
4. Billioux AVK, Anthony S, Alley DStandardized screening for health-related social needs in clinical settings: the accountable health communities screening tool. NAM Perspectives. 2017:19. Available at: Accessed April 24, 2020. Google Scholar
5. Thomas-Henkel CSM, Schulman MScreening for social determinants of health in populations with complex needs: Implementation considerations2017. Available at: Accessed April 24, 2020. Google Scholar
6. NYC Health + Hospitals. 2020. Available at: Accessed February 2, 2020. Google Scholar
7. Lindau ST, Makelarski J, Abramsohn Eet al. CommunityRx: a population health improvement innovation that connects clinics to communitiesHealth Aff (Millwood)2016;35(11):20202029. [Erratum in Health Aff (Millwood). 2017;36(2):384] CrossrefMedlineGoogle Scholar
Go To Source

Want to know more?

Request a Demo