Duration: Apr - Sep 2022
Team: Mengzhe Ye, Nuoran Chen, Bing Liu, Jiayuan Wang, Anyi Li (PM),
My Roles & Responsibilities: User Research & Analysis, User Flow, Information Architecture, Wireframing, Prototyping, Coordinating with Developers
Service Ecology
- What is the Cancer Care?
When it comes to radiotherapy (RT) treatment, the patient's electronic health records play a crucial role. Approximately half of the cancer patients undergo RT treatment (a process that utilizes advanced algorithms and machinery to deliver precise radiation to the specific target area). During this process, doctors must carefully examine their patients' profiles to create personalized treatment plans.
After in-depth research literature review to have deeper understanding of the cancer care domain. It is important to note that the RT treatment involves three stakeholders: oncologists, physicists, and therapists, each with distinct requirements for the patient's records.
The client, Memorial Sloan Kettering Cancer Center(MSK) is currently developing the new patient information system - WRAP, some portion of the backend and data have been integrated into for the purpose of automating the RT workflow. Our design scope is to revamp the current dashboard to support the needs of a growing number of patients.
Problem Breakdown
Together with PM, we identified these problems
Excessive number of internal systems which creates an unfriendly and complex environment for doctors when reviewing patient information.
Lack of efficient communication always caused some information delays about patients' treatment progress.
Solution
Working closely with UX researcher, project managers and data science team, l identified three design improvements that could potentially help us improve the doctors' work efficiency.
New dashboard navigation for better discoverability and fewer clicks to search for patients' records.
.svg){kind=small}
Improve readability of medical reports to support diverse lab environments
Responsive UI and dark mode to ease cognitive load
Impact
5min
AVERAGE VIEW TIME
The time spent on reading the patient records reduced from 45mins to 3 mins.
5+
AVERAGE DIAGNOSIS RATE
The radiation specialists can diagnose and identify high-risk patients much faster.
Research
Competitor
Analysis and structured interviews
I collected previous research in collaboration with the key researcher, I went through a research archives to find qualitative insights on what are common product feature offerings. I analyzed four main competitors: Phlips, Elekta, RO Dynamics and Varian to gain insight on how they handle the complex workflow and document exchange.
Synthesis
To better understand the process, we utilized the experience map to identify key stages and users' actions throughout the process of RT treatment and what types of the patients' health record they need.
After gathering all the information, we synthesized user data to create three personas. This helped us understand the expectations and motivations of each role during the RT treatment process.
Insights
After conducting user interviews with four pathologists and reviewing research provided by the research team on cancer care team workflows, we synthesized the findings by analyzing interviewees’ quotes and drawing the following insights:
Multiple switching between apps causes context fatigue
Many internal systems create an unfriendly and complex environment for document reviewing
.svg){kind=small}
Hard to search for patient info for urgent diagnoses
A patient's medical progress cannot be tracked efficiently, and frequent stage switching often leads to information delays.
.svg){kind=small}
Low legibility and scannability of Patients' medical reports
Valuable data points are often buried in unstructured text within doctors’ notes and the documents are usually formatted in PDF
Hard to exchange data between departments
Poor system integration makes it difficult to share patient info, leading to delays, repeated work, and potential errors.
How might we empower the doctors to efficiently navigate, access and evaluate different patients medical records?
Brainstorm with the team
I facilitated collaborative sessions with stakeholders to gather insights on improving the dashboard experience and also aligning with the business goals: Phase out outdated systems to streamline operations and enhance cross-departmental collaboration to eeduce redundancy. Using findings from user interviews and market research to support the workshops, we defined the core problems and voted on the most impactful features. Based on these priorities, I worked with the design director to outline design requirements for the development phase.
Introduce intelligent search and filters to surface urgent case details quickly
Transition away from unstructured document by introducing structured data entry
Build patient profiles with clear schedule, documentation and notes across departments
Ensure interfaces are responsive and optimized for difffernt lab environment
Information
Architecture
Based on the patient's raw data provided, I developed the information architecture (IA) that could efficiently organize and analyze the significance of patient data that is provided. After having the conversations with development team, we collected some insightful feedbacks and finalized the IA structure.
Layout Exploration
After built the information architecture for the patient insight dashboard, we then moved on to work iteratively to develop different versions of the dashboard design based on the information structure.
The reasons why we chose this direction because:
✔
A smooth navigation to switch between diverse patient profiles.
✔
Tab system allows switch between historical documents and current records.
✔
A tool bar helps to efficiently organize and navigate through a large number of files.
User Testings
We then conducted the 10 structured interviews to test our our lo-fi wireframes.
I summarized the interviews results to the below user feedbacks.
The current default view only show minimal report details. Practitioners want to read more treatment related information in the dashboard.
Practitioners suggest that the patient list panel takes too much space in the dashboard, but they usually don't switch patient profiles back and forth.
Practitioners suggest searching, filtering and prioritizing patients' data is difficult due to information overload on the dashboard.
Iteration
Then we moved to the ideation phase, I created multiple wireframes and tested them with practitioners in MSK to ensure the features we identified will deliver value. As we developed our final design, we conducted 10 user testings and consolidated the users' feedback. Then we update our design versions of key features.
Design Challenge 01 - Redesign of RT Report Screen Reader
Some users have reported that they would like to see more detailed information about the RT report in the default view, as they feel the current information is too minimal. They have also expressed concern that the toolbar and progress bars are taking up too much space.
🤯
Chance of flips and errors during work since it is against Fitts’ law.
🤯
Split screen causes the information overload.
😊
Similar to users’ habits of other reading application.
Design Challenge 02 - A drop-down patient list to switch multiple patient
Practitioners suggested patient list occupy a significant amount of space on the dashboard seems unnecessary and impractical. Therefore, it might be worthwhile to consider optimizing the dashboard layout by minimizing the space allocated to the patient list.
🤯
Difficult to search with profile images.
🤯
Hard to navigate through the banner.
😊
Concise and clear.
Design Challenge 03 - Key Indicator and Notification Center
Practitioners mentioned that the information in the dashboard is overwhelming, making it challenging for them to access valuable patient data like risk factors.
🤯
Floating Notification bar interfere with the usability of the user interface.
🤯
The presence of a floating toolbar reduces the available screen space.
😊
Highlight critical patients’ medical record and deliver timely updates.
Design System
Based on the patient's raw data provided, I developed the information architecture (IA) that could efficiently organize and analyze the significance of patient data that is provided. After having the conversations with development team, we collected some insightful feedbacks and finalized the IA structure.
Take Aways
Through my experience collaborating with a cancer care team, I gained a deeper appreciation for the significance of understanding the mental models of users. In our discussions, we uncovered that practitioners possess diverse responsibilities and place a high value on working together. The goal of the patient insight dashboard is to simplify the work flow and provide a clearer patient data structure. While I created the information architecture, I realized the structured patient data system guided our future design and became an efficient communication tool when discussing with various stakeholders.
