Diagnoss
PAVING THE PATH FOR AI IN MEDICAL CODING
Overview
Diagnoss is amongst the first to introduce artificial intelligence into the medical coding space.
As the team UX/UI Designer I was tasked with leading a three-part design.
This included the landing page, the SaaS analytics interface, and the medical coder interface.
Deliverables
The Problem/Hypothesis
Until recently, healthcare institutions as we know them were not equipped to leverage big data. Artificial intelligence is introducing countless opportunities in healthcare. As healthcare institutions explore the implementation of AI software solutions, a few common questions stand... How can we (the healthcare institution) quantify the expected benefits we will receive from this software? Will our staff experience a smooth transition with the implementation of this technology?
The Question
The Goal
The How
How can we best communicate the value of the software that we are providing?
Identify and design for user needs.
Three part design:
Design a concise landing page sign up flow. (Identify the most necessary and intuitive user flow.)
Design a clinical management interface consisting of components that track efficiency metrics over a given period of time.
Design a coder interface for maximized efficiency. (Components must include code suggestions and claim status.)
Conduct usability studies, monitor website analytics, collaborate with the cross-functional teams to design for user needs.
The Research
We kicked off research with stakeholder interviews.
I conducted a SWOT Analysis to gain an understanding of business needs.
I conducted user interviews with clinical management as well as medical coders to gain an understanding of user needs. I then synthesized the affinity map into a collection of key insights.
Upon gaining a preliminary understanding of business and user needs, I proceeded to conduct usability tests across the three product platforms…
Landing Page
I conducted landing page usability tests with 6 clinical managers. After multiple iterations, our redesigned user experience yielded a 12% increase in conversion.
Key Insights:
User flow and UI content.
Clinical Management Interface
I conducted usability tests with 5 clinical managers. We prioritized the features and metrics based on both user and business needs.
Key Insights:
Value of integrating the entire digital environment with the analytics interface.
Prioritizing analytics models with both user needs and technical feasibility in consideration.
Feature prioritization
Heat Map
Medical Coder Interface
I conducted usability tests, A/B tests, and heat mapping with 16 medical coders. After multiple iterations in collaboration with developers, we designed the ideal medical coding interface yielding a 14% increase in efficiency (submission time, accuracy, and CSAT).
Key Insights:
Fine tuned the user flow, component arrangement, and code descriptiveness.
Explainable AI (XAI) methods were used to provide users with context (quantitative and qualitative) pertaining to code suggestions.
Implemented dark UI in order to counteract eye strain.
Lo-Fidelity Medical Coder Wireframes
After multiple iterations the design was ready…
The Design
Landing Page
Management Analytics Interface (NDA Limited Preview)
Medical Coder Interface (NDA Limited Preview)
Thank You!