Transcript: Point of Care Testing

Transcript

TRIONA HENDERSON: Introductions out of the way. So good afternoon. My name is Triona Henderson. I am a clinical pathologist and the facilitator of this ECHO Model™ project. I extend a warm but rainy welcome from the Division of Laboratory Systems at the Centers for Disease Control and Prevention in Atlanta, Georgia. Thank you for joining us today.

The topic for this interactive web discussion is point of care testing. And I know for me, there are no coincidences. And when we were planning this almost a year ago, Dr. Yu had a choice of many months to choose from, and she chose this month. And what an opportune time for us to be talking about point of care testing.

So Dr. Elsie Yu will present today's topic. Dr. Yu is board certified by the American Board of Clinical Chemistry and currently serves as the medical laboratory director at Geisinger Health System with a system leadership role in clinical chemistry, immunology, toxicology, and point-of-care testing. She is also a clinical associate professor at Geisinger Commonwealth School of Medicine. Thank you for joining us today, Dr. Yu.

Here's some technical information. Before introductions and the presentation, here are some technical details. If possible, please use your video. This is really an interactive session, and it's really helpful to put a name to the face and will enhance interactivity. All your microphones are now muted. During the discussion section, please mute yourself.

If you are connecting by phone only and would like to take part in the discussion section, please announce yourself by name and organization before beginning to speak. If you experience technical difficulties during the session, please send a private chat message to Johanzynn Gatewood, labeled as DLS ECHO Tech. She will do her best to respond to your issue.

And finally, we design relevant sessions based on your evaluation responses. We do take the time to read them. We would like to encourage you to complete the post-session evaluation. For those requesting P.A.C.E. credits, they are being offered for this session and all ECHO sessions this year. Either a certificate of participation or a P.A.C.E. certificate will be issued upon completion, depending on your selection.

And if you have further comments or any suggestions, please send a private email message to Johanzynn, or you can email us directly at dlsecho– that's DLSEcho@cdc.gov.

So how are these ECHO sessions different? So they are different from teleconferences and webinars in that it's really based around the discussion of cases. Or in our specific case, clinical laboratory challenges being the main feature. Subject matter experts hope to share some of their work that can be translatable to all of you in your individual laboratories.

These ECHO sessions will focus exclusively on clinical laboratories in the United States and US territories, although we know we do have international participants with us on every session. Once again, we value the discussion amongst all of you that ensues, and we encourage you to share your own experiences and challenges on this topic.

So here's an overview of the process. We're going to have a case presentation by the subject matter expert. Then I will provide some clarifying questions to the SME. Then we open up the floor for ideas, shared experiences, and comments, from the SME, closing comments, reminders, and then we will adjourn this session.

Today's session is being recorded. If you do not wish to be recorded, please disconnect now. Closed captioning will also be provided, and you can find the link in the Zoom chat box, which is being placed right now. The audio and transcript will be shared on our ECHO website after this session, and we will also provide slides for the presentation via email to registered participants.

If the link was shared with you and you did not register through TRAIN, please send us an email at dlsecho@cdc.gov, and we will manually register you and get you those slides.

Here is the bios sketch of Dr. Yu. Dr. Yu received her PhD in Cellular and Molecular Biology at the University of Wisconsin Madison and completed her clinical fellowship at Boston Children's Hospital. In 2016, she was named 40 Under 40 Top Five by the American Society of Clinical Pathology.

Since she joined Geisinger in 2010, Geisinger has grown from a two to over 10-hospital system with over 80 clinics. During this rapid growth, she has integrated different hospitals and clinics with the Geisinger Health System through consolidation and standardization to enhance testing services. Her main interest is to improve laboratory test utilization and operational efficiency.

Much of her work requires extensive collaboration with clinical providers to achieve system standardization and clinical effectiveness, my personal passion. Two such examples include her efforts to, first, curb the opioid epidemic with enhanced urine drug testing program and decision support. And secondly, to improve ED throughput and transform cardiac care by implementing high-sensitive troponin.

Now we will invite Dr. Yu to begin her presentation. Please be thinking of similar situations that you've encountered or how you might apply Dr. Yu's experiences at your institution. We expect this to be a robust discussion. Please offer your questions and feedback in the discussion period. Dr. Yu, welcome.

ELSIE YU: Thanks, Triona. So as Triona was talking about, I'm going to talk about a point-of-care testing today. With the COVID pandemic, I think it's no surprise that a lot of us are talking about why point-of-care testing is so important.

So I'm actually not going to focus my talk on why we need point-of-care testing, because I think most of us know. And that was my only slide on COVID-19. We can talk more about COVID-19 or SARS-CoV-2 point-of-care testing at the end of the talk, but I wanted to focus on some other topics at this point.

What I want to do today is to talk about what our role from a lab perspective in point-of-care testing. So I know teleconference is hard. Webinar is hard these days. So I set this up so that we can do some activities together through Poll Everywhere. So right now, I invite you to join the activities that I have set up. There are two ways that you can do that.

One is in the web, you can go to pollev.com and then enter LCU674. And then it should give you a message that you have join my activities. And then when the activities show up, you can simply type in A, B, C, or D.

Alternatively, you can use your cell phone and text LCU672 to 22333. So again, 22333 would be the number that you would dial into or text to, and then this is the message that you put in. So I will give everyone a moment to do that while I do that myself as well.

Again, once you join, it should give you a message that says you have joined.

So let's do one simple question to make sure the poll works. If you have too many meetings today– it seems like people know what they're doing– then put A. If not, put B. So it looks like this is working. Now I'm going to ask you real questions. Oh, nope, hold on a second. People are still answering.

So a majority of people have too many meetings, which kind of seems to be the life of us. So I'm not surprised. But I block off my days, at least part of it, so that I can do this. So not too bad. 25% of us don't have too many meetings.

Now let's do a question set more relevant to my discussion today. So the question is, do you think point of care should be oversight by the lab? A is no; testing is so simple, the providers and nurses can just handle this themselves. B, eh, it depends. Yes and no, it depends on how complex the test that they are doing. C, absolutely, without a lab there would be many issues.

So I'll give people some time to do this. This is fun. When I was testing, I can't see all these changes going on. I'll give people another moment.

So I can see the majority of us think that the lab should absolutely oversee point-care-testing, which is nice. But I can tell you that historically I kind of swing between B and C. There are times that I do think that it's not as critical for a lot of lab oversight, maybe consultation. And we can certainly discuss more of that later.

So really, today what I wanted to focus on is to talk about point-of-care testing and to explain to you that it really is not all that simple. And then to talk about how we, from a lab perspective, can ensure properly utilization to improve patient care. Before I go into that, I thought I will tell you where I work. Some of you may not know where Geisinger Health System is.

It is in central Pennsylvania. And when I say central, I really mean central. If you put a dot in the middle of the map of Pennsylvania, you probably are close to one of the Geisinger hospital. So here is the map of part of PA, just to orient you. Philadelphia is right here if you can see my arrow on the screen. And that is in the bottom right.

And Pittsburgh would be off the screen to the left here. So Geisinger Health System is really in the middle of them. We currently have about 10 hospital, which in this map, you can see that is the H blue sign here. And we have over 80 clinics, which is notified here in the red triangle or the green triangle here.

We serve about 4 million people in over 45 Pennsylvania counties, even though we're in the rural area setting. In terms of point-of-care testing, we have about 9,000 point-of-care testing operator. A majority of them luckily only perform waived testing, make our life a little bit easier. But we do have about 1,000 operators who perform non-waived testing.

And because of recent scrutiny or regulatory scrutiny on provider-performed microscopy, we have really limited the microscopy to only dermatology and to only 45 provider at this point. Our volume is about 1.5 million each year. A majority of them are on glucose or i-STAT testing, which probably is the case for most of the people on the call today.

So let's look at one of the cases that involve i-STAT since it's one of the device that we use a lot. And I'll see what you think about this. So, in this case, the nurse was being asked to perform i-STAT chemistry test. She reached into her pocket and couldn't find a plain syringe. Luckily, she had another syringe, but this one has saline in it. To save time, she expelled the saline out from the syringe and then dropped the patient's specimen for i-STAT testing.

So the question is, what do you think is going to happen to the lab result? A, the lab results were unaffected. B, only the electrolyte results were affected because it had saline in it, so it would be low potassium, high sodium, and high chloride. And C, all results would be affected.

So I see the majority of you think that all results are affected. I can see that most of you are probably from the lab. That's why you answer the questions correctly. I actually put in A just to make sure the poll worked.

So yeah, you guys got it right. The results are affected. Again, obviously, this is one that is not difficult for lab people to know. But for nurses and providers, I'm pretty sure if they know it would have affected the results, they would have spent some time to get a plain syringe.

So what do you think happened next? Do you think the nurse and doctor, A, suspect device error, and then contact the lab when they see the results and send a sample to the lab for testing? Or B, they think the results make sense because the patient was nonresponsive, so they look at the results and treat the patients accordingly. Or do you think the nurse actually realized her mistake when she saw her results, and so she just repeated it without even telling the doctor?

This is interesting. So in the ideal world, I think both A and C would be a pretty good outcome. But unfortunately, in this case, that we have, what actually happened is B. They actually did treat the patient and transduce the patient. So it's not really a good situation.

So the point being that a lot of time, providers don't know what they don't know. They don't understand that a small deviation will have major negative impact to the results and patient care. They may have the best intention. The patient is nonresponsive. They wanted the results right away to find out what happened. But then they end up making a bigger mistake.

So it is our job, from a lab perspective, to make sure operator competency is done and train them on how to collect specimen correctly, how to run the samples properly. And typically, at least for us because we have so many operators, we do that through a partnership with nursing. And I'm not going to actually spend a lot of time talking about operator competency.

Heather Duncan, one of the presenters for ECHO, actually did a really good job in February. So if you're interested in this topic, you should go to the ECHO website and look at the transcript or listen to the transcript.

Now let's look at another simple case. This one has to do with glucose meter because that's another one that we do a lot of. So in this case, a sales rep dropped off a home glucose meter and test strips for a clinic to try. The question is, can it be used to perform patient testing at the clinic? A, I'm not really sure, but since it's free and we don't really have a lot of money, that's better than nothing. B would be, no, it should not be used because it violates Stark Law.

C, it should not be used because it's not FDA approved for hospital use. D, why not? I just saw the commercial yesterday on TV. It must be good.

So that's really good. Most of you actually picked the correct answer. No, you shouldn't. But I guess I kind of set it up to tell you that it's not a good idea. I do want to talk about, for a moment, some of these commercials on TV make me laugh in a bad way. They will tell you how great the glucose meter is just because they measure it twice.

From my standpoint as a labratorian, if it's a good device, you don't need to do it twice if it's accurate. If you have to do it twice. It's basically saying that it's not a precise device. But anyway, yeah, they're a different category of glucose meter. But fundamentally, the reason that FDA have different category is because not all glucose meters are the same.

In the hospital setting, we have more critically ill patients, so the device needs to be a little bit better in that they are not interfered by hematocrit or some of the common drugs that we use. But another thing that we need to think about is infection control, which, again, with the pandemic, I don't think that's a subject that you haven't heard of. But infection control is an important thing with glucose meters.

If you think about it, the nurse may be carrying the glucose meter from one bed to the other. You definitely don't want to transmit any disease because it's on the meter. So it has more to do with it than just the glucose meter, but providers don't know what they don't know.

And it is really not uncommon for me to see that clinic sites have a home use glucose meter when I work on acquisition projects. Or even people who ask me, hey, can I use this glucose meter instead of the one that we recommend because it's so much more expensive? So they really just don't understand that there's a difference.

So again, from a lab perspective, when we look at point-of-care testing, we focus a lot on test accuracy and whether it's clinically effective. So let me spend a moment on test accuracy. When we look at a device, first of all, we look at whether the test result is accurate. But that's not the only thing. We also look at to see if the device is durable or reliable.

I don't really know what happens on the floor, not being a nurse. But we have a device come back very broken. So it's important to not get any device that is flimsy. It has to be pretty sturdy, so to speak. And it's also important when we look at a device or test procedure; we also try to put it in the head of the user to make sure that it's not too many sad. It's not too error prone.

One of the device that we look at before, after you collect the swab, you have to mix it with a different buffer. That's not really easy for a nurse. It's not that they cannot do it, but it's not easy when they have so many other things to do. So to us, that's error prone, and we try to avoid those. So that's mostly on the device.

The other one is operators, which we kind of discussed already with the first case that we talked about. We really want to make sure that the operators are competent in knowing how to do patient testing. And a majority of the error, as most of the laboratory on the call would know, is in sample collection. So we do spend quite a lot of time focusing on that and also to make sure that when they are performing the test, they are doing it according to the procedures.

In terms of clinical effectiveness, that we mostly work with the providers. One of the very important things is to make sure they understand the testing limitation. I would say there's no tests that are perfect. Very few tests are perfect. I don't even know what I would say about being perfect. 100% perfect is very rare.

So understanding testing limitation is very important. And based on that, that would determine how the test should be used and what follow-up should be done. So to just give you an example of what I mean by that, let's talk about flu testing, especially since flu season is coming up.

There's often a lot of debate on whether we should have the rapid antigen test versus the molecular test because, well, a lot of the microbiologists in the lab will say that they don't want the antigen test because they are less accurate. The molecular test is 20% to 40% more accurate. So that's true, and I support that.

But at the same time, we need to think about it. This test is more expensive, the molecular test. So that's a burden to the health system, either the insurance or the patient. And then the other thing is the device itself is more expensive, so then the lab probably can't afford to have this device in every single clinic. So then you slow down the turnaround time for a patient to know whether they have flu testing.

So then I would argue that there are places to use the antigen test, but the important part is to know that they're not as accurate. So the key point is, what do you do if you have a negative result and you have a patient who may have flu or RSV or other respiratory syndrome?

I would say that you still need to follow up with molecular testing on those patients. So, again, it really is important to understand the test limitation, the accuracy of the test, and then you can do proper follow-up.

So at Geisinger, the lab provides oversight for a majority of point-of-care testing. When any point-of-care testing is needed, the provider needed to submit the request to us. And then the first questions that we ask is whether the point-of-care testing is clinically warranted. And what I mean by that is, how is point-of-care testing going to improve patient care?

And once we learn that it actually is going to improve, say, GD throughput or help diagnose a patient quicker and hence treat them faster, then we work with the provider to determine which device or kit would be most suitable for their clinical need, and then discuss the testing limitation with the provider, and also, of course, have to perform some cost analysis to make sure we understand the impact.

So here's an example of what we do. So we started to have a medication therapy management clinic. It was a clinic that is managed by pharmacists on various medication. And in this particular case, it's pharmacists managing a patient who takes warfarin, who requires PT INR measurement rather frequently to adjust the dosage of warfarin.

And we use CoaguChek as our device. So part of the training that we focus on is to make sure the pharmacists understand that this cordon of the PT INR results on the CoaguChek compared to the lab. Because when we do the accuracy study in our lab, we know that the results on the CoaguChek are only comparable to core lab results 20% to 25%. And as the PT and the R go up, the results become even more discordant.

So that's important for the pharmacist to understand. And based on that knowledge, when they see the patient, they would know when they needed to recheck the result with the core lab. Say, if the patient is taking their warfarin on schedule, they didn't adjust the dose too much, their diet didn't change too much, but their result changed too much, that would be a sign that the result is not making sense and they need to recheck.

So, as a result of training and, of course, the pharmacist also having knowledge, a lot of knowledge in warfarin. Our warfarin treatment program is safer than national average. When we do a study, we have less side effects, whether it's bleeding or thromboembolic events. And so we are able to continue use of warfarin, which is a cheaper drug compared to some of the other newer medication like Eliquis.

And this is kind of the core of our lab 2.0, which, for those of you who are not familiar with this concept, you may want to look at Dr. Crawford's presentation that is in the archive on ECHO. He did a presentation of lab 2.0 in March.

So, again, back to our model at Geisinger, once we have the device picked out and agree with the provider, then we, the lab, actually implement and manage to test for them. And that includes building the test in the electronic health care record and also interfacing the device if the device is interfacable so that they can document the results properly.

Again, we work with them to do user training and competency. We wrote a procedure for them, which is important. And we validate the that device to make sure the device actually works. And we also have a lot of standard quality management systems in place. Simple things like just making sure they have the forms to document QC is actually important, as well as when they're not change for reagent, we make sure the reagent is performing the same as before.

So the lab actually handles all those things for them. So really, when the provider gets the device, they know that it is a good device and the people are already trained. And once we set all this up, we standardized testing across the whole system and all the hospitals and clinics throughout the system.

And that really make it easier from a lab management standpoint. For example, when we troubleshoot device, when we get a phone call, it's easier for us to troubleshoot because it's very unlikely that it's a new problem. Typically we can find in our log that this has happened before, and we can just use the same method to solve the problem. Or if we cannot fix the problem right away because the whole system uses the same device, oftentimes, we do have loaners available, and we can just lend them a device while we work on the broken one.

So it really helped in terms of troubleshooting. And also helped in terms of inventory management. What we do is that we order supplies centrally so we can have the same lots of reagent and QC, and then we distribute to the floor so that we don't have variation of reagent and see that maybe one lot doesn't work as well as the other.

And of course, when you buy more in one purchase agreement, you usually have greater leverage. But it's not just an advantage to the lab. It's also an advantage to end-users. If you think about health care provider, especially for us who have over 10 hospitals and 80 clinics, providers do sometimes go to different hospitals. But because we standardize it, they know that if they order a point of care glucose, it's the same glucose meter. It's not a different meter.

So it's the same level of service. And knowing the limitation is important for them. Or for our users, like the MTM pharmacist that I was talking about with the CoaguChek, they practice in more than one clinic. But because we standardize it, they only have one device to learn, so again, it's easier. And for our patients, they can show up at any of our hospitals and clinics, and they'll have the same level of service and know what to expect.

So to make this all work, it's because I have a very good team. And this is kind of serve as my acknowledgements slide, even though I'm trying to tell you how to structure the team. The way we structure our team is based on the region, hospital size, and test complexity.

I have a very good left and right-hand person, Beth Ambrose, who is the system manager who worked with me. She essentially managed everything from an operations side so that when we divide up the team– so, for example, in the central hospital, which is the one in Danville, the shorthand is GMC, I have four people, one person being the lead. That's about a 500-bed hospital.

And then the Northeast I set up kind of the same way even though they have three hospitals. That's because two of the hospitals are about 200 bed, and one of them is, I think, about only 50 beds. So they're not as big. So I can have the same number of people manage it in the same region so they can get to the location easily.

In the southern region, the hospital size is a lot smaller. The complexity is a lot less, doesn't have a lot of the complicated device, so I only need two people. And then with a lot of the smaller hospitals that I have– when I say smaller hospital, they are around 100 beds– and this one is actually only about 20 beds. So I can group them all together have one person take care of it.

The problem lies typically with our clinics. Currently, we don't have a super good structure on how to support all our clinics. Is kind of a makeshift process, where they call in, and whoever is on call will deal with the problem. But it's really not effective. A lot of times, we couldn't actually get to it till the next day because, again, unfortunately, hospitals take priority because we have more critically ill patients.

So we're working on a better structure to support the clinics, and I would love to hear any suggestions from other people if they have advice later in the talk. So, again, I think this topic is really important, partly because point-of-care testing is fast growing and also changing. We know that there is a lot of devices that are coming out that are not like an i-STAT.

We talk about mobile devices using, plugging into your Android or iPhone. So this is a fast-growing and changing area. And recent surveys show that 70% of all the clinical labs only perform waived testing. That amounts to about 187,000 labs.

And unfortunately, the example that I showed you earlier on things that we kind of screw up is not really just us. A report by CDC 15 years ago– I know it sounds a little bit dated, but I think it's still true based on experience– shows that really going back to the point that providers, they just don't know what they don't know. The lab is not their specialty. It's our specialty, so we need to help them.

So briefly, in that survey, they showed that 21% of waived testing labs, they don't really check the patient insert update, which could be important. They may tell you that they need to change the methodology, how to perform the test differently.

45% of waived tests do not have proper documentation on expiration dates, or testing dates, or lot numbers. So if they have a problem, there's no paperwork to go back to trace. And 21% of sites did not really perform quality control. So technically, we don't even know if the device is working. 31% of waived tests did not have testing record, and 9% of them did not have a requisition.

So if you don't remember anything that I said in the talk, I hope you remember this one slide, which is my final slide. I really strongly believe that it's important to have a good partnership between the lab and the providers. And that applies to not just point-of-care testing. That's for every test. And it's important for us to truly understand what the clinical need is, and from then on, figure out how the lab can support them.

And by working on it together, I truly believe that we can enhance the clinical utility of lab testing and improve operation and regulatory compliance, which is a big hurdle in point-of-care testing and also ensures accountability for quality care and cost. Because ultimately, the goal really is to serve our patients and improve their care.

So with that, that's all the slides I have. And I don't know if Triona wants me to put up the other slide yet.

TRIONA HENDERSON: Not yet.

ELSIE YU: OK.

TRIONA HENDERSON: Thank you so much, Dr. Yu. Before we open up the lines for discussion, I do have a couple of questions for you. Can you share a case involving point of care at your institution that kind of highlights that it may not be as simple as it seems? I've heard so many not necessarily complaints, but frustrations from microbiology directors and anybody in the laboratory with the push for bringing on any type of point-of-care testing right now, where the providers maybe didn't know what they didn't know.

ELSIE YU: So you're asking me the questions or the audience?

TRIONA HENDERSON: You, you, you.

ELSIE YU: Oh, OK. I think that's a lot. We can talk about COVID, but let's talk about something else. So one that I get a lot is drug testing. Urine drug cup, I did a lot of work on that. There's a big push on having the urine drug cup in the clinic so they can tell right away whether the patient is taking the drug.

The problem with that is fundamentally, providers already have a really hard time understanding urine drug testing. I have a complete different slide back on that. The survey that shows that only about– actually, if we give them a test to do, none of them answer all five questions correctly. That's what the surveys show, which is really concerning because it wasn't asking anything complicated.

And typically, what I told them is, because when I first came here, I get a lot of phone calls on them interpreting the results wrong or not knowing how to interpret. So I tell them that that was the situation, which is why I set up a completely different way of doing compliance testing or pain management, which is them telling me what the patient took, and I order all the right tests for them and give them the reinterpretation. That's how I end up solving the problem.

So to go backward to give them a cup is to reopen the can of worms that we have, which is they look at the result. Say a patient is supposed to be taking hydrocodone. The opioid is not positive. It's not positive, more or less, because it's mostly just looking at morphine and codeine, not hydrocodone. And then they dismiss the patient.

That's not right. But I know we have done that before—the same thing with benzos. So we don't know what the benzo strip is testing. It may or may not be able to detect lorazepam. But the patient's taking lorazepam, so you can use the cup. And they're not going to know. It's too much information for them. So I would say that's the number one that I got a lot.

TRIONA HENDERSON: Perfect. And you described your utilization process. Is it like the usual utilization we think of where you have a representative? You have clinicians. You have a laboratory. You have administration. And the providers come in and say, hey, I want to bring this on? How does that process work?

ELSIE YU: Yeah, we actually just changed that process. So we're still trying it out a little bit. But basically what happens is they fill out a form, a point-of-care testing request form that tells us why they need the test, what's the expected volume, who is using the test. Yeah, I think that's the two questions that mainly is the thing.

And then a majority of them are really simple. Like they would need urinalysis advice in the clinic for urology. So that doesn't need a lot of discussion. That one is pretty simple. And we usually would just approve it so long that the volume makes sense and they have people to operate. And we usually only ask them to join us in the meeting for discussion if it's one that is new and we have questions for them. Yeah.

So it's not always with the administrator. It just depends. And the cost analysis will be done with the administration.

TRIONA HENDERSON: OK. And I know Geisinger System is unlike any other. And so if it's approved, is that rolled out to everyone or just to the section that requested it?

ELSIE YU: It depends. Some of it is system initiative; then, it will roll out to everyone. Some of it is not. So standardization is something that I probably can spend another hour talking about. It makes sense only to a certain level, especially when you look at our hospital system, how we're set up.

Not all of them have 500 beds. Some of them only have 50 beds. Some of their ED is next to the lab. Some of the ED is half a tube system to us directly. So the situation actually changes, and it does vary. We standardize as much as we can is really what I mean, and as much as it makes sense. So, in that case, if the ED is next door to the lab, I highly discourage putting an i-STAT there.

TRIONA HENDERSON: Awesome. Thank you. So now, we're going to open up the floor for any feedback or any questions from the group to Dr. Yu. Or to the group, maybe you want to share a case that you have or some concerns that you have with your institution managing point-of-care testing. Please remember to unmute yourself before speaking. Alternatively, you can use the chat function. If your camera is on, you can also raise your hand, and we'll identify you before speaking.

Somebody asked an interesting question. Will this lecture be on the Penn State ECHO YouTube channel. So, Darrell, this initiative is actually stemming from the CDC and not, I guess, from the Pennsylvania side of things. But I'm sure if they wanted to share it on their site, we will be linking the transcript and the video on our DLS website.

If you scroll up into the chat, it's cdc.gov/labquality/echo. So that's where you'll be able to find this after this session and previous sessions. And for the new ones who have been here, we actually started this project in January of 2020. And so I really thank Dr. Yu even for pulling in the previous presentations, and so you can find a lot of them on our website, lab 2.0, and lab quality, and other presentations.

So I can kick off another. Do you have any advice for institutions in how they can manage point-of-care testing utilization? I think even when I remember when I was a microbiology director for the last place that I worked at, it was different– oh, perfect, Heather is on– having multiple hospitals.

Just like how you guys acquired hospitals, a lot a lot larger academic institutions are acquiring hospitals. And it's, I think, challenging to standardize across when places have been doing things for 20 years. And they're like, OK, you've acquired us, but we still want to continue doing what we've done.

ELSIE YU: Yeah. That is tough. And I went through a few of those already. Again, there's no perfect solution. It really starts off with a discussion with the provider to really talk about whether it's clinical needed. Sometimes it helps to pull in data to show them why it has not worked. We have those situations.

For example, like I said, glucose meter actually is a big one. A couple of the places that we have acquired were using those home glucose meters that whoever sales rep dropped off. That one is not too difficult because we have regulation, and we can just show them that the regulations say that this is not OK. And also clinically, here's the reason that's not good. So that one just takes time to work through.

The other one, mostly, I would say is using clinical reason to help them understand that that may not be the best idea. The nurses are busy. Let us help you do it in the core lab. You will get the same results in the end– maybe a 20-minute delay, but that's not going to make any difference. So it really depends on the situation.

TRIONA HENDERSON: Perfect. Other questions from the audience? Heather, Duncan. You turned your camera off, but your volume is also still off.

HEATHER: Well, I had a little trouble with my Zoom today. So I know you mentioned clinical utilization and a few different, I guess, benchmarks that you use on the simple form. Do you have other benchmarks that you specifically use when you're looking at new requests for tests that are coming in? Or is it very, I guess, customized to each request?

So do you have specific benchmarks that you've already outlined in advance for test requests? Besides volume and cost.

ELSIE YU: Yeah. It really depends, I would have to say. To give you an example, we recently wrote our point-of-care A1C device. I know that's not a new idea, but that's something that we recently wrote out. And the way that we set up maybe a little bit differently than other hospitals is that we are trying to target on patients who is uncontrolled diabetics and really want to bring down those numbers on those A1C.

So that one, it's not so much that they need to convince me that it's a good idea. It has been out there. I know it's a good idea. I think we were collecting data to try to show that we are implementing it correctly, that it really was helping patients. So what we did with that one– and we're still ongoing tracking– is the, A, for the uncontrolled diabetic patient, are they getting the A1C collected in a more timely manner versus before when we told them to go to the lab?

Even though it just down two rooms, they decided not to go. So that was one thing that we tracked to see if more people are being tested on A1C in a timely manner. But another one that I have requested data on is to really see, from a population standpoint, is the A1C number going down? And luckily– I didn't show it in our slide today. Luckily we were able to show that with the implementation of A1C point-of-care device, we were able to dramatically reduce a high A1C patient, showing that it was effective.

So I think it really depends. And in terms of the ED, some of the things that we sometimes would measure is really, can we help them with their turnaround time, so to speak? Not the lab turnaround time, but them being able to triage patients quicker because they have A1C. so then they are not as busy. And that's one of the conversations that we actually have a meeting set up down the line with both flu and COVID going to make it– pre that flu, COVID season, we're going to talk about how we can help them so not so many patients are just going to be in the ED.

HEATHER: We are having the same conversations.

ELSIE YU: Yeah, so we should share notes.

HEATHER: Absolutely.

TRIONA HENDERSON: Thank you, Heather. We have a question from Jeffrey Driscoll. In my experiencing visiting hospitals and physician groups, the places with the best point-of-care testing programs always have someone from the lab dedicated to just making sure all those point-of-care tests around the facility are being stored and used correctly. I've run into a number of places where the staff isn't uniformly trained or remember proper cleaning of devices between uses. Do you have any suggestions for things you always like to check for and frequently see wrong?

ELSIE YU: I think the two things that I will say for those, sometime it's just really general training and competency. That you have to put that in your box, that you have trained them, you have discussed that.

But another that I want to talk about is more that if you have people that keep forgetting things, I am one of those that I don't believe in sending a bulk email to the whole world is going to solve the problem. I am one of those that actually are going to be like, hey, user A, you have a problem. We need to have a discussion. And your supervisor is going to know that this is not OK.

So it a little bit depends on your approach. I find the latter more successful if it's a few user situation rather than a general problem.

TRIONA HENDERSON: Great. Thank you. We have another question. Any recommendations for SNF facilities to start performing COVID point-of-care testing?

ELSIE YU: That one, I think, is kind of similar to what I said in the flu testing slide. It depends on what COVID testing you are doing. I know there are a lot out there right now. And I don't know if they talk about antigen testing or molecular testing. I would say if you're doing not molecular testing, antigen testing, for example, or antibody, which won't be useful, so I'm just going to focus on antigen and molecular.

If you're using antigen testing, I will say if you have a negative test result, it needs to be followed with a molecular test. And I would even go further to say that if you're doing molecular testing, if you have a patient that you really suspect have COVID, it needs to be retest.

We have had cases that, quote-unquote, is "false negative." But we don't know if it's really false negative or if the sample wasn't collected correctly. And we also know that there is variation depending on which molecular platform you use, that some is more sensitive than the other. So it really depends on your scenario, what you're doing.

Are you doing it for public health kind of surveillance, or are you doing it so that you know which patient to isolate? Because if you're talking about isolation, then if you have patients who have symptoms, you should isolate them, even if you have a negative antigen test, unless you can really confirm it's negative with molecular.

And of course, it also depends on the onset of symptoms. We know that the first couple of days, you may not have enough virus or antigen to pick up.

TRIONA HENDERSON: Thank you for that. I'm sure everybody's heavily preparing for the new season. Jumanji level 22. Another question. So how do you educate your point-of-care coordinators on connectivity or programming of all the different devices? It seems like a constant challenge.

ELSIE YU: It is. I will say they educate me on that instead of me educating on that. So, again, let me see. I'm gonna go back to my slide here. Beth really has been doing point of care for many, many years. So she kind of knows it in and out. So luckily I don't have to train them. They kind of train me on that.

I have more a general understanding on how it works and help them if things need to be escalated instead of me actually connecting anything myself.

TRIONA HENDERSON: Perfect. Continue dropping the questions. Either you can– we'd love to hear you. Otherwise, you can drop them in the chat box. I have another question for you. Is there any point where a point-of-care request is really a bad idea?

ELSIE YU: The urine one was a bad idea, the urine drug testing one. Oh, the using PT INR in the ED for a stroke patient is also a terrible idea. But that one is also easy to tell them that that's not a good idea from a clinical standpoint. So it wasn't too bad. And I also have good relationship with the ED. I think that helps, too.

Which is why I go back to saying that I think partnership is the most important part. If you have a good relationship with them, I think it works better. Yeah.

TRIONA HENDERSON: Perfect. And I noticed that you stratified the operator and then the provider. When you're doing training, do you break it out separately so you have all the providers together? Or do you do it team based?

ELSIE YU: Training typically is only with the people who actually do the tests, not with the provider. The provider is usually at the beginning of the discussion. Yeah.

TRIONA HENDERSON: OK. Any other questions before we wrap it up for today? I'll give you guys 30 seconds. And thank you, everyone, for the participation. We had a really robust discussion today, as I expected. I'm sure everyone is concerned about COVID point-of-care tsting.

ELSIE YU: And I'm glad the poll worked. I was a little concerned about that.

TRIONA HENDERSON: That's always my favorite part. In the meantime, I guess we can advance to the next slide. A little blurb from us from CDC. So my division, Division of Laboratory Systems. I know a lot of you in the group are interested in point-of-care testing, and I wanted to bring to your attention that CDC will shortly be coming out with a new web page specifically on point-of-care testing.

So currently there's a short section on point-of-care testing on the interim guidelines for biosafety and COVID-19 page, which our division takes care of. But we're developing a more robust page as it continues to be a topic of discussion for laboratories and testing facilities. So the page should be live by the end of the week. If you want to receive updates on this page, please sign up for CDC Laboratory Outreach Communications System, shortened LOCS, by sending an email to locs– L-O-C-S– @cdc.gov.

It's also where you'll get the updates for the clinical laboratory COVID response calls that are happening bi-weekly now. And then we can go ahead– let me just check. There are no more questions. Advance to the last slide.

So we're excited to have next month Dr. Bryan Dangott. It's going to be on clinical informatics. And he has a wealth of experience. He's also a hematopathologist that has covered all subspecialty sections of clinical and anatomic pathology. He's now at the Mayo Clinic in Jacksonville, Florida.

Please visit the DLS ECHO website to register for the session and to view the final sessions for the year. I can't believe we're almost at the end of our pilot project, which has been scheduled for 12 months. Hopefully, we'll have updates about what we'll be doing next.

Thank you for taking part in our discussion today. We hope you found it very valuable and important in the work that you're engaged in. And we look forward to having you participate in the final sessions for 2020. And we'll keep you posted on any changes.

We will adjourn. Thank you, Dr. Yu, and thank you, everyone, for your participation in this session, and have a great day.