We describe the development, implementation, and feasibility of universal screening for SDOH at a large, urban academic medical center. Through funding from the Centers for Medicare and Medicaid Services Innovation Center’s Accountable Health Communities (AHC) Model, NewYork-Presbyterian Hospital at Columbia University implemented universal screening by using the AHC Screening Tool for SDOH in four hospital-affiliated, community-based primary care practices that offer pediatrics, internal medicine, and obstetrics and gynecology services. Screenings were completed on tablets via NowPow,⁴ a SDOH screening and referral platform, and were coupled with clinical screenings, including depression, substance use, and asthma, to encourage provider buy-in and ensure sustainability of the model (NewYork-Presbyterian Hospital used the Centers for Medicare and Medicaid Services Innovation Center’s Health Related Social Needs Screening Tool to screen for SDOH; the Patient Health Questionnaire-2/9 for depression; the Alcohol Use Disorders Identification Test, Drug Abuse Screening Test, Screening to Brief Intervention, and Car, Relax, Alone, Friends/Family, Forget, Trouble for substance use; and the Asthma Control Test for children 4–11 years or 12 years and older for asthma).

The program took place in four primary care practices located in northern Manhattan, New York. The pilot screening program was launched in May 2018 with full implementation in September 2018.

Northern Manhattan is home to largely Latino neighborhoods where nearly half the population is foreign-born, 18% live below the poverty line, and 40% are limited English speakers.⁵ All dual-beneficiary patients were eligible for screening. A screening frequency logic was established by discipline and built into the electronic medical record (Appendix A, available as a supplement to the online version of this article at http://www.ajph.org).

This program was driven by a hospital mission to improve population health and address health equity in surrounding communities. Given the large role that SDOH play in the health of populations, identifying unmet needs through a universal screening program is a promising tool to begin to address root causes and improve overall health.

The program was implemented in a stepwise approach to establish proof of concept before scaling to all sites. Multidisciplinary teams consisting of physician champions and practice administrators were formed to develop site-specific workflows, taking into consideration each site’s culture, workforce, and other competing priorities. Clinical screens were incorporated into workflows, and results were integrated into the electronic medical record.

We chose a pediatrics primary care site to pilot the full model, and we used quality improvement Plan-Do-Study-Act cycles to inform implementation (Appendix B, available as a supplement to the online version of this article at http://www.ajph.org). Once screening was implemented in the waiting room without jeopardizing clinical workflow, the program expanded to all four sites and three disciplines over a period of twelve months. To prepare practices, a series of lectures were given to physicians and practice staff that focused on the role of SDOH in health care and the importance of identifying and addressing social needs.

A cohort of volunteers assisted sites with the screening process, which took approximately 10 minutes to complete. Volunteers helped patients with computer and health literacy issues, provided community resources to patients based on identified needs, and collected data to support ongoing performance improvement.

Program and site leadership met regularly to discuss implementation challenges such as information technology and clinical workflow. Run charts were developed to track progress at each site, and screening data were shared regularly with site leadership to monitor and improve performance. Each discipline was supported with their unique challenges with adoption. Elderly internal medicine patients struggled with computer literacy, and the high volume of obstetrical patients limited the time available to complete screening before the clinical encounter.

After screening, NowPow automatically stratified patients according to risk level and produced customized referrals to social services for identified needs (Appendix C, available as a supplement to the online version of this article at http://www.ajph.org). High-risk patients (defined as having two or more emergency room visits in the last 12 months and at least one social need identified through the AHC Screening Tool) who consented to social services navigation received closed-loop referrals to collaborating community service providers through information technology connectivity established by NowPow (Appendix D, available as a supplement to the online version of this article at http://www.ajph.org).

Preliminary data revealed that between September 2018 and August 2019, 13 273 patients were screened across four sites and three disciplines: 1939 patients were identified with previously undetected needs, and 944 were enrolled in navigation to address social service needs (Figure 1). An accurate screening rate has been difficult to determine because of data extraction limitations. Of the population, 27% screened positive for food insecurity, 25% screened positive for housing insecurity, 12% screened positive for transportation needs, 8% screened positive for utility needs, and 1% screened positive for safety needs (Figure 2). Of the population screened, 82% identified as Hispanic, 14% identified as Black/African American, and 68% identified as female. The average household size was 3.6, with an average household income of $24 000.

FIGURE 1— Number of Patients Screened for Social Determinants of Health Each Month: Northern Manhattan, NY, September 2018–August 2019

Note. OBGYN = obstetrics and gynecology. The total number of patients screened was 13 273.

FIGURE 2— Percentage of Need by Discipline: Northern Manhattan, NY, September 2018–August 2019

Note. OBGYN = obstetrics and gynecology.

Implementation brought to light the complexities of a decentralized clinical delivery system. Despite hospital mandates for standardization across sites, it became clear that to facilitate practice ownership of the screening program, each site would require the flexibility to change and adapt its workflows. Technological problems between the screening platform and the electronic medical record interface resulted in workflow disruptions.

Patients who seek care at these practices often feel overwhelmed by the health care system. Introducing a new program in the waiting room that requires computer and narrative literacy adds to this burden. Volunteers were encouraged to approach patients and offer assistance with screenings, to minimize the potential embarrassment that comes with actively seeking help. Staff and volunteers also were encouraged to explain the importance of screening to patients but to emphasize that SDOH screenings were voluntary.

The decision to embed SDOH screenings with routine clinical screenings has been key to sustainability because the latter are mandated by myriad regulatory bodies governing the hospital. Incorporating technology, despite difficulties, ensures that all screening efforts can be tracked and offers data that can be used to understand drivers of health care use and cost.

In response to resource gaps identified by the screening process, two groups were formed to work with community partners on food insecurity and housing instability. As practice staff learned more about the needs of the community and the potential effect that addressing these needs could have on patients, they became increasingly motivated and positively engaged in the screening process.

Screening for SDOH has enabled NewYork-Presbyterian Hospital to identify the social and environmental conditions affecting its patients and address needs that extend beyond standard clinical care. The data collected over the course of implementation were the most robust and standardized social and demographic information the hospital had compiled to date. They have been used to inform programming and interventions at both the practice and the community levels, including engaging a nutritional referral program for families identified as food insecure. In addition, SDOH screening has helped to identify a not previously detected population of “rising risk” patients, who are socially but not yet medically complex.