Embedding a Community Resource Desk in Family Medicine Clinics: Lessons from Portland, Oregon

By Rachel Smith, Providence Health & Services

While visiting her primary care provider, Louisa informed her doctor that she was struggling to pay her rent. She had a job, but was putting most of her income towards housing. Her doctor referred her to the Community Resource Desk (CRD) which is coordinated by a co-located social service agency, Impact NW, on-site at her clinic.

On the same day, she met with the CRD and was referred to an organization that offered affordable rental and housing options.  She also learned that she qualified for energy assistance, which could help offset some of her limited finances.

It was clear that Louisa’s financial situation was impacting her health, but situations like hers are often difficult to quantify. How do we move from knowing that programs and partnerships are making an impact on patients’ day-to-day lives to knowing that they impact their health? For Providence to answer that, we partnered with the Center for Outcomes Research and Education (CORE), an independent research team under the umbrella of Providence that examines the intersections between health care, social services and life experiences on individuals and communities’ health and well-being.

Through funding from the Social Intervention Research & Evaluation Network (SIREN), we have been able to assess:

AIM 1The impact of community resources desk on participant reported outcomes
AIM 2The impact of a community resource desk on utilization and costs of care

We decided to tackle these aims by conducting a longitudinal study comparing clinics with and without an embedded CRD on-site; we then followed patients with an identified social need for one year. We selected three intervention sites with a CRD, the same as our Bridging the Gap clinics, and four control clinics that did not have a CRD on-site but had similar patient demographics (e.g., payer mix, primary language, race and ethnicity) and neighborhood characteristics. In the study design, we also included pre- and post-surveys and a claims component.

We began with a baseline survey to measure health, social determinants of health and other patient characteristics. This was seemingly straightforward, but as we worked on the design and methodology we encountered several questions:

  • How do we successfully connect with this population, knowing that low-income populations can be harder to reach?
  • How do we reach patients without burdening clinic teams or unintentionally bias our results by having CRD staff distribute our survey?
  • What do we do if someone shares they have a social need? We felt that ethically we needed to have strategies to address basic needs once the need is known.

We did not have all the answers, but we decided that part of innovative work is to try new processes. We incorporated more electronic survey options so that we did not have to rely only on mail surveys. We added a QR code to mail surveys (with $5 incentives) and used e-mail reminders with survey links. This was a success. We had an approximate 40% response rate, with over 60% of respondents having an income of $30,000 or below – so we knew we were reaching our target population.

We found that 54% (n= 1,263) of respondents had at least one social need. Of the 14 needs we inquired about, we broke the categories down into social resource needs, more akin to basic needs like food, housing and transportation, and health-related resource needs, like dental care, health insurance, counseling/behavioral health.

Overall, 64% indicated that they had at least one social resource need and 85% indicated at least one health-related resource need.

Similar to national trends, our initial findings show that patients’ health is often connected to their social needs. More than 30% of respondents said they had poor or fair physical, mental or dental health with approximately 50% self-identifying as experiencing anxiety or depression and over 40% shared they had high blood pressure.

Disappointingly, the number of individuals who attested they had visited a CRD in the last year was lower than we had hoped; only about 12% of respondents had used the CRD. This makes it challenging for us to assess whether short solution-focused interventions can influence outcomes and utilization. It also pushes us to explore new questions like:

  • What are some of the reasons people with social needs do not go to the CRD?
  • Is our approach to addressing social needs working for those who need it the most?
  • What can we improve?

These questions and more, have led us to take action on a few things. A 2020 priority for the program is to do an assessment of each CRD site. We plan to utilize client data, shadow CRD staff to learn from their processes, review our protocols, and conduct feedback sessions with frontline staff, patients, and other stakeholders. While we strongly believe in this work, we also want it to be as effective as possible, and sometimes that means taking a step back to ask the hard questions. 

The prevalence of the needs in our clinical settings reminds us that we still have work to do. Resources like the CRDs, partnerships with organizations like Impact NW, and programs to improve integrated care like Bridging the Gap are critical to supporting infrastructure that helps our staff not just identify needs, but find the best strategies to address them as well.

If interested in learning more about this work, contact Rachel Smith at rachel.smith@providence.org.