Goal: End Racial Disparities in Diabetes Care

By Steven M. Willi, MD, Director, the Diabetes Center 

Back in 2015, we wanted to test the premise in the medical literature that it was socioeconomic status (SES), not race, that accounted for poorer outcomes in non-Hispanic Black children with type 1 diabetes when compared to white children. We examined a large cohort of children from across the country with type 1 diabetes, including Black patients whose families earned more than $100,000, and confirmed that non-Hispanic Black patients had lower insulin pump use and higher mean hemoglobin A1c than white patients, even when we controlled for SES.

In response to this evidence of racial disparities, the Diabetes Center at Children’s Hospital of Philadelphia (CHOP) changed our practice in two ways:

1. We moved to a team-centered approach for ongoing care, with the same endocrinologist, nurse practitioner, nutritionist, diabetes educator and social worker seeing the same set of patients throughout the year.
2. We accessed an internal grant program called the Chair’s Initiatives to fund and implement a community health worker (CHW) model to give extra support to families that struggled to control their child’s T1D.

To our knowledge, it was the first CHW program in a pediatric diabetes center in the country.

Were our efforts helping?

We studied our own patient population at CHOP, reviewing more than 1,300 electronic health records (EHRs) from 2018 and 2019, examining healthcare utilization (appointment attendance, ED visits, hospitalizations), technology use (insulin pumps, continuous glucose monitors [CGM]) and hemoglobin A1c (HbA1c) for each race/ethnicity. We stratified data by insurance type (private/government) as a proxy for SES.

Again, we found disparities. When compared with White children, Black children were 5 times more likely to have HbA1c levels higher than goal, 4 times more likely to attend the Emergency Department and 8 times more likely to be admitted to hospital. They were also 4 to 5 times less likely to be treated with insulin pumps or CGM, regardless of insurance status. One encouraging data point: There was no difference in number of appointments attended.

We wanted to do better

The numbers showed we needed to do more to erase racial disparities in care.

We turned again to CHOP Chair’s Initiatives, our hospital’s incubator for innovative care improvement ideas, and received funding to take the next steps to address racial disparities. We consulted with CHOP’s Center for Health Equity and the hospital’s quality improvement group as we refined our tasks and goals.

Our project proposed to:

• Carefully examine our systems, behaviors and practices around diabetes care to identify sources of persistent disparities.
• Gather insights from families of children we serve (via focus groups and surveys) and from our clinical care team members (through surveys, the Implicit Association Test (IAT) and randomized case study assessments).
• Align and implement EHR metrics to track care outcomes in a way that facilitates quick reporting for the Diabetes Center overall and by individual provider as well. The data will enable delivery of ongoing feedback on each provider’s specific care decisions.

In the first two years of this project, we’ve made progress in our implementation, but we still are not far enough along, technology-wise, to be able to turn on automated monthly scorecards for each provider. However, Diabetes Center leaders can access scorecard reports.

First steps

Here’s what we’ve accomplished so far:

• Redesigned documentation workflow in the EHR to streamline diabetes care documentation and enabled more efficient and accurate data collection to monitor and track health disparities.

• Deployed the Racial Disparities in Diabetes Qlik Sense dashboard to monitor and track health disparities health outcomes and diabetes technology use across clinical care teams and individual care providers.

• Assessed CHOP Endocrinology provider and staff implicit bias through the IAT and undertook an assessment of the attitudes, beliefs and personal experiences of providers that may contribute to implicit bias.

• Assessed CHOP diabetes provider implicit bias in clinical decision-making through use of a randomized case report survey.

• Engaged caregivers of non-Hispanic Black children with diabetes, through five focus groups and a parent advisory panel, to better understand the diabetes care experiences and perspectives of these families — with a particular focus on structural racism.

• Assessed the relationship of hypervigilance and social needs to the Diabetes Center care experience of caregivers of non-Hispanic Black children.

In general, families of color told us that they generally received excellent care and they felt respected. However, when there were problems with diabetes management, families felt discouraged and overwhelmed by the negative communication with their care team members during diabetes appointments. They also reported being hypervigilant in their daily lives, e.g., having to watch what they say, and preparing for possible insults during interactions with others.

Next steps

We aim to build on the relationships we have with our families to recruit a diverse group of caregivers for an advisory panel to keep communications channels open going forward. Based on what we learned from families, we will work with providers on improving communication when diabetes management is not going well, and become aware that the need for hypervigilance is an added stressor for marginalized families. 

We need to encourage the remainder of providers to take the IAT and then analyze the data. Is there a relationship between IAT results and patient outcomes? Do staff need additional training or awareness building?

Preliminary dashboard data shows that some of our Diabetes Center teams and individual providers have a higher degree of disparities between racial groups than others. We see this as an opportunity for improvement. If data show that certain providers are, for example, not consistently ordering lipid levels for Black patients, we can take note and send the providers information, including specific patient names as needed, to bridge the gap.

Providers will have access to their individual scorecards, and during our weekly patient review meetings, we can discuss any disparities and strategize how to overcome them, using successful interventions as a springboard to additional improvements. We want to meet providers where they are in this journey and work with them collaboratively to create solutions.

The early focus of this project has been on patients with T1D, and we will broaden the scope to include the growing number of patients with type 2 diabetes.

We plan to share our learnings across the institution, where we believe successful strategies can be replicated, especially in divisions that manage children with chronic diseases and that may also see racial disparities in outcomes from the care they provide.

We are unwavering in our goal: To ensure equitable care for every patient in our care to improve diabetes management outcomes across all races. Well-managed diabetes now will set these children up for a longer and healthier life.