Coming into Centene fresh, I learned that doctors, health workers, and analysts had a common need: a dashboard that would allow users to see a full list of every member, in any provider network, in any part of the country. Existing solutions would only have the data from a single provider network, or in a specific region, but did not unify all of these member profile in one location, leading to a fragmented data environment that left key details out when analyzing populations to design interventions.
It was made clear to me that this was an initial exploration leading to a 'proof of concept' which would serve to convince leadership to green-light this project for a full design and build process. So, I made adjustments to my process to optimize for speed and to give a general sense of what this dashboard might look like.
Research
In order to make sure the design is based on real users' problems, I asked to speak with the group of several doctors, health workers, and analysts that had been advocating for a new dashboard with these capabilities, and I was given license to interview them.
I had a series of conversations with these key stakeholders, conducting a lean canvas investigation, keeping track of what features were mentioned, what tasks they would like to use this dashboard for, and what their current workflows look like and where the pain points are.
One of the central themes for this project was the "scorecard". According to my group of stakeholders, there was a desire for a summarized view of a person's health, but that summarized view might have up to 100 pieces of information.
We received the following graphic as a representation of the information that would be helpful to have in the scorecard.
We started asking ourselves... How can we make sense of this scorecard? Which sections or pieces of information are most important? How can we show a summarized view of a patient, when the summary could be fairly long and detailed?
Process
When I was hired, this project had already been building momentum, and upon joining, I was asked to join a timeline that boiled down to two weeks to have a wireframe, and one week to have visual designs, in order to have a visual that would be shown at a quarterly meeting of VP's.
With information from our lean canvas exploration piling up, I was able to start sketching. Key priorities here included: clear section for KPIs, clear area for refining the target population, an executive summary of health information, and full health record information that can be explored if needed.
One of the most important metrics that users wanted to see, front and center, were
care gaps.
A care gap "refers to the percentage of individuals who require treatment in a country or a defined community but do not receive it due to various reasons" (
source). This is a metric that allows doctors and care providers to quickly understand whether there are any gaps in their care, whether it is a mammogram that is overdue, an asthmatic person who hasn't renewed their prescription for a long time, or a colonoscopy that was ordered but never completed.
So, how can we optimize this dashboard to display care gaps, show the scorecard, provide KPI's, and avoid being overwhelming to look at?
With my key priorities and concepts outlined, I created this first, early sketch, which helped define the overall structure of the dashboard:
Sketching
That initial sketch was completed on a Friday, and over the weekend, I continued thinking it through. On Monday, I continued sketching and produced several concepts, one layer deeper.
For the main page / list view, I positioned the filter at the top of the screen, put the KPI section underneath the filters, and outlined a large scrolling table taking up the bulk of the page. This would be the main view that allows health workers to select a population and quickly see important information about that population, both at a high level (KPIs) and in detail (scrolling table).
I provided two alternative concepts for how to arrange the detail view. The first version has an expand & collapse pattern, and the second concept is structured around using tiles or cards to provide an overview of member data that could help to separate information thematically.
Collaboration for Visual Design
I communicated my thoughts that led to these sketches and received positive feedback from our internal team. Due to the tight timeline and its status as a 'proof of concept', we didn't further validate these sketches and moved straight into visual design.
We decided that the expand & collapse pattern made more sense for the member detail page. What made this clear to us was the concept of the 'scorecard', which was discussed as a 'card' that would show the 'highest priority information' right up top. Based on the needs communicated, we felt that we could structure the collapsed view of the member detail page in a way that would keep the most important information permanently displayed.
Unfortunately, since I was still new to the organization and my work computer was having some technical issues, I needed to partner with a fellow UX designer who had full, unimpeded use of their computer to make the designs in Figma. Kajal, you're the real MVP!
Working with Kajal was a delight, and based on my screenshots, she created an initial draft of these visual screens, which we then edited and refined together. As a team, we worked to ensure that our refiners and KPIs were logical and told a story that was relevant for health workers. We added a secondary filter, 'Refine List', which gave prominence and more ability to cross-compare to our important metrics: Quality Gaps, Risk Adjustment Gaps, and Clinically Potentially Addressable Issues. One of my contributions at this step was to insert the urgency tier number into the scrolling table, bringing prominence and providing clinicians an immediate way to sort the list based on systematically determined need.
Here on the member detail page, you can see how the "scorecard" area expands and collapses, and the urgency tier is maintained as a prominent piece of information. On this page, I worked with Kajal to develop logical groupings of sections, and also provided feedback that led to a clearer presentation of the "SDOH Issues / Other Programs" area.
Implementation & Impact
We presented these three screens and our thinking around them to a small group of executives and decision makers, and received positive feedback.
One of the biggest takeaways was that the information in the scrolling table on the main list page was not nearly enough information. They said, "will it scroll to the right? There might be dozens of additional collumns that we want to add."
However, this proof of concept served its purpose, helping to generate excitement, conversation, and momentum for a need that has been developing over many years at Centene.
Unfortunately, I was laid off only a few weeks after completing this project, well before the anticipated 12-month contract was over. My manager cited 'budget issues' and, sadly, both me and Kajal, as UX contractors, were let go from the company.
Learnings & Future Path
From my time at Centene, I learned a large amount about successful dashboard design. I worked with MicroStrategy, Tableau, and Power BI, and was able to take several company-sponsored trainings about working with data, visualizations, and data storytelling. I learned about many of the challenges of managing data at a large healthcare network, and got a window into how such an organization runs their Business Intelligence "Center of Excellence", which operates as a consulting body for the more than 2,000 dashboards that serve the 70,000 people working at Centene to provide quality healthcare.
Despite the unsatisfactory end to my time at Centene, I leave with an enlarged amount of respect for data scientists and data designers, more knowledge about how to design industry-leading dashboards, and with renewed interest in the healthcare space which is tackling some of the most complex, important, and interesting challenges in any industry today.
.png)
