Pulmonx Corporation (LUNG) Earnings Call Transcript & Summary

So the music has died down, which tells me, it's time for us to start the next presentation. I am still Joanne Wuensch from Citibank. And this time, we have with us the management team from Pulmonx. So welcome to New York City.

Thank you for having us.

It's interesting as I try to follow or I do follow the Zephyr Valve launch and uptake. And we've spent a fair amount of time talking about the eases or obstacles to product adoption, particularly launching a new product during the pandemic. So I'd like to sort of start our conversation today from a big picture vantage point. Where are we in terms of the product launch? And where are you maybe in terms of moving some of those obstacles out of the way.

Well, we can take no credit for moving the biggest obstacle out of the way, which was COVID. So -- but it's nice to see that we're in a fundamentally different place. And I think we talked about that in our third quarter earnings report how across the first -- or the second and the third quarter of last year, we had the first uninterrupted 6-month window into our business and we learned a few things and it modified our strategy and focus as we execute. Taking a step back, as we look at our business, we're building it in 3 different ways, and we have to be careful the order of operations of these 3 things. One is that we're trying to be very selective on the accounts that we choose to go into in the United States, we're targeting on the order of 500 accounts. We're into nearly 300 of those 500 accounts. And we're selecting those accounts based on their capabilities, their geography, the surrounding COPD populations and very importantly, the sort of keep treating physician. And as we've talked about in the last couple of quarterly calls, we're very much focused on ensuring that they're getting to a point where they're doing the procedure routinely. And what that requires is for them to be doing it efficiently. So they're adopting best practices. They're looking into their own networks. And ultimately in collaboration with our marketing departments looking to tap into the referring physicians that are out in the periphery. We estimate 80%, 90% of the patients are being managed outside the field of view, if you will, of the treating centers. So -- and probably 70% to 80% of the patients are coming through referring physicians. So the second area of focus is that the COPD physician, the referring physician, we know who they are because the patients themselves have a distinct pharmaceutical fingerprint. We can capture data on the prescribing habits of that combination of drugs, and we can know who's managing the majority of patients in the geography that surrounds our treating centers. Of course, turning that second area on before that treating hospitals are up and running is not practical or logical or rational. And so we want to make sure we've got good treating centers. We then try to make sure that the referring physicians both have the information that they need and are introduced and well acquainted with the treating physician because we're talking about referring some of the most fragile patients that these doctors have. And then ultimately, something that we've also talked about a lot is our direct-to-consumer activity. So we've got a couple of hundred thousand patients that we've engaged with on -- through social media on a global basis. About 150,000 of those have opted in to communicate back and forth with us on e-mail, but we have to be very careful if 70% to 80% of those patients are going to go through a referring doc to make sure that we don't create a wave for those referring docs before they understand the magnitude of benefit that we're delivering and the people that are doing the procedure. So those 3 elements have to be ignited in the proper sequence in a given geography, and we're sort of advancing things in staggered parallel in certain geographies, we're ahead of others. So as we look ahead and as we are primarily currently focused, we're making sure that we've got the right hospitals up and running in a given geography and in parallel then engaging with the referring physicians in that chart.

And how do you identify the right hospitals in the geography? And is it possible for you to say, okay, we train the physician here, but he or she is not adopting it. We're moving our energy elsewhere?

Yes. So we're very selective on the hospitals that we are targeting. So I think I said 500 hospitals in the United States, which represent something on the order of 8% of U.S. hospitals. So our patients are going to drive by 5 hospitals on the way to getting to the treating hospital, which is actually very attractive to the treating hospitals. Typically, those hospitals that are doing the procedure have a lung business, if you will. They're doing -- they've got thoracic surgery there. They've got pulmonologists that are there in the hospital. They tend to be closer to 350-plus bed hospitals, not 90-bed hospitals, typically, although we do have in certain geographies like billings or what have you, we've got smaller hospitals that are doing the procedure as well. But in any case, that's how we identify the hospital itself and then we're looking as well for a team that exists in that hospital. Typically, there's an interventional pulmonologist. Maybe 500 fellowship-trained interventional pulmonologists probably a little bit less than that in the United States. We expect about 800 doctors to be ultimately trained to do our procedure. And we very much prefer to have a center that's doing many as opposed to 2 per year. And so we're trying to be -- we're pursuing sort of a classical center of excellence approach. And we've got sales reps around the country. The entire country is covered. And when we're growing out that base, and we're about roughly 2/3 of the way to get into the ultimate number of hospitals that we're going to target.

Are there any positions that you've had to pull back?

Yes.

Okay. So talk me through, you train Dr. X and, I don't know, January 2020, and he or she are not doing what you think they should be doing. How do you assess and how do you retrain?

So when we assess, we decide whether we want to retrain or whether we're just going to walk away. Often it's from a doctor, not from a center because the centers here are generally interesting. The only situation where we would walk away from a center would be if that we made a bad decision on selecting a doc. They just wanted to get on to our website. They just wanted to tell their colleagues that they were doing the procedure, but they're not really doing the procedure. And they happen to be a controlling influence at that specific site, i.e., typically, you go -- if you're at a teaching hospital or something and you pick the wrong doc, you just find another one in that center. At the Cleveland Clinic, they've got 10 people that can do this procedure. So you just go to the next person and so forth. And typically, it's not a problem in the person who thought they were going to do the procedure who's not doing it typically doesn't have a problem. So there's relatively few hospitals that we walk away from. There have been some and the physicians we typically work around. But it is super important to us is -- we have some number of our accounts, about 25% are exactly where we want them to be. There's another -- if you take the balance, there's some number of -- nearly all of them were ultimately going to get to the place that we want them to be. But in the immediate term, we have to choose who we want to invest in. So I know you asked a question about divestiture. There's really not that much divesting going as on an account level, but there is in that next group a very active process going on as to where do we get the best return on our investment of time in terms of that middle group of hospitals that we're trying to take into the most productive cohort.

How do you think about the procedure reaching a tipping point?

What do you mean tipping point?

Either the patient it's obvious he or she should be doing this or it's like the physician that's not trained yet as what I would call procedural coma like I need to know how to do this.

Yes. So we're pretty close to -- I mean like I said, we're about 2/3 of the way there to getting the doctors and the sites set up. So it's really about the referring physicians and where are they? And at what point does this become I don't know. I mean, to put it in the most -- well, I'm not going to use this phrase, but I mean there is a certain element of not only should I be doing this, but is there a consequence to me if this is the standard of care. It's in all the national and global guidelines, and I'm not even talking to my patient about it. It's an information thing. When we go out to physicians, there's some number very uncommon for pulmonary specialists, people who have a big COPD practice to have not heard about valves. It's quite common in the early conversations to learn that those people don't have complete information. So for example, have you ever heard of valves, yes. Are you referring your patients to a local center to get valves done? No. Why? Well, maybe that they don't know who's doing the procedure. These are among their most fragile patients. It's -- they not only -- they need to know what information am I going to get on before the procedure in the periprocedural phase. After the procedure, are you going to round trip that patient back to me? You're a pulmonologist, I'm a pulmonologist, am I giving you my patient. There's that whole conversation. And the interventional pulmonologists are subcontractors, so they're not going to be stealing any of these patients. But they've got to go through that. The most common reasons why people -- when we ask them, are you referring and they don't refer, they'll say something like, well, I'm waiting for the data to be published. Well, we've got 4 randomized controlled trials. One of them was published in the New England Journal, the other 3 in the American Journal of Respiratory and Critical Care Medicine. So that's an information thing. We're waiting for them for this technology to be recognized in the guidelines. Well, we're in the global guidelines, national guidelines, the rating on the quality of our data is A, which is the highest level randomized controlled trials, waiting for this to be reimbursed. It's broadly reimbursed. 99% of patients essentially can get covered. 75% of our patients are Medicare patients. There's no question that people are getting paid for this. So they'll come up with all kinds of reasons. And I would say the top 5 reasons are just education. It's like, wait a minute, that's done. That's done. That's done. And they go, "Oh, well, that's interesting. I don't know, but there's 2 things. I think once you get that out of the way, it's okay, we're delivering this magnitude of benefit. The best you can do on drugs alone is this. Do you -- why not talk to your patient about their interest in turning back the clock some number of years to possibly enable them to do things that don't have them tethered to their couch with an oxygen -- with oxygen. Maybe get back to walking up to your backroom. Spending time out in the yard with your grandchildren. Going to the mall. We've had examples of -- one woman said, "My life is my horse. And now I get to go back on my horse." Another woman was a professional musician. Had to give it up and now is back in the band. Those kinds of things. that they can turn back the clock. So it's such that education on the magnitude of benefit. And then as I mentioned before, the introduction to the physician and understanding who's going to be doing this procedure because this patient is fragile. Let's talk about your experience, what could go wrong? What are you going to do about it? How experienced are you in managing that? And what's the probability that patient is going to come back to me not happy with me because I sent them to you.

How do you think about sort of the other side of the claim, the patient coming in with the Google print out saying, this is what I want. I mean you're starting to see the push versus the pull aspect of adoption?

Well, 20% to 30% of our patients on a national basis are going to the treating sites themselves with that Google print out. So there is some amount of that -- we have a talk to my doctor program. So they are going to their referring doctor, they're referring doctors is going, "Oh, that sounds new. Oh, we got to wait for that to get published." And so the patient is like, wait a minute, I'm on this website with 100,000 other people, and they all think that this is all taken care of. So we have this program called Talk to My Doctor, which is a fully compliant program, where the patient says, "Listen, my name is Jane Smith, I'd like you to go to Dr. Jones, use my name, tell them I asked you to come and go see what you can do to get them some additional information." That's about the warmest lead you could possibly have. Our sales rep follows up on those types of things in that specific geography virtually every time they get in front of the physician. It's a very powerful way to enter. And when we then -- either us or just the doctors themselves get connected, the treating physician and the referring physician, 50% to 70% of the time patients start to flow.

Okay. Do you think that there's a backlog of patients? And how do you define backlog if the answer is yes?

There used to be -- there were always backlogs coming out of COVID waves, which were pretty obvious. And the ways that we could very clearly quantify them was that they were patients that were scheduled and then the state of Texas says all elective procedures stop or a given hospital said, if there's any chance that this patient -- I mean I don't care how small might end up in an ICU bed, we don't want you to do that procedure unless the patient is going to die tomorrow type of things. So those were situations where it's like we could see the scheduled patients and then know that the bump that you'd see often, we would show it in a quarter coming out of a COVID wave. Those kinds of -- that's what we viewed as pent-up demand or backlog or whatever. We're now 3 quarters away from those types of things shutting down our business. And as a result, we have a lot of patients in process. So this is not a procedure where you walk in, in the morning and the procedure is done later that day. there could be months between when that patient walks in the front door and when they've gotten through all the pretesting and ultimately get the procedure. So we've got a lot of patients that are in process. We know what's scheduled out in front. We know we do these quantitative CT analysis software assessments to identify the patients that are most likely to benefit. If there's a green light, we call them a StratX green light, that's software StratX green light. So we can see those StratX green lights accumulating in a given geography. We have calls into our reimbursement services group. We can see those building, which why are they calling to get preauthorization to have a procedure, if you're not planning on having the procedure. So we do have certain ways to see what's coming on the horizon. But there is no thousand patients that are 1 million patients or whatever that are "sitting by themselves in a room waiting to be called for their procedure."

Number 223?

Yes.

Okay. So all of this, when you put together your 2023 guidance, and we were talking about this after your earnings call, how did you piece it all together? Because Pulmonx is one of the few companies that gave guidance above the consensus estimate. Most companies safely bracketed it.

Yes. Well, so I appreciate the question. I mean I think that the answer is going to be similar to what we've always said, which is the guidance that we give is our best estimate of what we can do and what we feel is achievable at the time that we give it. And so we feel good about the $63 million to $65 million that we guided to. I mean, I think the components -- digging a little deeper into your question, the components of that guidance, first off, I'd say that our assumption is that the macro environment is really just stable as it is. So there's no assumption one way or the other that the macro environment needs to improve, say, for us to achieve that guidance. We believe that we can get there primarily by driving increased account productivity and throughput through the year, and that is something that we think we will be able to do internally and it is within our control. So that's the primary lever, I would say, of guidance, both to the upside and to the downside as well, which is really, this year, we're focusing on driving throughput, driving same-store sales and productivity. We're going to open up new accounts in the U.S. We've talked about opening up 40-to-50-or-so new accounts. But really, our focus because it's going to be on driving those same-store sales, and it's going to be our ability to do that, that I think drives upside or downside to our range.

Okay. I do want to speak a little bit about, because Zephyr is still new. And you touched upon this a little bit. But the pathway of the patient getting off of their couch, getting in to get the valve and getting back to their life. And there's a number of touch points that, that patient needs to go through. If you could just remind us of those steps.

Yes. Typically, as I mentioned, they're either going to go talk to their physicians who manages their day-to-day care more often. The probability that somebody is seeing a pulmonologist goes up with disease severity. We're treating patients with severe emphysema. So though, there's a lot of internal medicine docs, most -- about roughly 50% of the patients that are out there are being seen by pulmonologists. So that's -- 70% of patients are going to go ask that individual. What do you know about this? I just learned something about it. And the rest of the patients are going to go directly to the source. The University of Chicago is doing the procedure. They're going to go and start knocking on doors or they saw something on TV, they go to talk -- try to get an appointment with the physician that was featured in that. And then whether the referring physician starts doing the work up or whether they simply refer to the treating center to start the work up, there's questions that have to be answered. One is that we're treating patients that have severe symptomatic emphysema. And some of those patients may -- by the assessment of the physician who's treating, may not yet be sick enough. That's a progressive disease. So if they're not ready for the procedure today, perhaps some time in the future, they will be. So in that case, that's a fairly quick assessment. Their pulmonary function testing, which they'll do, maybe stronger than what would be considered normal to do the procedure. We treat patients that we solve for air trapping in the chest. So our procedure, imagine you have a 1- or 2-liter bottle taking up space in your chest cavity of fixed volume, and we're trying to pull that space-occupying inefficient air out of the chest to allow healthier tissue to expand. If the patient is symptomatic and yet they don't have discernible air trapping that is considerably greater than what would be considered to be normal, they're probably not a good candidate. Now 80% of the patients who have shortness of breath, severe emphysema are going to have some kind of air trapping. So that's not a normal stopping point. We have to be sure that these patients are candidates for anesthesia. So if they have significant cardiac comorbidities or if their disease is just beyond where we can really reasonably help them, that they're too sick to withstand the insult of the procedure. Now just to put into perspective, the insult of the procedure, last year, we did 5 or 10x as many procedures as the surgical analog that the lung volume reduction open up the chest. So it's -- this is not a one-to-one kind of thing. We've already lapped the surgery. And now what we're doing is we're going after the great number of patients that aren't candidates for surgery because of the insult. And our procedure can be -- can treat a whole array of disease distribution within the lung. Surgery, you have to -- basically, you can only target on the edges, right? Because you're not going to go into the middle of someone's lung and take out a baseball size chunk of lung tissue. You're targeting a lobe, and you want to make sure that you can get at where the problem is. In our case, we can be very surgical in our approach, if you will. Whereas with surgery itself, you can only deal with the edges. So in any case, there's a number of reasons why somebody would be included or excluded in a number of tests that allow them to determine whether those patients are good candidates or not.

And where does StratX end? I'm sorry, I just planted...

Chartis, yes.

Chartis. StratX and Chartis. I think I just combined them, StratX. Sorry, StratX and Chartis fit into this?

Yes. So taking a step back. So 5 or more years ago, everybody, we were thinking about treating. We needed to anesthetize and stick a tube down their throat, blow up a balloon at the area that we were going to target and measure -- basically simulate the procedure to understand whether that patient was collateral ventilation negative. Meaning not -- didn't have significant communication of air between the target lobe and the adjacent lobe. If we go to a patient who has a lot of collateral ventilation, and we put our valves in an effort to try to evacuate the target lobe, it just refills through the back channels from the adjacent lobe. So what was happening was about 50% of the time -- this is like 5 to 10 years ago, 50% of the time the procedure will be done, the patients would be woken up, and they'd be really disappointed. "I'm sorry, you're a collateral-ventilation positive. You're not a good candidate." And so what we did was we said, "Hey, wait a minute, all of these patients are being diagnosed with the CT scan." And basically, you show up at your doctor's office, "I'm really short of breath doing this activity that I used to be able to do routinely, and I can't do it anymore." Are you a current or former smoker? Yes. Let's do a CT scan." They look at tissue density of the lung relative to normative values for somebody of your age. And if you have considerably less density of your tissue, you're diagnosed as having emphysema. We take that very data from the CT scan, and we run it through this quantitative CT analysis software called StratX. And we're able to rebuild the fissure or tissue division between lobes. And if it's at least 80%, we say, green light. We're going to go ahead and schedule a procedure. If it's less than 80%, the odds of them being CV-positive is quite high. So it's not worth it. Those patients today, and we can talk about AeriSeal in a moment. But those patients today are not candidates for our procedure. So we -- the patient gets anesthetized. They go in with a bronchoscope, and they do that little balloon catheter. Let's-- that's Chartis. So Chartis catheter. It's connected to effectively a computer. It measures changes in pressure and flow. If the calculated volume of the target lobe is 1 liter, and 2 liters flows through that Chartis catheter, we know that errors coming from somewhere other than the target lobe, there must be collateral ventilation. If the amount that flows through is say -- as I said before, let's say there's a liter of air in there. And at 500 ccs of air that flows through the catheter, pressures start to rise in that lobe, that's an indication that, that's a CV negative patient because you would think, as you evacuate that space, the pressure in that space is going to go up. And so that's a green light. 80% of the time we anesthetize somebody to undergo our valve procedure, we put in valves. 20% of the time, today, they are woken up and disappointed to learn that they were CV positive and, therefore, not good candidates.

Do you charge for those 2 tests?

We do not charge for StratX. Our cost is on the order of $75 to $100 a piece. We'll do 2 or 2.5 of those for every patient we treat. So we're making, in the United States, $11,000 a patient at 75% gross margin with, whatever, $150 for StratX that we give away.

Okay. What about for Chartis?

Chartis, we do charge for. We have the Chartis controller. We sell or we charge a per use fee. And we -- if that patient gets woken up and they are CV positive, they didn't get valves and they're immediately flipped from an inpatient DRG to an outpatient procedure, and the hospital gets reimbursed for that outpatient diagnostic test, and we get paid. We get paid about $1,000 for the Chartis catheter. And I think on the order of $650 if they rent the use of a Chartis. And then the hospital gets paid for that outpatient procedure.

Do you want to talk about...

AeriSeal?

Yes.

Yes. Well, it kind of follows, right?

Yes.

So I talked about, okay, we're going to -- 80% of the time we anesthetize the patient, we end up placing valves. 20% of the time, the patients get woken up and they're disappointed. We have a technology called AeriSeal, which is a 2-part injectable polymer that we reported on -- we put out a press release, I think, in the fourth quarter that...

Fall.

Or in the fall, yes. And so we had some data on a study that's a multicenter, multinational study out of Europe that looks at whether you can do 2 things. One, can we go into these patients that are -- that have -- that are essentially Chartis red lights? These patients that we go in, we want to treat them. We can't treat them because they're CV-positive and use AeriSeal in a procedure to make them CV negative. So like a tire-patch kit. You just -- we got an interruption here. We can see it on our StratX scan exactly where it is. We know the airways that feed it. We go in with AeriSeal, deliver it into those airways. Will that take a patient that is objectively CV-positive and make them CV negative? And the answer -- that was -- that came out of the one -- there were 2 things that were reported basically around the third quarter. One was the interim convert data, which was on the order of 40 patients. It indicated that 80% of the time, we can move those -- take those patients from CV-positive to CV-negative. And then the second question, which makes the first question irrelevant, unless the second question is the right -- turns the right way, is when you put valves in those patients that you've just made CV-negative, do they behave the same way as patients did in the prior studies that are the basis for us having the product on the market? And the answer is yes. And we saw that in these -- both the interim data from the CONVERT trial, which we reported late last year, and there was a single center study out of Australia that looked at something on the order of 14 patients or something to show the same thing. So it's very encouraging. If we -- if this continues to play out, we ultimately are looking to get AeriSeal utilized, both outside the United States and in the United States. And if in doing so, we would increase our total addressable market by 20% to 25%, because now these patients that we would normally be waking up, we would actually be doing something for them. As I mentioned, how and where we place the AeriSeal is informed by StratX, which we're already doing. So it's a little bit different analysis, but it's not inconceivable. We know that -- or the plan is that the doctor who's doing the procedures talks to the patient, says 80% chance, we're going to wake you up and you're going to have valves. 20% chance, we're going to have put in AeriSeal, and an 80% chance thereafter that, that's going to make you ineligible for valves in the future. And they're all -- so there's about a 96% chance that the patient is going to wake up happy, like you did something to help me, and I'm CV-negative. So let's see how the valves work. So anyway, that's the plan.

And what is the timing of FDA approval?

So we have an IDE trial in front of us. So we expect to enroll our first patient in an IDE trial before the end of this year. So we're a few years away from U.S. market entry. We have the CE mark on AeriSeal. And we expect that with this CONVERT 1 trial that we provided an interim data on late last year, we'll have all those patients followed up and published certainly by next year. And we would expect that sometime next year, we'll start seeing some revenues coming out of CE marked countries.

And where do you think you are internationally? And where do the can go internationally?

Well, last year, 60% of our revenue was in the U.S., 40% was outside the United States. Between 90% and 95% of our global business is spread across 6 countries. So Australia, plus 4 European countries, which are the -- our largest OUS country is Germany, then France, then the U.K., then Australia, then Switzerland. And so that -- we expect that the U.S. will continue to grow faster than international. So we estimate maybe 65% of our 2023 revenue will come out of the United States, but we continue to see growth opportunities outside the United States in those markets. We're in about 25 different markets. As I mentioned, the bulk of the business comes out of 6 countries. 97% to 98% of our business is direct. So one of the things that I think differentiates us as a little company and, frankly, allows us to stand shoulder to shoulder with any company that's in the interventional pulmonology space is that 97%. We've got our people on the ground. We've got people in China, Australia, Japan and all the major European countries. So we've got a good footprint to take advantage of the opportunity that's out in front of us with the technology that we have with that, which is under development and a cornerstone for any other technologies that we may develop or putting our bag another way.

Question for you Derrick. As you think about gross margins, what do you think peak gross margins are for the company? And I'm going to add a little bit more to that. As you add AeriSeal expand geographically, reach this maybe, maybe not tipping point utilization in the U.S., how do you think about your need to ramp SG&A?

Both great questions. So I think from a gross margin perspective, we expect we can get into the high 70s at scale. So today, this year, we'll be operating around 73%, 74%. We continue to see meaningful gross margin expansion opportunity that's primarily going to be driven by increasing our output and driving overhead absorption. So we'll be holding sort of production relatively flat through this year and then start ramping up production and utilization again end of this year into next. And as we do so, we expect to see a steady, moderate increase in our gross margin to kind of that high 70s range. So that's where we think we'll be. As we build the capacity, I think we'll also demonstrate what we certainly will plan to demonstrate operating leverage, which maybe is kind of getting to your question on SG&A. So this year, we're guiding to growing our revenue 20% on average. We are guiding to grow our cash OpEx spend, that's OpEx excluding stock-based compensation, by about 10%. So already this year, you'll be seeing sort of that operating leverage move forward. We're certainly going to be continuing to grow our SG&A or our OpEx spend, if you will. Some of it will come in the form of R&D as we continue to ramp up AeriSeal, et cetera. But we expect that our OpEx spend will continue to grow at a meaningfully lower rate than our revenue. And so we'll be seeing some nice operating leverage there.

What about R&D?

So primarily, our R&D increase over the next couple of years is going to be the AeriSeal program that Glen talked about. So some of the 10% roughly increase in R&D spend this year is going to be that. I think you'll continue to see our R&D increase on that order over the next couple of years. And then certainly, we'll have other internal projects internally, externally. There's going to be R&D spend that's going to continue on. But I would say, for the next couple of years, the R&D growth primarily will be a lot of clinical study, regulatory spend related to the AeriSeal program.

Do you have a vision of where you get to cash flow breakeven?

Yes. I mean just rough numbers, we've been saying between $100 million and $200 million in sort of annualized run rate. If you want to pick a number, call it $150 million or so-ish is where I think we can get there. We burned $44 million in cash last year. I expect that burn to be flat to down this year, and we expect to show and demonstrate cash flow -- our cash burn leverage moving forward. One of the metrics that we're now referring to the Street is our adjusted EBITDA loss. And that, I think, is a clean metric -- a cleaner way that's -- for us to look at our operating leverage, is it sort of adjust out some of the noncash expense. Our adjusted EBITDA loss last year was around $41 million. And again, our target would be to see that -- or demonstrate that, that can move down even this year over time.

Excellent. We were together a year from now, what do you think we will be talking about?

Well, we're very confident in the plan that we're on. We see it as a multiyear plan. Obviously, there's some number of accounts that we've yet to go into that we're going to have to be bringing up. We see our accounts, even in the pre-COVID phase, taken a solid year to go from sort of doing their first patient to getting close to being a routine account. We expect that we'll be showing meaningful increased productivity in the accounts that we're in. Obviously, we have always thought of -- and we think about new accounts as being very opportunistic. It's not central to our growth strategy. What is central to our growth strategy is same-store sales. And so we're going to -- we measure productivity on average in our accounts across a given quarter, and those that are established have been up and running for a while. And we talked about in the last quarterly call, the idea that we delivered 4 to 5 patients per quarter. I think the average in the fourth quarter was 4.8 patients. We would expect to see, on average, those numbers getting into that -- exiting this year into the sort of 5 to 6 patients per quarter in established accounts. So those are the kinds of metrics that, I think, we would be expecting to see moving in a positive direction.

Wonderful. Glen and Derrick, thank you for joining us today.

Thank you very much.

Thanks, Joanne.