Healthcare

Modernizing Medicaid Data Collection Applications

Team Coforma helped the Centers for Medicare & Medicaid Services modernize and unify state data reporting applications used to support healthcare services for more than 90 million people.

The Challenge
Our Approach
Results

The Challenge: Develop More Robust Data Reporting for Healthcare Programs

Client Name(s):: Centers for Medicare & Medicaid Services
Partner Name(s):: Aquia, Inc.; BlueLabs Analytics, Inc.
Delivery Date:: February 2023

The Centers for Medicare & Medicaid Services (CMS) wanted to centralize and scale the applications they use for all data collection related to state Medicaid and CHIP programs.

Medicaid and the Children’s Health Insurance Program (CHIP) are the largest programs providing health coverage to low-income Americans. They’re administered by states, according to federal requirements, and serve more than 90 million people across the country.

CMS relies on data sets from state programs to support their work and maintain public transparency. Their existing data reporting applications made it difficult for state users to submit data, and for state health administrators and experts to parse. In addition, CMS’s data collection needs were growing and they needed stewards for their existing applications and support to launch a new one.

During this year-long project, Team Coforma—comprised of Coforma as the prime contractor, supported by Aquia, Inc. and Blue Lab Analytics, Inc.—worked on four different applications, providing human-centered design, research, technical analyses, and engineering along with operational support across multiple workstreams. This suite of products was built using consistent, familiar design patterns and systems for its users to support the continuous improvement of existing CMS data collection applications, as well as rapid scaling to compatible new applications.

Our Approach: Human-Centered Design and Research in Tandem

We rapidly iterated on designs based on user feedback and content audits to modernize existing applications, build a new tool, and develop an intuitive digital form flow.

Our research efforts focused on understanding the needs, goals, and pain points of report contributors in state Medicare agencies. To surface these insights, we hosted 40 primary research sessions with participants from 18 states. We also held weekly subject-matter expert (SME) discussions to understand the information they sought to gather from reporting indicators. Blue Labs Analytics supported recruiting, interviewing, and market research efforts and provided customer relationship management (CRM) expertise. Because the user archetypes for all the applications we worked on were similar—state health administrators and SMEs that plan, collect, review, and finalize the data collaboratively—we had a strong research foundation to apply to each workstream.

Our research team collected and synthesized insights from qualitative interviews, a customer experience survey, and evaluative testing usability studies, our product design team used that data and a detailed content audit to iterate on designs that met business and user needs while conforming to established design system patterns, reporting requirements, and delivery constraints.

We also analyzed existing CMS applications and used our observations to inform our approaches to modernize those applications and build a new tool to achieve the desired functionality. We took inventory of content structures and components from CMS’s design system, conducted a heuristic analysis, and performed accessibility testing to discover items we wanted to replicate going forward, as well as issues to fix.

Iteratively building and evaluating prototypes let us establish interaction patterns and content structures. This knowledge allowed our team to translate data structures defined in Excel into a digital form flow that would facilitate the entry, validation, review, and submission of reporting data.

A graphic shows an abstraction of a transformation from a busy and complicated web form on the left to a clean, simple web form on the right

Improving Existing Form Design

State report contributors were already familiar with a set of data reporting templates, so we kept their content but focused on improving how it was structured across three existing applications. We mapped out the information hierarchies and content inventories, then used these assets to build out and digitize updated form designs.

Each section went through an iterative design and validation process to ensure the updated design met business requirements and was usable for report contributors. We also improved the way information added to the template was recorded in CMS’s system. The result was an updated template-to-system workflow that made it easier to reuse data.

A 3 by 2 grid of 6 measurements used in heuristic analysis written in blue with grey interlocking circles around each word

Application Modernization

We used a heuristic analysis and accessibility testing to evaluate the CHIP Annual Reporting Template System (CARTS), an existing application that states had been using for more than five years. We kept the familiar look and feel of the system while rewriting its back end to optimize data structures, rely on more modern infrastructure, and enable easier future updates.

An abstraction of a slide that displays, in green and red, respective metrics and rankings when doing an accessibility assessment

Improving Accessibility of Existing Applications

Reports posted on Medicaid.gov must be Section 508-compliant by law. We used Acrobat Professional’s accessibility checker to review the PDFs auto-generated using CMS’s data applications. When our accessibility testing uncovered major issues in the PDFs, CMS worked with the system’s original creators to make system-wide fixes while we provided guidance on what to look for and how to fix it.

Our engineering team used the program Prince to pass several document elements through Adobe to reduce the amount of manual remediation necessary. However, automated document remediation doesn’t catch everything. Some testing and remediation must be done manually. For what couldn’t be done through software, our design team identified manual fixes to achieve full 508 compliance.

An abstraction of a generated report in a grid tilted at a 30 degree angle

Building a New Tool

We took lessons from existing applications to design a new tool to collect data related to managed care plans, using a similar design system and incorporating feedback from previous research phases to create something familiar for users.