Never let a pandemic go to waste: applying an equity-focused quality improvement framework to close gaps in patient portal activation

Background Studies show that historically marginalized populations, such as racial/ethnic minorities, those living in poverty, and those with limited English proficiency are less likely to utilize patient portals that are becoming increasingly integrated into clinical care. In March 2020, nearly all in-person clinical operations for our large primary care network were ordered to cease due to COVID-19. In order to pivot quickly while ensuring our most vulnerable children and families were not left behind, we leveraged an equity-focused quality improvement (QI) framework. As we rolled out increased telehealth capacity we concurrently sought to reduce existing disparities in patient portal activation and utilization. Methodology We executed our intervention in a large primary care network that cares for over 270,000 children in southeastern PA/NJ, including over 90,000 Medicaid-insured children. To determine baseline data, we queried our data warehouse and identified the patient portal activation status of all patients with a primary care visit between January 2018 - December 2019. We then stratified the data by sociodemographic variables: race/ethnicity, insurance type (private or government), preferred language, and the percent of households in their neighborhoods (approximated by census tract) living below the federal poverty line. This analysis revealed disparities in activation status, which prompted assembly of a multidisciplinary task force to ensure that barriers were quickly removed for families to access telehealth. We used an equity-focused QI approach to carry out cycles of sequential interventions. Discussion During the two-year pre-intervention period, nearly 300,000 patients were seen in our primary care practices. Notable disparities were observed across all four demographic categories examined (Figure 1). Differential patient portal activation status was noted by: race/ethnicity (69% among non-Hispanic white vs. 42% among non-Hispanic black patients); insurance types (67% among privately-insured vs. 42% among Medicaid patients); language (60% among English speaking vs. 22% among of non-English speaking patients); and poverty (66% in the lowest poverty neighborhoods vs. 38% in the highest poverty neighborhoods). After iterative innovative interventions to remove technological and process barriers to portal utilization, rates of activation for the entire primary care population increased by 16% overall. Additionally, we observed marked improvement in activation rates among Medicaid-insured children (26% increase); among Spanish-speaking families (36% increase), and African-American families (23% increase). Conclusion We demonstrated significant narrowing in baseline disparities in electronic patient portal utilization by race, ethnicity, insurance type, and language. This improvement was accomplished through use of an equity-focused QI framework that resulted in interventions focused on removing as many barriers as possible and creating a value-add for patients. Next steps include demonstrating sustained improvement, evaluating patient satisfaction, and ascertaining which of our interventions were most contributory to our improvement.