ANTONIO BONILLA: An astronomical amount of money is spent not only on recovering for medical errors, but their ripple effect – treating the initial ailment as well as whatever complications may have arisen due to the error itself. Our mission is to improve a product’s design to reduce errors in the field. By doing so, human factor design also can improve workflow efficiency.

JP STEPHENS: When you have engineers designing a piece of instrumentation, they have an idea of how that hardware should work, but the customer also has a mental model of how a piece of hardware should work in their laboratory environment. Those two mental models are not always the same. Our job is to help product developers, often engineers, bridge the gap between their vision of a new product and how the customer will eventually use it.

KIMBERLY DEBAISE: In ideal situations, we're actually testing at the laboratory, bringing the software to them, talking to them, finding out what they want, helping them identify their needs. We observe them as they go through their laboratories’ workflow, rather than taking them out of their own setting in order to understand their needs.

JP STEPHENS: Exactly. Throughout the iterative design cycle, we run our evaluations, collect feedback and comments and then iterate what we’ve learned and observed back into in the design, whether it's safety- or design-related. Then we move into concept, where we take those requirements and ideas and start building hardware. Human factor design is integrated from that very first initial concept all the way to on-market and even end-of-life.

ANTONIO BONILLA: Yes. We try to have an initial understanding of a laboratory’s needs and its workflows. First, it’s about listening and observing. Then assessing where there's risk or opportunities for operational efficiency. Next, validating that the system is working as intended in achieving the design requirements. Finally, once we've released it, we invite customers back to show them new iterations of software and get their feedback. It’s like a feedback loop for future iterations of software and advanced features – listening, risk assessment, validation and continuous improvement.

JP STEPHENS: We do an analysis to understand a laboratory’s orkflow and where we can reduce unnecessary tasks to improve it. We look for pain points: where they have to perform workarounds, where they have handwritten notes on things and where they have makeshift setups, among others, so we can incorporate mitigations for them in the design. We look for areas we can we improve to make their life better.

ANTONIO BONILLA: I visited a laboratory in Texas. They had sticky notes on their analyzer filled with lot numbers of commodities used because they needed to keep track of them. They commented to us that the software didn’t allow them to record those numbers, so they did what they had to do. We took that observation back to our engineers and they designed a way to incorporate a user interface that would track those numbers through the software, saving time and reducing the chance for error.

Another example is that a customer commented to us that they did not like having to stop the analyzer, putting the system in idle and lift the lid to replenish reagents, and then restart the system back up to start testing again. It was inefficient. They were losing time, stopping testing because of one issue, starting again and then having to stop again for another issue. We were able to address this issue on the new analyzer, so when these events occurred they didn't have to stop the system anymore. This design approach enabled laboratorians to replenish reagents without disrupting sample processing.

KIMBERLY DEBAISE: Another example is from a large validation we did where we saw that one of the customer pain points was identifying out-of-range QC numbers. We implemented a software change to supply a red highlight to them. Just that small change, from a black font to red, served as a starker visual cue to the laboratorian to identify QC out of range. It made things more efficient and helped us, and the customer, feel more confident from the safety perspective.

JP STEPHENS: One of the first discussions we had was about how to navigate these new waters and interact with customers when we're supposed to stay away from people. After all, our No. 1 data collection tool is to be next to somebody and observe what they're doing. We continued to perform usability testing, but incorporated new measures to ensure safe interactions, like setting up tents in the parking lot.

KIMBERLY DEBAISE: Our customers conduct essential work so we had to continue to develop and refine our products for them despite the pandemic. Plus, they wanted to continue participating in these evaluations and provide their input. Because we were involved in the testing for new COVID-19 assays, customers immediately saw the urgency behind getting those to market quickly and the need to work together to overcome the challenges we faced.

ANTONIO BONILLA: It does point to the fact that the customers understand the symbiotic relationship we have with them and how their feedback is important in getting products to market that they can then use in their labs to better improve people's lives. I thought there would be a great decline in the number of customers that would be interested in participating with us, but we haven't seen that at all.

JP STEPHENS: We're always looking at ways to perform the same test using less reagent, less sample, less commodities, and then make the overall unit test less expensive. One of the directions that medical devices are heading is that everything is getting smaller, more electronic or app-based, so you don't necessarily need specialized hardware. You can do more from your phone, more from a tablet. That should lower the cost of healthcare because you're not investing as much into hardware, and everybody's got that technology readily available to them. It should also help bring down the cost of running those tests and provide value across the system.