Krishna Ram, we get many questions about where coagulation testing should live—in a core lab on an automated line or in workcells of automation. There’s a lot to think about here.
Krishna Ram (Sysmex): Yes, and the question is, what does automation mean to each laboratory? Because it can mean many things. Ultimately, it should mean minimizing the time needed for mundane, manual tasks.
As a former biomedical scientist, I was averse to total lab automation because for some institutions that have a total lab automation solution, it becomes, “I’m now spending more time doing things I wish I didn’t have to spend time on.” And that is, “Where is this sample? It’s somewhere on this track.” So it has to be the perfect balance and must enhance each laboratory’s workflow efficiencies. And that depends on what solution is brought forward. Is it TLA that you require? Or is it a workcell, i.e. an island track automation solution? Or will a standalone analyzer be more than sufficient for your needs? Then, what tests do you wish to encompass? With all this there has to be a balance of what is vendor-led training versus what is a clinical decision. It’s scary to think where this will go in terms of artificial intelligence and decision-making from a clinical aspect.
Kristi Smock, how have you dealt with these questions? Give us a brief history of your approaches to organizing coagulation at ARUP.
Dr. Smock (University of Utah): We have both worlds at ARUP because we have our home institution, the University of Utah, where dedicated hematologists are doing hemostasis, thrombosis, and other benign hematology. From our reference clients, we talk to people across the range of practice from primary care to oncologists and so forth. I get the struggle people have with being up to date on everything.
In our hospital sites throughout the University of Utah system, we have automation in coagulation and instrumentation that looks at preanalytical variables. It’s relatively new. In the hemostasis/thrombosis lab, where more specialized testing is done, we’re not utilizing much automation. It’s very high volume and it’s been the most efficient for us to do batch testing and, say, do factor assays now or antithrombin testing now with all the samples we have.
We continue to think about whether adding automation into the lab and running samples as received would make sense. We’re still doing manual decision-making to ensure sample quality and make sure, for instance, what we’ve received is citrated plasma and not an EDTA sample. We still do a lot for sample quality, but it’s not very automated.
Ashley Sanchez, can you comment on this? You see a lot of this sort of testing in places large and small.
Ashley Sanchez (Siemens Healthineers): With the staff shortage, many technologists are shying away from what they call specialty testing, when it really isn’t. For example, anti-Xa was for a long time categorized as specialty testing. Getting technologists today to be comfortable with that kind of testing and be able to utilize automation or their standalone system is something Siemens Healthineers tries to support—getting technologists to perform what would be called specialty testing, moving on to factors, for example, on our systems, but simplifying it without taking away the sophistication so they are reporting the right results.
Is the availability of middleware an important part of this equation?
Ashley Sanchez (Siemens Healthineers): Absolutely. Our middleware offers different algorithms to guide technologists, such as generalists, to take the next step for reflex or in a panel. For example, lupus anticoagulants—what is the next step? The middleware offers guidance and tailors it to each laboratory’s SOP, because each laboratory might have a different lupus anticoagulant algorithm. Our Atellica Data Manager has hemostasis-specific algorithms built in to help simplify that testing.