April 2019

cover story One

Social determinants of health

Customizing patient care

ur work in health care is only part of what contributes to overall health. Social Determinants of Health (SDoH)—people’s education, employment, finances, legal status, literacy skills, English fluency, transportation, housing, and, most importantly, their neighborhood—can be just as important. SDoH factors contribute to health disparities in Minnesota; our communities with high SDoH burdens experience higher rates of disease and disabilities. Social inequities also contribute to disparate medical quality scores; those of us who see patients with high SDoH burdens have lower scores.

The challenge is to identify these issues in our patients and determine how to respond as individual physicians and as members of health care institutions to promote health. Struggling with these issues at Minnesota Community Care (previously known as West Side Community Health Services), we have taken some steps to quantify, explore, understand, and respond to SDoH.

About our clinic

As a Federally Qualified Health Center (FQHC) with 17 locations in Ramsey County, Minnesota Community Care (MCC) provides comprehensive primary health care services to a population that disproportionately shoulders the burden of health disparities in our community. In 2017, of 36,338 patients, 98 percent had incomes below 200 percent of the federal poverty level, 71 percent were women and children, 42 percent were medically uninsured, and over 56 percent did not speak English as a first language. In addition, 86 percent were from communities of color, predominantly Hispanic/Latino, Black/African American, and Asian (mostly Hmong). MCC services are available to all; patients with incomes less than 300 percent of poverty level are offered a sliding fee program. No one is turned away for lack of insurance or inability to pay for services. Anecdotally, we know that many of our patients struggle with SDoH issues, but other than race/ethnicity, preferred language, and country of origin (often known collectively as RELO), we had not collected specific data to quantify these issues.

Working with these underserved populations led us to seek additional information and develop a focused response. In 2016, we began to develop a data system that provided direction to more effectively impact the communities served and to identify strategies in specific clinical indicators and health systems that could overcome the barriers that facilitated inequitable health care. Through the Disparities Leadership Program at Harvard University, we decided to expand our identification and understanding of SDoH that may be affecting our quality metrics and contributing to our health disparities and health inequities. We adopted the Kotter Model for Leading Change as a blueprint for change and to translate understanding of disparities into realistic solutions. Dr. John Kotter observed organizations execute their strategies for over 40 years, extracted the success factors, and developed them into a methodology. (See www.tinyurl.com/mp-kotter.)

The Kotter Model shows eight steps toward leading change:

  • Create a sense of urgency.
  • Build a guiding coalition.
  • Form a strategic vision and initiatives.
  • Enlist a volunteer army.
  • Enable action by removing barriers.
  • Generate short-term wins.
  • Sustain acceleration.
  • Institute change.

This model challenged us throughout the improvement project, from assessing varying accomplishments—or lack of progress—and as we completed the project. While we initially assessed that the system resided at “generate short-term wins,” the team later determined that, in some aspects, the first step of “creating a sense of urgency” had not yet been created. The Kotter Model was instrumental in guiding our progress toward a deeper level of cultural assessment and uncovering system barriers that critically impacted the communities we served.

We are looking for new, innovative ways to partner with other organizational and societal advocacy groups.

Preparing for PRAPARE

With the system now ready for change, the first goal we set included developing a sophisticated data infrastructure to identify social determinant disparities with the hope that we could target our services to identified special populations. Prior to this goal, we had used a data structure that produced system-level data to support reportable measures, such as diabetes, asthma, and cancer screenings. This system did not include the infrastructure to include disparity reporting, including race, ethnicity, language, and country of origin (RELO), and SDoH data with insecurities such as food, housing, transportation, legal services, safety of household, and neighborhood.

In an alignment of timing, the National Association of Community Health Centers (NACHC) began dissemination of the PRAPARE (Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences) tool. The tool consists of a set of national core measures as well as a set of measures for community priorities. It was informed by research, the experience of existing social risk assessments, and stakeholder engagement. Core measures include personal information about education, employment, income, literacy skills, and safety; insecurities in housing, utilities, food, and transportation; and factors such as social integration and stress. (See www.tinyurl.com/mp-prapare.)

We customized the PRAPARE tool based on patient engagement and stakeholder assessment and created an adjusted tool to collect the measures. Data collection began in June of 2017, focusing on adult populations as a pilot project and then expanding to all family practice sites. Because the PRAPARE tool has not yet been validated in pediatric populations, we continue to focus on adults.


The results to date quantify the extent of SDoH for 3,756 adults who have completed the form. Most of the adults who completed the form were middle-aged women with a range of ethnicities/races, education, and insurance. About 10 to 18 percent of respondents reported having insecurities in housing, utilities, food, clothing, childcare, and phone service. About 9 to 18 percent had difficulties accessing needed medical, mental health, and dental care. Many respondents reported low levels of contact with people whom they care about and are close to: 41 percent said they interacted with others just one to three times per week, while 48 percent said their contacts occurred less than weekly. When asked about stress, 55 percent reported they were somewhat stressed, and 22 percent had a lot of stress. To explore how these SDoH factors may affect quality of care, we are examining how these issues are related to our quality metrics, and, therefore, may be affecting our quality measures. [Note: MCC reports to the federal government using the Uniform Data System (UDS), and to Minnesota’s Statewide Quality Reporting and Measurement System (SQRMS) through MN Community Measurement.]

Data on two fronts

For this article, we examine one preventive health service measure (cervical cancer screening) and one chronic disease measure (A1C as a measure of control for diabetes mellitus).

Our overall cervical cancer screening rate is 55.2 percent, which is the percentage of women 23 to 64 years of age who have had pap smear in the previous three years. The highest rates are in Hispanic White women (65 percent), younger women (57 percent <40 years of age), and those who are uninsured on our discount program (54 percent). The lowest rates are in women who report a lot of stress versus no stress (48 percent versus 59 percent); in women who connect with people they care about less than weekly versus more than three times a week (53 percent versus 61 percent); and in women who report unstable housing versus stable housing (35 percent versus 56 percent). There are no differences in rates for education and literacy.

We Minnesota physicians continue our dedication to improving health of Minnesotans.

Our overall diabetes control rate is 29.1 percent, which is the percentage of adults with diabetes mellitus type 1 or 2, 20 to 75 years of age, who have had an A1C <9 percent in the past year. The highest rates are adults >60 years of age (70 percent), women (64 percent), and non-Hispanic Whites (64.9 percent), with about equal rates of insured and uninsured. The lower rates are in people with a lot of stress (62 percent versus 79 percent) and with transportation challenges (60 percent versus 70 percent). There are no differences in rates for housing, education, and social integration.

Follow-up action steps

What do these results mean, and what do they signify?

These results are limited in several ways. One, these results do not represent our entire population. While MCC serves over 36,000 people, only 10 percent have completed a PRAPARE questionnaire so far. Two, these descriptive statistics are not analytical statistical assessments. They are the beginning descriptions that afford us additional insights into SDoH for our patients. Three, they represent slices of people’s lives, and do not measure other aspects of health, health status, and the many other factors that impact their ability to achieve healthy outcomes.

Despite these limitations, the PRAPARE data lay the foundation for ongoing clinical improvements. Understanding the SDoH connections more clearly than we did before we collected the data, we are designing internal quality improvement efforts to align our clinical care with identified disparities, deficiencies, and gaps, in order to improve clinical and operational outcomes. And we are looking for new, innovative ways to partner with other organizational and societal advocacy groups to address social inequities.

For pap smears, these results seem to indicate that we are doing well with young women, and Latina women, regardless of insurance status. Our Spanish-speaking staff and providers, our discount program, our connections with the SAGE Cancer Screening program, and our relationship with young women seem to be working well. We are looking at designing specific clinic approaches to reach out to older women, non-Hispanic women, stressed women, women who attend our mental health services, and women with unstable housing in new ways.

For diabetes control, we seem to be doing well with older people, women, and non-Latinos. Our current discount program for medical care and medicines may be helping people without insurance have rates equal to people with insurance. To improve control, we could design special efforts to target younger people, to identify and approach stressed people, and to create telemedicine programs and use mobile health technology to reach people with transportation barriers.

For these two examples and for our other quality metrics, we are examining how to outreach beyond our clinic walls to partner with and refer to local community organizations that can address SDoH issues. To this end, we are connecting with NowPow, an internet-based tool that, through the patient’s EMR, can connect patients with targeted/needed community resources. And we are planning on instituting a Medical Legal Partnership to support patients with legal needs.


For physicians and organizations interested in exploring, identifying, and then targeting social factors that influence health, we recommend a process similar to our own:

  • Use the Kotter Model for Leading Change to assess and guide your organizational process. Create a sense of urgency by connecting with others who care about this and work with your leadership to create a coalition of people to carry the work forward.
  • Decide whether you want to collect data to affect population health or individual health—or both, as we did. The tension is inherent: we are asking individuals to report on their personal situation, so we can collate population data, but in doing so, we decided to also respond to individual needs by adding a question to the survey: Do you want to see a social worker today?
  • Choose a data collection tool, then adjust it to fit your organization’s goals and your populations. Consider types and number of questions, wording of questions, length of time to complete, language, literacy, and your local context.
  • Plan and implement a pilot program, review process results, and adjust as needed.
  • Review the results—who answered, what they answered—and relate the results to measures that are important to your organization’s goals and quality improvement projects.

Finally, reach out to others who are attempting to identify and respond to SDoH. You are not alone in this effort, as we Minnesota physicians continue our dedication to improving health of Minnesotans.

Kathleen A. Culhane-Pera, MD, MA, is medical director of quality at Minnesota Community Care. She is a family physician, having received her medical doctoral degree from Michigan State University and completed her family medicine residency from the University of Minnesota. She is also a medical anthropologist with a master’s degree in anthropology from the University of Minnesota.

Chris Singer, MAN, RN, CPHQ, is chief operating officer at Minnesota Community Care. She has over 20 years of experience in clinical care leadership focused on improving quality for patients in a variety of health care settings. She has a clinical background as a registered nurse and holds an MA degree in nursing from Bethel University with a focus on health systems leadership as well as certifications in health care quality and leadership.  


PO Box 6674, Minneapolis, MN 55406

(612) 728-8600


follow us

© Minnesota Physician Publishing · All Rights Reserved. 2019


cover story One

Chris Singer, MAN, RN, CPHQ, is chief operating officer at Minnesota Community Care. She has over 20 years of experience in clinical care leadership focused on improving quality for patients in a variety of health care settings. She has a clinical background as a registered nurse and holds an MA degree in nursing from Bethel University with a focus on health systems leadership as well as certifications in health care quality and leadership.