Cardiac arrest and deep T wave inversions

The paramedic swung the stretcher into the resus bay, and started giving report. As the team of RNs, techs, and residents swung into action, I noted that the young adult patient didn’t look very sick at all. Confused, yes, and perhaps a bit anxious, but this seemed like an over-triage.

“Paramedic Battistelli,” I called out, “why is this 38 year-old female patient here, instead of in fast track?”

“Hey Dr. Walsh! We were called for a seizure, but she looked fine when we got to the house. She denied any problems, but family said she was just lying on the bed, no warning, when she started convulsing. Vitals and sugar were fine.  Listen, I didn’t think I should call in the cath lab team, but I didn’t like the looks of this.”  And he handed me his 12-lead:


“Well.,” I responded, “look at these deep T wave inversions in V1-V4. This looks like Wellens syndrome. We better get the cath lab rolling! When did she stop having chest pain?”

“Doc, she denied chest pain, pressure, burning – everything.”

“Okay. Then this is probably the anterior T wave inversions you see with a massive pulmonary embolism! We might need to give her tPA. Was she very hypoxic?”

“No, in fact she never so much as coughed. No trouble breathing, sats were great.”

“Well, it’s kind of rare, but you can see these sorts of inversions with Takotsubo cardiomyopathy too. Let me guess,” I asked loudly, “she must have had a terrible scare right before the seizure. Or maybe a nasty argument, or some other emotionally charged moment?”

“Nope. She had been napping on the bed with her mother.”

I asked to repeat the ECG, figuring that the leads had been misplaced, or there was some artifact, or the EMS monitor had some bad filter setting..


Crud, this looked worse.

“Medic Battistelli, this shows clear signs of Wellens at this point. As you know, women often don’t feel any pain with their MIs. I’m activating the cath lab.”

“Ok, sounds fine,” he interjected, “but I saw something else on the ECG. It sounds crazy, but…”

“Dr Walsh!” The RN grabbed my arm. “She’s seizing again – and look at the monitor!”



Palpation of her carotid confirmed pulselessness. CPR was started, and the a single defibrillation restored sinus rhythm, as well as consciousness. “Looks like that was VT triggered by ischemia,” I told the team, “let’s focus on getting her to the cath lab as quickly as possible.”

“Doctor Walsh, I was wondering,” asked Mike, “should we try one thing before sending her to the lab?”

What was Mike suggesting for therapy?


Magnesium, as therapy for Torsades de Pointes (TdP).


The arrest rhythm was a wide-complex tachycardia, and thus overwhelmingly likely to be a form of ventricular tachycardia. This VT does not have the usual monomorphic morphology, however, and is instead called polymorphic VT (PVT). One might be tempted to leap to calling this TdP based on this rhythm, given the dramatic appearance of the “twisting points,” but it’s crucial to remember that you can only diagnose TdP when the QTc is long. If the QTc was not prolonged here, this would be just PVT, which is usually caused by cardiac ischemia.


A look back at the EMS ECG shows that the QTc is quite prolonged, > 600 ms, and so we can diagnose TdP.

Why does this matter for therapy?

Well, as mentioned, PVT is usually caused by cardiac ischemia, so evaluation and treatment should first focus on ACS. TdP, by contrast, often has a metabolic (± genetic) cause, and these should be sought and corrected.


Most importantly, distinguishing between PVT and TdP is crucial when selecting an antiarrhythmic medication.

  • For PVT associated with cardiac ischemia, beta-blockers and amiodarone are considered a class I intervention.
  • For TdP, however, magnesium plays a central role (after defibrillation!), while amiodarone has no established or theorized benefit.

Interestingly, many clinicians have been taught that amiodarone is dangerous in TdP, since it is well-recognized as causing a prolonged QTc. There seems to be little evidence for this, and it appears that amio only rarely is implicated as a cause of TdP.


Continuous Chest Compressions vs. 30:2 – Does It Matter? Depends On the Quality!

seattle fire engine 2 resus

Seattle Fire Department Engine 2. The rescuer on the left has just placed a metronome on the ground.

I will eventually get around to composing a more thoughtful blog post about my experiences at the Resuscitation Academy but in the interim I wanted to share something about the ventilation strategy in King County, Washington.

There isn’t much controversy in the fact that ventilations are probably unnecessary in the first 4 minutes of sudden cardiac arrest. The idea is that the arterial system is full of fresh, oxygenated blood at the time of collapse. However, after that time period has elapsed things get a bit murky and this is where there are differences of opinion.

instrumented manikin

Instrumented manikin at the Resuscitation Academy.

There is a high rate of bystander CPR in King County (about 1/2 the time dispatcher-directed) so bystanders are performing continuous chest compressions for 4-6 minutes prior to the arrival of EMS. I think we can all agree this is a good thing. However, you can also argue (and Peter Kudenchuk, M.D. does) that ventilations are not so easily omitted when 4-6 minutes of continuous chest compressions have already taken place.

Once EMS arrives at the scene, expertly performed chest compressions (rate, depth, and recoil) are initiated. They train with instrumented manikins in King County, and one thing we discovered is that every single one of us “leaned”. The smallest amount of leaning on the chest destroys recoil, preventing the negative pressure gradient responsible for blood return to the heart.

(See also: The Science of CPR by Peter Kudenchuk, M.D.)

As a side note (not to stray too far off topic) the communication is excellent in the Medic One system, so problems with rate, depth, and recoil are corrected quickly because a fellow rescuer will point out the problem. It’s a part of their culture and professionalism that everyone observes what is happening.

There are slight differences between King County EMS and Seattle Fire but they are not too caught up in whether or not you shock as soon as possible or at the 2-minute mark once you have reached the patient’s side. However, the emphasis is on chest compressions and defibrillation. For example, if there were only 2 rescuers on scene, chest compressions and the first shock would take priority over using the BVM.

Once more help arrives, or assuming there is a third person, a BVM is deployed. As we have previously discussed, King County EMS uses a 30:2 strategy initially, while Seattle fire performs “BLS Continuous” where they ventilate at a 10:1 ratio without interrupting chest compressions.

(See also: Trial of Continuous or Interrupted Chest Compressions During CPR)

This is where I want to point something out that is critical to understand. They are only bagging with 300-400 ml of volume! That is a very small bag squeeze. Without the instrumented manikin I wouldn’t have believed it was adequate. In the absence of chest compressions it causes a noticeable chest rise, and that’s all they interested in — just enough for air exchange and to prevent atelectasis.

Resuscitation Academy logo with stylized heart, the tree of life or knowledge, 10 branches for 10 steps to improve resuscitation, and 4 stars for the chain-of-survival (prior to 2010). Two stars are larger because they are more important (CPR and defibrillation)

Resuscitation Academy logo with heart, the tree of life (or knowledge) with 10 branches for 10 steps to improve survival, and 4 stars for each link in the chain-of-survival (prior to 2010). Two stars are larger because they are more important (CPR and defibrillation)

There is no doubt that we have been over-bagging in my system, and I can see why cardiocerebral resuscitation confers a benefit in systems that bag overzealously with 30:2 or delay for too long to deliver the breaths.

In the Medic One system, when 30:2 is used, the rescuer on chest compressions is “divorced” from the person on airway. They deliver 30 perfect chest compressions, pause for 2 seconds, and then deliver another 30 chest compressions. They don’t “wait” for the person on the airway.

As you can imagine it takes training, and re-training, to achieve this level of performance. But that’s sort of the point. They have been perfecting their craft since the Apollo program and each iterative change has been made thoughtfully and using the best available evidence (mostly derived from their own measurements).

Still, they readily concede that resuscitation is, and perhaps will always be, an unsolved puzzle. So whatever you do, don’t just be good at it; be absolutely phenomenal at it. If you are successful in cultivating a culture of excellence and continuous quality improvement your EMS system will get where it needs to go.

Further reading

Ten Steps for Improving Survival from Sudden Cardiac Arrest


Nichol G, Leroux B, Wang H et al. Trial of Continuous or Interrupted Chest Compressions during CPR. New England Journal of Medicine. 2015;373(23):2203-2214. doi:10.1056/nejmoa1509139.

Kleinman M, Brennan E, Goldberger Z et al. Part 5: Adult Basic Life Support and Cardiopulmonary Resuscitation Quality. Circulation. 2015;132(18 suppl 2):S414-S435. doi:10.1161/cir.0000000000000259.

Ewy G. Cardiocerebral Resuscitation: The New Cardiopulmonary Resuscitation. Circulation. 2005;111(16):2134-2142. doi:10.1161/01.cir.0000162503.57657.fa.

The 12 Rhythms of Christmas: Paroxysmal AV-Block

This article is the ninth in our latest series, The 12 Rhythms of Christmas, where we examine a different rhythm disorder with each new post. It’s a continuation of the theme behind last year’s 12 Leads of Christmas. And, just like last year’s series, I’m rather late getting the final articles out, but the end is in sight.

Hope you had a good Valentine’s Day—let’s talk about some heart stuff. Today I want to discuss a form of AV-block that many providers don’t even realize is its own unique entity:

Paroxysmal AV-Block

What differentiates this arrhythmia from the other AV-blocks is that it occurs in discrete, self-limited episodes—or “paroxysms.” You patient will be hanging out, minding their own business, when out of nowhere they suddenly drop two or three or forty P-waves in a row until the AV-node just as suddenly recovers. It tends to give you a pretty good wake-up.

Paroxysmal AV-block

Figure 1. Paroxysmal AV-block. I suggest you click to enlarge.

The strip above was run on an 80 year old male who was being treated for pneumonia when he unexpectedly appeared to have a short seizure. It wasn’t a seizure—it was syncope—and this strip is our proof.

Most folks would call the above rhythm “Mobitz II AV-block”  because there are dropped P-waves without an apparent increase in the PR-interval preceding them, but as we discussed in our article on high grade AV-block, we shouldn’t really be labeling something a type II AV-block if two or more consecutive P-waves are dropped.

Well, then it must surely be “high-grade AV-block?”

Not quite. That term implies a fairly fixed conduction disturbance. A true high-grade AV-block will typically last somewhere on the time span of hours to forever, whereas the conduction disturbance we see in Fig. 1 lasted only 20 seconds or so and is—you guessed it—paroxysmal.

The patient never showed any dropped P-waves in the two hours before that spell, and he didn’t show any more over the next two days until he received a permanent pacemaker. In hindsight he recalled experiencing two unexplained “falls” in the days preceding his admission, so based on his inability to recall how or why he fell, it seems likely that similar paroxysms of AV-block had struck him then as well.

That’s the interesting and scary thing about paroxysmal AV-block—it can strike without warning in patients whose resting 12-leads are fairly unremarkable.

Figure 2.

Figure 2. This is the 12-lead of the patient from Fig. 1. It shows sinus rhythm with some non-specific irregularity and a left anterior fascicular block—nothing too concerning.

Figure 3.

Figure 3. This is the same patient’s ECG from two years prior and it is identical to the tracing in Fig. 2.


How about at some more cases?

Similar to the last patient, Fig. 4 is from an 80 year old female who fell at home and may or may not have experienced syncope; she was unable to recall.

RBBB + LAFB (bifascicular block)

Figure 4. Sinus rhythm with a right bundle branch block and left anterior fascicular block (bifascicular block).

Also similar to the last patient, she has a left anterior fascicular block (LAFB) on her 12-lead, but this time there is the addition of a right bundle branch block (RBBB); together they produce a bifascicular block.

While most bifascicular blocks are not an acute issue, in the patient presenting with syncope it can be a major red flag that he or she has extensive conduction disease and could be experiencing paroxysms of AV-block.

Our patient was asymptomatic in the ED… That is, until she experienced another episode of syncope while in bed.

Paroxysmal AV-block

Figure 5. Paroxysmal AV-block.

She had an uneventful night in the hospital and received an implanted pacemaker the next day. Are you noticing a pattern?

This next patient is an 84 year old male who was brought to the ED because his family claimed that he had experienced a couple of brief “staring episodes.” They said that while he never fell or “passed out,” on a couple of occasions over the past two days he had stopped responding for about 10 seconds and just stared into the distance. Fig. 6 is his 12-lead on arrival in the ED.


Figure 6. Sinus rhythm with a RBBB. Pretty unimpressive.

Much to the annoyance of his family and staff, the bedside monitor seemed to have trouble tracking his QRS complexes and occasionally set off a crisis alarm of ASYSTOLE even though the patient, under constant watch of his loved ones, was fine.

While I have changed many details of the case, this next part really happened…

So, being the guy who troubleshoots all the monitors, I got called back to figure out why it kept alarming indiscriminately. My first step was to see what kind of events it was alarming for and I pulled up the following strip:

Paroxysmal AV-block

Figure 7. A short episode of paroxysmal AV-block. It mimics sinus arrest but there are subtle P-waves marching through the pause.

Turns out it wasn’t a false alarm. Then, literally, while I was looking at that strip, I heard the machine crisis alarm again and looked up to see the following rhythm:

Paroxysmal AV-block

Figure 8. A rather long episode of paroxysmal AV-block.

Turns out the patient had experienced a couple of short, asymptomatic paroxysms of AV-block while in the ED that were setting off the monitor’s alarm but were not immediately noticed. Of course, we all picked up on that last episode. What looks like V-tach at the end of the strip is actually artifact from seizure-like activity he exhibited as the AV-block subsided—a new development for our patient. Have I mentioned that syncope and cardiac arrest are often mistaken for seizures? I feel like I have…

While he experienced several more short and long spells of paroxysmal AV-block in the ED, he responded well to transcutaneous pacing and went on to receive an implanted pacemaker the next day without further issue.

With that, I leave you for today; go forth and call paroxysmal AV-block by its proper name.

Check out the rest of The 12 Rhythms of Christmas! (updated as new posts come out):

The 12 Rhythms of Christmas: Sinus Tachycardia
The 12 Rhythms of Christmas: Sinus Bradycardia
The 12 Rhythms of Christmas: Atrial Flutter
The 12 Rhythms of Christmas: First Degree AV-Block
The 12 Rhythms of Christmas: Type I AV-Block
The 12 Rhythms of Christmas: Type II AV-Block
The 12 Rhythms of Christmas: 2:1 AV-Block
The 12 Rhythms of Christmas: High-Grade AV-Block





You’re still here?



It’s over,

go home.






Well, if you’re still around, you must be interested in some of the gritty details of paroxysmal AV-block. You know, minor stuff like “What triggers paroxysmal AV-block?” Let’s dive in.

First off, you may have heard the terms “phase 3” or “phase 4” block used in reference to paroxysmal AV-block—or maybe you haven’t—either way, those are probably misnomers. For a full explanation of why, I highly suggest this paper by El-Sherif and Jalife (I’m not going to re-hash the whole thing). For now just know that I’m going to refer to the former “phase 3 block” as tachycardia-dependent paroxysmal AV-block (TD-PAVB) and the former “phase 4 block” as pause-dependent paroxysmal AV-block (PD-PAVB). They’re big names but the concepts are simple and we’ll look at some examples in a second.

In the most basic sense, TD-PAVB is a paroxysmal AV-block that’s initiated by an early beat or increase in heart rate and constitutes most paroxysmal AV-blocks, while PD-PAVB occurs when a run of paroxysmal AV-block is triggered by the normal heart rate slowing or a pause in the sinus rhythm, and it is comparatively less common. Since nothing is ever that simple or dichotomous, sometimes PAVB strikes without any appreciable change in the sinus rate—in which case we just call it plain old paroxysmal AV-block and try not to think too much about it. There may have been a subtle change in the patient’s physiology that set it off, but at the bedside we can’t even begin to fathom what exactly that might have been.

I haven’t seen a great many papers on this topic and my bookcase of textbooks is not at all helpful, but it seems that the block often occurs at the level of the bundle of His—even in patients with infra-Hisian conduction disturbances (bundle branch blocks). Its mechanism seems most closely related to that of the classic type II of AV-block, however I have seen a couple of purported examples of supra-Hisian (AV-nodal) PAVB. I guess what I’m saying is that we should approach this like high-grade and 2:1 AV-block and not make any assumptions whether the underlying mechanism is type I or type II based on the surface 12-lead—leave that to the electrophysiologists with their His-bundle electrocardiograms.

Let’s examine which flavor of PAVB each of our patients above experienced…

Paused-dependent paroxysmal AV-block (PD-PAVB)

Figure 9. This is an excerpt of the strip from Fig. 1.

Despite being the least common form, it looks like our first patient in Fig. 9 is experiencing pause-dependent PAVB as evidenced by the longer-than-normal PP-interval (803 ms) that triggers the AV-block. Interestingly, it’s not a huge slowing of the sinus rate that sets off the PAVB—it’s not even the longest PP-interval on this short strip—and while we can’t guarantee cause-and-effect from this solitary episode, that’s the #1 contender for a diagnosis at the moment. I’m about 60% confident it’s really PD-PAVB.

How about the patient from Fig. 5?

Tachycardia-dependent paroxysmal AV-block (TD-PAVB)

Figure 10. TD-PAVB

It’s rather tough to see exactly what’s going on here because artifact obscures some of the atrial activity, but it appears that the AV-block in this instance is instigated by an acceleration in the sinus rate (PP-interval of 639 ms); so it is likely a tachycardia-dependent PAVB.

Finally, what about that patient with the sneaky pause in Fig. 7? It turns out that I have a bunch of strips from him…

Tachycardia-dependent paroxysmal AV-block (TD-PAVB)

Figure 11. This is the same tracing as Fig. 7, just cropped.

Tachycardia-dependent paroxysmal AV-block (TD-PAVB)

Figure 12. Another strip from that patient.

Tachycardia-dependent paroxysmal AV-block (TD-PAVB)

Figure 13. And another strip from that patient.

Tachycardia-dependent paroxysmal AV-block (TD-PAVB)

Figure 14. This is the same tracing as Fig. 8.

Tachycardia-dependent paroxysmal AV-block (TD-PAVB)

Figure 15. Finally, yet another strip from that same patient.

Hopefully I don’t need to march out the PP-intervals this time—it should be apparent that each of these paroxysms of AV-block are precipitated by a PAC. As a result, they are all examples of tachycardia-dependent paroxysmal AV-block. I didn’t post the full strips, but the runs of AV-block that resulted from each PAC lasted anywhere from a couple of seconds to over 20 seconds.

Before we conclude for real, I have one more mind teaser for you. The following strip is from the same patient as we saw in Fig. 7–8, and Fig. 11–15, but it looks a bit different from the earlier rhythms.

Is it still paroxysmal AV-block?

Figure 16.

Figure 16.

That big deflection about 2/3 of the way through the tracing is from the transcutaneous pacer kicking in. While we just established that this patient has been experiencing TD-PAVB, this pause doesn’t seem to be preceded by an apparent premature complex. Maybe the PP-intervals are subtly diminishing and the heart rate increasing?

Figure 17.

Figure 17. This is the same strip as Fig. 16.

Odd. There’s no significant variation in the PP-intervals… Perhaps it’s just plain old paroxysmal AV-block this time, with no increase or decrease in the atrial rate? The only issue—and it’s a big one—is that I don’t see any dropped P-waves marching out either, so it can’t even be true AV-block…

Maybe something is up with the sinus node? Could this patient have combined SA and AV-node dysfunction?

Another interesting finding is that this pause is much shorter than any of the other episodes of PAVB we’ve seen in this patient. Perhaps the transcutaneous pacer kicking in is allowing the heart to recover sooner?

Figure 18. Evidence of false-capture.

Figure 18. Evidence of false-capture. This is the same strip as Fig. 17.

Well, that can’t be the case, because the pacer isn’t capturing! As we see in Fig. 18, there is a native QRS buried in what should be the absolute refractory period following the pacer spike. If the pacer was actually achieving electrical capture, the myocardium would still be refractory at that point and unable to generate a QRS.

So, enough Socratic jabber, what’s really going on?

Don’t forget your basic arrhythmia training!

Figure 19. Non-conducted PAC.

Figure 19. Non-conducted PAC. This is the same strip as Fig. 18.

The most common cause of a pause is a non-conducted PAC—keep that in mind even in patients with wacky AV-blocks! I know it’s super subtle and you should certainly be skeptical of my explanation at this pint, but here’s some more examples from the same patient to prove that what we’re seeing this time really is non-conducted PAC’s.

Figure 20.

Figure 20. The arrow is pointing to a non-conducted PAC and the circle highlights a buried native QRS complex, proving that the pacer is not capturing. It looks very similar to Fig. 18 and 19.

Figure 21. It looks nearly identical to Fig. 20, but this tracing was obtained two minutes later.

Figure 21. It looks nearly identical to Fig. 20, but this tracing was obtained two minutes later.

Figure 22.

Figure 22. Again, this is nearly identical to Fig. 20 and 21, but this was actually an hour later.

Figure 23.

Figure 23. This time there is an additional blue arrow that points out a sinus P-wave sneaking in just prior to the transcutaneous pacing spike.


For real, we’re done now.

That's All Folks


71 year old male with “seizures” that prove to be ICD shocks

EMS is called to the residence of a 71 year old male for “seizures.”

On arrival the patient’s spouse meets the ambulance outside and hurries the paramedics along saying “Come quickly! Please help him!”

The paramedics arrive at the patient’s side just in time to see him receive an ICD shock.

They ask how long this had been going on.

“That was my 15th shock!”

The patient states that he “felt himself going faint” just prior to the first shock.

The cardiac monitor is attached and the following rhythm strips are recorded.


The patient appears anxious. His skin is pink, warm, and moist.

Numerous skin tears are noted to the patient’s arms which the spouse states are related to convulsions induced by the ICD shocks.

Vital signs are assessed.

  • RR: 20
  • HR: 60
  • NIBP: 108/72
  • SpO2: 98% on room air

Past medical history:

The EMS crew learns that this patient survived two sudden cardiac arrests prior to receiving his first ICD in 1992. The device was replaced in 2008. The patient does not have his device ID card but knows that it was made by St. Jude Medical.


The patient states he takes several medications but he can only remember one of them: Coumadin.

The EMS crew contacts Online Medical Control and receives permission to apply a ring magnet to the device. The magnet is applied and taped in place. The tape doesn’t hold and a FF is assigned to hold the magnet over the device.

A 12-lead ECG is obtained.


And another.


A bigeminal rhythm is noted on the monitor.


The patient is loaded for transport.

IV access is achieved.

En route the the hospital serial 12-lead ECGs are obtained.



Vital signs are re-assessed.

  • RR: 18
  • HR: 76
  • NIBP: 102/70
  • SpO2: 99% on room air

The patient feels much calmer and says, “Please don’t let that thing shock me again.”

A final 12-lead ECG is captured on arrival at the hospital.


What do you think the patient would say was the most important thing the EMS crew did for him?

For a full discussion of inappropriate or ineffective ICD shocks, including when and how to apply a ring magnet, see the following links.

Inappropriate or ineffective ICD shocks: Part 1

Inappropriate or ineffective ICD shocks: Part 2

Inappropriate or ineffective ICD shocks: Part 3

Updated 01/06/2016

90 year old female with abdominal pain and wide complex tachycardia

EMS is called to the residence of a 90 year old female who awoke to an “uncontrolled bowel movement” that corresponded with sudden onset abdominal pain.

On EMS arrival, the patient is alert and oriented to person, place, time, and event. She has a grimace on her face and appears acutely ill.

When asked the exact location of her pain she points to the epigastric area.

Past medical history: “Cardiac problems”

Medications: Numerous (but list unavailable)

Vital signs are assessed.

  • RR: 18 shallow
  • HR: Too rapid to count
  • NIBP: 118/60
  • SpO2: 96% on room air

The cardiac monitor is attached.

A wide complex tachycardia is noted with a rate of 194.


The patient is immediately loaded on the gurney and relocated to the back of the ambulance where she is placed on oxygen, an IV is established, and the combo-pads are placed.

Breath sounds are clear bilaterally.

A pacemaker is noted in the upper-left chest.

A 12-lead ECG is captured.

Atypical RBBB morphology in lead V1 (left bunny ear taller than right bunny ear) with QS complex in lead V6 and right axis deviation.


wellensWellens’ Criteria

Wide and fast rhythms should be considered VT until proven otherwise. Regardless, in this case the morphology strongly favors VT over SVT with aberrancy.

See also: 60-Second Soapbox: Wide Complex Tachycardia at Academic Life in Emergency Medicine.

At this point the patient’s skin appears grayish, pasty, and moist. Her level of consciousness is diminished and she stops responding to verbal stimuli.

Synchronized cardioversion is performed at 100 J.


The patient is successfully converted to sinus rhythm and immediately becomes more responsive.


Another 12-lead ECG is obtained on arrival at the hospital.


Sinus rhythm with demand pacing at a rate of 70.

The patient was given amiodarone and admitted to the ICU.


Wellens HJ, Bar FW, Lie KI. The value of the electrocardiogram in the differential diagnosis of a tachycardia with a widened QRS complex. Am J Med 1978;64:27-33.