Hello, this is Dave Callens. I'm a professor of medicine at Perlman School of Medicine at the University of Pennsylvania. Today I'm speaking about complications of EP procedures, avoidance, and management. These are my disclosures. It's important to recognize that complications are a part of all EP procedures. I love this study. It's a very old study. Back before EP was interventional, back before device implants were performed in this particular lab, back in the 80s, they looked at complications of EP procedures in a thousand patients. And despite the fact that no one was doing anything all that complicated, death, vascular injury, systemic arterial embolism, pulmonary embolus, cardiac perforation, were still part of this experience. So complications happen and it's important to realize that doesn't make us bad people. We should still strive to make that happen as little as possible. And that's part of what this lecture is about. The incidence of procedures changes over time, particularly getting worse because we're doing more and more intense interventions on patients who are sicker over time. There's an increased incidence of complication with procedural complexity, with severity of patient illness or comorbidity. And so those parts are not, you know, not something that we can do something about easily. The other part, increase in incidence with inexperience, both of the operator and also of the team is something that can be improved with increased experience. There's a marked variation in complication type and incidence based on procedure type. Prompt recognition and treatment is vital. And we'll go over this in slides to follow, but there is kind of a signal in studies of procedural complications that the complications actually lead to a downhill spiral in the trajectory of patients who have them. So just to make the point that complications are very important and must be managed very effectively. This is a very nice study from the University of Virginia, a prospective study of 4,300 procedures performed in a university hospital from 2010 to 2012. Major complications were observed in this experience in 3.6% of procedures. Importantly, only about half of complications were discovered during the hospitalization. Half of them were discovered after discharge, making the point that, you know, when you do a procedure on someone, the responsibility continues onward in time, even after the procedure is over. The 30-day mortality was 0.6%. There were 24 deaths in the experience and five were directly procedural. Most of the rest of them were related to worsening of comorbid conditions. When we look at this as a breakdown of which EP procedures had how many complications, there's a general sense that as things get more complicated, the number of complications increase. So on the left hand of the screen, during ablation procedures, SVT procedures are the simplest and they're the least complicated. This goes upward to AF procedures, to endocardial VT procedures and to epicardial procedures with kind of a huge increase in incidence in this experience. On the right side of the screen, the same thing is seen with device procedures. Although they're kind of an unexpectedly high incidence of pacemaker complications, for the most part, simple things have less complication than more difficult things such as lead extractions. So you can use this to kind of see where your complication comfort zone really works and not do more complicated procedures on sicker patients, unless that's what you're used to doing. So I wanted to have an overview and I'm going to break this down kind of artificially, procedure by procedure, what complications are involved. So complications that can be expected in all procedures are problems with vascular access, hematoma, retroperitoneal hematoma, AV fistulas or pseudoaneurysms, pneumothorax for device procedures, DVT or PE. There are catheter or lead related complications, perforation and tamponade, thromboembolic complications, damage to surrounding structures. In all procedures, the potential for infection, for radiation exposure, or for anesthesia related complications, all factored in. I'm not going to talk so much about these last three in this lecture. So complications related to vascular access. There was a very nice study done by Parikh Sharma that demonstrated, mainly looked at, was factors that increase the likelihood of complications. So complications for vascular access are more likely to occur in female sex of the patient, advanced age or obesity, patients with peripheral vascular disease. It's more likely to have vascular access complications with arterial versus venous catheterization. Complications increase with the number and size of the sheets that are implanted and in the presence of anticoagulant or antiplatelet agents. This study, at least for me, really put vascular access in the context of the vascular ultrasound on the map. So in the video, you can see there's a wire in the vein and existing and we're imaging as the needle perforates the anterior wall of the vein and allows for access for the second sheet to go in. This greatly decreases vascular complications to image what we're actually doing with ultrasound. Also fastidious management of the access site after withdrawal of sheets cuts down a great deal on pseudo-aneurysm formation. There is no clear effect of vascular closing devices or figure of eight sutures, which we happen to use in our lab more for patient convenience and safety. So in this study, it was a prospective non-randomized sequential design. 360 patients who had kind of the control contemporary use for access, 360 patients who had vascular ultrasound guided access. The major complications of access related type in the non-ultrasound group occurred in 1% of SVT procedures, 1.8% of AF procedures, and 2.2% of VT ablation procedures. And as you can see in the table, there's a marked reduction in vascular access complications going from the non-ultrasound to the ultrasound group. So all complications occurred in 5.3% of the non-ultrasound group and 1% of the ultrasound group. The use of ultrasound reduced incidence of access complications was fairly spectacular. There was also a reduction in access complications in studies that focused on just access for atrial fibrillation ablation, particularly in reduction of important bleeding. Pericardial effusion and tamponade. So there are a number of causes. And often when you're deconstructing, when you have a complication, it's difficult to know what part of the procedure actually led to pericardial effusion or tamponade. So frequent causes are catheters that are placed in the right ventricular apex. It's a very thin wall of the heart. And often we forget about that even being there. And that can lead to perforation. There can be damage during the transeptal puncture, particularly the posterior wall or the left atrial appendage. Any degree of mapping with an ablation catheter in thin walled areas, such as the RV free wall, the left atrial appendage, again, the roof of the left atrium, the area around the right superior pulmonary vein. That's tricky because of pushing the catheter up. And often it gets stuck in these recesses that you can think of the right superior pulmonary vein. They're actually just a recess in the septal portion of the right superior pulmonary vein itself. And if you push, you will go through the wall there. Myocardial rupture can occur secondary to steam pop during ablation, which is another cause for pericardial effusion. Pericardial effusion and tamponade are exacerbated by anticoagulation, which is obviously necessary for many of the procedures that we do and probably worse in patients with advanced age. I've been doing this a while. So remember what pericardial effusion and developing tamponade used to look like in the quote old days. So typically these patients were lightly sedated and they would start to get a little bit rammy. They'd be restless and move about the table. And sometimes I would yell at them for moving and they'd complain of vague chest discomfort, which I usually realized only too late what they were really trying to complain about. Then there was a slow reduction in blood pressure over the course of the next 30 minutes to an hour when suddenly then there was a dramatic reduction in blood pressure. And even worse, if the heart rate goes down, if the sinus rate goes down, that's a sign of impending doom. We reacted with pressors. We did a chest wall echo to confirm the problem. We waited for an interventional colleague to come for a pericardial synthesis. Sometimes in that waiting period, there would be CPR being performed. So this is the wrong way. What we do now is ice imaging to allow prompt detection of pericardial effusion. And a lot of times, and again, this is the pericardial effusion often forms in the most dependent portions of the heart. So posterior to the left ventricle is the most dependent. And if you can see this pericardial effusion, then you can stop what you're doing and reverse interprocedural anticoagulation. And usually that's enough for the bleeding to stop. When we discontinue anticoagulants, usually that's enough to get out of this problem. I personally think that electrophysiologists should be skilled in pericardial synthesis. This is not required in the HRS ACC 2017 cardiac EP lifelong learning statement. But I think between being able to deal with this complication very promptly, as well as for allowing epicardial access for PT ablation, I think everyone should know how to do this. This is a nice slow guidance in Yamana and K catheter ablation of cardiac arrhythmias. So starting with recognition of cardiac tamponade, cascades of medical intervention on the left, procedural intervention on the right, procedural interventions beginning with pericardial synthesis. I think the only thing that I would add to this is that as soon as you recognize that there's an important pericardial effusion, I think CT surgery backup, at least knowing that there's a problem and alerting them to this is an important part of the cascade. And then subsequent to the publication of this, we've also gotten reversal agents for non-vitamin K dependent anticoagulants. And this should really be part of this flow diagram as well. And knowing how to actually obtain these things when you need them in the context of your own hospital is very important. Okay. So I'm going to move on to specific procedural types. And it's not like individual complications can only occur within individual procedure types, but there are some that tend to congregate more with certain procedures. Starting with SVT, AV block is a very important complication of SVT procedures. Certainly not as dramatic as things like cardiac tamponade, but AV block in a young, healthy patient is certainly life-changing. This is a picture on the right of the anatomy that's pertinent to slow pathway or mid-septal bypass tract ablation. When I saw this the first time, I thought, this can't be true. Someone must be joking. There's such a big separation in distance between the trajectory of the histonal catheter and the coronary sinus catheter. Do you mean to tell me that if I put my catheter at that star on this picture, I would cause heart block? And how close that is to the ceiling of the coronary sinus. But that is the reality of everything. So the ceiling of the coronary sinus is really an absolute barrier of how high you can place the ablation catheter. Any higher than that has a very real chance of causing heart block. So this should happen extraordinarily unusually. There was a expert registry experience of AV node modification for AV node reentry by Katrizas that showed a 0.1% chance of this happening with a slow pathway modification. And that feels like that's a target that we really should be aiming for. It's also observed with mid-septal accessory pathway ablation with BT ablation of septal BTs. The danger increases significantly above the ceiling of the coronary sinus ostium. Possible solutions when that neighborhood gets very constricted, when there's very little room between the ceiling and the coronary sinus and the histonal recording catheter, cryoablation, not because cryo is inherently safer than heat energy, but because cryo allows for cryoadherence. So at least the catheter is not going to move once you start ablation. We've used jet ventilation again to prevent breathing motion from destabilizing the ablation catheter. And for at least for accessory pathway ablation, positioning the ablation catheter in a very ventricular position so that there's no atrial recording on the unipolar tip electrogram that would allow for right bundle branch ablation, but not AV node ablation. This is a very old picture of RAO, LAO catheter assessment for AV node modification. And the point of this is that the ablation catheter is for slow pathway modification, again, cannot cross the vertical plane of the coronary sinus ceiling. And it should be under the hist catheter in both orientations. So the point of this, we used to put CS catheters in through a superior approach, and it's actually better to define the ceiling of the coronary sinus. And it's actually a lot more comfortable for the patients to put the coronary sinus catheter in from an inferior approach, because then the elbow of the catheter kind of pushes up against the ceiling. And it's very easy to understand this. In the old days, and this is also sometimes helpful, placing the ablation catheter within the coronary sinus and then flipping it out and understand where that rotation is can also be very helpful to appreciate this anatomy. AF ablation complications, again, we would expect more of these because there are more complications than there is, because they're relatively long procedures. They have multiple vascular sheaths, dual anticoagulation, and the requirement for transeptal puncture, which has its own hazards. We expect that pericardial effusion tamponade happens in series between one to 2%, probably should be less than 1% currently. Axis complications are more frequent because of the multiple sheaths and axis on anticoagulation. Thromboembolic complications are less than 1%. But there's also this thing of asymptomatic cerebral MLI that can be recognized by MRI imaging of the brain, but the impact of them is unclear. It would probably be best to have less of these rather than more of these and different ablation technologies have more or less of these. There are some relatively unique AF ablation complications. Air embolism, all those can be seen with transeptal no matter what the procedural goal is. Pomeranian stenosis is certainly very specific to this. Phrenic nerve injury, atrial esophageal fistula, and proarrhythmia, and by this I mean atypical flutter. And this kind of cuts both ways, but we really think that ablation encourages atypical flutter to happen, particularly because it's been noted that the more atrial substrate that's ablated, the more likely the incidence of atypical flutter will follow. There've been two voluntary multicenter registries. This is the second one. The registry was collected over 2003 to 2006, so it's already pretty dated and published in 2010. This is one way of looking at kind of a benchmark for what ablation complications should look like. And again, I'll make this point again a little bit later, but you should aim to measure your own frequency of complications and really judge yourself relative to these benchmarks. So if we look at the table, things like death, tamponade, death of about one in 1000 is how this is usually related. Tamponade, 1.3%, arterial complications, about again, 1.3%, and stroker TIA at 1%. So major complications were noted in 4.5%. There was no improvement from prior survey. This will be a feature that we often look at over time. So there are limitations to voluntary registries. They probably significantly under-report complications. And they're also kind of over-represented with expert programs. So this is really a high benchmark to aim for, but I think we all should strive for that. This is a relatively contemporary look at AF complications in a Japanese study, which was a insurance database registry study of over 135,000 patients with atrial fibrillation, fibrillation from 2012 to 2018. They noticed major complications of 3.4% with a rate of tamponade of 1.2%. It was a very low in-hospital mortality. The point of the graph is that this study and other studies like it have noticed an increased complication rate with advancing age and with the development of comorbidities. So comorbidities and age certainly ran together in this experience and most experience, but overall complications for patients less than 60 was 2.5%. And for those greater than or equal to 85% rose to 6.8%. So again, you have to know your comfort level and what patients you're willing to take on for atrial fibrillation ablation. There's more complications in older, sicker patients. That makes perfect sense. This is a story that I continually am shocked at no matter how many times I review it. This was a shot heard around the world and then kind of promptly forgotten and certainly nothing was done to improve upon this. So this was a study looking at the national inpatient sample. 20% of hospitals throughout the United States were sampled to obtain 93,000 AF patients treated with catheter ablation from 2000 to 2010. In-hospital complications were assessed by ICD-9 codes. Importantly and astonishingly, 81% of ablation procedures in the sample were performed by operators who perform less than 25 AF ablations per year. The in-hospital mortality rate was 0.46%. Again, compare this to registry data of one in a thousand, compare this to the Japanese study that I just showed you. Acute complications at 6.3%. And importantly, inversely related, the commercial is inversely related to operator experience and program volume. So this is what this looked like. It's the same for program volume, but looking at operator volume. Complication rate was nearly 7% in operators that did less than 25 procedures per year and dropped to under 2% in operators that did many more procedures. This was kind of emphasized in a little bit different sense from a very nice Cornell study of over 60,000 patients, again, taken from a nationwide readmissions database insurance product. This was performed from 2010 to 2015. There were major complications in 6.7%. Early mortality was 0.46%. Again, pretty similar to the prior experience. The quarterly risk of mortality actually increased over observation time. So that risk went from 0.25 to 1.35%. This was blamed on not so much experience of the operators, but instead that patients who were referred for AF ablation became sicker and had more comorbidities over this period. The factors that were associated with complications and with early mortality were procedural complications with an odds ratio of four, heart failure, or low volume hospitals. So again, pretty much the same message as the Deshmukh study. And it's kind of incredible that so many of these procedures are done by operators that are not very familiar with the technique. I've put together, and these are all just my opinions, my reflections about best practices for preventing complications. Here, looking at AF complications. So to prevent procedural stroke, I think that continuous oral anticoagulation is really important and helps to cut down on both stroke and also vascular complications, even though it doesn't seem like that would work out that way. It's much better than bridging was, which is what we used to do when we were using cumulative anticoagulation. I think that procedural anticoagulation for an ACT of greater than or equal to 350 seconds is important. It's not clear, it's been wondered about if irrigated ablation cuts down on the incidence of stroke, but this is performed in most labs in any case. To prevent pericardial effusion, I think that ice-guided transeptal puncture is important. I think the detailed real-time understanding of cardiac anatomy, either from pre-procedural imaging or ice guidance is really important when you're trying to figure out where your catheter's going at any particular second. I think that force-sensing catheters probably improve safety against perforation. Careful ablation power titration. We use impedance drop to titrate power and it prevents kind of overpowering and may decrease the risk of steam pop. And finally, prompt detection can make pericardial effusion much easier to treat and make it not be kind of a downhill spiral. So this is ice guidance, a transeptal procedure. You can see how simple that was. We can look and make sure that the transeptal apparatus goes cleanly over the fossil ballast and into the left atrium. And there's lots of room around it for it to be guided into and not encounter any of the walls of the left atrium or the left atrial appendage. Air embolism is related to air entrapment in sheaths either at the initial placement or due to suction when catheters are removed and exchanged. Air embolism can cause RCA ischemia or infarction, shock, neurologic symptoms. It's prevented by meticulous sheath management. If it does happen, it's pretty easy to recognize and usually transient. But if RCA ischemia or neurologic symptoms continue, treatment is high intensity supplemental oxygen, placing patients in the Trendelenburg position and sometimes even hyperbaric oxygen. RCA essentially retired as a potential complication, but it still occurs. I think it's kind of a double-hit hypothesis. It's caused by ablation within the tubular portion of the vein, but then probably some difficulty that that specific patient has with healing. It's prevented entirely by understanding the PV anatomy. And again, that's largely an imaging thing, whether you do this imaging job with ice imaging or pre-procedural imaging with integration to the MAPIC system. Either way is fine. It's just, you really have to understand where the pulmonary veins are. Treatment is anticoagulation to prevent the impact of subsequent thrombosis, further narrowing the stenosis. PV stenting can be performed and time, so over time, collateralization will help with this. In the early days, we had several patients with PV stenosis and that condition is much more symptomatic than even the most symptomatic of atrial fibrillation. Atrial soft gial fistula, very rare. We think that the best estimates of this are somewhere around one in a thousand, but it's a little bit murky. They can be prevented by understanding the soft gial location, again, based on pre-procedural imaging or real-time ice imaging. The role of continuous intraluminal temperature monitoring is unclear. We use this in our lab because we think it at least prevents stacking of lesions. So if you have high temperature after one lesion, then you know about it. At least you would not ablate in that same territory immediately thereafter. I think there's an insistence on lower power or lower duration ablation for posterior wall sites. The role of protein pump inhibitors after procedure, active movement of the esophagus, cooling of the esophagus is all yet to be determined. Phrenic nerve injury, the right phrenic nerve is near the septal aspect of the right superior pulmonary vein in its carina. It's encountered very frequently with catheter ablation and atrial fibrillation, in particular with balloon-based ablation procedures. We prevent phrenic nerve injury by pacing at each site prior to ablation with rate of frequency ablation. With a balloon-based ablation, we pace the phrenic continuously from a catheter that's placed in the SVC or in the right subclavian. And if you notice any diminution in the pulse of the phrenic pacing, then balloon ablation has to be interrupted immediately. Proarrhythmia, again, talking about atypical atrial flutter. I think that atypical flutter is enhanced by gaps in linear lesions. And even though acutely, it seems like these are very easy to form, very often after healing, there are gaps in linear lesions. I personally think that it makes sense to reject empiric linear lesions. If they're done, it's essential to verify completeness. Moving to VT ablation complications. This is an interesting study performed by the Mayo Clinic. It was a meta-analysis of the administrative database of over 14,000 patients involving clinical trials, randomized controlled trials, and observational trials. And this is what this looked like. They found major complications in eight to 10%, but as can be seen in the bar graph, the major complications varied greatly depending on the source of the information. So randomized controlled trials had much less complications than non-randomized trials than administrative claim studies. So this can cut both ways. So certainly randomized trials are kind of a euphemism for expert operators, whereas administrative claim studies are due to the general population of operators. But there's also a failure to report in single center studies, and it's impossible not to report in administrative databases. So again, when you're working on benchmarking, it's kind of hard to know where to aim. There's no one perfect truth for this. Nonetheless, complications, again, varied considerably by data sources we just talked about. I think this really reflects mostly operator experience more than anything else. So another study using the National Inpatient Sample from 2002 to 2011, 4,600 patients with ablation procedures for VT in the setting of healed infarction. These investigators noted a total complication rate of 11.2%, vascular complications in nearly 7%, cardiac complications, mostly referring to pericardial effusion or cardiac tamponade in 4%, and neurologic with stroke or TIA in 0.5%. They noted an in-hospital mortality of VT ablation of 1.6%. Interestingly, again, there were no change in complication or mortality over time. What I think that this means is maybe better experience facing a worsened sample of sicker and sicker patients over the course of the experience. So VT ablation complications are more frequent because the procedure is much more complicated and the patients are much sicker than in other catheter ablations that we perform. There are long procedures. There are multiple sheets. Again, often this involves arterial catheterization, exposing sick patients to anesthesia is a big deal. Sometimes patients with heart failure are holding on to their blood pressure with just that last bit of sympathetic stimulation that they can muster. And when anesthesia, particularly general anesthesia, wipes that out, hypotension and downhill spiral can ensue. There are often VT induction in VT ablation procedures and then often cardioversion shocks that follow, both which cause a stunning effect on hemodynamic tolerance. There are added complications with VT ablations that include epicardial axis and ablation. There are several, again, not quite unique, but VT ablation complications that I'll focus on. Acute hemodynamic decompensation, and we'll define that in the slides to follow. Cardiac tamponade is higher in VT ablation than in other procedures. A benchmark is somewhere around 1.8%, but this is higher with epicardial axis procedures. Monocardial infarction, coronary injury, refractory VT, and cardiogenic shock. This is work done by my partner, Pasquale Santangeli. When he looked at this concept of acute hemodynamic decompensation during VT ablation. And what he meant by that is that sometimes during VT ablation, even though patients are fine at the outset, eventually they develop so much hypotension that they either require a marked increase in pressor agents to continue, or hemodynamic support with devices, or cessation of the VT ablation. So he looked at risk factors of this, and very little of it had to do with the procedure itself. Most of it had to do with the comorbidities that the patients brought to the lab with them. Things like COPD, advanced heart failure, presentation in VT storm, LVEF less than 25%, ischemic cardiomyopathy. Those things are not negotiable. We also found that general anesthesia, probably for the reasons we talked about on the last slide, also increased the risk of this acute hemodynamic decomposition. When Pasquale arranged this into turtiles of risk by this risk score, he noted that in advancing risk turtiles, the incidence of hemodynamic compromise went up to a height of 44% of procedures in the highest risk turtile. What was the meaning of this? So clinically, acute hemodynamic decompensation had an important effect on mortality. So in this graph of follow-up, freedom from death being measured on the Y-axis, patients without hemodynamic decompensation during the procedure did fairly well. 14% lived for mean follow-up of 20 months. But patients who had acute hemodynamic decompensation during the procedure had a mortality rate of 58% in 20 months. And when everything was adjusted, acute hemodynamic decompensation remained an independent variable related to mortality after the procedure. So again, this idea that procedural complications can downhill spiral into having important effects on the trajectory of patients' perspective health after the procedure. So when we looked at these results, we made some changes based on how dramatic that really looks. We approached patients who need VT ablation with setting the advanced heart failure with a heart team approach, which allows for optimal pre-procedural heart failure management, exploring the possibility for advanced heart failure treatment options if the procedure goes poorly. We also noticed that there's a big difference between prophylactic support, here mostly with ECMO versus rescue support. So we have a series of patients who required rescue support, either prior to the procedure because of VT storm and hemodynamic decompensation or within the procedure itself. Rescue support was attendant with an 81% 30-day mortality. So this is really a non-starter. The heart team approach and trying to plan ahead really makes a lot more sense. There are some procedural variables, even though like we just talked about, a lot of these things the patients bring with them. So avoidance of cardioversion shocks seems prudent. We wonder about the avoidance of general anesthesia, if that's at all possible for the procedure. Finally, in Pasquale's experience, it really took quite a while during the procedure for things to start to go wrong. And that was about somewhere between four and six hours of procedural duration. So if you can keep a procedure short, it exposes patients that are on the balance and ready to tip over. It keeps them safe. But an aside that I think we think about ablation as never being strong enough, and we're always kind of struggling to do more and to get deeper. I think it's really important to realize that irrigated radiofrequency ablation in scar or on top of papillary muscles is really difficult and we can never get enough ablation. I think this is because of both catheter stability and what that scar tissue is made of. However, ablation in normal myocardium is much more powerful than we give it credit for. So this is a procedure that we're guiding by ISIM and it wasn't my procedure, thankfully, but it was a procedure done at my institution where we're doing papillary muscle ablation for theopathic PBCs in a patient who's had bypass surgery. So you can see the position of the catheter right now. It's completely surrounded with muscle. The tip of the catheter has nowhere to go. All that heating is going to go into tissue because it's wedged between the inferior border of the posterior medial papillary muscle and the inferior wall, both of which were normal. Ablation here eventually resulted in this complication. So a big loculated pericardial effusion, loculated because of the prior bypass surgery rethink, and this was very difficult to deal with. So irrigated RF in normal myocardium is very powerful. We have to resist the kind of the impulse to keep going more and more and more and more. And also I think it helps to monitor impedance drop. A 10% impedance drop is more than enough, particularly ablating in normal tissue. In addition to normal VT ablation procedures, there are unique complications related to epicardial approach and ablation. There's mostly the need for general anesthesia. I've done several procedures without general anesthesia, but it takes a pretty stoic patient to get them through this. There's management of procedural anticoagulation. If you're anticoagulating for endocardial access to the left ventricle and then switching to epicardial access, you have to be really certain that you eliminate all that anticoagulation. The epicardium has an amazing amount of insulating epicardial fat, which can complicate measurement of voltage for voltage mapping. It also produces smaller RF lesions. There's the potential for coronary artery occlusion and damage, for phrenic nerve damage, for persistent pericardial bleeding, for pericarditis, and also increased risk of pericardial effusion. This really suffers from being an expert experience, but this was a multicenter retrospective analysis of complications of epicardial access in three very, very experienced centers. So this benchmark is very hard to meet. 156 epicardial procedures with eight complications, 5%. Most of these were epicardial bleeding that was eventually self-limiting. One was coronary stenosis. There were also three delayed complications. Pericardial inflammatory reaction, which can be prevented by steroid injection after leaving the pericardial space. Delayed effusion with tamponade, very frightening. Delayed coronary stenosis with acute myocardial infarction. There were 20 repeat procedures in this series, and all but one of these repeat access was uncomplicated. There are some ways to try to prevent epicardial complications. One of these is to make epicardial access less traumatic with a micropuncture needle. And this is kind of a telescoping thing that displays the micropuncture needle within a needle. This is helpful, but it makes things a little bit more complicated. Micropuncture needle admits a different wire than you would be used to if you were using the traditional tui needle. So it just makes sense to get comfortable with the kit that you're using before it's kind of game time and you're doing the procedure. The way that we really manage epicardial access is with management of respiration. So if patients are under general anesthetic, you can control their level of respiration. It's very important not to have just natural ventilatory breathing during the time that you're trying to do epicardial access. If you're in that space and then there's suddenly a forced ventilation from the mechanical ventilation, that can move the heart relative to the needle and cause a tear in the right ventricle. Coronary injury is another possibility with epicardial ablation. I think it's essential to do coronary angiography. And we've kind of had this weird 0.5 centimeter rule that you can ablate with at least a margin of half a centimeter away from the artery. And this has been quote, established for avoiding acute occlusion. I think this has often been like, I can do this with one centimeter. I can do this with half a centimeter and so on and so on. Almost like a bravado sorts of thing. I think it's important to have respect for the coronary arteries. And we've really kind of established this just for avoiding acute occlusion. We don't know if we ablate at 0.5 centimeters, what happens to the artery over time. It also kind of presumes that the catheters aren't moving and that the heart's not moving, neither of which are true. Phrenic injury, it's important to understand the left phrenic anatomy. And usually this is encountered with procedures of epicardial ablation for non-ischemic cardiomyopathy at the lateral base. It's important for pacing before ablation, just like we would for AF ablation outside of the right superior pulmonary rate. Last focus is on implantable device related complications. It's a very nice recent study called the Pointed Study, which was a prospective observational study in six high volume European centers of the incidence of an effective complication in de novo CIED implants. It was greater than three-year follow-up. 283 complications occurred, which was 10.5%. Again, the incidence of complications increased with the increasing of device implant complexity. Importantly, CIED implant complications are associated with an increased risk of CV mortality and early complications are associated with an increased risk of all-cause mortality. Just to again, make the point that complications can start a downhill slide for patients that are at risk. They're best avoided and at least best impacted on quickly. So this is a table of this experience. We see that device complications, including infection, hematoma, lead failure, all increase as the complexity of the device that's implanted goes up. There were both early and late complications in this experience, and all of these need to be dealt with. This is the central illustration from this paper, just demonstrating that complications caused less good outcomes than patients who did not have complications in blue over 96-month follow-up. This makes two points though, actually. There's kind of a dual message. One, some of this difference in outcome is the impact of acute complications, just as it was in the VT ablation story. But also to be factored in is the impact of patient comorbidities. So patients who have complications are more likely to be sicker, and that sickness also influences the trajectory of their plight towards cardiovascular death. Interestingly, these comorbidities factored very greatly for increased cardiovascular mortality as well. So complications did, but comorbidities did as well. And they're the usual suspects. Age, decreased EF, diabetes, atrial fibrillation, chronic renal failure with dialysis, COPD. Device infection as a late complication also had a very prominent impact on cardiovascular mortality. Another plug for vascular ultrasound access. So when we first started doing this, it was one of our fellows that kind of took the initiative to do this. And I thought it looked really foreign to have this object in our operative field, but now we do it on every single procedure. It gets away from so many potential problems. The idea of reaching the axillary vein before the central circulation very much cuts down on the possibility for pneumothorax. The clean venous stick and avoidance of the artery cuts down on bleeding quite a bit. So this really is an important feature in reducing complications for device procedures as well. Hematomas, as I've noted before, are a very big deal and particularly both for morbidity of the patient and for the development of infection after the hematoma. A very important trial was the bruise control study. And this looked at continued warfarin versus heparin bridging for implants in patients with a greater than 5% annual risk of stroke. It was actually terminated early for an increased clinically relevant hematoma in the bridging arm. So this really put bridging on the map as a really never do sort of thing. And now we continuously do all of our device procedures on people that can't safely stop anticoagulation on completely uninterrupted anticoagulation. The follow-up of this study showed two interesting things. They looked in bruise control two at continued DOAC versus discontinuation and for implant patients, again, with the CHA2DS2-VAS coordinator than or equal to two and it was terminated due to futility. But look at the incidence of complications. So in the bridging arm, it was 16% in bruise control one versus 3.5% in continued warfarin anticoagulation. It was 2.1% in both arms, meaning that continuing anticoagulation is perfectly safe and things are hopefully getting better over time as people get more experienced with this idea. Hematomas can lead to infection. Infection can occur spontaneously. The rapid trial using an absorbable antibiotic salooning envelope in generator replacement, pocket revision, upgrade to Bi-V or an initial Bi-V implant was a markedly positive trial. It looked at almost 7,000 patients. The primary endpoint of infection requiring extraction or revision, the need for long-term antibiotics or death at 12 months. You'd see that in high power devices or in generator replacements, there was a marked improvement going from a control rate of infection of 2.2% to 1.5% of the envelope. This is a major impact and a very easy to obtain reduction in this complication. So to summarize, in general management of complications in the EP lab, complex procedures have less complications when performed by operators and their teams and their hospital support who have a great deal of experience. So I think everyone has to understand what they're comfortable doing, what they're good at doing. For example, I'm not trained to do lead extraction. So it would be ridiculous if I were going to decide that I'm going to do a lead extraction on a sick patient. It doesn't make any sense. We should be comfortable with what we're doing day in and day out. And when we have that experience, we'll have much less complications. Management of acute severe complications often requires a team. So I know that I am very well backed up by my team of acute care nursing and techs in the EP lab. We have a great anesthesia support. Our CT surgeons are on call all the time. We have heart failure physicians that can help us manage sick patients. So again, my tendency is to go ahead and proceed with some kind of risky procedures because I know that I have all the support. If you don't, then maybe you would make a different decision. It's important to recognize risky procedures and risky patients. And it's really just a kind of an idea of knowing your patient and thinking about what's going to be necessary in the procedure well before the procedure actually starts. Patients with complications need more care afterwards. More care to keep that downhill spiral from occurring if possible. And I think also just to make doctors human beings, I think facing patients with complications and telling them what happened, it's important. And I think that humility and honesty are very useful in this interaction after a complication. Finally, I'd like to challenge you. I think it's important to track personal complication rates with each individual procedures and compare it to standards. I said one available because there's a problem with all of our benchmarks. But I still think that you should have a feeling for where you rank according to these benchmarks and maybe make decisions about which procedures you're doing well or maybe not so well and where there's room for improvement. Thank you for your attention. Thank you.

complications

electrophysiology procedures

avoiding complications

managing complications

procedural complexity

patient illness

operator inexperience

complication prevention

complication management