The visually guided patient sample transfer
A tool that improves patient sample transfer accuracy and efficiency by guiding users and eliminating manual tracking errors.
Project overview
Challenge
During the end-to-end sample workflow, a batch of samples must be transferred manually from one container to another, and the lab operator manually tracks the source and target container positions with the help of spreadsheets and physical notes. This may result in human errors like pipetting patient samples in wrong target wells (positions).
My role
Senior UX Designer
Time
3 months
Solution
A feature within the application that guides the user visually to transfer samples from one or more source containers to target containers was developed. This completely replaced spreadsheet tracking, and supported users to choose the pipetting method and recommended target positions.
Tasks
User research, Wireframes, prototyping, usability testing
Tools
Figma & Axure RP
Pictures of the simulated lab environment that includes cardboard, containers, and consumable canisters.
Project Goal
At a certain point in the sample sequencing workflow, patients' samples need to be manually pipetted from one output container to another input container to proceed further. For this, the users organize the source containers in a rack and write down the sample ID/ source position in the target container position to ensure that the intended samples from the source reach the designated position in the target container. This manual method creates a high risk of human error and a big cognitive load on the user. The goal was to provide a solution for sample transfer through the application that should guide the user and handhold them throughout the process which could replace the planning and execution of the manual pipetting process.
Persona
In crafting the persona of Alex Reynolds, a Senior Lab Researcher, the goal was to define a role that reflects the challenges and responsibilities of a typical sample transfer process in the NGS labs
User Interviews
We identified a set of users from the lab to understand their end-to-end workflow during a sample transfer process. This helps us visualize their lab environment, the setup they use for sample transfer, the tools, common practices, and the challenges they face during this process.
​
Decoding user interviews based on the user interviews, two things were very clear.
​​​
The user’s mental model
The interview helped ensure that our planned designs aligned with the user's mental model.
​​
User needs and challenges
The interview also helped us understand the critical activities the user needs to perform for a successful sample transfer. We also could understand the user's challenges during manual sample transfer and the support they need during this process.
Concept explorations
Before getting into design concepts, we wanted to explore different user flows that align with the real-time sample transfer process in the lab. We had multiple internal stakeholders comment on the flows and narrowed them down to a flow.
Different options explored that could be possible user flows
Different attributes being central
Design explorations
The first step was to get the base layout right. Based on the user research, we understand that the users' set up manual pipetting by placing the source containers on the left and target containers on the right. The transfer happens from left to right. We wanted to replicate the same experience to the user
The layout is planned in such a way to reflect the user's lab setup which is source containers laid on the left side and the target containers on right.
Designing target containers
Designing target containers
Detailed design of the 8 tube strip and 96 well plate as target containers with different position statuses.
Pipetting orientation
Based on the user interviews, we understand that the users use different pipetting methods using different pipettes. There are multi-channel dispensing methods where the users will be able to pipette 8 or 12 samples at the same time. The multi-dispense 8-channel pipettes place samples vertically, while the 12-channel pipettes place the samples horizontally. The single-channel pipette allows the user to choose and place samples individually.
​
The design allows users to choose the pipetting methods they want to use, and based on the selection, the system decides the pipetting orientation in the target 96 well plates.
A drop down was added in the target container section with different pipetting options. Each option changes the way the samples could be pipetted
The end to end workflow is demonstrated through a workflow diagram which involves users and the devices that they interact with in a lab
The critical part of sample transfer was to get the container visuals right. We were dealing with two types of containers, a 96-well plate and an 8-tube strip. For each container type, we want to indicate the status of the position. The possible statuses were an empty position, a suggested position, a filled position, and positions that are not available for pipetting.
Formative usability testing
A formative usability test was conducted on the design with 3 users to gauge their mental model against the design. This also allowed us to validate how we translated the user interview learnings to design. We got additional inputs like adding sample level comments for each sample to take notes on the sample input volumes, concentration, and buffer details.
Users being tested with the clickable prototype to capture their mental model.
Results
Following the formative usability evaluation, we conducted a comparative analysis of the time required to perform sample transfers manually versus using the visually guided digital tool. Results indicated a notable decrease in the time taken, with an average reduction of 28.20% in completing the sample transfer process. Additionally, the accuracy of the transfers showed improvement, with an average increase of 13.45% across three users.
28.20% Faster & 13.45% More Accurate
Learnings and next steps
The user interviews gave us a lot of valuable information that could solve users' day-to-day problems. This exercise also gave us an understanding of the terminologies used by users in their environment so that we could reflect the same in our designs. The follow-up sessions also confirmed that the designs reflected the user's mental model.