03
New Product: Designing for Machine Learning
Product Team Collaboration, Research and Design for a Health Services Bundling Tool
Medium-fidelity, public-facing UI wireframes show the integration of “Related procedures” and confidence scores into the existing Medicare Cost Estimator tool.
Client
WETG
Challenge
Through machine learning, we can create bundles of health services that consumers commonly receive together. But how can we surface bundles in a way that addresses real healthcare pain points? And how might we ensure that bundles are surfaced to medical experts so that they can evaluate the bundles for usefulness and safety?
Approach
- Expert interviews & comparator analyses
- Solutioning workshops
- Content design
-
Research and design input on product decision-making
-
Front-end testing
Role
Sole researcher & product team member
Output
- Clickable prototype
-
Content mockups
- A working validation tool
The interface to the expert-facing validation tool allows medical experts who are reviewing bundles to search claims codes, add related procedures to a bundle, rate and comment.
Documentation
Public-facing
Expert-facing tool
Data
• EPIC Conference All Hands Presentation--Data Ethnography
• Data Cards Workshop
- Research plan
- Supporting document for the internal CMS Division of Research
- Health Literacy Workshop
- Knowledge Share: CMMI episodes of care vs. MCiEt bundling
Expert-facing tool
- SME Interview Agenda & Questions
- SME Validation Tool Usability Script
- Validation Tool
- Test flows for Validation Tool
Data
• EPIC Conference All Hands Presentation--Data Ethnography
• Data Cards Workshop