[1] Thank you all for coming to this session. I really appreciate your attendance, and I’m just going to try and walk through, kind of an approach to how to do troubleshooting. Can you all hear me in the back? No? Really? Wow! Can you hear me now? Yeah? All right. [2] Just a quick disclosure. I have done some speaking for Guidant and HMP Communications. [3] So when we think about ICD algorithms, you can break them down into two different categories, you can do brady algorithms and tachy algorithms and tachy algorithms. Brady algorithms consist of your basic brady pacing and CRT therapy as well and tachy algorithms can consist of detection algorithms, detection enhancements and tachycardia therapies. [4] So, if we think about this a little further, what types of algorithms or what types of troubleshooting problems do we run into when we think about these algorithms? Well, we run into basically four – failure to pace when expected, unexpected pacing, expected therapy not delivered and unexpected therapy. [5] I’m going to try and walk you through a few examples to show you how to approach troubleshooting some of these issues. You need to develop a systematic approach to do the troubleshooting. You need to start with your patient history. You always need to start with your patient history. You’ve got to figure out what’s going on with your patient because they have the best information to help you get started on this journey of troubleshooting. You did do a device interrogation and evaluate your program parameters. See how they’re programmed to give you a clue as to why, what happened or didn’t happen happened, and see if they’re tailored to your patient’s needs. You need to evaluate real time electrogram. You get a lot of valuable information from real time electrograms using them in comparison with episode electrograms to help you appropriate function. And then also you need to look at your episode EGMs and then again comparing them with real time and see if you can figure out why the device did what it did. And then you also need to finish up with determining normal device function, doing pacing and sensing threshold testing. [6] So if we start to look at some examples. We’re going to start with failure to pace when expected. This is a patient who was in the CCU with a Medtronic atrial defibrillator, obviously atrial therapies are programmed on and we got a panic call that said the patient had a pause. Now, I guarantee you that the patient was not the only person in the room that had the pause at the time when this happened. Now what this is, is that this is a surface ECG that they had saved for us, which they often don’t have saved for us, but this one they definitely did, and if you look at this atrial ATP being delivered. This is a pacemaker dependent patient and the atrial ATP is being delivered and you see no backup pacing. Is anybody familiar with this algorithm does anybody have any idea why this would have occurred? Okay, I’m going to give you one more clue. I’m going to show you one more beat over here. [7] This is a feature with the Medtronic atrial defibrillators that has an option to program backup pacing during atrial therapies either off, on at prior, or on always. This particular device is programmed on if prior. That algorithm works by looking at the previous four beats. If any of those four beats are sensed, it withholds pacing, okay? In this case, they’re pacemaker dependent but they actually had a PVC that was sensed on the fourth beat, therefore, qualifying it for withholding backup pacing during therapy, okay? So, when we look here, we had kind of an escaped PVC that came in here but we essentially had nothing in here. And then we have this little period here where we felt, okay fine that all makes sense. We’ve got the manual out. We checked it all out. What happened here? What took it so long to kick in and pace? We contacted the company and they told us this is what they call the device getting his act back together. So, it took just a little bit of time to regroup. It seems like an eternity, but as you can see it’s really not that long. Now, the current features and the current atrial defibrillators are called on, I believe it’s auto, but it still functions the same way. It still functions by looking at the previous four beats and then again withhold if any of them are sensed. So, what I always do, is I always program it on always, especially on pacemaker-dependent patients so that this would never occur, okay? [8] Here’s another example of failure to pace when expected. We were called to the floor after a new implant and this was a Medtronic EnTrust ICD and they said failure to capture. I see some heads nodding out there. Who knows what this is? Go for it. Shout it out. MVP – managed ventricular pacing. It is a feature with Medtronic, and I like to call it minimizing ventricular pacing, it helps me remember, because what we’re trying to do is promote intrinsic conduction. What the device will do and if you look at this example with markers, the device will actually let you drop a beat before it actually kicks in. So you see we have an A pace, V sense, A pace, V sense, A pace, V sense, A pace and nothing. The device will let you drop a beat. It will come in an AV pace and then it will look again A pace V sense and continues on. Now, if you, for example, dropped a second beat here, we’ll look again and let you drop another beat. Now, this would look even worse to the nurses on the floor, but still is indeed normal function. You guys are good. [9] All right, so let’s flip to the other side of that algorithm, then unexpected pacing delivered when we didn’t think it should be. This is a Medtronic BiV ICD. The patient’s in chronic afib. It was just implanted the day prior to this episode. RV and LV leads only obviously because it’s chronic afib. Program VVI 70. We got the call saying pacemaker is broken. It’s not working. It’s not pacing at 70. It’s pacing faster than 70. Any thoughts on this one? [10] This is a feature called conducted AF response. It was programmed on in this particular patient and the goal of conducted AF response is to increase the percentage of BiV pacing during afib, okay? So when they go into this atrial arrhythmia, they often have a little faster rates, and we don’t get the contribution of the CRT. So, it helps increase the percentage of pacing during CRT. It also helps make the V-V intervals more regular, hopefully, decreasing some off the symptoms that come along with that in our patients, and it does have a maximum programmable rate, which makes sense because we would not want our patients, CRT patient, at 160 beats per minute. So, it does have a maximum programmable rate. Again, increases the percentage of CRT during delivery of, or during atrial arrhythmias. Does that make sense? All right? [11] Okay, here’s another example of unexpected pacing. This is a Medtronic BiV ICD and we got the call saying, get here how. Come and see this patient. he’s having a lot of VT. So we looked at the rhythm strip, we see here that we have V pacing, V pacing, V pacing. We have a PVC right here that has a pacing spike in it, and then we have VT and the first period of VT here actually has pacing spike in it as well. This is a separate episode that happened shortly after and again, we have a PVC with a pacing spike in it. Anybody have any thoughts on that? Shout it out. I know you guys know some of this. This is a feature called ventricular sense response. What this does is it says whenever I see a sensed beat, I want to contribute BiV pacing to that beat. So, when it senses and it senses only in the RV in this particular device, it will then pace according to how you have it programmed. So, if you only had programmed RV pacing, it will RV pace. If you have it programmed BiV pacing, it will pace BiV and it will accordingly pace in the order you have it programmed. So, if you LV pace first, then RV, you have a V-V delay programmed on that’s how it will deliver that pacing. [12] We can further confirm that this was our case because if we look at the episode electrogram, we see that we have BiV, BiV, and then we have a BiV on top of a V sense. Again, we have a BiV on top of the V sense and then we’re off to the races here. The device terminates and then if you see over here, we actually have the same thing happen again here and here, [13] but then we have this ventricular sensed beat that came in that actually didn’t have a BiV pace. Does anybody have any thoughts as to why that might have happened? This particular device feature also has a maximum programmable rate and again, for obvious reasons, you wouldn’t want it to be throwing pacing at a very rapid rate, and that particular beat then exceeded the maximum programmed rate of that feature. So, it wouldn’t have thrown that, okay? [14] All right. So then if we get into the tachy therapy side, if we look at unexpected therapies. – a 79-year-old male with an ICD implanted for out-of-hospital cardiac arrest, CAD, low EF, single-chambered Guidant ICD, came into the ECU after receiving multiple shocks. Okay, he was awake for the shocks and we know we didn’t really wanted that to happen to start with. [15] We have to look at the rhythm. He’s programmed with a VT zone of 165 and a VF of 190, and backup pacing at VVI at 60. [16] Okay, so this is your electrogram, and now we’re looking at a single-chamber device here. So, it’s got two electrogram stored. And so, if you look at this, we have what we call a near field, which is V tip to V ring, and a far field electrogram which is can to coil. Now, your far field electrogram resembles more of your electrogram, okay? So, if we look at this, this is a more narrow complexes. There’s a very small sensing area and this is a little larger so it’s a little wider. So if we look at this electrogram, and this one’s called VT, the device calls it VT, do you think that’s VT? Everybody knew that one. Atrial fib. Why do you think it’s atrial fib? It’s irregular. Exactly. Unfortunately, the intervals became quite stable here and that’s stability and the device delivered therapy. [17] Okay, so we already know it’s afib and then we have therapy #4 delivered, and then we get this. Now, what do we have? What do we have? Do we have afib still? Do we have VT? We have VT, okay? Do you still think that? Why not? I’m making you work here. We still have some irregularity up here, don’t we? That tells us that this is still likely afib. [18] What happens is after you receive multiple shock therapies, it’s not unusual for your wave form to widen out. So, this is actually an afib with probably an aberrant conduction here. So this is not what we wanted our patient to get shocked for. Did the device malfunction? It worked according to how we had it programmed, right? We basically, because it met stability with the regular intervals. Basically, we need to control this guy’s AF, right? [19] Okay. Okay. Second example of an unexpected therapy delivered. This is a postop ICD evaluation of a GEM III DR 7275. It’s a dual-chambered ICD. It was implanted elsewhere so he is not familiar to us. Noncardiac surgery, I only say that because we didn’t obviously messed with the leads during the surgery. When I saw him preoperatively, I was able to do baseline assessment to determine normal function, but I wasn’t actually able to look at the episodes that were in his episode log book. So post-op, I wanted to look at his episodes a little more closely. [20] This is an episode then of VT that was in his device, and now we’re dealing with the dual chamber device now. So now, we’re looking at the two electrograms, an atrial electrogram, and a ventricular electrogram. These are both near field, so A tip to A ring, V tip to V ring. Both near field, both narrow complexes. So, when we approached this, we want to look at how many A’s for every V, right? So that’s the first step we do. Do we have one A for every V in this episode? We do, right? So what does that kind of lead us to initially? Some kind of a 1:1 arrhythmia, could be sinus tach, right? If we look at our markers, the device is saying tachy sense, A refractory, tachy sense, A refractory, tachy sense, A refractory. So, it’s calling that a retrograde P wave, right? Okay, so then it says, well if it’s retrograde, it’s what? It’s VT. Okay. So what would we want to look at for this particular patient? [21] I’m sorry? Do they have a first degree block and how would we look at that? [22] This is how he is programmed by the way you’re on the exact right track there. This is how it’s programmed and a particular feature that we’re looking at is this 1:1 VT-ST boundary. So, if we pull up his intrinsic electrogram, his intrinsic rhythm, you’re exactly right, he has a sinus rhythm with a first degree AV block. In fact, this AV, or his PR interval is probably 280 milliseconds here, [23] okay? So then if we further look at that, we have this programmed at the nominal setting of 50%. Everything that falls on this side is A-OK. Everything that falls on this side is considered retrograde and the device says that’s not okay, okay? [24] So, if we look at this further, this V-V interval is roughly 480 milliseconds. If we take that right in half it’s 240 milliseconds. This P wave falls just to the left, just to the retrograde side, okay? So the device calls it VT because it’s retrograde conduction. Did the device malfunction? It did not. It worked exactly how we programmed it to work, right? So what could we do to fix that? We would want to move that percentage, the oops! I’m sorry. I want to go back here. We would want to move that percentage over this direction, right? So, if we move this over to 66%, 66% of this 480 milliseconds turns out to be about 316 milliseconds. That ought to capture our P wave and make it fall on the correct side, and help us out there. Is there anything further you’d want to know before you could securely program that feature? [25] We would want to probably look to see if we knew what his true VT looked like just to make sure, right? And so, unfortunately, the device had an old VT episode in it that was able to interrogate. And if we look at this, this electrogram looks a little different than our intrinsic does, right? So that helps us. We also have definite VA dissociation here, right? So, this looks like his true VT, and that may be more comfortable in safely programming that to 66%, okay? The other thing that was real important, that I felt was really important to do is to follow up with his local facility since he wasn’t followed by us and wasn’t going to continue followup with us; to communicate with the local facility that we had made that programming change, so that they could continue to follow him and watch the setting, to make sure it was appropriate, okay? [26] All right. I think we already talked about that. [27] So if we look at the flip side of this, we’re going to show some examples of failure to delivery therapy. This particular patient was admitted with an atrial fibrillation, GEM III AT device. He had tried his patient activator to deliver a shock at home, but it would not deliver a program shock. He said that he doesn’t always know when he is in afib but he has not been feeling well of late. [28] So, we interrogated his device, looked at his parameters. Indeed he did have a patient activator shock programmed on. That’s all right. The other key piece of information to look at is his time to stop therapy. It was programmed to 24 hours, which means that after 24 hours of initiation of an episode, it stops delivering therapies. No more therapies can be delivered, okay? [29] So if we then interrogate his episodes, we see that he has this episode that started on October 12, and it says episode in progress. So this is his current episode. Is there anything on this programming sheet that tells you why it would not deliver that activated shock? You guys are so good. It say’s October 15, this is three days after, right? Our 24 hours long expired, right? [30] So, the patient said he didn’t activate initially because he thought he had a nighttime shock window programmed on. He had waited for that shock window and then held off on the patient activated shock. So, therefore, when he tried to self-activate just before he came into the hospital, it would not deliver. [31] So, what we did for him is we reprogrammed time to stop therapy to 48 hours. We did add a nighttime shock for him. When you do this, make sure that the device time is appropriate. We gave him a two-hour shock window, meaning from, example say for from 2 am to 4 am. You need to make sure that device time is appropriate because there’s nothing worse than the patient laying there waiting for the shock to come at 2 am and it doesn’t come until 3, or 4, or 5 am. And it’s a very big deal. These patients with atrial defibrillators are in very desperate straits when they are willing to get shocked for atrial fib. And so you need to make sure that you cover all your bases, dot all your i’s, cross your t’s to make sure everything goes as according to plan as possible for these patients. [32] The second example then of failure to deliver therapy when expected, this is a 63-year-old active male with ischemic cardiomyopathy, Guidant dual-chamber ICD. He was transferred to us after increasing episodes of VT requiring increasing doses of amio. Amio was starting to affect his LFTs and it was also making him even more chronotropically incompetent. He initially had a single chamber, and was upgraded to the dual-chamber just a year prior, and the amio then of course as we know had slowed his VT. [33] Quickly then review, you could see he’s indeed chronotropically incompetent. [34] Quickly, how he’s programmed, you can see he has a slow VT. [35] I was doing a V pacing threshold and this is the resulting rhythm. And now, in this particular case, it’s a surface lead atrial electrogram, ventricular electrogram. What’s this rhythm? Do we have a 1:1 AV relationship? No, we don’t. We have an atrial rate here, and we have a much faster V rate, right? This is clearly a VT. The problem is that the device is saying V sensed, PVC, PVC, V sensed, PVC, V sensed. Why didn’t the device treat this as VT? You’re all saying it. Shout it out. It’s too slow. It’s not, it’s below the programmed interval. [36] So the programmed interval was 130 milliseconds. The VT was probably about 108 beats per minute. This patient was then scheduled for VT ablation. After successful ablation, we were unable to induce VT and we were able to raise his VT zone and give him a higher sensor rate to make him feel a lot better. [37] In conclusion then, you need to have a systematic approach when you evaluate these devices. It’s becoming increasingly difficult to troubleshoot these algorithms without the use of electrograms and markers. These algorithms are becoming much more sophisticated. They’re definitely benefitting our patients, but sometimes become more troublesome to figure out what’s going on. You need to use your resources, device technical manuals, and when you’re in the heat of the moment tachycardia technical services is a great resource. We use them all the time. They know these devices inside and out. They’ve tested them. They’ve done all sorts of things with them. If they don’t know the answer, they know where to go to get them. So please use those in addition to your colleagues. Your colleagues are also excellent resources in troubleshooting. I thank you very much for your time. [38]

Amy Beeman 157ABC Troubleshooting Special Features and Algorithms Customization for Optimal ICD Programming: Troubleshooting and Case Presentations