Managing patient sample workflow
Patient sample entry to result generation - A complete end-to-end solution for patient sample tracking in labs
Challenge
Lab users manually track a batch of patient samples when processing them through an end-to-end NextGen Sequencing workflow. This results in human errors in sample tracking which in turn reduces the usage efficiency of the consumables and reagents.
Solution
AVENIO Connect, a patient sample workflow management tool, directly addresses users' pain points by aiding users from registering a batch of samples till generating patient sample results. The application orchestrates the whole sample management process and assist lab technicians by providing sample statuses during each workflow step.
The problem to address
Lab users in a next-generation sequencing (NGS) wet lab face difficulties in efficiently tracking and sharing critical information about patient samples, such as sample container IDs, Run IDs, and sample IDs, during the various workflow steps involving different robotic instruments. The current method of using spreadsheets and physical notes is time-consuming and prone to errors, leading to potential miscommunication and delays in the lab workflow.
The spreadsheets and physical notes used for tracking batches of patient samples
Our users - Personas
The primary persona is the lab technician who works in NGS (next-generation sequencing) labs and is responsible for preparing and analyzing DNA samples using advanced sequencing technologies. This can include extracting DNA from biological samples, running sequencing reactions, and analyzing data using bioinformatics software. They may also be involved in the maintenance and troubleshooting of NGS equipment. The technician likely has a background in molecular biology, biochemistry, or a related field, and has experience with laboratory techniques, and computer skills to work with bioinformatics software. They may work in a research laboratory or a diagnostic laboratory setting.
Design process
Understanding the problem
A contextual inquiry in the user's environment was planned to understand the user's needs. Considering the sanity and security restrictions, getting access to the wet labs involved lots of formalities, which delayed our progress in getting to the users.
Due to this, as a workaround, we created a simulated lab environment with cardboard devices replicating the ones used in the labs. We used actual containers and pool tubes wherever possible to give the users a natural feel of the lab. We added labels to make the props clear.
Key observations from user research that translated into application features
Application overview
With the help of user research, we were able to understand the users' day-to-day activities performing different tests. With the help of an experience map, we were able to visualize an entire end-to-end experience of the primary persona to accomplish a goal. This experience is agnostic of a specific test type or a device used in the lab.
We summarized users' interactions with different devices, LIS systems, and reports, and created this chart.
Frequency criticality chart
Based on the user research activities, we conducted a session with stakeholders to plot the activities that are more critical and frequently used. We were able to identify the critical tasks that deliver the most value to our users and the use cases that are executed frequently.
This exercise helped us to understand the foundational user journeys that make the application valuable and typically capture 90% or more of our users' actions. This facilitated prioritizing user needs and helped to avoid scope creep or the introduction of any extraneous/ancillary features.
Most importantly, it helped our team to build and optimize applications capabilities and features that deliver the most value to our users and drive the key metrics.
Whiteboard sessions
During our in-person office days (and even now) we made sure to involve our core team in whiteboarding exercises before committing to building anything. This is because we knew that it is pertinent to involve the team from the inception and have them aligned with what we were pursuing, as this was the key to scaling our application in the right direction.
These sessions not only act as a way to discuss design challenges but also as a tool that allows other team members — such as developers and product owners — to raise their own ideas and give them the confidence to participate in the design process, so the solutions the team agree on are grounded based on different perspectives.
Feature set discussion
Whiteboard brainstorming session
Concept explorations
Even before we gave a visual form for any solution, we made sure that all possible flows that were in scope were addressed and that there were no ambiguities or loose ends. This requires inputs from various teams so that we can clarify any open queries. The MIRO board helped us to visualize the concepts and task flows through boxes and arrows for an effective demonstration of our thought processes. This also became a medium for us to exchange ideas and clarify queries with different teams.
Concept creation - Batching samples
As creating a batch of samples is the very first step at the end-to-end workflow and after exploring different concepts, we narrowed our focus to two potential user flows for defining the batch.
The first solution allows the technician to first select the samples and then assign them to batches. The second solution allows the technician to define batches first and then assign samples to them. Both options aim to make it easier for the technician to define and manage batches, reducing the chance of errors and improving the traceability of samples.
However, to ensure that this leads to more efficient and accurate analysis, we must evaluate these options with actual users.
Evaluating concepts with users
We interviewed four users and walked them through both solutions, asked them to weigh the pros and cons of each, and explain their decision-making process. The majority of users preferred the option where they would select the samples and define a batch out of it - this option was much closer to their existing mental model.
Wireframing
Registering patient samples
Once the lab receives patient samples from the collection center, those samples need to be registered in the application. The users preferred manual sample entry for one or two samples and preferred a bulk order registration when there are more samples. The application supported two ways of order registration based on user needs.
Manual entry
A provision to manually enter sample details as a form. This is helpful for users to register samples in an impromptu way.
Bulk upload
A way to register a large number of samples in one go. The application supports a CSV file upload where the users can download a template, add data, and upload to register multiple samples at once.
Samples scanned with barcodes
Most of the wireframes and visuals were based on findings from the user research findings. During the user research, the lab users preferred to scan the samples to include them in their systems. This helped us to take a design decision on having a provision to scan samples.
The test quality is evaluated by a control sample that confirms the accuracy of other patient sample results. Control samples are introduced during the workflow and the application allows the user to scan the control sample barcode.
Sample history and downstream steps
There are three critical statuses of a sample the user is interested in. The past workflow steps that were processed, the current status of the sample, and the downstream workflow steps that are pending to be performed.
Design documentation
The design, created in Figma, includes annotations to clearly communicate the flow between elements and make it easy for anyone who interacts with the design to understand. This documentation is especially helpful for ensuring that all team members are on the same page when it comes to the design flow, such as the example shown of a user selecting a primer plate for assigning Sample IDs to specific positions. Each selected primer plate has a unique plate configuration and a number of sets.
Formative usability
A formative usability evaluation session was conducted with internal lab users. The target lab users were Lab directors, Lab operators, and a Lab manager. The test script was prepared with 8 tasks that need to be performed on a clickable prototype that I created. The users were given the context, and each task was explained.
Lab testing
User feedback
Prototype walkthrough
Results & Impact
A comprehensive formative usability evaluation was conducted with 8 lab professionals (average 10 years of laboratory experience) across Roche's IVD labs. Participants performed end-to-end Capstone IVD workflows using the application over a 4-week evaluation period.
Industry benchmark: 68
Key findings
- Above-average usability score — The System Usability Scale (SUS) score of 75.9 exceeded the industry benchmark of 68, indicating a good overall user experience
- Easy learning curve — Users expressed they understood how to perform tasks after just a few uses, even those new to the application
- Core workflows rated straightforward — Library preparation, pooling, and sequencing preparation were consistently rated as smooth and intuitive by participants
- Pooling rated highest in ease of use — Participants felt comfortable after their first attempt, with the learning curve described as easy