Cohort Definition Builder
Redesign of Atlas, a powerful data retrieval and analytic tool, in tandem with developing a comprehensive design system.
Project Background
Project duration:
Sept 2023 - Dec 2023.
My Role
In my role as the UI/UX designer, I undertook the redesign of the Cohort Definition Builder to enhance user experience and streamline the cohort creation process.
Responsibilities
User research, UI design, low and high-fidelity prototyping, conducting usability studies and iterating on designs.
The Problem
The overall user interface is complex or lack user-friendly features, making it harder for users, especially those new to the platform, to navigate and understand.
The Goal
Simplify the user interface to make it more intuitive for users at all skill levels.
Streamline the cohort definition process to reduce the time and effort required from users.
The product:
Atlas is a web-based tool developed by OHDSI for exploring and visualizing data in the OMOP CDM. It provides a user-friendly interface for researchers to interact with standardized health data and perform analyses. Atlas supports a variety of analyses, including cohort identification, characterization, and population-level estimation.
Atlas allows users to define cohorts of patients based on specific criteria, such as age, gender, clinical conditions, and medications.
Design Process
Empathize
Define
Ideate
Prototyping
Testing
Empathize
User Research : Summary
The potential users encompass a broad spectrum of people with varying needs, preferences, and health-related requirements. Some key user segments might include:
-
Healthcare Researchers: These are professionals conducting research using health data. They may be epidemiologists, data scientists, or researchers interested in analyzing and deriving insights from large-scale healthcare datasets.
-
Healthcare Analysts: Analysts working with healthcare data, who use Atlas to explore and analyze data patterns, define cohorts, and perform observational analyses.
-
Data Scientists : Individuals with expertise in data analysis and modeling, leveraging Atlas for extracting meaningful information from observational health data.
-
Tech-Savvy Consumers: Individuals comfortable with technology and mobile applications, who appreciate the convenience of accessing healthcare services through their smartphones.
-
Clinicians: Healthcare professionals, such as physicians or nurses, who may use Atlas to explore data patterns, identify patient cohorts, or contribute to clinical research.
-
Biostatisticians: Professionals specializing in statistical analysis within the field of biomedicine. They might use Atlas to perform statistical analyses and generate evidence.
-
Educators and Students: Individuals in educational settings, such as universities or research institutions, who use Atlas as a teaching tool or for learning about health data analytics.
User Interviews
-
What are your primary goals when creating cohorts within Atlas?
-
What are the biggest challenges you face when using the current Cohort Definition Builder?
-
Have you ever abandoned creating a cohort due to difficulty using the tool?
-
Is there anything else you'd like to share about your experience using the Cohort Definition Builder?
-
What are your ideal features for a user-friendly cohort definition tool?
Stakeholder Interviews
Understanding Business Goals - We aim to increase user adoption by making the cohort creation process simpler and more intuitive.
Defining Product Vision - The tool should support researchers in creating cohorts effortlessly, without needing deep technical knowledge.
Identifying Pain Points - Users struggle with the current interface due to its complexity and disjointed workflow, often leading to drop-offs.
Setting Success Metrics - Success will be measured by higher cohort completion rates, fewer support queries, and improved user satisfaction.
Technical & Resource Constraints - The redesign must work within the current technical infrastructure of OHDSI Atlas, with minimal backend changes.
Empathy Map
User persona
.png)
User Story
Aa a health care researcher, I want a tool that streamlines the cohort building process, allowing me to focus more on analysis and less on navigating complex interfaces
User Journey Map
User journey map is a illustration user goes through to achieve their goal. I created user journey map which is based on the persona, where we discovered emotions, behavior and feeling experienced by the users while reaching their goals
Persona : Dr John Reddy
Scenario : Creating a cohort
Let's have a look at old design - Heuristic Evaluation Results
Secondary Research
Competitor Analysis
Research Insights
Through research we learned that:
-
User flow is complex for old atlas, user need to switch between three tabs search, concept set, cohort definition for cohort generation.
-
Users need minimalist cohort build workflow
-
My initial thought is that we should start by combining the "Search," "Concept Sets," and "Cohort Definitions." sections into one unified workflow on a single page.
-
While discussing with stakeholders , they were interested a design like a logic builder, kind of hierarchical structuring.
Define
Problem statement
Sarah, a researcher short on time, needs a way to easily find the right group of patients for his studies using Atlas, Because, The current builder is confusing and clunky,
Hypothesis statement
If we redesign the Cohort Definition Builder in Atlas to be more user-friendly with clear navigation, simple filtering options, and informative data visualizations, then researchers and analysts will be able to define accurate and relevant cohorts more efficiently, leading to a reduction in time spent using the builder and an increase in the overall effectiveness of their healthcare research
Value proposition
-
Reduced time spent using the builder
-
Increased efficiency in defining cohorts
-
Improved effectiveness of research
-
User-friendly interface for a smooth experience
Ideate
Ideation
Brainstormed design concepts focusing on simplifying the interface, removing unnecessary complexities, and improving overall usability. We created bunch of ideas using "how might we" exercise.
After creating possible solutions we evaluated the ideas, and we moved forward with logic builder interface for the builder.
User flow
We created user flow using figfam illustrate how the user will navigate.
Initial wireframes
Created wireframes before starting the design, it helped to outline the structure and layout of your cohort definition builder in a simplified manner.
Visual Design
-
New Color Scheme: Introduced a fresh color palette to enhance visual appeal and create a more engaging user experience.
-
Typography : Implemented a new typography style for improved readability and a modern aesthetic.
-
Ui components : Designed and developed custom UI components to ensure consistency and cohesiveness throughout the Atlas platform.
Voting
We organized a design voting process to select the best design for a cohort definition builder. That's a great way to involve your team or community in the decision-making process and ensure that the final design meets the needs and preferences of the users. Voting process
Voting Results
Option 2a was selected, based on voting results and feedbacks.
Mockups
Landing Page
-
User will see 4 sections (Entry, Inclusion, Exclusion, Exit)
-
The section has a dropdown for the user to change the logical operator :All, Any, None, At Least, At Most
-
criteria main bar starts as a search bar for the user to search desired item
-
Users can add additional single or group criteria as needed
-
Users want to combine criteria in order to narrow down the cohort for their research
-
When user types a string and List of auto-complete entries appears.
-
Select single concept/set from the autocomplete list.
-
Add additional criteria's , it will minimize the former one.
-
When user types a string and List of auto-complete entries appears.
-
Select single concept/set from the autocomplete list.
-
Selected concept/sets is entered as bubble in the search bar
-
Configure criteria attributes
-
Configure criteria attributes
Screenshot of full designs
.png)
Collaboration
Throughout my design process, I engaged in collaborative efforts with both product managers and stakeholders. By actively working together, we ensured that the design aligned seamlessly with the overall product vision and met the diverse requirements and perspectives of key stakeholders. This collaborative approach allowed for a holistic understanding of user needs, market dynamics, and business goals, fostering a shared commitment to creating a successful and user-centric design solution. Regular communication and input from both product managers and stakeholders played a crucial role in refining the design iteratively, resulting in a more robust and well-informed final product.
%201.png)
%201.png)
%201.png)
Testing
Testing and Iteration:
Conducted usability testing with a diverse group of Atlas users, using interactive prototypes. Gathered feedback on the redesigned cohort building interface, custom UI components, and iterated on the design based on user insights.
Impact
The redesigned cohort building functionality in Atlas not only achieved improved efficiency and user-friendliness but also introduced a visually refreshed interface with custom UI components. Positive feedback from users highlighted the success of the redesign in providing a modern, engaging, and cohesive experience for cohort definition within the Atlas platform. The case study underscores the importance of both functional and aesthetic considerations, including the creation of custom UI components, in UI/UX design.