O1027: To Keep Us on Track: An Equity Dashboard for the Emergency Department

Purpose/Objectives: Disparities are often addressed only within the context of specific projects. To address disparities over time, we need equity metrics central to operations. Dashboards represent a method for visualizing healthcare data, identifying crucial patterns, guiding quality improvement efforts, and improving patient outcomes. Most disparity dashboards were static and only sometimes updated after the final versions were released, suggesting the demands of maintaining disparity dashboards. This approach is one step to mitigate inequities in care through rapid assessment, illumination, and action.

Design/Methods: A multi-disciplinary team created an interactive dashboard using the business intelligence tool, Power BI, using data from our institution’s electronic health record. The dashboard included our existing emergency department (ED) operational metrics, e.g., time from patient arrival to 1st provider, ED length of stay, and patient disposition. The dashboard examines each metric is examined from the lens of REaL (race, ethnicity, and language) data in subgroups found commonly in the literature for similar exploratory data analysis. The dashboard includes other filters for each metric, based on variables with known or hypothesized associations with each metrics, e.g., arrival location and triage level.

Results: The dashboard identified several potential equity gaps for ED patient populations, primarily based on ethnicity (Hispanic versus not) and primary language (Spanish versus English):
- The average time from patient arrival to the first provider was 14 minutes longer for patients identifying as Hispanic than for non-Hispanic patients: 87 and 73 minutes, respectively.
- Average time from patient arrival to 1st provider was 18 minutes higher for Spanish-speaking patients compared to English-speaking patients; 91 and 73 minutes, respectively.
There were also areas where equity gaps for historically underrepresented groups did not obviously appear present:
Average time from arrival to 1st provider for Black or African American patients and White patients was the same (73 minutes).
Average ED length of stay was lower for populations of interest (Black or African American, Hispanic, and Spanish-speaking patients) than their respective reference groups (Whites, Non-Hispanics, English-speaking patients).
Equity gaps have an important seasonality attribute. The most pronounced differences occur during the Fall months when the EDs experience high patient volumes. Additionally, triage level appears to be a significant factor; equity gaps are primarily present in patients with triage levels 3, 4, and 5.

Conclusion/Discussion: Factors such as sample size, reference group selection, and filtering characteristics make it challenging to understand the scope of equity gaps in patient populations. However, leveraging business intelligence tools for data exploration of historically underrepresented populations can identify potential areas for further analysis and options for targeted QI projects. Thus, a visual dashboard is one step to search for effective interventions that promote health equity.