Bobbi Pritt, M.D. (00:04):

Hello, I'm Dr. Bobbi Print, a clinical microbiologist and laboratory leader at Mayo Clinic and your host for today's episode. Dr. Morice is away this week, so we're going to look forward to his news updates next time. For now, we're going to go straight into the deep dive. We're going to go beyond the headlines with Dr. Brad Karon today to discuss laboratory stewardship. Dr. Karon is a pathologist, division chair for Mayo Clinic's Clinical Core Laboratory Services, and a member of our laboratory and pathologist stewardship team. He's going to share with us insights into how to ensure clinicians use the right test at the right time. Thank you for joining me, Brad.

Brad Karon, M.D., Ph.D. (00:42):

My pleasure, Bobbi.

Bobbi Pritt, M.D. (00:44):

So let's start with a basic question. Can you share why lab stewardship is so important?

Brad Karon, M.D., Ph.D. (00:51):

Yeah, sure. Happy to. I think certainly if you look at laboratories and healthcare systems, controlling costs has been essential to survive. So, some labs do generate revenue, some have become cost centers. Either way, the need to control cost is essential and that's really, probably a primary driver. Beyond that, staffing is just hard to find. So, if you're struggling to staff your lab and everybody is working so hard just to keep the enterprise afloat, you don't want to be performing tests that are simply not needed. And then third, I would say, is that downstream effect. The needs of the healthcare system, the need for efficiency, the need to improve, the ability to care for more patients, is everywhere. And it becomes very easy for a provider to order a test. But then there is time downstream to look at that result, to get that critical call or alert for an abnormal, to act on it. So, the downstream effects of over-testing on utilization of provider time downstream, I think those are the major motivations.

Bobbi Pritt, M.D. (02:05):

It also impacts our patients, right? It could have a negative impact if you do unnecessary tests.

Brad Karon, M.D., Ph.D. (02:11):

Yeah, absolutely. There is a patient effect and that can come play out in kind of two ways. We've heard some time, we've been in the internet era for a while now, that patients are coming to their appointments with the Google searches and the AI and asking about these often quite esoteric tests. And sometimes, the easiest thing for a provider to do is simply to order them. And laboratory can provide that support downstream about saying, "Hey, this is an expensive, esoteric test. There's not a solid interpretation of this." So, just a little support to maybe go back and say, you know, this test may not be needed, and that provider would go back to that patient with that information. But we can't forget that that is one segment of patients. For many patients, an abnormal result produces tremendous anxiety and fear and concern. And so even if it's not a patient generating an interest in a test that maybe is a little esoteric or not needed, a provider who simply orders many things that aren't indicated at a given time will cause great concern because it is essentially universal that patients are seeing their results almost always before the provider has had a chance to review them through a portal or a system. And so the patient anxiety is the primary concern as well.

Bobbi Pritt, M.D. (03:32):

Yeah, absolutely. And I suppose it could even prompt additional unnecessary testing. So, you've covered a lot of important reasons. So, let's talk about strategies. What are some strategies that you've worked on and that have worked well for promoting lab stewardship at Mayo Clinic?

Brad Karon, M.D., Ph.D. (03:49):

Well, I think when we talk about utilization strategies, of course one that we've done and published on and many others is the decision support, the dreaded popup order, for a given test or combination of tests in a given scenario. Our system and many are moving away from that. They can be very effective; they can be intrusive and unpopular, and there's that alert fatigue that's going on that hampers their use. But it can be successful in the right situation. So, we've got some quite general ones that have been successful with sed rate, CRP, magnesium, phosphorus, in an ICU environment. But I think here and in other places, they're becoming more tailored. So a very successful recent one was ammonia testing based on diagnosis or ICD-10 codes. So yes, if you have liver disease or liver transplant patient, you're gonna need that frequent testing, but many other diagnosis not indicated.

So refining it to the right situation to have that pop-up saying the test may be unnecessary and provide some direction. One that's often overlooked and is the most definitive and often easiest is to simply obsolete tests that don't work so well anymore, are just dated. So past red cell folate, CKMB have been obsoleted. Or changing methodology. Rheumatologists will tell you, "If I can get the immunofluorescence ANA test, I'd have more information." Reality is they would, but it just took too much time and labor and money, it couldn't be supported. So for utilization, we had to come up with, "Well, most people will get in alignment on here's when we'll do an IFA and for some places here's where they might send out an IFA." So, obsoleting or replacing methods. And the last one I'll mention is, rules looking at the standing orders or future order frequency. And that's one that we've done and other places have done that's probably been harder to do because a lot of places use Epic or EMRs, a lot of preference lists, and providers are busy. But there's been some really successful work done in limiting the duration of future orders. A recent nice publication on an order set that alternated CBC with and without diff. just made it easier not to order large panels indefinitely. So that would be a third one that's worked here and in other places.

Bobbi Pritt, M.D. (06:16):

Those are great examples, Brad, and I'll mention one from my lab, too, is just renaming a test so it makes more sense. Sometimes we have a long list of tests that are available, and the provider doesn't know which one to order because they all sound the same. So sometimes just changing the name of the test to be more intuitive is helpful. These are all great practical things that people could be doing. So, what are some things if we expand out beyond Mayo Clinic that every laboratory leader should be doing to improve test utilization?

Brad Karon, M.D., Ph.D. (06:46):

I think there's some things that probably many or most places have done. A lot of places do start with their send-out tests and you know, there's reasons for that. Send-outs are often expensive and there's a big budget line item, but there's reason beyond cost to perhaps start there or to do that in every lab. So you know, your referral tests, they are a source of nonstandardization. There's extra effort. I'm sure when that provider comes back from the conference or a new provider comes to a practice, and they say, "Hey, I heard about this great test," they're not thinking about, "Wow, somebody in the lab is going to have to maybe even call this place, understand what do I draw, how do I transport it?” And then those results will come back and somebody might be typing them in or scanning them in.

And that report format, maybe that provider's the only person who's seen that report format. So beyond cost, there's this practice standardization and quality aspect to controlling send-outs and identifying preferred referral labs. Because again, that provider may know how to interpret that test, but the next provider may totally misunderstand, not understand the report format, never seen it before. So, I think that send-out process is a big one to look at. I think reviewing the standing order set in  the hospital practice. There's admission orders and very specialty orders. So again, that's a common one you want to get involved with, who's making these order sets and how many lab tests are going in? Is there a system where somebody from the laboratory, in laboratory medicine and pathology, can get involved in at least reviewing them as they get renewed or made? Every place has a process to do that.

And then, I would say look at panels or care pathways instead of individual tests. So this has worked well here, again, for some of the Lyme disease and microbiology testing, where the individual tests are confusing, but you put them in a panel, and there's directions about, you know, start here, this is screening, this is confirmation, what do you do. Oftentimes, processing instructions will go along with that. And then, probably the most effective is the sort of care pathways or problem rates. So, and again, we have order sets for a lot of individual coagulation. We do highly esoteric coagulation here. But you know, you can, if you go search long enough, find all of these tests. But, it's much easier to just order a prolonged PTT work-up or a thrombophilia work-up, and t lab decides what testing is needed. So these problem-based or care pathway tests where the providers don't have to go through and pick, they just tell us, "Hey, what's the problem?" And the laboratory decides what testing to do. So those are all good ways that places can start.

Bobbi Pritt, M.D. (09:33):

Yeah, I really like those, Brad. And that last one you mentioned really allows our laboratory leaders to use their expertise. Of course, you know, it has to be done with certain approvals from your organization, but it's a great way for the laboratory leaders, pathologists, and other laboratory scientists to be able to be involved in the testing process. And we're the test experts with what to order when. So a lot of times those nuances don't appear to be clear to our ordering providers. So, looking ahead, how do you expect stewardship efforts to evolve, especially with emerging AI solutions and, of course, the increased pressure for cost-effective care?

Brad Karon, M.D., Ph.D. (10:14):

Yeah, that's a great question, Bobbi. If I start first with where I would see two amazing opportunities, I would say, looking at high-volume testing and the patterns, it's something that probably everywhere it’s said we need to address utilization. You often start with inpatient, because it's most often not reimbursed, so you start there. But these are often very high-volume tests, and you start by saying, "OK, I want to dig down and get a snapshot of how frequent certain ordering patterns are, what groups: inpatient, outpatient, environment service." But that takes time and effort, and you've got a snapshot for one time. And so we have tools, right, to just be continually monitoring that and to find patterns that we might not even go look for. Just like, "Hey, something's changing. We're seeing these test ordering patterns from this area of the hospital or that."

So, I see that as a great opportunity just because many places do like to start with inpatient testing. That is, you know, not reimbursed generally. And we should have the ability to be constantly surveying and identifying what are potential intervention strategies. The second one, again with potential, I would say, is intervention in really complex patients. So, there are groups, like Lapasada and others, with diagnostic review teams, and they go out to the bedside and do a consult and here's what test to order. The reimbursement is tricky around that. It'd be great to cover more of it, but I think with AI and tools such as that, we should be able to be constantly monitoring it and have an AI program looking at, "Oh gee, there's been somebody getting a lot of blood products and a whole bunch of tests that we don't normally see from that particular area of the hospital.

Something's going on.” Someone from the lab or transfusions can reach out there and say, "Hey, how can we help?" And if we can make that a consult to get reimbursement, that's great. But even if not, the cost savings for the hospital would be enormous. And I think there'll be great motivation to do that. You know, the barriers to both are the limits of the privacy and operating in a lab information system, or EHR. Some of the things happening now, we're seeing certainly the use of AI for predicting disease with common tests. So there's many publications of common panels of tests that predict outcomes and acute coronary syndrome, liver disease, renal disease, others, as well as using information from the lab that we don't normally report or use. So, most common examples of that are all these parameters and cell counters that are kind of research use only or hiding in the background.

And, you're seeing all these publications with mean platelet volume and inventory reticulocytes and combinations, these parameters that aren't reported on a routine CBC that may predict diseases. So, I see that that is emerging, it's happening. It's an interesting utilization opportunity and potentially dilemma. It may, on the one hand, it may say to providers, "Oh, let's order these large panels because it's going to tell me something." So, it could be bad for utilization. If we can use it right though, it could help us prevent the use of esoteric tests to say, "Oh, based on the patterns of tests, even with these parameters in the lab, we're not actually releasing to the EMR under normal circumstances that there's a low risk for something.

So, let's not do the flow cytometry or molecular testing for hematologic malignancy. Because all of this data, even the cell morphology in a slide or a digital slide, is telling us that's a low risk.” So, we in lab medicine may not be developing these, but I think we'll be responsible for understanding how they can be used and making sure that it isn't just a reason, "Oh, let's order everything because maybe we will get some information at the other end." But understanding well, what in practice can actually be done, what's actionable, and how can this be used? I think it's an exciting opportunity and challenge for us in the laboratory.

Bobbi Pritt, M.D. (14:23):

Yeah, it is exciting and hopefully we can be at the table when some of these decisions are being made about bringing in some of these AI tools because it does use the data that are generated by our laboratories. It's not always the case, but we always advocate on this podcast for lab involvement. Well, this has been a great discussion, Dr. Karon. Thank you so much for joining us today.

Brad Karon, M.D., Ph.D. (14:45):

Yes, my pleasure. Happy to, Dr. Pritt, and thank you for having me. And I hope those who listen will get something out of this and start working, or continue working, in their own institutions.

Bobbi Pritt, M.D. (14:58):

I'm sure they will. Let's wrap up with the key takeaways and how to learn even more on the topics we discussed. In our deep dive summary, Dr. Karon discussed the importance of laboratory stewardship, sharing some practical advice and his thoughts on how stewardship will change as lab medicine evolves. If you'd like to explore this topic further, we've included links to a case study, an article from our outreach team, and a white paper on maximizing the value of the hospital laboratory. Thank you for joining us today. If you haven't already, make sure to subscribe so you never miss an episode. I'm especially excited for our next deep dive. Dr. Ellie Theel is joining me to talk about one of my favorite topics, vector-borne diseases. I hope you can join us.