Thursday, November 12, 2020

Extreme shock and cardiac arrest in COVID patient

This is a 30-something healthy patient presented with COVID pneumonia who presented to the ED.  He was moderately hypoxic.  He had the following EKG recorded:

Low voltage, suggests effusion.
(see Ken's discussion of low voltage below)
There is a QS-wave in V2.
There is minimal, probably normal STE in V2-V6.

A bedside cardiac ultrasound was normal, with no effusion. 

He had troponins ordered, and the first returned at 72 ng/L (Abbott Architect hs cTnI; URL for males = 34 ng/L).  An elevated troponin in a COVID patient confers about 4x the risk of mortality than a normal one.

He was admitted on oxygen and was doing fairly well with saturations of 100% on 2 L nasal cannula. 

By the next day, the hs-cTnI was up to 1827 ng/L.  A Rising Troponin

That afternoon, he complained of increased shortness of breath and was noted to have oxygen saturations in the 70s, prompting a mini code to be called. 

The physicians found him to be in shock, with very poor O2 saturations.  He was intubated and then went pulseless.  He underwent CPR, and regained a pulse after epinephrine, with an organized narrow complex rhythm at 140, but still with severe shock.  

Bedside echo at this point showed a flat and collapsible IVC, so he was given fluids, and also pressors.  He lost pulses again, and after chest compressions again had ROSC and was put on more pressors.  Lactate was 20, POC Cardiac US showed EF estimated at 30%, and formal echo showed EF of only 15%, and a normal RV. 

Assessment was severe sudden cardiogenic shock.

They recorded an ECG:

New ST Elevation.  What is it?

There is STE in V2-V6. The QS-wave in V2 persists.  There is also STE in I, aVL, and II, but it almost impossible to discern because of the tiny QRS voltage (and tiny overall voltage) in the limb leads.

Notice the ST elevation is flat: the T-wave is not higher than the STE.  In acute MI, the T-wave is large, and the T/ST ratio is high.  This is much more typical of myocarditis.  

But we diagnose myocarditis at our peril.  And so it is wise to look at the coronary arteries.  

Angiogram was negative.

He remained hypotensive and in shock.

He was started on Extracorporeal Life Support ("VA ECMO")

Here is the ECG on ECMO:

Very low voltage

On Day 3, the EF recovered (that seems quick!) and the patient was converted to veno-venous (V-V) ECMO due to persistent pulmonary insufficiency.  He remained supported on an intraaortic balloon pump. This is the ECG on V-V ECMO:

Very little STE

He had a lot of troponins measured.  Here they are:

Learning Points:
1. You diagnose Myo(peri-)carditis at your peril.  This ECG certainly looks like myocarditis, and was due to myocarditis, but missing acute coronary occlusion is not acceptable.
2. Suspect myo (peri-) carditis on the ECG when the T/ST ratio is very low.  Acute MI generally has a large T-wave, unless it is beginning to reperfuse and thus invert.
2. Beware elevated troponins in COVID patients.  Below I write a mini-review on the topic.

For an exhaustive review of troponin in COVID, see this paper:
Kavsak PA, Hammarsten O, Worster A, Smith SW, Apple FS. Cardiac Troponin Testing in Patients with COVID-19: A Strategy for Testing and Reporting Results. Clin Chem [Internet] 2020;Available from:

Smith mini-review:

Troponin in Emergency Department COVID patients


Cardiac Troponin (cTn) is a nonspecific marker of myocardial injury.  In normal times, the most common use of cTni is in diagnosing, or ruling out, acute myocardial infarction (AMI, a subcategory of acute myocardial injury.  However, in multiple studies, even in the absence of AMI, both acute and chronic myocardial injury (as diagnosed by any elevated cTn) are powerful markers of adverse outcomes in both the short and long term.1,2  New data on acute myocardial injury associated with COVID-19 again shows very strong independent association between elevated cTnI and disease severity, including mortality, and a correlation of increasing severity with increasing troponin levels.3–8 


Shi et al.3 studied 416  patients hospitalized with COVID in China, of whom 82 had an initial cTn(I) above the upper reference limit.  Those with elevated cTnI, compared to those without, developed more severe disease on multiple measures, including mortality: (42 of 82 [51.2%] vs 15 of 334 [4.5%]; P < .001).  Those with elevated cTnI had far more cardiac comorbidities, but by Cox regression model cardiac injury independently associated with a higher risk of death, both during the time from symptom onset (hazard ratio, 4.26 [95%CI, 1.92-9.49]) and from admission to end point (hazard ratio, 3.41 [95%CI, 1.62-7.16]). 


Guo et al.4 studied 187 patients admitted with COVID.  Mortality was 7.6% for those without cardiovascular disease (CVD) and a normal cTnT, and up to 69% for those with CVD and just a single elevated troponin.  In those who died (n= 43), troponin (and NT-proBNP) levels rose steadily throughout hospitalization; for those who survived (n= 144), the levels remained at a plateau.


Of 3047 patients hospitalized in 5 New York medical centers between February 17 and April 12, 2716 had at least one troponin I (0.030 ng/mL) measured within the first 24 hours. 1601 (62%) had an initial troponin within the normal range; 54 had an initial level > 2.0 ng/mL.  Median age was 66.4 years; 40.7% were over age 70, and 60% were men.  CV disease was more prevalent in those with higher troponin.  Higher troponin correlated with more history of heart failure, diabetes, and hypertension, as well as higher D-dimer, and nearly all inflammatory markers.  An initial cTnI between 0.030 ng/mL and 0.090 ng/mL (n = 455, 16.6%) had adjusted hazard ratio for death of 1.77 (1.39-2.26), and cTnI > 0.090 ng/mL (n= 530, 19.4%) was associated with HR of 3.23 (2.59-4.02). Authors do not report whether the troponin in question was the initial troponin, the peak troponin, or whether more than one troponin was measured.7


These 3 studies, as well as 1 smaller meta-analysis,6 and another small study,8 make it clear that troponin is associated with increased severity and mortality in COVID when adjusted for multiple other variables.  The mechanism of myocardial injury associated with COVID, however, is not certain, and could be due to acute coronary syndrome, myocarditis, direct damage from inflammatory mediators/cytokines, microvascular damage, type 2 MI due to hypoxia or tachycardia, microvascular damage due to diffuse intravascular microthromboses, direct entry of SARS-CoV-2 into myocytes by using ACE2 receptors, or other unknown mechanisms.


Elevated troponin does not, however, guide any treatment strategies to minimize myocardial injury.  It is uncertain whether the myocardial injury is simply a marker of severity, or whether it contributes to severity of illness and mortality.  At this point in time, elevated troponin not due to acute MI does not guide therapy, but remains essential to diagnose acute MI, which may be co-existent with COVID.  For this reason, some argue that it should not be measured in patients unless acute myocardial infarction is on the differential diagnosis.9  However, because troponin is a clear marker of disease severity and a powerful independent predictor of adverse outcomes, it may be quite useful in the ED disposition decision: if troponin is elevated, then outpatient management should be reconsidered. 


When cTn is elevated, is there a way to differentiate AMI from Non-AMI myocardial injury?  There is no highly accurate method, but the ECG and echocardiography are helpful in determining need for angiography.  In patients without COVID, troponin elevations in non-ischemic acute myocardial injury are substantially lower than those of Type 1 AMI (though similar to type 2 AMI);1,10,11 in contrast, Guo’s data show that the values of cTnT in very ill COVID patients with myocardial injury rose to very high levels.4  Unfortunately, this article provides no electrocardiographic, echo, or angiographic data, so it is not certain that these high levels were in the absence of acute MI. 


In a series of 18 patients with COVID and ST elevation, 8 were diagnosed with STEMI, 6 of whom had an angiogram and it showed obstructive coronary disease.12   All STEMI patients had very high cTn typical of STEMI (cTnT > 1.0 ng/mL or cTnI > 10 ng/mL), but 7/10 patients not diagnosed with STEMI had similarly high cTn and mortality was very high in both groups.  Since many COVID patients present with symptoms consistent with AMI, both STEMI and NSTEMI appear to be very difficult to differentiate from other myocardial injury among very ill COVID patients.  Nevertheless, if the ECG is typical for acute coronary occlusion (especially STEMI), immediate intervention is recommended if it is available, and if the patient is not terminal13.  When the ECG is nondiagnostic for coronary occlusion, or the patient is suspected of having a non-occlusion MI, consider echocardiography to inform the decision for angiography.13





1.     Sandoval Y, Smith SW, Sexter A, et al. Type 1 and 2 Myocardial Infarction and Myocardial Injury: Clinical Transition to High-Sensitivity Cardiac Troponin I. Am J Med [Internet] 2017;130(12):1431–9.e4. Available from:

2.     Sandoval Y, Thygesen K. Myocardial Infarction Type 2 and Myocardial Injury. Clin Chem [Internet] 2017;63(1):101–7. Available from:

3.     Shi S, Qin M, Shen B, et al. Association of Cardiac Injury With Mortality in Hospitalized Patients With COVID-19 in Wuhan, China. JAMA Cardiol [Internet] 2020;Available from:

4.     Guo T, Fan Y, Chen M, et al. Cardiovascular Implications of Fatal Outcomes of Patients With Coronavirus Disease 2019 (COVID-19). JAMA Cardiol [Internet] 2020;Available from:

5.     Bonow RO, Fonarow GC, O’Gara PT, Yancy CW. Association of Coronavirus Disease 2019 (COVID-19) With Myocardial Injury and Mortality. JAMA Cardiol [Internet] 2020;Available from:

6.     Lippi G, Lavie CJ, Sanchis-Gomar F. Cardiac troponin I in patients with coronavirus disease 2019 (COVID-19): Evidence from a meta-analysis. Prog Cardiovasc Dis [Internet] 2020;Available from:

7.     Lala A, Johnson KW, Russak AJ, et al. Prevalence and Impact of Myocardial Injury in Patients Hospitalized with COVID-19 Infection. medRxiv [Internet] 2020;Available from:

8.     He XW, Lai JS, Cheng J, et al. Impact of complicated myocardial injury on the clinical outcome of severe or critically ill COVID-19 patients. Zhonghua Xin Xue Guan Bing Za Zhi [Internet] 2020;48:E011–E011. Available from:

9.     of Cardiology AC, Others. Troponin and BNP Use in COVID-19. 2020;

10.   Sandoval Y, Smith SW, Sexter A, Schulz K, Apple FS. Use of objective evidence of myocardial ischemia to facilitate the diagnostic and prognostic distinction between type 2 myocardial infarction and myocardial injury. Eur Heart J Acute Cardiovasc Care [Internet] 2020;9(1):62–9. Available from:

11.   Sandoval Y, Thordsen SE, Smith SW, et al. Cardiac troponin changes to distinguish type 1 and type 2 myocardial infarction and 180-day mortality risk. Eur Heart J Acute Cardiovasc Care [Internet] 2014;3(4):317–25. Available from:

12.   Bangalore S, Sharma A, Slotwiner A, et al. ST-segment elevation in patients with Covid-19—A case series. N Engl J Med [Internet] 2020;Available from:

13.   Mahmud E, Dauerman HL, Welt FG, et al. Management of Acute Myocardial Infarction During the COVID-19 Pandemic. J Am Coll Cardiol [Internet] 2020;Available from:


MY Comment by KEN GRAUER, MD (11/12/2020):


There are 2 unique features to this case: i) Dramatically increased troponin in this critically ill patient with COVID pneumonia (which Dr. Smith has discussed above in detail); andii) Extreme Low Voltage on ECG.

  • By use of the first 2 ECGs in today’s case — I focus my comments on the ECG finding of low voltage. For clarity — I have reproduced these first 2 ECGs in Figure-1.

Figure-1: The first 2 ECGs in today’s case (See text).

MY Thoughts on ECG #1:

As per Dr. Smith — the patient was a 30-something man who presented with hypoxia from COVID pneumonia. At time of his initial ECG ( = ECG #1) — he was moderately hypoxic — but responded to 2L oxygen (administered by nasal cannula) by restoration of 100% O2 saturation. Bedside Echo at this time showed normal LV function and no pericardial or pleural effusion.

  • POINT #1: There is baseline artifact in ECG #1 — which is most marked in the limb leads. We know that the source of this artifact arises from the RA ( = Right Arm) connection — because: i) Artifact is maximal in standard leads I and II — with virtually no artifact in lead III. This suggests the Right Arm as the “culprit” extremity — because by Einthoven’s Triangle, this RA electrode is equally involved in electrical derivation of leads I and II; ii) There is no baseline artifact in lead III. The reason for this — is that the RA electrode is not used in electrical derivation of lead III; andiii) Artifact is maximal in lead aVR of the 3 augmented leads (on which the RA electrode is placed) — and the relative amount of artifact in the other 2 augmented leads is approximately half the amplitude of that seen in lead aVR (which corresponds precisely to the equation used for derivation of the electrical potential in these other 2 augmented leads)NOTE: For those interested — I review in detail determination of the artifact “culprit extremity” in My Comment in the September 27, 2019 post of Dr. Smith’s ECG Blog.

  • POINT #2: Overall QRS amplitude (voltage) is dramatically reduced in ECG #1. Of note — P wave amplitude appears to be normal, almost satisfying ECG criteria for RAA (Right Atrial Abnormality) in lead II, in which the P wave is pointed and ~2 mm tall. This P wave amplitude is as large (or larger) than QRS amplitude in 8/12 leads on this tracing. Therefore — there is diffuse low voltage.
  • Otherwise in ECG #1 — the rhythm is sinus at ~95/minute — and PR, QRS and QTc intervals are normal. The frontal plane axis is indeterminate (ie, lying in the upper right quadrant) — because other than aVR, the tiny limb lead QRS complexes are all either isoelectric or predominantly negative. There is no chamber enlargement.

Regarding Q-R-S-T Changes:

  • There are QS complexes in leads V1 and V2 of ECG #1. Because of tiny QRS amplitude — it is difficult to tell whether or not there might be a Q wave in lead II. I believe the other limb lead complexes all manifest an initial positive deflection.
  • R wave progression is appropriate after the QS complexes in leads V1 and V2. Of note (and consistent with what we see in the limb leads) — QRS amplitude is again very small in the lateral chest leads.
  • There is straightening of ST segments in multiple leads — with slight ST elevation in several of the chest leads. Disproportionately peaked T waves are seen in leads I and V3. That said (as per Dr. Smith) — the overall impression was that ECG #1 did not suggest findings suggestive of OMI.
  • Clinically — despite an initial 2-fold increased troponin, the normal bedside Echo was reassuring against OMI or pulmonary embolism.

MY Thoughts regarding ECG #2:

The patient’s clinical condition worsened the next day. Troponins increased greatly — and the patient became increasingly hypoxic. He became pulseless on more than 1 occasion (presumably in EMD) — but responded to resuscitation + treatment with pressors, by regaining a pulse with a spontaneous rhythm. During this period — ECG #2 was obtained (Figure-1). Clinically — the patient was felt to be in cardiogenic shock.

  • The ECG in this follow-up tracing (ECG #2) again showed diffuse low voltage, with comparable overall QRS amplitude as was seen in ECG #1. The limb lead artifact (from the “culprit” RA extremity) had resolved. The rhythm was again sinus — albeit at a faster rate than was seen in ECG #1. This sinus tachycardia (at ~130/minute) — is consistent with the patient’s worsening clinical condition, with development of cardiogenic shock.
  • Of interest (though I’m not sure why ...) — P wave amplitude has decreased in most leads in ECG #2 — and the frontal plane axis is no longer indeterminate.
  • As per Dr. Smith — a remarkable change between ECG #1 and ECG #2 is development of subtle-but-real ST elevation in multiple leads (ie, in leads I, II, aVF; and in chest leads V2-thru-V6). There is also accentuation of a negative “dip” after the elevated ST segments in a number of leads (slanted BLUE lines in leads V3-thru-V6, with a negative “dip” probably also present in leads II and aVF). Whether this surprisingly broad negative dip represents terminal T wave negativity (with resultant QTc prolongation) or possibly inverted U waves is uncertain — but this finding clearly wasn’t present in ECG #1.
  • POINT #3 ( = Dr. Smith’s PEARL! ): A point worth repeating that was highlighted above by Dr. Smith is the shape of the new ST elevation seen in ECG #2 — which is flat! As opposed to peaking and increasing T wave size that commonly occurs with acute MI — the surprisingly level ( = flat) shaping of ST elevation seen in multiple leads in ECG #2 without T wave prominence is much more characteristic of acute myocarditis than OMI.

COMMENT: Cardiac catheterization was negative — and supported the clinical presumption of myocarditis with marked troponin increase. Adverse prognostic implications of markedly elevated troponin in this acutely ill Covid-19 patient have already been covered in detail by Dr. Smith. But the ECG finding of low voltage is worthy of further mention:

  • I discussed Low Voltage in My Comment at the bottom of the page in the January 24, 2020 post of Dr. Smith’s ECG Blog.
  • True Low Voltage is defined as a QRS amplitude of ≤5 mm in all limb leads. In an unselected population — true low voltage is a relatively uncommon ECG finding. Of note — there may (or may not) also be low voltage (≤10 mm) in all chest leads (Figure-2). Criteria for generalized low voltage (ie, present in both limb and chest leads) — are easily satisfied in both of the tracings shown in Figure-1.
  • POINT #4: The differential diagnosis for “low voltage” that is frequently put forth by many providers tends to begin-and-end with pericardial effusion. But a look at Figure-2 reminds us of a long list of additional entities to consider! This list takes on new relevance given the ongoing Covid-19 pandemic — which predisposes to acute thrombotic events, stress cardiomyopathy (Takotsubo), infarction/ischemia and myocarditis. Any of these entities may result in myocardial stunning” (ie, transient marked reduction in cardiac contractility, that occurs in response to a major acute insult).

Figure-2: Causes of Low Voltage on ECG (See text).

POINT #5: I’ll finish by speculating a bit based on clinical correlation from today’s case with ECG findings in the 2 tracings shown in Figure-1. Low voltage was noted in the initial ECG (ECG #1) — despite bedside Echo at this time showing normal LV function.

  • In addition to the adverse prognostic implications of dramatically increasing troponin in sick Covid-19 patients — perhaps the ECG finding of diffuse low voltage serves as another harbinger of a reduction in LV function that will soon be occurring?
  • Is there a lag between diffuse low voltage on ECG and Echo indication of depressed contractility?
  • Is low voltage + increasing troponin in the setting of today’s patient a harbinger for Covid-related myocarditis?
  • Is preserved P wave amplitude (as was seen in ECG #1) in the face of shrinking QRS complexes an indication of soon-to-develop myocardial stunning?
  • IF the patient in today’s case were to fully recover — WHAT will be the time course for recovery of reasonable QRS amplitude on ECG and improved contractility on Echo?

Stay Tuned for answers!


  1. is there limb lead reversal in ECG#1? tall R in aVR and indeterminated axis

    1. @ Unknown — I also considered lead reversal. Reasons why I did not think there was lead reversal in ECG #1 were: i) the P wave was upright, and of maximal size (compared to other limb leads) in lead II, as it should be; ii) the P wave and T wave in lead I are both upright (as they should be); iii) the P wave and T wave are both negative in lead aVR (as they should be) — and there ARE times with normal lead placement when you can see a predominant R in lead aVR; iv) I thought the tiny QRS amplitude and pathologic process (ie, Covid myocarditis) were sufficient explanations for the limb lead appearance we see in ECG #1; and v) significant S waves (relative to the tiny R wave size) were also seen in the 2nd and 3rd tracings in lead I in this case.

  2. very interesting, and not a little bit depressing, especially in light of the fact that we had more than 160,000 new cases on thursday, and averaging over 1000 deaths a day.
    it seems (am i correct?) that the damaged involved myocytes (myocarditis)simply cannot generate the voltage for healthy QRS complexes, and thus the low voltage we see. Amal Mattu just gave a 16 hour ecg course for EMCases (our toronto friends), and devoted the last 60 minutes to covid related cardiac issues.
    and it looks that the need for our awareness in this pandemic is not going away any time soon.
    thank you both, Steve and Ken.

    1. Thanks for your comment Tom. Simplistically, my understanding is as you say — that with processes such as large MI, myocardial “stunning”, and myocarditis — that injured myocytes are unable to produce sufficient organization for effective cardiac contraction — leading to low voltage. The effect of COVID appears to add another contributing factor.

  3. Obviously, Cardiac Troponin (cTn) is a nonspecific marker of Myocardial injury,in the normal clinical situation, the use of (cTni) is for diagnosing or to rule out acute Myocardial infarction.

    1. I'm not sure what you are trying to say. There is no "normal" use of troponin. The most frequent use is for ruling in or ruling out MI. But it has myriad uses in risk stratification of all cardiac disorders.

  4. Although it was only present in V4-V6, the ST segment elevations reminded me of the spiked-helmet sign seen in other critically ill patients. To my knowledge though, the spiked-helmet sign is typically diffuse?

    1. @ Jeffrey — Excellent thought (re the QRST in leads V4,5,6 in ECG #2 of today’s case)! We discuss (and illustrate) the Spiked Helmet Sign in the June 28, 2020 post of Dr. Smith’s ECG Blog (Please scroll down the page to My Comment for illustrated discussion of this phenomenon — at — ).

      As I discuss in this June 28, 2020 post — the QRS-ST segment with the Spiked Helmet Sign typically shows 3 specific components; i) Some elevation of the isoelectric line that begins before the QRS complex (ie, making up the first half of the helmet); ii) Then sharp ascent of the R wave (ie, the “spike” in the helmet); and finally; iii) Coved ST elevation of varying degree, that may mimic an acute ST elevation MI (ie, the second half of the helmet).

      In contrast, in today’s case — ECG #2 DOES indeed show the 2nd & 3rd components of the sign = a bit of a spike in the R wave (that wasn’t present in ECG #1 in leads V5 and V6) - PLUS - the coved, unusually-shaped ST elevation that mimics an infarct (which wasn’t present in ECG #1) — but it LACKS the 1st component (ie, I do not see any elevation in the isoelectric line that begins before the QRS complex (as I illustrate in the June 28, 2020 post). Whether what we see in ECG #2 represents an early = “evolving” Spiked Helmet Sign, that would have developed that 3rd component had ECMO not been started I think is unknown.

      Thanks for suggesting this interesting consideration!

    2. I see what you mean, but the spiked helmet sign has "raised shoulders" on BOTH sides of the QRS, not just one. This is ST Elevation. See this example:

    3. Thank you for the clarification! Always appreciate the insight on this blog.


DEAR READER: I have loved receiving your comments, but I am no longer able to moderate them. Since the vast majority are SPAM, I need to moderate them all. Therefore, comments will rarely be published any more. So Sorry.

Recommended Resources