Literature Review March 2026 (Part 1 of 2)

NTSP Literature Review Podcast: March 2026 (S7, Ep3.1) Part 1

This months papers cover:

• Next Generation Sequencing of Tracheal Aspirates
• Robotic PDT
• Bronchoscopy and PDT
• Staff roles and view of PDT
• Machine learning limitations for predicting PDT

Brendan is joined by Dr James Orr to explain all about sequencing techniques to identify bacterial fragments and what that means for clinical care.

Link to supporting PDF: https://tracheostomy.org.uk/Podcast-Resources

The UK National Tracheostomy Safety Project (NTSP) is committed to providing education, information and resources to improve patient safety and the patient experience for those with tracheostomies and laryngectomies. All of our resources are housed on our website www.tracheostomy.org.uk, accessed by over 30,000 visitors each month from around the world.

This is the only podcast to bring you literature reviews, hot topic discussions and interviews with healthcare staff, patients and families.

Our goal is to improve the safety and quality of care for patients with tracheostomies and laryngectomies through education. We work closely with patients, families and healthcare professionals to develop new resources to improve care. We’ve collaborated with key stakeholders in tracheostomy care since 2009, and developed freely accessible resources, supported by online learning developed with the UK Department of Health. We’ve worked with the Global Tracheostomy Collaborative since 2012 to improve care for patients and their families everywhere.

We are funded by grants, donations and in partnership with medical device companies through unrestricted awards. This podcast series is supported by unrestricted education funding from the Atos Learning Institute. The funding supports the professional production of the podcasts and videos, and the medical device companies that support us do not have any creative influence over the content that we record. All of our work is undertaken by volunteer healthcare staff, patients and their families. 

Most of our content is supported by videos. You can access our training videos and resources for Basic Care, Emergency Care and Vocalisation & Swallowing. Download and print bedhead signs and emergency algorithms from our resources. 

You can support our work by watching or clicking any of the advertising links that appear via the NTSP YouTube Channel. You can also donate directly to the NTSP through the NTSP website, or by clicking the Buzzsprout podcast hosting "support" links. You can support our work by watching or clicking any of the advertising links that appear via the NTSP YouTube Channel.

Transcript

SPEAKER_01 0:07

Hi, it's Brendan McGraw from the NTSP. Welcome to our in-depth look at the literature in the world of tracheostomy and laryngectomy that's caught my eye in March 2026. We've got next generation sequencing of tracheal aspirates, we've got three papers that cover diverse aspects of PDT per trache insertion, and finishing up with the power of AI. Is it any good at predicting who needs a trache? So let's dive in. The first paper is a pediatric airway paper looking at next generation sequencing, or NGS as they call it, in children with tracheostomies. So next generation sequencing of tracheal aspirates, a prospective case control study. That's the title of the paper. It comes from Pia Bresning, is the first author, and she is from the University Hospital Essen in Germany. Now, one of the biggest ongoing challenges, I think, in pediatric tracheostomy care is working out the difference between colonization of the respiratory tract, which is bugs that you just find there, and that the respiratory tract also includes any artificial airways there, and trying to work out whether the presence of bugs equals infection. Anyone who looks after children with tracheostomies will know this problem really well. So tracheal aspirates often grow bacteria, sometimes multiple bacteria, they grow yeasts, viruses, all sorts of different things. But trying to work out whether they're actually causing disease or they're simply just living there quite happily can be incredibly difficult. And in health we've got loads of stuff living quite happily in our airways, and so this distinction of is it friend or foe really matters because it directly influences antibiotic prescribing, hospital emissions, and whether we're going to escalate care up or down. So this study explored whether next generation sequencing could provide a clearer picture of the airway microbiology in these children. So the authors performed a prospective case control study on 45 children, 28 of them had tracheostomies, and 17 of the children were controls, just undergoing elective surgery, and they didn't have any particular airway problems or lung disease. So importantly, they looked to exclude children that had any evidence of acute infection, judged by clinical parameters and blood work, for example. And so what they were really looking at was the baseline presence of bacteria and other things in the patients rather than acute illness. So the tracheal aspirates are just samples where you take a sample of suction from the proximal trachea via the tracheostomy tube or via the opera way. And these aspirates were analyzed using three different methods. Standard bacterial culture, where you basically take the sample and you stick it on an agar plate and stick it somewhere warm and see if anything grows. Something called PCR testing, and that was done against a panel looking for respiratory organisms that are pretty common. And then finally the next generation sequencing using what's called cell-free DNA analysis. So this is a really good chance to take a moment to step out of this paper and get some help understanding these techniques. So I'm gonna go and have a chat with an expert. So here we are on the intensive care unit at Windsure, and I was looking for an expert in microbiology techniques, and I found James. So James, would you be good enough just to introduce yourself and explain uh what you do as a day job?

SPEAKER_00 3:31

Yeah, sure. Uh my name's James, uh James Orr. I'm I am an intensive care registrar and an anaesthetist as well, similar to Brendan, um, and I do have an interest in tracheostomies from a research perspective and from the perspective of being someone who helps put them in and looks after patients with both acute and long-term tracheostomies clinically. But also, when I'm wearing my research hat, I am currently doing a PhD in relationships between what we call the microbiome, which we use next generation sequencing techniques to analyse, in particular the lung microbiome, which is the bugs that live in your lungs and how they relate to how the immune system within your lung tissue is actually behaving in response to infections. So papers like this one are really relevant to what I do in that respect, and I thought this was really interesting.

SPEAKER_01 4:21

Yeah, so James is perfectly uh qualified to uh be our expert here and try and just untangle some of the uh sort of nuance of this paper. So you and I both read the paper and we both thought it was really interesting. And you know, for for me as a very much a job in clinician, you know, we we base our microbiology on stuff that we sort of scrape up from place in the body and then try and grow in the lab. So, what what from your perspective uh do we mean when we talk about next generation sequencing?

SPEAKER_00 4:51

So um it's it's an interesting term. I think the first thing for for us all to be aware of in the ITU world and the tracheostomy world is that um the next generation and next generation sequencing isn't necessarily talking about microbiology and diagnosis and infection. It's actually it actually harks back to when PCR uh was first developed in the 20th century. Uh, and this is the same technique that's used on forensic crime shows and things like that. And the and when we talk about next generation sequencing, it's actually the sequencing technology that's next generation, and it's the next generation from that more historic PCR-based targeted sequencing technology that we mainly use for human DNA in clinical practice, um, or for PCRs for viral screening. Next generation sequencing takes it to another level. What you do is you take all of the genetic material that's in a sample, and you can target some particular genes that you'd expect to be there, or you can just sequence efferfine. Um this was this one was one of this was one of the studies that's done efferfing, but just looking at DNA, so bacteria and DNA-based viruses. Um, this is quite different to what's existed before because it gives you an idea of every single microorganism and actually every single DNA-based human cell that's present in a sample. Okay. Um, and you can, depending on the exact technique you use, look at their relative abundance to each other. And you can uh at the moment we can only estimate in someone who's ill what what bug we've sequenced is likely to be the cause to the organism. This is completely different from culture-based techniques where we're depending on what will grow in an agar plate in a lab or in a blood plate in a lab, which as we all know isn't reliable. We we've all done sputum cultures, blood cultures, stool cultures on patients who we can see clinically are septic. They should have a septicemia. And either through bad luck or because the patient's been partially treated with antibiotics, we don't grow anything, which means that our cultures are not informative and we end up stuck on empirical treatment. Next generation sequencing by taking all of this genetic material in a sample, not just target genetic material and sequencing effrofen, potentially offers us a solution to this problem where we sometimes struggle to target antibiotics at not just empirically at the organism that's likely to be the problem, but also target any antimicrobial resistance genes that are being expressed either by the bug that's actually causing the infection or other ones that are hanging around. Because that's one of the other things that we have a problem with is we give someone a broad spectrum penicillin antibiotic for their chest infection, and they're someone who's been in and out of hospital, they've picked up and been colonized by some resistant bugs, and we don't realise they're there until we give a penicillin and we select for them. Right. Next generation sequencing potentially offers the opportunity to avoid that as well.

SPEAKER_02 7:59

Okay.

SPEAKER_00 7:59

Yeah.

SPEAKER_01 8:00

So how do you begin to try and work out what's what when you sequence these samples, which are full of loads of different things? How do you begin to try and make sense of those results?

SPEAKER_00 8:12

So that's the other side of the coin, because I think I've painted a very rosy and optimistic picture of of the potential of the technology, and it is still very much in its infancy. Okay. Which is why I don't I'm not aware of any hospitals that are routinely using this kind of sequencing as their diagnostic technique. There's lots of research happening, including here at Withinshore with the credit study, but um no one's using it routinely yet. And there's a few issues with it. The first is that you don't, unlike a culture where you put your put your put you put your crud on a plate that me or you has got on got on with a bronchoscope, or that, or that one of our physiotherapy colleagues, like Barbara, has managed to get through induced sputum. Um you put it on an agar plate, you grow, you grow it, and the microbiologist can often tell you what the different bugs are by looking at them macroscopically, not even under a microscope, because they look different on the plate. You get these lovely discrete populations, ideally. That is not how sequencing works. What you have is you have a mess of different lengths and strands of DNA sequencing or RNA sequencing. Um, you do not know what organism those sequences come from. You also don't know if they're human. Okay. Um, you can probably infer the latter, but you don't know it. So when you have this huge, once once you sequence this huge library of sequences from every single organism in the sample, you then have to rearrange them using very complicated software, which I would not claim to be an expert in yet. Okay. Hopefully I will be in a few years' time. But you have to rearrange all these sequences into what are reasonable and sensible suggestions of what bugs are there.

SPEAKER_02 9:58

Okay.

SPEAKER_00 9:58

And I think I've underplayed how complex it is. The sequences are based on the sequences that you compare your sample to to build to build this final data set. They come from bugs that you've grown in isolation and then sequenced. Okay. But there are issues with it. There are papers where that rear that rearranging of the material into a set of individual organisms that we think are present has been done incorrectly. And there's been major papers that have been retracted because that process hasn't been done correctly.

SPEAKER_01 10:27

Okay. So I'm getting the sense that you're trying to sort of make a jigsaw out of all these pieces, but you haven't necessarily got the instructions or the reference picture.

SPEAKER_00 10:35

I think that's the perfect way to put it, Brendan. That it is like it is like trying to do a uh a thousand-piece jigsaw without the picture on the front of the box.

SPEAKER_01 10:43

Okay. But with a bit of, well, I guess, either perseverance or uh clever software, you can actually build up a reasonable portion of that jigsaw and and that's the potential then to make it useful.

SPEAKER_00 10:54

Yeah, imagine it's your classic uh picture of a steam train jigsaw, you can sort of start to see that you're building a train. Yeah. And if it looks like a train at the end, that makes sense. Okay. But um obviously it's not quite that simple. And you can and if you're not doing it carefully and you're not when you're not using well-resourced, carefully made reference libraries to analyze your samples, you can make mistakes.

SPEAKER_02 11:18

Okay.

SPEAKER_00 11:18

Saying that, going back to this paper specifically, I think I think we'll go into this paper specifically in a bit more depth in a second, but one of the most important things to sense check the results of uh next generation sequencing is what have you grown? Does it make sense that you would find this in a healthy volunteer if that's who it's from, which is a lot of the work I do, or from a tracheostomized patient? And the results they got here, they're three main bugs, including uh multifilia, which was one of the ones they found in slightly higher proportions in the end in the next gen compared to culture. It makes sense. And I think that's the fundamental thing is that the results that they found here make sense. They haven't said that they've discovered that they're not trying to claim all sorts of obscure bugs that we've never seen clinically prior to NGS being normal are actually present in tracheostomised children. What they're saying is that three quite common bacteria can be identified more sensitively with this technique.

SPEAKER_01 12:13

Yeah.

SPEAKER_00 12:14

So I think it passes the basic sense check there. Yeah. Okay, yeah.

SPEAKER_01 12:17

Well, uh uh as I read that paper, you know, I was struck by just that that there's a load of stuff that we know grows in our patients all the time, and trying to make sense of A, is it there, B, is it going to cause us a problem, and then I guess C, then do we need to do something about it? I mean, how does these sort of uh newer techniques, I mean, how how how could that fit into say our everyday clinical practice? You know, if we've got a child out there with a tracheostomy and and we can take a sample, yeah, uh, we can do perhaps what you're doing. In your studies at the moment, where you we we do some standard techniques, but we also have the opportunity to perhaps try some of these newer techniques. I mean, how could that fit into clinical practice in there? Hopefully, near future and help us out.

SPEAKER_00 12:59

So I think looking at this population of tracheostomised children specifically, I think the way it could fit in for these for this population is you would have potentially an accurate, more reproducible and traditional culture and more rapid way of assessing exactly what they're colonized with. And that's a key thing that the paper mentions briefly, they talk very briefly at the start, they give actually quite a nice overview of bacterial ecology in in humans, which um which I think is necessary at the moment because this is a technique that is still in its infancy. The importance of that is these samples you with with traditional culture you don't always get reproducible growth, you don't always grow necessarily grow the bug that's actually the problem, and it takes longer. These samples with NGS next generation sequencing can be returned within one or two days as opposed to up to five. And if you use full in-depth sequencing, looking at individual gene expression, you can also tell what antimicrobial resistance genes or AMR genes as we call them are present. Which means if this child becomes unwell, you can immediately put them on the right antibiotic regimen. The second thing, going back to colonization and that introduction to bacterial ecology that the authors of this paper gave us, which I quite enjoyed, full of all Rin, these children are colonized with bacteria and often fungi as well, though they didn't do the right kind of sequencing to grow fungi, that's a whole other thing. It's much, it's even more in its infancy than bacterial sequencing. Healthy people aren't colonized with anything, or shouldn't be. To the point where there are people working in the lung microbiome field who feel that actually, if you discover a healthy volunteer appears to be colonized with something, because they volunteered that they're not really healthy, they've got bronchiectasis or something like that, and they just didn't know about it. Okay. These children are colonized with things, and that's the second thing. So we've talked about rapid results and targeted antibiotic therapy when they're unwell, but we also know that children with uh long-term tracheostomies are very likely to be abnormally colonised with various species, and you can monitor what's going on, and it doesn't mean that you're going to jump in and give them antibiotics just because they're growing this type of staff or that type of strep, but it means that when they do become unwell, you can immediately step in and give them the right treatment. Okay. And also you can also spot concerning patterns. So if a child comes in and then the acute admission screening actually shows a completely different bug, then you know that you might have a superimposed infection of a of a bacterium that's not part of that child's normal microbiome, which might be more concerning.

SPEAKER_01 15:56

And so, based on the sort of turnaround times, then at the moment, it it sounds like this is more about understanding what's in the child at baseline or you know, and the community is a surveillance tool. Um, but it sounds like there's promise as well, if the child does become acutely unwell, you can use the same technique to to say, look, something new is is is happening here. Yes, exactly.

SPEAKER_00 16:19

What is that sort of turnaround time at the moment then? At the moment, I would say um with the techniques that are available, you're looking at 48 to 72 hours.

SPEAKER_02 16:28

Okay.

SPEAKER_00 16:28

There are this isn't the device that they used in this study, they use the key agent device here, but there is the Oxford Nanopore, the current generation of the Oxford Nanopore device is potentially promising real-time readouts for sequences as they're derived.

SPEAKER_01 16:42

Okay.

SPEAKER_00 16:43

But obviously, turnarounds as short as 12 hours in that situation. But I would hasten to add that that's very novel technology that probably needs to be used in observational studies in both children and adults repeatedly before we can start saying confidently that um that it's accurate enough to use clinically. Not that I'm trying to criticize them, I think it's it's great stuff, but I'm sure the people developing that technology would agree that it's it's in its that is very much in the early stages of development. More established techniques like the Kiagen one that they've used here for double-stranded DNA or the various short read techniques, or what we call taxonomy, which is targeting bacterial RNA in the mic in uh bacterial mitochondria, they are more accurate. I think their accuracy is established, but they don't offer the rapid turnaround that would make this truly game-changing. Yeah. But I would say 48 to 72 hours is still considerably faster than five days waiting for an agar culture to come back.

SPEAKER_01 17:46

Okay. And how far off do you think the sort of next generation techniques are from sort of being, I guess, more routinely available than me did? It sounds like to be truly routinely available is probably quite a stretch off, but you know, for someone turning up to hospital who's unwell, do you get a sense for how far off we are?

SPEAKER_00 18:08

I think my optimistic prediction, um, looking at with a bit of knowledge of what studies are being planned by different groups, including our own, I think we'll be looking at moving towards observational trials of using next generation sequencing, not just to observe and compare to what traditional microbiology has grown in in ill patients, which is what we do here at Withenthrott at the moment. I think the initial studies looking at it as as a diagnostic tool and actually using it for diagnosis rather than just comparison after the treating team have done whatever they were going to do anyway. I think those are probably going to start happening within four to five years.

SPEAKER_02 18:51

Okay.

SPEAKER_00 18:52

And I think if those studies demonstrate that it is safe, uh, that it can act as a replacement's probably the wrong word, because I don't think we're going to replace culture, but that it can actually be produce results that you can act on. I think you'll probably see quite rapidly after those papers are done that it becomes an adopted technique in larger teaching hospitals, probably confined to developed countries. We're probably still five to ten years away from it being a widespread technique in the same way that traditional culture-based methods are what are widespread techniques. I think we're still a long way off that. And I think it's not just whether science can establish that it's reliable, it's infrastructure and cost. Not all microbiologists are trained in sequencing. Most research microbiologists can do both, but not necessarily all hospital diagnostic microbiologists are trained in next generation sequencing. And even if they did do it because they did a master's in it or a PhD in it, where they did it, or they did it through their undergraduate training, they don't necessarily have the equipment in their hospital to do it. And that's probably the big barrier. Because you can teach people who are already scientists to provide these techniques, but you need a £30,000 sequencer to do it. Yeah. And that's probably the barrier, the cost will be a barrier. Yeah. Because I expect that understandably the capital investment that's needed to provide this technique, we will need a lot of really good evidence that it that it works before uh the purse strings are loosened. And I think that that is fair enough. Yeah.

SPEAKER_01 20:30

Okay. Yeah. Well, James, thank you for lifting the lid on this uh paper and these sort of promising uh complicated techniques underneath. You've done a great job of explaining it. Thank you very much. Hope so, thank you. So we'll dive back into some more of uh the papers that have caught my eye this month. So thanks, James. Thank you. Okay, so I hope that helped. Back to the paper. So in this study, as you would expect, NGS detected far more organisms than either culture or PCR alone. That included additional bacteria, viruses, and fungi that conventional methods simply missed. One of the headline findings was that Tracheostomana children had significantly higher colonization rates with Pseudomonas, Staphylococcus, and Stenotrophomonus, and another bug called Moroxella. So Pseudomonas, in particular, stood out. It was identified as a dominant organism in almost half of the positive samples. That probably won't shock anyone working in long-term airway management because Pseudomonas has long been recognised as a common and difficult organism in chronic airway disease and tracheostomy populations. What was interesting here was the ability of NGS to identify not just which organisms were present, but which organisms appeared dominant within this complex polymicrobial colonisation. And potentially that's clinically useful. Because as we've learned, one of the big problems at standard cultures is they often produce a long list of organisms without helping clinicians determine which, if any of them, Actually, driving symptoms. The study also confirms something we've seen in previous microbiome work that tracheostomized children often have reduced microbial diversity. Now, when we're talking about alpha diversity, we essentially mean how varied the microbial environment is within an individual airway. So lower diversity is generally thought to represent a less healthy, more disrupted microbial sort of ecosystem. And that's often associated with chronic inflammation, repeated antibiotics, and airways disease. So while diversity was lower overall, the actual microbial patterns varied hugely between patients. So there wasn't just a single tracheostomy microbiome that you could identify. Each child appeared to develop their own distinct colonization profiles. And clinically that makes sense. I mean these children have all had very different underlying diagnoses, they've got different airway anatomy, ventilatory requirements, antibiotic exposure histories, and different interactions with healthcare systems, which we know can change your fauna and flora quite radically. But the paper also highlights that one organism in particular that is difficult to identify using conventional microbiology, which is stenotrophimonis, and specifically a subspecies called multifilia. So NCS has substantially more ability, it's a lot more sensitive to detect this organism in the routine culture. And that's important because stenotrophomonis is a problem, and we see that in our clinical practice and the ICUs that I work in, and it's an increasingly recognised bug in chronic airways disease populations. So despite all of this sort of positive news, the authors are very careful not to oversell the technology because detecting organisms is not the same as diagnosing infection. One of the major challenges with these highly sensitive sequencing techniques is that they may identify lots of organisms that are simply part of normal upper airway fauna and flora, or palmless colonization. This study really enforces that the clinical context is absolutely essential. You still need to assess the patient's symptoms, inflammatory markers, blood works, oxygen requirements, radiology, and overall trajectory to get a sense as to whether these bacteria being present is actually a problem or not. So NGS is not a magic answer that suddenly tells you whether you need antibiotics or not. And there are some practical important limitations here. NGS is relatively expensive, it's technically complex, and it's slower than some standard tests in many healthcare systems. So it doesn't provide a direct antibiotic resistance information either, which conventional culture still does better where you grow the bug and you grow it in the presence of different antibiotics to see which one has the effect. The study itself was also relatively small, single centre, the groups were certainly not perfectly matched or balanced in any way, but it does give us an interesting glimpse into where airway microbiology is probably going to be heading. So rather than simply ask what's growing, we can ask sort of more specific, more precise questions about this microbial ecosystems, what's the dominant pathogen, and are there any patient-specific colonization patterns that have occurred or that have changed over time, which give us more of a clue as to what might be going on. So the authors conclude, I think, very reasonably that next generation sequencing is a useful supplementary tool for analysing tracheal aspirates in children with tracheostomies, and it's got the potential to improve pathogen identification and maybe eventually reduce unnecessary broad spectrum antibiotic use. They also emphasise that much larger studies are still needed before this becomes routine clinical practice. So the take-home message here is that as diagnostic technologies become more sensitive, the real challenge really shifts from detecting information to interpreting what that information actually means. The second paper this month is, I think an interesting, engineering and innovation paper looking at brand new PDT device called a trackie pen. This is written by Wang Tang, who's an engineering student working with my excellent engineering colleagues here at the University of Manchester. So this paper was published in PLOS 1. Now, listeners will know that although PDT is fairly routine practice in many intensive care units and some operating theatres, the procedure is still quite mechanically complex. And when you break it down, there's multiple sequential steps, multiple instruments, and quite a lot of force involved in actually puncturing and dilating the tissues of the neck. And that matters because procedural complications like bleeding, false passage creation, posterior tracheal injury or loss of control during the dilatation are often related to instrument movement or the transmission of force, and if you're not pushing in the right direction, you can suddenly cause lots of problems. From an engineering and robotics perspective, this paper asks a fairly simple question. Could we simplify PDT by combining puncture and dilatation into a single device? The result was a prototype engineering design called the trackie pen. And essentially what this does is combines the needle puncture system and the dilator into one compact instrument that can be used for either manual or future robotic assisted PDT. The device contains the retractable needle, which is inspired by a ballpoint pen mechanism, alongside expandable dilator arms. So instead of multiple device exchanges into the hole that the operator creates, what the tracky pen does is it causes the needle puncture, then it retracts the needle safely and then expands the dilator arm to create the stoma. So in theory that reduces the number of procedural steps, the amount of instruments involved, therefore the amount of instrument movement, and potentially could control the forces transmitted into the airway. The authors tested this device on portine or pig tissue models and compared it with standard Blue Rhino dilator, which is probably the most commonly used PET system internationally. I thought the results were quite interesting, although clearly I've got a vested interest in this. So Trachee Pen generated significantly lower insertion forces than the Blue Rhino system. The average insertion force around 53 newtons, which compares a 69 newtons for the conventional dilator. So to put that into context, that's a difference of about 1.5 to 2 kilograms of weight applied to the neck during the procedure. So if you're in the UK, our bags of sugar are usually a kilogram. So if you pile two bags of sugar on top of your hand, that's a sort of force that you're talking about that you could save if you use this mechanical technique compared to a manual technique. In practical terms, lower insertion force potentially means less tissue trauma, it's less uncontrolled advancement of the needle and the dilators, and the theory is that that may cause fewer procedural complications. The trache pen also created overall a smaller stoma size, which could reduce tissue disruption and bleeding risk, although that would need proper clinical evaluation in humans rather than isolated bits of pig. One of the particularly interesting bits of this paper, I think, is its focus on robotic assisted tracheostomy. So this device was specifically designed to work with robotic systems, and the authors actually tested it using a robotic manipulator arm. So robotic air procedures still sound fairly futuristic at the moment, but there's plenty of robotic assisted airway surgery that occurs right now. And I think there's an interest in robotic assistance for these very precise procedural tasks, particularly where force control and movement accuracy are really important. So if you think about PDT mechanically, it probably does lend itself reasonably well to robotic support in the future. That said, the authors are very realistic about the challenges. One of the biggest issues is the positional guidance. The robot still needs to know exactly where the trachea is, where the blood vessels are, how much force is safe during puncture and dilatation. And the tracheal target, as anyone has ever done it, per trache knows, is often pulsatile, it moves with respiration, it compresses when you squeeze it, it moves when you apply pressure, and its position varies significantly between people and even up and down the neck. The depth of the trachea relative to the skin changes quite a lot. So future systems are likely to need integration with maybe ultrasound, bronchoscopy, force sensors or other imaging technologies to help tell the robot essentially where to stick the tracheostomy. This study also included a feedback workshop with experienced clinicians, which I think adds to its validity. The clinicians generally felt that the concept was feasible and potentially useful, particularly because it might reduce axial force during dilatation and it can simplify one-handed operation when you're doing the PDT. But the clinicians also highlighted practical concerns around sterilization, force feedback, and the importance of actually getting the needle in the right place, so development of accurate guidance systems. One of the recurring lessons in airway innovation is that clever engineering by itself is never enough. Devices have to work reliably in the messy, high-pressure, real-world ICU environments. So with this paper, there's obviously some make limitations. It wasn't a clinical study, it was an early stage engineering and feasibility project using porcide tissue models. Human neck anatomy is much more variable and much more complex, with tissue thickness, obesity, calcified tracheal rings, abnormal anatomy, and bleeding risks, all creating challenges that the lab simply cannot reproduce. But importantly, I think this study tells us nothing about clinical outcomes, complication rates, or actual patient safety. But despite that, I think it's a really interesting proof-of-concept paper, because I think it reflects a broader shift that we're increasingly seeing in procedural medicine. We're combining engineering, robotics, sensing technologies, and clinical practice to make invasive procedures simpler, safer, and more reproducible. So the authors conclude that TrachePen demonstrates the feasibility of combining puncture and dilatation into a single PDT device, and that's got the potential for future applications in robotic-assisted tracheostomy. So while there's clearly a long way to go before anything like this reaches routine clinical practice, it's an interesting glimpse into the future to see how airway procedure design continues to evolve. So that's it for this episode. We've cut a lot there, we've gone through tracheal aspirates, we've looked at PDT insertion, and we've covered a few other things in between. Please follow us on our various platforms and social media and feel free to comment on the discussions. As ever, the views are mine and don't represent any of the organisations for which I work. So that was Tracky Talk for this month, edited and produced by Simon Williams. Thanks for listening and look forward to seeing you next time. Bye for now.