Implementing respiratory testing algorithms: Insights for diagnostic laboratories

(audience applauding) All right, welcome. Thank you guys for joining us today. As Jonathan said, we're going to be talking about diagnostic stewardship and how we can apply those tenets to developing respiratory algorithms. And I am very lucky to have two panelists from Vanderbilt today. So we have Alisha Ezell, she is the Enterprise Point of Care Manager for Vanderbilt. And then we also have Dr. David Gaston, and he is the Medical Director-- Hey everybody. - For the-- - Molecular Infectious Disease Laboratory, MIDL. But it's part of the clinical microbiology laboratory at Vanderbilt. - Yeah, thank you.

And so today, what we're going to do is kind of get their perspectives about some recent changes that they had with their respiratory testing algorithm, some of the considerations that went into it, some of the successes and maybe challenges that they're experiencing, and hopefully provide you guys with some good feedback on maybe how you can apply these tenets to your institutions as well. So let's start with a kind of a bird's eye view. Why are we even talking about this?

So, of all of the really negative things that happened during the pandemic, one of the really great things that came out of it is there was a lot of innovation when it came to molecular diagnostics, especially for respiratory infections. And there's lots of different types of technologies that we can fit underneath this umbrella. But what we're going to talk about today, specifically, is PCR. And PCR is, but a slice of that, but it's a very rich slice that comes in lots of different flavors. So we have singleplex testing that detects one pathogen. We have targeted a multiplex that expands a little bit to have two to five pathogens. Some vendors are also coming out with flexible multiplexes that kind of allows for customers to define their own adventure and choose whatever panel that they want that will specifically meet those patient needs. And of course, we have the expanded panels, oftentimes called syndromic, that usually test for 20 plus pathogens, typically.

The other thing is that these tests are not just performed in the central lab anymore. You know, we've had a lot of awareness and demand for bringing these closer to the patient. So you also have these molecular point of care technologies. And I think that the big problem statement or the big question here is how do you choose? As you're a provider, as you're working in the lab, how do you look at this really complicated landscape and figure out how to optimally position each one of these solutions to meet the patient needs for all of the diverse care settings that you have represented within your institution? And so that's what we're going to talk about today, is how Vanderbilt has approached this, whenever it comes to redefining their post pandemic algorithm.

So we're just gonna, first to start out, Alisha, can you just tell us a little bit about Vanderbilt Health, what kind of institutions are representative and what your patient population looks like? - Absolutely. I think that's a good place to start. You need to understand, you know, what your patient population looks like and what the care looks like that your organization gives. And so we have a pretty large health system. We have seven hospitals. Our main campus consists of an adult, a peds, a psych and a rehab hospital. And then we also have three community hospitals. At main we have a level one trauma center, a regional burn center. We provide high level trauma for pediatrics as well, and a very large NICU. Our psych hospital and behavioral health is quite extensive across our enterprise. And then, like I said, we have rehab and community hospitals. We have a lot of ambulatory care, and it's spread out within about a 2-2.5 radius of Nashville. And so very unique in terms of urgent care to large multi-specialty clinics that encompass a lot of different specialties. And then we do quite a bit of, you know, solid organ transplant, admission from ER, and then bringing higher acuity patients from those community hospitals into our main hospital. So a very large patient population and very diverse across Middle Tennessee.

Yeah. So you have a massive catchment, you're making lots of impact. So David, what are some of the considerations that you have as you're trying to figure out how to design an algorithm to meet each one of those patients at every kind of institution in which they may be presenting to? - Yep. You have to, at least what we've done is, start with what you don't know and walk into it not thinking, "Oh, I know what every patient is going to need", or "I know whatever use case is gonna be." And kind of approaching with a bit of humility. I come from an infectious disease background. I'm an infectious disease provider by training, who's found his way into laboratory medicine. So grateful to be here with all of y'all. So I have some perspective there, but I don't know what individual providers are going to be needing to provide the best care for their patients. And we've had lots of conversations, with many providers across the institution, to try and craft something that fits, but doesn't cause so many restrictions that they say, "Hey, we have to go back and now we can't use it." So, being flexible and building algorithms that are not so honed into one use case that others that are just as significant get pushed to the side or get worked out.

So, as we'll go through our algorithms here in a moment, the algorithms are designed to be in line with institutional standards, but also allow flexibility for providers to say, "This is a unique situation and I need to pursue a different testing route." And so we'll allow that and go for it. But there are lots of conversations, lots of moments where I've sat back and said, "You know, I never would've thought of that use case." Like, "that makes a lot of sense." And just having to kind of have that moment to say, "All right, let's go back to the drawing board. Let's figure out how we can fit this in and make it work."

Yeah. Yeah, I think that that conversation between the clinicians in the lab is especially important to be able to provide guidance but also give them autonomy, right? - \[David\] Sure. - So that each person on the side of that aisle is learning from each other. I think that that's what you guys have done really well. - Appreciate it. Talk to the providers, see if they think we've done it really well. - Yeah. So do you wanna talk a little bit about kind of the recent lab move and maybe how that contributed to some of these changes that you have as we go into algorithms?

Yeah. Yeah. So our laboratory has been based for, I don't know how many years, there's some pictures around the laboratory dating back almost 100 years in Vanderbilt. A lot of fun to see just this one small room with some glass flasks to say that was, you know, the micro lab way back. We grew, outgrew the space where we were maybe 20 years ago, but have remained in that space. So Vanderbilt is very committed to laboratory medicine, very committed to anatomic and clinical pathology in a remarkable way. Have started with a laboratory transformation project that has included moving the laboratory to an offsite location. That is not the entire transformation project, but that's one portion of it. And that laboratory move was finalized in March of this year. So we're in a gorgeous facility, right at the bend of the Cumberland River. You could walk out back and take a walk along the greenway and see the goats that are chewing the grass back there, the chew crew. But that has led to a lot of other changes, particularly with our laboratory information systems. And so as we were moving, we had the opportunity to say, "Let's revamp this testing." We're at a different stage in the pandemic and as an infectious disease doctor, we're not outta the pandemic, but we're at a very different stage in the pandemic at this point. So, how can we take some of the testing practices that had gotten away from a diagnostic stewardship focus, how can we realign them as the lab is moving and as we're able to really have the ear of the hospital and a lot of providers to do a bit of a reset. This testing algorithm is one part of that transformation. The move as part of the transformation is going on for years in the future.

Nice. And Alisha, did this lab move have any impact on your point of care? Were you kind of, I guess, impacted by this or was it independent? - Oh, absolutely. I mean, anytime you move your central laboratory into an offsite space, you know what happens to the main campus and the inpatients that are impacted, you know, what does your outpatient flow look like at that point? And I think it gave us a chance to reset, you know, as David alluded to, a chance to step back and go, "Okay, are we really meeting patient needs?" Should we change, you know, work on testing utilization and other ordering practices with this move to really meet the needs of our patients as we move in change to make sure that we're providing the best patient care. And point of care, we're certainly at the table with that. 

And I think that was a big part of the success of this is integrating, you know, MIDL's perspective and point of care and coming together with our clinicians and saying, you know, "Are we making changes that make sense for the entirety of our patient care and all the populations that we serve?"

- Yeah, definitely. Okay, so let's jump in. David, do you wanna talk a little bit about kind of, what your prior state for respiratory testing was and how maybe some of the challenges that you experienced that made you want to update it to what we have currently?

- Sure. The respiratory testing algorithm that was in place, so with the prior state, began during the height of the pandemic. This was before I was at Vanderbilt, so, I'm speaking historically, you had providers not knowing what the best test to order was for a given situation. This is, you know, also early in the pandemic where, you know, designing algorithms to say, "Okay, well we need to test for so many other respiratory viruses." Turns out everything was SARS-CoV-2 for a solid, you know, two years. So the algorithm was built and then it expanded and then the lab would bring in new technologies and new testing panels and it would expand more and then new technology, new testing panels and would expand more. And so the algorithm became a bit of a hodgepodge, where we're looking at the left side of these slides, we're at a point of breaking it down between immunocompromised patient, immunocompetent patient, favoring broader panels for immunocompromised patients, narrow panels for immunocompetent patients, back in the, you know, the height of everything. If you remember, if you needed to walk into a coffee shop, you needed to get SARS-CoV-2 screening and show your, you know, it was a different age to be able to say, "Hey, we just go through." At a conference like this, you wouldn't have been able to attend the conference without saying, "Hey, you know, here's my vaccine card", or going from there. So lots of asymptomatic testing was occurring, that's primarily gone. And then looking at patients who were positive trying to say, "Okay, have they become negative?" That's more for placement, particularly for long-term care facilities or other places in the hospital that would require some sort of documentation of a negative test. And then for transplant patients, who are going to be undergoing certain transplants. 

So that was the prior state, quite honestly, what it had come to was most providers ordering broad multiplex testing for patients who were being admitted to the hospital, regardless of any of the immune status, any of- People just became very accustomed to ordering that. And that was, you know, we'll talk through diagnostic stewardship and how that matters later, but that's really where the algorithm came to. So we said, okay, we've gotta do a bit of a reset. Let's work with the institution to say "What are we seeing?" So we went back through the data over the past few years and said, "All right, now we're in a spot where SARS COV2 is less prominent but still there. Influenza A is very prominent, and RSV is very prominent." 

So showing providers that data was very compelling to say, "Perhaps broad respiratory testing for all patients is not needed. Let's focus on a more focused panel." And then designing the epic side, or utilize Epic as our EMR at Vanderbilt, to work in decision support tools, so that when a provider is ordering respiratory testing, they just have to make one or two clicks and it aligns them with the institution to have focused testing, instead of having someone have to look through and say, "Well, I think I can you know, I'm just gonna make it easy. I'm just gonna order everything." We try to make it easy for them, but also provide those situations where they'd say, "No, we do actually need broad testing." "We do need to assess a patient for a less routine respiratory virus because of their unique situation." 

So still providing that, which is what's really shown in the current state. And also having certain tests that have a much faster turnaround time, particularly for patients that are being admitted to wards with negative pressure rooms, where patients could be at risk if they're being admitted and have a respiratory virus that would be easily transmissible. So hematopoietic stem cell transplant wards, that sort of, not point of care, but almost as close a point of care that you can get in a laboratory testing environment, is something that we've incorporated to be able to help patient flow and go from that. So this is, we have, this is the downtown inpatient slide, the next slide with our regional hospitals is very similar. So we've aligned our testing algorithms so that what's available at the regional hospitals is also available in the main campus. And there are just a few tests that would be sent from a regional hospital to us, including some of the rapid testing that would be used for patient care and patient management and rooting through the hospital system. And then the next slide with our outpatient, incorporates a lot of the point of care system. So, turn that over to Alisha.

[Alisha] Yeah, so I mean, I kind of wanna back up for just a second and talk about inpatient, you know, making that shift from prior state to current state. You know, we had to think about what we had to do, kind of, scrambling at the beginning of the pandemic to really just provide testing in general and bring it together and really look at what the workflow and the testing volume was like in the main laboratory and how could point of care support that. And so, I mean, for a while the ER was running testing, it was kind of all hands on deck with whatever testing we could do, and we really stepped back and took a look like, does that make the most sense with their time and resources and is that something that the lab can take on and through waves, what does that look like? And then really reinventing the current state workflow to really match with what we had learned from the pandemic and where we could support through point of care, but bring as much back to the lab as possible, to allow providers and clinicians to get back to what they were doing, almost to pre pandemic, at least to let the lab handle the majority of testing and only use point of care when it was really relevant and helpful in the situation to provide care right at the time that it was needed. And so outpatient obviously, is more point of care centric as you would expect, but really in the conversations that we had about building the algorithm to its current state was, you know, where does point of care, long term support MIDL's workflows and considerations for our patients, but then also looking at testing utilization and are we effectively providing what we need across the board, whether it's point of care or our lab. 

So I think that we were more involved here in this conversation just because, you know, we found that really providers were kind of getting that baseline on those patients that were really important. Do I have a result that I can treat off of and send the patient home, or can this wait and be sent to the lab? And I'd like you to expand a little bit more on how this impacted MIDL. But from a point of care perspective, I think that we've kind of found our place and we're kind of level set with, you know, our job is to provide that timing for those patients that are really gonna maybe be admitted or they have (indistinct), and the provider is really concerned about next steps for that patient. And then if it's self-limiting or something like that, or they can wait for a result, they're just gonna go ahead and send that to the main campus and it becomes part of that workflow.

Yeah. Quality was also a very large portion of some of these decisions at the, you know, the height of the pandemic when, you know, everybody's just holding on and some of these tests were going out near patient care. There were some quality issues, because laboratory testing was being put in a place where people were not accustomed to laboratory testing and, you know, contamination events occur and, that's something that we wanna avoid. So Alisha did some really fantastic work in terms of teaching so that, you know, we pulled back some of the testing and those that came back into the laboratory, but then through some great educational efforts that testing is being able to go back out and being able to get near patients to decrease the turnaround time and also increase the availability of the results to the patients and the providers. But that focus on quality is, you know, always in our mind, we wanna make sure that we are offering the best tests and getting the right results and getting that in absolutely as quickly as possible, without sacrificing any, you know, anything that we're doing from a quality perspective.

Yeah, that's really helpful. So one thing that I wanna focus on, on this outpatient, as you know, so far have talked about what the relationship has been like with the two of you as you're trying to figure out this algorithm, but in the end it's the clinicians that are ordering, excuse me, ordering these tests. So in the event where you have multiple tests that are attached to the same patient population, what are some of the considerations or guidance that you give to clinicians to ensure that they are ordering the appropriate tests, when they have so many things to select from?

So for instance, right here, if we look at the slide, your immunocompetent folks, there's potentially three different types of tests that could be ordered. So how do you help clinicians kind of choose which of those may be the most appropriate?

Yeah, we- I just want to give a shout out to the teams at Vanderbilt, the Epic team as well as the Cerner team. So they work so closely with us, and do a fantastic job and have just allowed this to work. So, having conversations with clinicians, designing how that needs to look at Epic. What we do is we sort of set the institutional standard to say, "Start with fourplex testing." So SAR COV-2, flu A, flu B, RSV, that'd be really where we think most people should start.

And something that we maybe we'll talk about in a moment, at this moment we're not changing based on respiratory season, because we haven't had a normal respiratory season for the past four years at this point. And maybe this year will be slightly more normal, but there really has not been an ability to change testing and turn off certain tests and turn others on. So we've just said, "Okay, start with the fourplex", because we've seen a lot of flu and a lot of RSV.

But we present that in the Epic order set, where when you open that order set, that's preselected and if they're inpatient, then all the contact precautions are preselected and as needed, contacts with infection prevention, that's all automated. So the provider doesn't have to go through and say, okay, which one do I want? Do you want this? Do I want that? We found many providers would open the order set and then just click order. So when a broad multiplex panel is the first thing that comes up, that will be what's ordered. And that may be what a provider wants, but it may not be.

So pre-selecting what we're actually seeing the most of, and then pre-selecting what a provider has, we're not forcing them to make that selection, but we're saying, "Hey, this is already available for you." And then to click out of that, there are other dropdown menus that you're able to see singleplex testing, or broad multiplex testing. So a provider has to look a little bit more to be able to get to what they want, but over time they say, "Okay, I know where this order is, I know what I'm looking for, so I'll add an additional click and go in."

From the click count too, this is something that our epic team helped us with a lot, changing from the prior respiratory algorithm to the current respiratory algorithm, decreasing clicks by anywhere between three to five clicks, which for providers that matters a lot. So, you know, there's an element of uptake there, where someone's not having to go through, "Okay, now I have to click this and select this." Just say, "Nope, it's preselected, this is what you're going for." So same thing with some of the immunocompromised or sorry immunocompetent panels where sort of put the institutional standard out there and then if there's a real reason for someone to say, "Yep, we won't follow that", then that's the clinician's decision and we leave that autonomy as you were saying, Alisha.

Yeah, great. I think that this is a really nice kind of transition, as we talk into what this implementation process looked like. So I know that as we were preparing, Alisha had a really great quote that was, as you're thinking about which stakeholders you need to involve to make sure that the implementation process goes as efficiently as possible, is that "everyone who needs a test is a stakeholder." And this starts from the very top and goes all the way down to the kind of patient that's getting those tests, right? So can you walk us through a little bit, in addition to all of the education and kind of, the LIS and considerations that you had, is who did you bring into the table and what are some of the resources that you made or even some of the conversations that you had, to reach out and like, have those contacts with those stakeholders so that this implementation process went smoothly or smoothly as possible? I guess.

[David] Yeah, it's just, it was a lot of meetings and a lot of cold calls, emails to say, "Hey, this is changing. We need your perspective for the emergency department, we need your perspective for the acute care oncology clinics. We need your perspective from OBGYN, we need your perspective from hospitalist medicine." And you know, just trying to pull as many of those perspectives together and find something that didn't work for everyone, but at least had the flexibility so that anyone could utilize it and not encounter roadblocks in the restrictions. So yeah, a lot of discussions got- Again, only been at Vanderbilt for two years. This was a wonderful place to be able to get to know the community and be able to really, really build, you know, build bridges with a lot of other groups outside of laboratory medicine as well.

And so Alisha, you know, in the point of care space it's more unique. And we've talked a little bit about, you know, some of the challenges like quality. How did this roll out or how did this change kind of impact you? Did you identify any kind of specific challenges within your decentralized settings that made this difficult to implement?

I think part of it is back to that stakeholder piece, right? You know, who's a stakeholder? And it starts with the clinicians and what they're seeing and what the impact the algorithm will have when we make those changes. But also boots on the ground. Sometimes we forget that our nurses, who are right there with the patients, are also involved in this process and that changes we make to that algorithm are gonna impact their workflows. And so, you know, understanding what they're seeing and experiencing and really having them inform us, so that we can escalate that to the larger group to say, you know, make sure we're considering if we make this change, how does that impact this entire workflow for all of our peds after hours clinics or, you know, all of the adult urgent cares or what does that look like, because we don't wanna make it harder for them to do their jobs and add more to their workflow. And so really having them a seat at the table and just really sitting down and saying, "How is this impactful?" Because if we pull them away to do more work to get the lab results, then we're pulling them away from their patients. And so we wanna make sure that they're considered and part of that process, and as David said, there were a lot of meetings and making sure that, you know, we're including people. We might have the ID doc that covers peds, but peds is a big group with a lot of different dynamics. Or do we include our psych folks, because sometimes we don't think about what that looks like for admitting patients to that group and just using and leveraging our relationships with most of the community that, you know, point of care is ubiquitous at Vanderbilt. So using those relationships say, "Did we catch everyone? Did we cast the wide net?" And make sure that they're all involved. And then how does that look for points of care downstream as we support it going forward into whatever the next phase of this journey is for Covid.

Yeah, I think it's a really important one, kind of recognizing your own limitations and then pulling in those stakeholders that compliment that. And I think that anytime that you bring people on board early in the process, they feel very much more invested and receptive to those changes, than just stating what the changes are and then hoping that it works, right? So I think that you guys did an excellent job with that.

And making changes now. So since it's live, I've had many discussions with providers that say, "Can we change the wording here? You know, that it's causing providers to order a test that's not institutionally aligned." We say, "Yeah, that wording looked good to us initially, but you're right when you see it in that perspective." So we'll pull back and we'll change some wording. We've changed around the way that tests are ordered that come up in the order panel, the way that they're presented. So we've gotten great feedback from providers who are very much engaged and very much see this as something that they want to contribute to. And then making those kinds of changes, rolling them out and then having providers then write back to say, "Hey, saw you made the change, thanks. I really appreciate that." That helps us build a lot of trust and helps build those relationships that are continuing on.

Perfect. Well, I think that we have four minutes for questions. Jonathan has the mic, so if anybody in the audience has a question that they'd like to ask Alisha and David, please just raise your hand and Jonathan will come over with the mic.

Okay. Awesome. One question, you guys moved to an offsite location and you streamlined your ordering. What did you notice about inventory? Like was it easier to maintain your inventory of all your tests or any impact from that standpoint?

That's a wonderful question. The inventory for the lab itself went from being distributed from individual laboratories around to then being consolidated within one group that manages the inventory. So, from the lab transformation lab move prospect, it's been great. A guy named Jerry Trujillo oversees a lot of that and he does a fantastic job.

So, but that was distinct from some of the respiratory algorithm changes. I think had we not moved and we were just focusing on those changes, it's an ongoing impact on how those changes impact ordering and then decrease or increase the amount of testing that we're doing with individual assays. So something that has not entirely been able to be predicted. We have seen our fourplex testing go up remarkably, our multiplex panel testing has not decreased in the same way. So there are a number of other providers that will have an order and then they will say, "Okay, that's negative. Now we're going to order multiplex respiratory testing", when a patient arguably may not need that, but it's something that they're accustomed to doing. So there has been a change, it hasn't been one-to-one.

So much of this in making algorithmic changes and provider changes is the human, you know, human nature and human activity, which if you can predict it, good on you. - But from a point of care perspective, you know, we have 200 sites of care, maybe about 80 or so are performing this testing. And as we change the algorithm, we're looking ahead, we know it's coming. So we're communicating to, again, those boots on the ground, those people that are ordering in those different locations for their clinics, "Hey, this is coming, what impact do you think this is gonna have based on the providers in your clinic and what they may start ordering more or less of and shifting." 'Cause we all know spaces are tight, especially in our ambulatory sites, and they may not have enough, you know, refrigerator spaces that's necessary or shelf space if that's necessary. And so just accommodating, you know, getting them ahead of the curve on the change and saying, "Hey, let's be prepared for changes in behavior" and monitor it closely until you guys equilibrate and you come out to a place where you feel like you're good with ordering. So just kind of keeping them well informed. And I think that was kind of the theme of everything is different ways of communication and things like that so that everybody's in the loop about what's changing.

Great discussions, David and Alisha. So my question is, how much did the reimbursement landscape shape what solutions you place in different settings and also what panels you put together? - Yeah, yeah. Also a great question. So we dissuade outpatient providers from ordering the broad multiplex panel testing, because that is not reimbursed by any providers right now, except in very limited situations. And we don't want that to be passed down to the patient. And many, many providers don't understand that. They don't know if it's over five or six analytes? I'm on stage and I don't remember the exact number, five, thank you. You know, if it's over five analytes, then that's not, you know, that's not gonna be covered. So, some of that comes with education through this process. Some of it also just comes with trying to put more barriers around the broadest testing that people have become very accustomed to. Within the hospital environment the reimbursement strategies that we're focused on are really saying, we wanna make sure that the patient's getting the right test as fast as possible and we go from there. So some of it comes more down to turnaround time. Some of it comes from working with the clinicians to be able to say, "What do you need and how do we get that to you quickly?" With working with the infectious disease, infection prevention teams are able to say, "How do we incorporate diagnostic stewardship here?" So that was not as much of a driver for the inpatient changes, but it was a bit more for the outpatient changes.

[Jonathan] Any additional questions for our panel here? All right, we will wrap that. Thank you again, Alisha. Alisha and David. Let's give them a round of applause again. (audience applauds)