Thursday, March 31, 2016

Altered Mental Status with Agitation and Tachycardia

This was originally posted on May 8, 2014.  It was one of the FOAM posts used to evaluate the ALiEM AIR Rating for FOAMed resources, just published online at Annals of EM.  It got pretty good ratings from the eight educators who evaluated all the posts, so I thought I would repost it.  Here is the list of posts that were evaluated.

A young male with unknown past medical history presents with AMS and tachycardia. EMS was called by a roommate after the patient was noticed to be nonverbal and lethargic. He reportedly took meth and had a recent drinking binge, but has not had alcohol for the last 2 days. His HR was 160 on arrival of EMS, and they gave him adenosine 6 mg and 12 mg and 500 cc NS, but with no response.  The preshospital ECG and strips are not available.  The patient was restless, agitated, and nonverbal on arrival to ED, with elevated HR at 150.  Here is the first ED ECG:
What is the likely diagnosis?

There is sinus tachycardia, a prolonged QRS (computer read it as 114 ms, previous ECG with 102 ms).  There is a large R-wave in lead aVR.  These findings are nearly pathognomonic of, or at least highly suspicious for, tricyclic antidepressant (TCA) overdose.

In addition, the QT interval is 360, and the Bazett-corrected QT interval (QTc) is 540 ms.

One can read about management of TCA overdose in many places, and I have put some links to lifeinthefastlane below.

A few important points:

1. An R-wave in lead aVR  greater than 3 mm, or an R/S ratio greater than 0.7, is highly suspicious for sodium channel blockade, which is the most important of the many toxicities of TCA overdose.

2. In 49 patients with known TCA overdose, a maximum limb lead QRS duration greater than 100 ms was 100% sensitive for detecting patients who will seize, and seizure is a harbinger of cardiovascular collapse.  At this cutoff of 100 ms, however, the specificity was not perfect.  Of 11 patients with an initial QRS duration of 100-119 ms, 2 (18%) had seizures, and of 22 with an initial QRS duration of 100-139 ms, 4 (18%) had seizures.  Of 14 with initial QRS of 140 ms or longer, 8 (56%) had seizures.  No patient with a QRS of less than 160 ms had ventricular dysrhythmias.

2a.  In an unselected population with suspicion of overdose, a minimally wide QRS (less than 110 ms) will be much less specific; furthermore, it is likely that frequent serial ECGs, by detecting an increasing QRS duration, will detect those at risk of toxicity.  On the other hand, administration of bicarbonate, the antidote, is relatively safe compared with a seizure.  If the diagnosis is unclear, narrowing of a widened QRS on ECG following sodium bicarbonate administration (1-2meq/kg) adds further support that pharmacologic sodium blockade is present.

3. There are multiple mechanisms of toxicity of TCAs:
a.   Blockade of cardiac fast sodium channels (leads to wide QRS, R-wave in aVR, R' wave in V1, Brugada pattern ECG, ventricular dysrhythmias.)  Sodium channel blockade in the CNS leads to seizures.
b.   Blockade of potassium channels which leads to long QT (as in this case) and torsade
c.   Antagonism of central and peripheral muscarinic acetylcholine receptors (leads to delirium)
d.   Antagonism of peripheral alpha-1 adrenergic receptors (causes hypotension)
e.  Antagonism of histamine (H1) receptors (may contribute to sedation)
f.  Antagonism of CNS gamma-aminobutyric acid (GABA) A receptors (increase risk of seizures)
g.  Exaggeration of therapeutic effect of inhibiting central serotonin re-uptake. (may cause/contribute to serotonin syndrome)
h.  Exaggeration of therapeutic effect of inhibiting central norepinephrine re-uptake (increase risk of seizures)


When it became clear that the patient had sinus tach and not PSVT, his presentation was recognized as an overdose or drug toxicitiy.  But because the physicians were so focused on his tachycardia, meth use, and rhythm, they did not look for or appreciate the findings of TCA overdose.  We in emergency medicine obtain ECGs in overdoses mostly to look for TCA findings.  So when you get an ECG in this situation, look for them!

Much later, the roommate called and reported that 20 amitryptiline (unknown mg per dose) were missing.  Fortunately, the patient had not had any adverse outcome by that time.
He was given multiple amps of bicarb and a bicarbonate drip.  He remained delirious and was given 3 mg of physostigmine (after pretreatment with 2 mg of lorazepam to prophylax against seizures).  His delirium greatly improved and he was then able to follow commands.

He had a prolonged stay in the ICU requiring days of bicarbonate.  The tox screen only showed amitryptiline.

Articles on TCA

More TCA ECGs from Dr. Smith's ECG Blog

More on TCA overdose, with ECGs, from life in the fast lane.

More still on TCA overdose, from lifeinthefastlane.

Tuesday, March 29, 2016

Wide Complex Tachycardia

An otherwise healthy woman in her 20's with no past medical history presented with tachycardia.  She had experienced palpitations and called 911.  Prehospital rhythm strips were at a rate of at least 200 (unavailable) and the medics gave adenosine at both 6 mg and 12 mg with no effect.  She was stable, with no CP, SOB, hypotension or evidence of shock.

Here is the initial ED ECG:
What is the diagnosis (this is pathognomonic)?  See below. 
(Notice that the computer incorrectly read ***Acute MI***)

1. The rhythm is irregularly irregular, therefore it is atrial fibrillation
2. The complexes are wide (so one might think of atrial fibrillation with aberrancy, in which case you should see RBBB or LBBB pattern, which is not there)
3. It is very fast (200 bpm)
4. The shortest R-R interval (between complexes 12 and 13) is about 240 ms (very short)
5. The complexes look bizarre and multiform. They are not uniform, as they would be with simple aberrancy.  Thus, these represent differentially pre-excited ventricular myocardium.

This is atrial fibrillation in the setting of WPW, and is a dangerous rhythm which can degenerate into ventricular fibrillation.  It is more likely to degenerate if the physicians gives AV nodal blocking drugs, including adenosine (so the medics were lucky they did not make things worse), but especially calcium channel blockers.

This was first reported with verapamil (a calcium channel blocker), which actually increases conduction down the accessory pathway, increasing the heart rate, and frequently resulting in ventricular fibrillation.

There is another example of this rhtyhm at this post.

How should this be managed?  It can be managed with medications that convert atrial fibrillation to sinus, such as procainamide or ibutilide (and others), but when you have a wide complex very fast tachycardia, it is best to use electrical cardioversion.  It is the safest, and keeps you from having to make a definite diagnosis.   As long as you can manage procedural sedation, which is very easy in the case of cardioversion because you only need seconds of sedation and amnesia, then cardioversion is the safest method.

Even in the ED, the pattern was not recognized as atrial fibrillation with pre-excitation, but as atrial fib with aberrancy.  These were very smart and very experienced physicians, but anyone can make a snap judgment followed by premature closure (this is another reason why electricity is the safest - you don't need to have the correct diagnosis).

Subsequently, the physicians gave the patient first diltiazem and then esmolol, with no ill effect but also little beneficial effect.  They were lucky they did not make things worse.  Either of these medications, but especially the diltiazem, could have resulted in Ventricular Fibrillation.  AV nodal blockade is particularly dangerous when there is a shortest R-R interval of less than 250 ms (as here), and especially if less than 220 ms.  As I mentioned above, adenosine is also contraindicated.

After dialing up the esmolol, the patient spontaneously converted to NSR and had the following ECG:
Sinus rhythm with short PR interval and large delta waves seen best in precordial leads, and confirming Wolff Parkinson White (WPW) syndrome of pre-excitation down an accessory pathway.  There can be positive, negative, or isoelectric delta waves depending on their own axis.  WPW may greatly change both depolarization (in this case with large upright R-waves in right precordial leads because the accessory pathway is left lateral, depolarizing the ventricle from left to right) and repolarization (see these cases for acute MI mimics due to WPW)

Followup, and what is "Concealed Conduction"?

Her charts revealed 3 previous visits for palpitations, and in all cases the ECG was interpreted as normal.  Here is one of them:
There are very subtle delta waves which were not noticed by the computer algorithm or by the reading physician.    The computer read a QRS duration of 106 ms which is borderline long and due to these subtle delta waves.  The PR interval is normal because the delta waves are so minimal.

When the delta wave is absent, it is sometimes called concealed conduction.  In this case, it is nearly concealed.  That is to say, the presence of an accessory pathway is not obvious (at least to most observers) on the baseline surface ECG.  (There are times when it is truly concealed, and there are no delta waves even in retrospect).  It is important to know about concealed conduction so that if you suspect WPW as a cause of tachycardia that is now resolved, you will not rule out the diagnosis by a normal baseline ECG.

There are two mechanisms of concealed conduction:

1. Conduction through the accessory pathway is retrograde only (mechanism unclear)
2. The impulse reaches the AV node and gets through to the ventricles before it gets to, and through, the accessory pathway.

The second mechanism applies in this case: Look at the first two ECGs above (those with abnormal conduction). Notice that the R-wave in V1 is very large, as it would be in RBBB.  This is because the impulse is going down a left lateral bypass tract and then proceeding through the myocardium from left to right, resulting in a large R-wave in V1.  Thus, the bypass tract (accessory pathway) is to the left lateral of the left atrium, which is far from the sinus node (right part of right atrium).  When the AV node is conducting fast (such as with anxiety, low vagal tone, high catecholamines, etc.), then the impulse gets to and through the AV node and through the Purkinje system before it makes it down the accessory pathway and therefore there is no (or minimal) delta wave.  On the other hand, if the AV node conduction is slower, then the delta waves will  be evident.

However, even in WPW with concealed conduction, the accessory pathway is always available to cause trouble!  

In both types of concealed conduction, it can result in orthodromic reciprocating supraventricular tachycardia (re-entrant down through the AV node and up through the bypass tract), and this cannot be differentiated from intranodal reentrant (standard) SVT on the surface ECG.

In the second type of concealed conduction, it can result in three abnormal rhythms: First, there can be orthodromic re-entrant reciprocating tachycardia.  Second, there can also be antidromic re-entrant reciprocating tachycardia (which is a regular wide complex tachycardia). And third, if atrial fibrillation develops, then it will manifest as this dangerous wide complex tachycardia.

Why did she not have concealed conduction on the post conversion ECG?   --Because she had received AV nodal blockade with diltiazem and the pathway down the AV node was slow.

Saturday, March 26, 2016

"Shark fin" ECG in I, aVL, V4 and V5. Which artery? Hint: patient is in shock and was put on ECMO

This was contributed by Rohin Francis (Twitter: @MedCrisis), a cardiologist from England and FOAM enthusiast.


A 55 year old lady initially presented to hospital with an acutely ischaemic arm. An embolic occlusion of her brachial artery was diagnosed by CT and treated with anticoagulation. The following day she developed sudden severe chest pain. This ECG was obtained:
Sinus rhythm.
The rather alarming appearance of the QRST may be mistaken for a broad complex QRS but,  in fact, her QRS complex can be clearly seen in V2, V3 and II and is narrow. What has manifested as triangular complexes is actually huge ST segment elevation seen in V2-5 and laterally in I and aVL. There is also ST segment depression inferiorly.  A proximal LAD occlusion can produce anterolateral ST elevation, but if the circumflex is also occluded, it is possible the anterior ST elevation might be attenuated and indeed here the ST elevation is less pronounced in V2 and V3.

Smith comment: Note there is ST depression in aVR, a true sign of Left Main occlusion.  Many authors state that ST elevation in aVR is a good ECG sign of left main occlusion.  This is erroneous.  (See this post for more explanation).

ST elevation in aVR is a sign of left main or 3 vessel insufficiency (not occlusion), due to diffuse subendocardial ischemia, which results in leftward and anterior ST depression (towards lead II and V5) and reciprocal ST elevation in aVR.
Left main occlusion is identical to simultaneous occlusion of the Proximal LAD (anterolateral STE: STE in V2-V6, I, aVL) and circumflex (posterolateral STE: STE in aVL, STD in V1-V4).  The opposing ST vectors of anterior and posterior can (at least partly) cancel each other out in V1-V4, so that the lateral has the most obvious STE, as in this case.  

Lead aVR is opposite an imaginary lead between leads I and II, often called (-aVR).  You can see that left main occlusion causes high lateral STE (aVL) with reciprocal STD (leads II and III); since aVR is reciprocal to this ST elevation, there is ST depression in lead aVR.

Case continued

She was immediately transferred to the primary PCI centre. On arrival she was in cardiogenic shock and haemodynamically unstable. Her BP was 80 systolic, GCS 13 and she had a grey appearance.

An intra-aortic balloon pump was inserted first and the cardiac arrest and ECMO teams were alerted. This was the initial shot of her angiogram:
The guide catheter was deliberately not engaged with the left main artery, as an occlusion was suspected. Here one can see contrast being injected towards the left main and a large thrombus sitting at the ostium. A small amount of flow is passing but not reaching further than proximal circumflex nor LAD. There is sluggish clearance of contrast from the aortic root indicative of poor cardiac output. An intra-aortic balloon pump can be seen inflating and deflating in the descending aorta.Smith comment: There is near total obstruction of the Left Main.  I am no angiographer, but I believe that this is TIMI-1 flow, which results in ST elevation.  It is not TIMI-0 flow (total obstruction/occlusion), as some minimal contrast gets through.   All thrombi are dynamic, however, and the only certainty one can have about the condition of the artery 45 minutes earlier, during recording of the ECG, comes from ECG analysis.

Passing angioplasty wires into her coronary arteries restored flow to the circumflex but not LAD. Her left main and LAD were opened using very careful aspiration and she received a drug eluting stent to her proximal left main.

Here is the aspirated thrombus:
Large red thrombus aspirated

Despite revascularisation and inotropic support, she remained in cardiogenic shock. An on-table echocardiogram revealed severe global impairment of her left ventricle. She was shocked out of VF on several occasions but continued to arrest so was placed onto femoral VA ECMO.

Peripheral ECMO increases afterload by introducing blood into the aorta and if there is no cardiac ejection, it can cause dilatation of the LV. Unfortunately despite maximal therapy in ICU, she continued to deteriorate. 

A trans-oesophageal echocardiogram revealed extensive thrombus in the aortic root and left ventricle. She died later that day. A post mortem was not performed but disparate embolic events in her arm and coronary circulation might point to an intra-cardiac source of thrombus. Thrombi tend to form in impaired ventricles, but it is unclear if the severe left failure seen on echo was pre-existing or simply a result of a catastrophic acute coronary syndrome.

Thursday, March 24, 2016

ED Diagnosis of Acute Coronary Occlusion by Speckle Tracking (Strain Echocardiography)

We just published this case report, the full text of which you can read here.  The ECG is diagnostic of coronary occlusion also, but many interventionalists might not believe the ECG.  Speckle tracking proves it.

A Pathognomonic ECG

This was sent to me by Mauro Cassazza from Italy:
What do you see?


It is tempting to think that the waves before the QRS are large P-waves, but in fact one can discern P-waves within these large waves (see V2 in particular).

Thus, these are very late T-waves.   Thus the QT interval is very long.

I measure the QT interval at about 520 ms.  Since the RR interval is about 560 ms, then the Bazett-corrected QTc is about 690 ms (very long).  

What part of the QT interval is long?   Is the T-wave itself very wide?  Is the time from T-wave peak to T-wave end (T-peak to T-end, or TpTe, long)?  No!  

TpTe is perhaps the best predictor of Torsade de Pointe (TdP).

The long QT is due to a very long ST segment.   This is pathognomonic of hypocalcemia.  The fact that hypocalcemia affects the ST segment and not the TpTe interval may account for the fact that TdP from long QT due to hypocalcemia is exceedingly rare, with only a few case reports (in which the etiologic link to low calcium is not even definitely established).  

In hypocalcemia (long ST, but NOT a wide T-wave, not a long TpTe), depolarization has a long duration, but repolarization occurs relatively quickly.  

The ionized calcium was 1.08 mg/dL.  I do not have information on the etiology.

Clinical history

This was an elderly woman who presented with altered mental status.  Vital signs were normal except for an oxygen saturation of 91% and temperature of 35C.
Exam had peripheral vasoconstriction, but otherwise normal
Bedside US: normal
Chest x-ray: normal
No tetany!
Calcium gluconate was given.

After the Ca was replenished to 4.09 mg/dL, this ECG was recorded:
The T-waves are somewhat flat, but you can see that the QT interval is now about 370 ms, QTc about 460 ms, and a normal ST segment.

The patient's mental status normalized.
The etiology of the hypocalcemia was not investigated.

Learning Points:

1. Hypocalcemia causes a long QT by lengthening the ST segment.
2. The T-wave duration, specifically the TpTe does not lengthen
3. It is a long TpTe which is most responsible for Torsades in long QT
4. Thus, long QT in hypocalcemia does not appear to have nearly the risk of long QT from other etiologies.

T-Peak to T-end (TpTe)
This comes from a manuscript that Dr. Ken Dodd and I have written.  Ken was my primary co-author on the derivation of the Modified Sgarbossa Criteria. 

"In patients with normal conduction, the TpTe interval has been noted to prolong during myocardial ischemia and correct after reperfusion [1, 2]. It has also been claimed to predict mortality after myocardial infarction and has been called the best determinant of TdP development in the setting of acquired bradydysrhythmias [3]. In patients with LBBB, prolongation of the TpTe has been shown to be independently associated with sudden cardiac death [4]."

[1]     Eslami V, Safi M, Taherkhani M, Adibi A, Movahed MR. Evaluation of QT, QT dispersion, and T-wave peak to end time changes after primary percutaneous coronary intervention in patients presenting with acute ST-elevation myocardial infarction. J Invasive Cardiol 2013;25:232–4.

[2]     Gupta P, Patel C, Patel H, Narayanaswamy S, Malhotra B, Green JT, et al. T(p-e)/QT ratio as an index of arrhythmogenesis. J Electrocardiol 2008;41:567–74. doi:10.1016/j.jelectrocard.2008.07.016.

[3]     Topilski I, Rogowski O, Rosso R, Justo D, Copperman Y, Glikson M, et al. The morphology of the QT interval predicts torsade de pointes during acquired bradyarrhythmias. J Am Coll Cardiol 2007;49:320–8. doi:10.1016/j.jacc.2006.08.058.

[4]     Panikkath R, Reinier K, Uy-Evanado A, Teodorescu C, Hattenhauer J, Mariani R, et al. Prolonged Tpeak-to-tend interval on the resting ECG is associated with increased risk of sudden cardiac death. Circ Arrhythm Electrophysiol 2011;4:441–7. doi:10.1161/CIRCEP.110.960658.

Monday, March 21, 2016

What, besides large anterior STEMI, is so ominous about this ECG?

This late middle-aged male had sudden onset of chest pressure.  Here is the prehospital ECG:
The rhythm is a bit hard to discern.  It appears to be sinus with PACs.
It is obviously an anterior (and lateral) STEMI.
What other ominous finding is present?

The cath lab was activated prehospital, and on arrival, the patient was in shock, with hypotension and an StO2 of 55% (this is a tissue oxygen saturation; normal is 75%.  55% represents shock no matter what the blood pressure).  He denied SOB but immediate bedside ultrasound showed B-lines of pulmonary edema.  Oxygen saturations were 94% on nasal cannula.  BP was approximately 90 systolic.

Here is the ED ECG:
Now there is atrial fibrillation with moderately rapid ventricular rate (126)
There is huge ST elevation in V2-V4 and I and aVL, diagnostic of proximal LAD occlusion
Here is the ominous finding: 

Right Bundle Branch Block (RBBB) + LAFB (left anterior fascicular block).

I have seen a dozen cases of STEMI with RBBB and LAFB.  All were close to death.  This article by Widimsky illustrates the danger of this finding:

See these cases:

Chest Pain and Right Bundle Branch Block

1. Aspirin 325 mg
2. Ticagrelor 180 mg
3. Atorvastatin 80 mg (small studies support this)
4. Heparin bolus
5. Fluid challenge
6. Because cardiogenic shock is likely to get worse, even after reperfusion (because myocardial stunning lasts many days), we intubated the patient.
7. Vecuronium paralysis
8. Ketamine sedation (to avoid affecting hemodynamics)
9. K replacement.
10.  Should have cardioverted, but did not

By this time, the cath team was ready.

A proximal LAD thrombotic occlusion was opened.  Here is the post-reperfusion ECG:
Sinus rhythm at a rate of 117.
Uncertain if conversion was spontaneous, or done electrically in the cath lab
RBBB and LAFB persist (not a good sign).
ST elevation is greatly improved (a good sign), though still very elevated (not a good sign).
There is terminal T-wave inversion, an early sign of reperfusion.

A balloon pump was placed.

Here is the ECG the next day:
Q-waves and poor R-wave progression.  RBBB and LAFB are gone.  (A good sign).

Highest troponin I was extremely high at 230 ng/mL at 10 hours after arrival.

Echo showed anterolateral wall motion abnormality and EF of 35%.

In spite of maximal supportive therapy, balloon pump, and persistently open arteries, the patient succumbed to cardiogenic shock 5 days later.

Learning Points:

1. Cardiogenic shock due to STEMI has very high mortality even if the artery is opened.  The mortality is approximately 50%, and, surprisingly, use of a balloon pump appears to not change that terrible prognosis. (1)
2. RBBB and LAFB are signs of very severe ischemia
3. Although I know of no supporting literature, I almost always intubate STEMI patients with shock and pulmonary edema.  Work of breathing may require 50% of cardiac output.  Positive pressure ventilation and paralysis takes away all that work and all that excessive requirment for cardiac output.

1)  Thiele H et al.  Intra-aortic balloon counterpulsation in acute myocardial infarction complicated by cardiogenic shock (IABP-SHOCK II): final 12 month results of a randomised, open-label trial.  Lancet.   Volume 382, Issue 9905, 16–22 November 2013, Pages 1638–1645.

Friday, March 18, 2016

A Middle-Aged Woman with 3 days of Chest pain and Posterolateral Injury on the ECG

A middle-aged woman called 911 for burning chest pain and dyspnea.  Here is her prehospital ECG:
There is clearly acute myocardial injury with ST elevation in high lateral leads.

The computer did not read STEMI, but medics were very suspicious and brought the patient to the critical care area.

An ED ECG was immediately recorded:
There is sinus tachycardia.  There is clear high lateral and posterior injury.

The cath lab was activated.


The patient states she has had 3 days of burning chest pain and SOB.  She thinks she has the flu and states that she is currently pain free.

These are her first vital signs:

A bedside cardiac ultrasound was done:

Here is the parasternal long axis:
There is a moderate size pericardial effusion and some collapse of the RV

Here is the short axis:

Technique is not optimal, but there appears to be a lateral wall motion abnormality and poor ejection fraction.

With the prolonged pain and pericardial fluid, there was concern for subacute MI with myocardial rupture, and also for myocarditis/pericarditis.

Clinical course:

The angiogram was normal.

Here is the post cath ECG:
Even more ST elevation, and ST depression reciprocal to injury of the posterior wall

A formal echo showed:
Decreased left ventricular systolic performance moderate .
Regional wall motion abnormality-lateral .
Regional wall motion abnormality-inferolateral .
Regional wall motion abnormality-anterolateral probable.
The estimated left ventricular ejection fraction is 35 %

The first troponin I returned at 33.7 ng/mL (33,700 ng/L)
3 hours:      30 ng/mL
8 hours:      32 ng/mL
10 hours:    43 ng/mL
44 hours:    16 ng/mL

ECG 2 days later:

Echo 2 days later:
Sinus tachycardia of 125-130 bpm.
Normal left ventricular size, mild hypertrophy and severe systolic dysfunction.
The estimated left ventricular ejection fraction is 25-30 % (visually estimated).

The patient became quite ill from cardiogenic shock due to myocarditis.

Learning Point

This case demonstrates that it is all but impossible to differentiate between acute MI and myocarditis.

Both may have:
1. ECG: Focal ST elevation with reciprocal ST depression
2. Echo: Focal wall motion abnormality
3. Very high troponin
4. Pericardial fluid (common in myo-pericarditis), possible in subacute MI if there is either postinfarction pericarditis or myocardial rupture

With the presence of fever, one might suspect myocarditis, but it would be dangerous to make any conclusions before studying the coronaries.

Recommended Resources