Koninklijke Philips N.V. (PHIA) Earnings Call Transcript & Summary

Welcome to the Royal Philips Analyst Conference Call on Tuesday, May 16, 2023. During the call hosted by Mr. Roy Jakobs, CEO; and Mr. Abhijit Bhattacharya, CFO, all participants will be in a listen-only mode. [Operator Instructions] Please note that this call will be recorded, and replay will be available on the Investor Relations website of Royal Philips. I will now hand the conference over to Mr. Leandro Mazzoni, Head of Investor Relations. Please go ahead, sir.

Thank you, and good morning, everyone. Welcome to today's conference call to update you on Philips Respironics Test & Research Program for the polyester-polyurethane sound abatement foam. We appreciate that you could join our call on short notice. I'm here with our CEO, Roy Jakobs, our CFO, Abhijit Bhattacharya; and Steve Klink, who is the spokesperson for the test and research program. Roy and Steve will take you through today's update. And after that, there will be an opportunity for Q&A. We ask that your questions be limited to the test and research program. The press release, slide deck and frequently asked questions on the topic were published at 8:00 a.m. CET on our Investor Relations website this morning. The full transcript of this call will be made available on the website as well. Over to you, Roy.

Thank you, Leandro. Hello, everyone, and thank you for joining us this morning. Today, we are providing an update on the latest testing results regarding the safety of the sleep therapy devices affected by the June 2021 field safety notice. Before I discuss the results in more detail, I would again like to take this opportunity to reiterate how important patient safety and quality are to Philips. As a company, we are driven by our purpose to improve people's health and well-being through meaningful innovation. This is at the heart of everything we do every day. Now let me turn to the set of results we are reporting today. After a thorough and comprehensive process, we now have a complete set of test results for the first-generation DreamStation devices, including tests and analysis for devices that have been exposed to ozone cleaning. The test and analysis show that the exposure to VOC emissions for devices exposed to ozone is unlikely to result in appreciable harm to health in patients, which is very encouraging. This is based on assessment of ozone-induced degradation from up to 500 cleaning cycles. This means that the outcome of all of the testing of the first-generation DreamStation devices is that the PE-PUR foam degradation, should it occur, is unlikely to result an appreciable harm to health in patients, which is very encouraging. In addition, test and analysis have now been completed for SystemOne and DreamStation GO devices and show that exposure to foam particles and VOCs is unlikely to result in appreciable harm to health in patients in those devices as well. We are also completing the analysis for SystemOne and DreamStation Go devices treated with ozone cleaning. Based on results to date, we do not expect different results compared to the first-generation DreamStation devices. Let me remind you that these devices contain the exact same of PE-PUR foam as the DreamStation1 devices. Together, first generation DreamStation, SystemOne and DreamStation GO represent approximately 95% of the devices affected by the recall globally. At this time, the overall guidance for health care providers and patients using devices that have not been remediated yet remains unchanged. The relevant competent authorities globally including the FDA, are still reviewing the extensive data and insights gathered over the last 24 months. We have incorporated their feedback to date and will, of course, address any feedback and questions that these competent authorities may have. I will now pass it over to Steve to talk about the testing processes and latest findings in more detail.

Thank you, Roy, and hello, everyone. I would like to start with a closer look at the outcomes of the testing of the first-generation DreamStation devices. Today's set of results built on the previous update we gave in December 2021, June 2022 and later December 2022. And on this slide, you can see an overview of the previous test results. We previously communicated to you that the prevalence of [ visible ] foam degradation is low. It's 0.5% in the inspected devices in the U.S. and Canada, the 0.04% of the inspected devices in Europe and none of the inspected devices in Japan. We also shared with you that both new and used first-generation DreamStation devices passed volatile organic compounds and particular matter emission testing. And in our last update in December 2022, we announced the completion of the biocompatibility tests for these devices. Exposure to particulates is unlikely to result in an appreciable harm to health in patients. Test and analysis have now been completed for first-generation DreamStation devices that have been exposed to ozone cleaning. The results published today indicate that potential patient exposure to foam particulates and VOCs from the PE-PUR foam in these devices is unlikely to result in an appreciable harm to health in patients. This builds on our previous findings that exposure to particulates from degraded foam with self-reported ozone use is unlikely to result in an appreciable harm to health in patients. Therefore, although ozone cleaning exacerbates foam degradation the test and analysis show that ozone-treated foam is unlikely to result in an appreciable harm to health in patients. We continue to advise patients to follow Philips Respironics' instructions and recommended cleaning and replacement guidelines for the sleep therapy devices and accessories. Ozone and UV light cleaning products are not currently approved cleaning methods for sleep therapy devices or masks and should not be used. Moving on to the SystemOne and DreamStation Go devices. Comprehensive third-party risk assessments have concluded that potential patient exposure to VOCs and particulate emission is also unlikely to result in appreciable harm to health in patients. This is based on particulate matter testing according to the ISO 18562-2 standard VOC testing according to the ISO 18562-3 standard as well as a bioassay evaluation, chemical characterization and toxicological risk assessment according to the ISO 10993 standard. As Roy said, we're also completing the analysis for SystemOne and DreamStation Go devices treated with ozone cleaning. Based on the results to date, we do not expect different results compared to the first-generation DreamStation devices. I would like to spend a minute on the rigorous methodology of the testing. We received a lot of questions from you on this topic, so it's important to explain it again. The test and research program has been conducted with 5 independent certified testing laboratories in the U.S. and Europe, and the results have been reviewed and assessed by third-party qualified experts as well as by an external medical panel. The testing laboratories conduct a test in accordance with the international Good Laboratory Practice or GLP standard, and they're making their own independent scientific assessment of the test results. The applied test methods comprising test planning, test execution and interpretation of the results for the completed risk assessments are in accordance with the applicable ISO standard. The design of the applied test methods was scientifically underpinned based on a thorough consideration and mitigation of test limitations that are inherent to any test standard and scientific research. The laboratories and/or qualified third-party experts ran very conservative toxicological risk assessments to assure confidence in the results. Worst case assumptions were considered. For example, assuming the theoretical consideration that all of the foam is degraded and the patient is exposed to all of the degraded foam. When we inspected more than 65,000 devices, in none of the devices, all of the foam was gone. So in all of the devices, there was still foam in the device. The VOC and particulate matter tests were conducted under different flow rate conditions. The VOC emission has the highest concentration at the lowest clinically relevant flow rate. Particulates are, however, stirred up via high flow rates, thus the maximum clinically relevant flow rate was used for those tests. Therefore, a 17 liters per minute flow rate was used for the VOC emission test and a 90 liter per minute flow rate for the particulate matter emission tests. We also checked testing at other flow rates, and they did not yield higher VOC or particulate matter concentrations. And ozone-induced degradation was tested up to 500 cleaning cycles. This resulted in foam degradation that was in line with the degree of degradation that we have observed in the expected -- in the inspected use devices with self-reported ozone. And let me please clarify that when we did the cleaning cycles, we have measured the VOCs and the particulates at various stages. For example, at 0, 20, 50, 60, 70, 80 and I can continue, so -- and then up to 200, 300, 400, 500 and even one at 1,300 cleaning cycles. And at each time the VOCs that were emitted were well below the safety limit. And at any time, it was almost a factor of 10 below the safety limit, which was taken very conservatively. Another important topic is that the sample size is not small at all for this type of study. We have spent 24 months inspecting thousands of devices and conducting hundreds of tests. As I indicated earlier, we officially inspected approximately 61,000 return devices from the U.S. and Canada, 2,500 devices from Europe and approximately 2,000 devices from Japan. For the devices from the U.S. and Canada, where the user indicated they have not used ozone cleaning only 164 out of 36,341 of these devices that is 0.5% showed significant visible degradation. So to do meaningful measurements on foam degradation, we need to primarily focus on those 164 devices and not so much on the other 36,177 devices. So automatically, you start with inspecting a big sample of devices of many thousands and then zoom in on the devices that show significant degradation. So -- and then testing was subsequently performed on a smaller sample size, comprising multiple use devices with deferring amount of patient usage and observed visible foam degradation. And on top of that, we have done tests on lab-aged foam that had been intentionally degraded to different degrees and to a degree that was worse than any of the devices that we have inspected. And this way, we have included the worst-case scenario for our measurements, and we have combined that, as I've already mentioned before, with very conservative assumptions for the risk assessments. So for example, that means that we have chosen fairly low safety levels. Following the December 2022 update, the Dutch competent authorities stated on their website that the research was conducted thoroughly and that the test design is apprehensible and the methodology is adequate. With regard to next steps for the Trilogy 100/200 and OmniLab Advanced Plus ventilator devices, the VOC and particulate matter testing continues as well as the chemical evaluation and toxicological risk assessment. We expect to provide an update on this in the third quarter of this year, as we indicated before. As a reminder, the device has passed VOC and particulate matter testing as well as several biocompatibility tests. New and lab-aged Trilogy 100/200 foam failed the ISO 10993 genotoxicity testing under laboratory conditions. And therefore, a weight of evidence assessment is ongoing to confirm or exclude potential risks for patients under the expected usage of these devices. I hope that you found the information that I just presented useful. And with that, I would like to give the floor back to Roy.

Thank you, Steve. Let me thank our patients and customers for their patience and our suppliers and partners for their continued support. The new test results reported today are again positive and reassuring. And as such, we are happy to present them. This means that the outcome of all of the testing of the first-generation DreamStation devices is that PE-PUR foam degradation, should it occur, is unlikely to result in appreciable harm to health in patients, which is very encouraging. The completion of the testing program as well as the remediation remain our highest priorities. And with that, we will now take your questions. Thanks.

[Operator Instructions] We will now take the first question. The first question comes from Ms. Veronika Dubajova from Citi.

Just would love to get a bit of insight to what the feedback has been from the FDA, as you have shown them this data, what their kind of thoughts have been? And when would you expect to have something more formal from them in terms of how they are perceiving this data and whether it's satisfactory to maybe alter some of the language around the recall of the risk assessment?

Yes. Thank you, Veronika. Along the whole process of testing, we have kept the FDA informed as we also shared earlier, and that also led to the inclusion of the notion that they are still studying the results. I also mentioned that we are now 24 months in the testing program. This is a very extensive testing program and as such, there are a lot of data involved. Now what I alluded to earlier is that it's important to know that FDA is, of course, not a testing house. So it's not necessarily expected that they will qualify the data or that they will kind of give them a formal stamp of approval. But they are, of course, looking into the data. And if they have further commentary then we can expect that they will provide that to us. So that's the latest of the FDA. And before publication of today's results, we have reached out to them, we have kind of included their commentary, and that's what we also did last time.

Okay. And can I just ask a follow-up, Roy? Just sort of surprised by some of the stats that you presented at the outset of the call about the proportion of devices where you've seen degradation. I'm just really surprised by the geographic difference, none in Japan, one in Europe and then so many in the U.S. Just -- are you confident with the data? And what explains it in your mind? Because humidity cannot be the only thing. And I'm just a bit surprised by the variability, it's quite striking and then I'll go back into the queue.

Veronika, this is Steve. The -- so indeed, for the U.S., it's 0.5%; for Europe, it's 0.04%. What I must mention is that for the U.S., it's self-reported no ozone use. And what we suspect is that among those that ozone has been used. Because -- and that leads to a higher prevalence compared to Europe and compared to Japan. I would say that Europe and Japan are similar at 0.04% or 0%. Yes, that is very close. And what we've also seen from Europe is that we have tested or we have inspected devices from various ages from between 1 and 5 years and from various parts of Europe, from France, Spain, Italy, the Netherlands, the Nordics. So all types of climates are in there, and that did not have an impact on the degradation of the foam and that is in line with our current understanding of the degradation process.

So in short...

[indiscernible]

Sorry. Yes, I think maybe to build on answer, as we have said before, ozone use does have an impact and that you have seen being used much more in the U.S. That's actually where you see the difference from. That's also why today's presentation results are so important because despite that ozone was used and as a result, more degradation happened even with that use of ozone actually, it's not leading to any appreciable harm to health in patients. So I think that is why today's update is really relevant and was an important step to kind of complete the whole test results on DreamStation1.

We will now take the next question from the line of David Adlington from JPMorgan.

Just to follow up on Veronika's question there. Just in terms of -- I think you said on the call, you have incorporated the FDA's feedback today. I just wonder what you meant by that? Some sort of clarity, some further color about what their feedback has been, how it's been incorporated in today's release? And sort of following up on that, just phrase appreciable harm. What do you mean by appreciable harm? Could it still cause some harm or how should we be interpreting that phrase appreciable harm?

Yes, maybe starting with the latter. So appreciable harm is, I think, the formal term that's used also as part of the standard. It, in essence, means extremely low probability that there's any harm done. So this is the technical term that's being used in case that your evidence shows that, in essence, there is no evidence harm. Then on the inclusion of the FDA's request, that is that they may come to a different conclusion based upon the ongoing study that they do. So that's what they have shared. And as you have seen earlier, we also included the limitations of our test results, and that was also kind of established together with the FDA. So you have this 1 pager, I think, also in the presentation and that shows that kind of any test has kind of limitations. We have also taken countermeasures against those limitations where we then also kind of led to an inclusion into these results. So I think that is the Slide 10 of the slide deck for the Investor Relations page. So those are the 2 elements that came back from the FDA.

[Operator Instructions] And we will now take the next question from the line of Sezgi Ozener from HSBC.

Congratulations on delivering the testing results. One question on the number of devices remediated. You've said you remediated 4.3 million in globally and a large part of this 2.3 million, I believe, in the U.S. When do you expect to complete the remediation globally? And can we now be ascertain about the total number of devices involved?

Yes. So indeed, we share the 4.3 million and the 2.3 million that are with patients. So what we see is that actually, we are making good progress. We shared earlier that kind of we have more than 95% produced. And depending on the country that we look at, we also are above 95% of remediated and back in hand of patient. But there are variances in it. And what we see is that we -- as we get now to the back of the remediation, there are certain patients or devices that are registered and that actually do not respond anymore to our outreach to replace the device. And that might not be too surprising as we know that 35% of the population actually drops out of therapy after 1 year. So it might be that we never get to that full number that we initially reported as totally registered number of devices. Now we're doing all our efforts to kind of reach these patients and to kind of make sure that they get it. We have them at hand, right? We have them produced, so we can send and install, but we do need, of course, to get the patient to share his details, both of the address as well as therapy. And that's kind of taking this last stretch. Now as we shared earlier, we expect that in Q2, we would really further progress on this in a significant manner. And then thereafter, there might still be some really tail end patients that might step forward in a later date. And then of course, we will, at any time, remediate as we have promised to do so.

Just as a follow-up, do you have an estimate of the number of people that may have dropped out after registering the devices initially?

No, we don't have an exact estimate because that's kind of -- it's hard to speculate on how many people do not return. But we do see that in certain areas, we see less responsiveness suddenly coming in. Yes, and then, of course, you start to kind of further look for contact, but that doesn't then come back. So we will keep attempting, but currently, we do see that there's a certain number that doesn't reply and I give you an example for the Netherlands, we're now at kind of -- we had more than 100,000 to remediate. We have now 97,000 remediated. Yes, and then it's the last few thousands that are hard to reach. So that's kind of where we are in this last percentiles that kind of we need to either reach or we will find that they might never come to us, yes, and then we don't need to remediate them.

And if you don't reach these patients, you set aside the provision, assuming that you will replace all of their devices registered. So if these people don't come back, is there a possibility that you might reverse some of that?

Yes. We will have to see at that point of time because the machines would have been remediated. So it could be either a repaired machine or a new machine. If it's a new machine, we probably would be able to sell it. If it's a repaired machine, maybe not. So these are things that we will have to estimate closer to the time of this thing coming to an end.

We will now take the next question from Mr. Robert Davies from Morgan Stanley.

I had a couple. One was just on, in terms of remaining testing. What's actually sort of left to do what data you're waiting to still get? And then the second part is just in terms of the options here from a legal standpoint, how are these testing results is going to impact the potential settlement or provision that you're setting aside? Has this -- has the outcome of this kind of testing had any impact on the size of the provision that being satisfied in your mind?

So this is Steve for the ongoing testing, so we are in the process of completing the testing for SystemOne and DreamStation Go devices that have been treated with ozone cleaning, but they contain the exact same foam as a DreamStation1 devices. So we do not expect different results there. And then, of course, we need to complete the testing for the Trilogy 100/200 and the OmniLab Advanced Plus ventilator devices. So that is among, say, the last 5% of the devices that are under the recall. A large part of this test have already been done, but there are still a number of tests that need to be done, and we expect to provide an update on that in the third quarter.

Robert, on the provision already set aside, the provision that we have set aside is for the economic loss. So it is not related to the test results. The provisions that we do not make or that we have not been able to make a reasonable estimate is on personal injury and that we will -- that will come, as we have said, next year. So the existing provision of 575 will not change because of the test results.

And this is, of course, very important news for the patients because we can assure them that there's no appreciable harm done to their health regarding the use of the device. Of course, if there are cases where people then claim differently, we will use this data to kind of have the dialogue around it.

We will now take the next question from the line of Wim Gille from ABN ODDO.

Yes. This is Wim Gille from ABN ODDO. Taking into account that no appreciable harm was done to at least 95% of the patients in scope, can you give us a bit of feeling on if there is any readthrough with respect to a potential outcome on the consent decree and the personal injury class action that is still out there? I.e., what do your lawyers tell you what the impact of these test results might be on a positive outcome on these 2 outstanding points as well?

Yes, Wim, thank you for the question. As we said earlier, it's hard to speculate on outcome, both on consent decree as well as personal injury claims, it's too early for that. Consent decree, in essence, anything to do with the testing. The consent decree of follow-up on the [indiscernible] that was given by the FDA after the filings of site visits, and that's being worked into a wider set of measures. So that's a separate topic than on the personal injury, as said, that's very hard to speculate on. So we will not do that in terms of what this will mean in or for any of these cases. As I said before, of course, the fact that this is not having any health impact for patients, that's the big news of today that we are very happy to share. And therefore, yes, it was important to complete ozone's part of the testing DreamStation1. So we now have very rigorously [indiscernible] all separate pieces over many, many devices, across the different test houses and with the independent views on it. So yes, we feel very encouraged to be able to share those and then also take them forward.

And then as a follow-up, taking into account that the use of ozone has basically increased the degradation of the foam quite substantially. Is there any, let's say, lawsuit or legal process going on against the company that is actually promoting this ozone cleaning or is that still not in progress at all?

No, we will not comment on that. I think we tested the ozone impact. So this was about the test results. And therefore, we are very happy to say that outcome was that there is no impact on patient safety and we will not kind of further comment on any regulatory or litigation action with other companies.