Friday, November 29, 2013

What Does this ECG tell us about the Electrolytes?

See description and answer below

Hint: this is a dialysis patient

1. There are peaked T-waves
2. There is a long ST segment, resulting in a long QTc (490 ms).  See the ST segments in inferior and lateral leads especially.

Diagnosis: hyperkalemia (6.0 mEq/L) + hypocalcemia (causing long ST segment)  (ionized Ca = 3.41 mEq/L)

Here is the patient's ECG when the K = 4.1 and the Calcium is also low:
ST segments and QT are long (hypocalcemia), but the T-waves are not peaked.

Monday, November 25, 2013

CHF Exacerbation with Old LBBB: Is There New Infarction or Not?

An very elderly woman with a h/o CAD with stents to the RCA and circ, mild aortic stenosis, h/o ischemic cardiomyopathy, EF 25%, and h/o heart failure, and with ICD for primary prevention, presented with a c/o chest and back pain intermittent for several nights, relieved by isosorbide.  On the evening of admission, she could not find her isosorbide and she became progressively SOB, in addition to chest and back pain.  She called 911, and she felt better on CPAP by EMS.  On arrival, her O2sat was 88%, then rose to 100% on BiPAP in the ED.  Her BP was 140/50, pulse 90.  She clearly had pulmonary edema.  She also had some peripheral pitting edema.  She was treated with intravenous nitroglycerin and furosemide. 

Here is her initial ECG:

What do you think?  Discussion below.  Multiple interpreters read this as left bundle branch block without any change or evidence of ischemia.

There was a previous ECG for comparison.
There is Left Bundle Branch Block.  All ST segments have appropriate and proportionate discordance.  There is left axis deviation, and the QRS duration is 179 ms (quite long for LBBB), both of which are associated with poor ejection fraction (EF) in LBBB.   There is a fragmented QRS, in this case "Cabrera's sign" (notch greater than 50 ms on the ascending limb of the S-wave in one of V3-V5) in V4 and V5.  A "fragmented QRS" is a sign of infarction, either new or old, analogous to Q-waves.  This ECG does not have "Chapman's sign," (notch on the ascending limb of the R-wave in I, aVL, or V6), but it does have, in addition to Cabrera's sign, other unnamed types of QRS fragmentation in many other leads: a notch in the ascending limb of the S-wave in III and aVF, and in the descending limb of the R-wave in I, II, and aVL.  V6 is also very fragmented.  

There is little angiographic data on fragmented QRS in LBBB.  The only one I know of was published recently as an abstract and had a small number of asymptomatic LBBB patients who otherwise were well match.  Among these patients, a fragmented QRS was highly specific for the presence of coronary artery disease.  In this study by Hands et al., QRS criteria were quite specific for MI of indeterminate age.

It is unlikely that a fragmented QRS on an isolated ECG with LBBB has good diagnostic utility for acute MI, but it is much more likely to be significant if there is NEW fragmentation

Here is the initial ECG again, annotated:
There is sinus rhythm and LBBB.  There is still left axis deviation and the QRS is now 193 ms.  There are multiple signs of new, acute ischemia.  
--The black arrow shows concordant ST depression in V3, and it is also clearly changed from the previous.
--The blue arrow shows even more concordant ST depression in V4.
--The red arrow shows an increase in QRS fragmentation.  Increased (new) fragmentation can also be seen in many other leads: Rsr' in I and aVL, Qrs in II, III, and aVF.
--The green arrow shows excessively discordant ST depression in lead V5.  This is also present in V6.

The concordant ST depression meets the second Sgarbossa criteria for STEMI in LBBB.  The other signs are supportive of MI.  The excessively discordant ST depression, as defined by a ratio of 30% of the preceding R-wave, had a sensitivity of 100% and specificity of 88% for coronary occlusion in our study of coronary occlusion in LBBB.

None of these signs were seen.  Fortunately, the woman was not a good candidate for, and did not want, invasive therapy. 

However, in a patient who is a candidate for PCI, these signs can be very helpful.

Her troponin I peaked at 28 ng/mL.  She improved on supportive and medical therapy.  Her subsequent echo showed new posterior and anterior (in addition to old inferior) wall motion abnormalities and a decrease in EF from 25% to 20%.

Here is her next day ECG:
The ischemia has resolved.  It is quite similar to the old ECG.

Finally, it is possible that the CHF exacerbation was the cause, not the result, of the ischemia; that there is severe stable CAD and that along with the hypoxia of a CHF exacerbation, she developed coronary ischemia.  ECG findings and positive troponin tell you there is infarction, but not definitively whether there is ACS.  However, concordant ST segments, and troponin values that high are almost always due to occlusion, which may or may not be transient.


In LBBB, there are many signs one can look for in order to diagnose, or strongly suspect, MI or coronary occlusion.   However, this is still not widely known by clinicians.  To illustrate this, one note said "LBBB, so no new ECG findings."

Had this been someone who could undergo angiogram and PCI, one would not want to miss these findings.  They could very well be signs of a coronary occlusion that could be opened and spare some myocardium.

Friday, November 22, 2013

You MUST recognize this pattern, even if it is not common

Disclaimer: I never state from where I get a case.  They come from all over the world.   Any case may or may not be from my own institution.  Do not make any assumptions.


A non-English speaking woman in her 60's with h/o HTN, type II DM, hyperlipidemia, CAD s/p CABG 16 years prior, and end stage renal disease on dialysis  presented to the ED at time 0.  She had awoken 7 hours prior with severe headache followed by upper chest heaviness and vomited x 2.   EMS placed an 18 gauge IV, gave 2 NTG, and aspirin.  She stated the pain was not similar to a previous MI.  BP was 200 systolic.   The patient stated that her chest pain was more of an issue than her headache.

Here is her initial ECG:

There is sinus rhythm, left axis deviation (probable LAFB), and new ST elevation in V1-V3.  What do you think?  See more discussion below.

What was not seen is that there is QRS widening (120 ms) and peaked T-waves diagnostic of hyperkalemia and a pathognomonic, though unusual, pseudoSTEMI pattern.  Cardiology was called and the cardiologist activated the cath lab.

The angiogram showed no acute disease.  However, following the procedure, during the sheath pull, she experienced VT arrest and was administered multiple shocks for wide complex tachycardia.  3 "amps" of lidocaine, 300 mg amiodarone, Sodium Bicarbonate x 2 "amps", D50 x 1 "amp", and calcium gluconate x 1 "amp" were all given.  She had return to sinus rhythm with narrow complex.   After resuscitation, she was comatose and was intubated.

It is uncertain when the K returned, but it was 7.5 mEq/L when it did.

Here is the post-resus ECG:
Now there is sinus bradycardia with an accelerated junctional rhythm and AV dissociation (no evidence of AV block on this ECG).  There is diffuse T-wave inversion, with massive T-waves, all but diagnostic of stress cardiomyopathy.

 The stress cardiomyopathy gets even more pronounced here (20 minutes after the previous):
Sinus brady with nodal escape and AV dissociation persists.  T-waves are massive and bizarre, with precordial T-waves inverted.

Echo confirmed tako-tsubo like apical ballooning of stress cardiomyopathy.  The stress was doubtless her cardiac arrest.

She was dialyzed, cooled (therapeutic hypothermia for comatose survivor of cardiac arrest).  One of my blog followers (an EM G2) was on the neurology service and recognized the initial ECG pattern as diagnostic of hyperkalemia.  He could do so because of these previous posts (and because he is very smart, hard-working, and attentive):

Hyperkalemia PseudoSTEMIs

HyperK pseudoSTEMI Post 1
HyperK pseudoSTEMI Post 2
HyperK pseudoSTEMI Post 3

Patient course

The patient awoke and went to rehab.  She was discharged with some minor neurologic deficits. 


1. Remember this pattern of ST elevation in hyperkalemia.  It is an unusual but pathognomonic pseudoSTEMI pattern
2. Potassium will hurt you and your patients if you are not always thinking about it!


Monday, November 18, 2013

Another STEMI in a Young Person with no CP or SOB: Why We Should Record an ECG for Unexplained Nonspecific Symptoms

There are many interesting facets to this case.  I have highlighted the clinical factors over the ECGs, but the ECGs are very instructive as well: even after successful PCI with good (TIMI-3) in the epicardial coronary artery, there are ECG signs of poor microvascular perfusion. 


A 31 year old male.  No known past medical history.  No cardiac risk factors.

Here is a great medical student history:

The patient presents to the ED with a complaint (through an interpreter) of a sore and itchy throat which led to feelings of generalized body pains that the patient describes as "cold pains" and feeling "tense/revved-up" or anxious and like he needed to "leave the room to scream, yell, and run around".  The throat pain is described as sore, tight, itchy and it resolved with the ending of each "episode", each of which lasted ~5-10 minutes and completely resolved. Similar episodes continued to recur throughout the day and would last ~5-10 minutes and resolve with periods of feeling normal in between each one.

He endorses similar episodes one time 3 months ago and 1 week ago.

He exercises regularly without exertional CP or dyspnea. He denies Chest Pain or tightness, palpitations, SOB, vision changes, diaphoresis, dizziness, or voice changes. He had an episode while in ED that included his usual symptoms + HA, mild epigastric discomfort (consistent with his previous "gastritis pain") and nausea. 

The faculty physician asked several times about chest pain or SOB and the answer was consistently "no."

Physical Exam: normal except that the patient had one episode while in the ED and had appeared anxious and kept changing positions.

Here is the ED ECG:
There is sinus rhythm.  There are anterior Q-waves, suggestive of previous infarction or prolonged infarction.  Inferior ST elevation diagnostic of inferior STEMI.  Convex ST elevation in V3 and V4 diagnostic of  anterior STEMI.  STE in V5 and V6 diagnostic of lateral STEMI.  ST depression in V2 suggestive of posterior involvement.

31 minutes later, while awaiting angiography, this was recorded:
The T-waves are inverting in V2-V4, suggesting that the artery is open now.

He was taken for angiography.  A 95% thrombotic occlusion (with TIMI-2 flow, consistent with the second ECG) of a type III (wraparound to inferior wall) LAD was found.  It was occluded proximal to the first diagonal to the lateral wall.  It was opened. There was much downstream showering of thrombi, some of which could be suctioned out.

Here is the post cath ECG:
There is persistent ST elevation.  There is less than 50%, and even less than 70%, resolution of ST elevation. This is a bad prognostic sign and suggests diffuse microvascular occlusion with persistent ischemia.  There is a high risk of development of LV aneurysm when the ST elevation does not resolve, especially when there are Q-waves already.
Here is the next day ECG:
ST elevation remains persistent.  No resolution of Q-waves.

The first troponin I was 2.3 ng/mL and the 4th was 95.3.  No peak was measured.  The echo showed anterior, apical, and inferior wall motion abnormalities.

This large possibly deadly STEMI could easily have been missed if the emergency physician was not attuned to the possibility of MI in patients with nonspecific symptoms.  ECGs are cheap and noninvasive.

The only problem with recording ECGs on such patients is not that the vast majority will be normal (they will be), but that many will have non-diagnostic findings such as non-diagnostic T-wave inversion.  One must learn which of the many nondiagnostic patterns are truly worrisome.

Mostly, in patients at very low risk and with very nonspecific symptoms, we are looking for the unequivocally diagnostic ECG.  When it is unequivocal, as in this case, it does not matter what the symptoms are: act on it.

Young patients: Young patients do have STEMI much more than is commonly recognized.  I have presented many including these two 20-somethings here and here.  When the cardiologist was called, the physician said, "OK, this is not straightforward, but hear me out." The cardiologist said "How old is the patient?" The answer was: "Thirty-one" and the cardiologist said, "Then it's not a STEMI."   

To his credit, he relented when he saw the ECG.

Door to EKG timethe door to EKG time was 3 hours in this case. But it is a miracle that any ECG was recorded.  Management of this case was exemplary because any ECG was recorded.  However, because it was recorded, it was recorded late, and thus the door to balloon time was long.  This sort of management could harm your statistics even though it helps your patient.  Institutions where a patient like this would not get an ECG would protect their DBT averages!!

STEMI patients without chest pain: In spite of a clear absence of chest pain or SOB on history done prior to the ECG, there were many mentions on the inpatient chart of the patient having had "chest discomfort" and "shortness of breath".  It is very common for physicians to impute chest discomfort after seeing a STEMI on the ECG.  They cannot imagine that someone can have a STEMI without any chest discomfort, or at least SOB.  But this does happen.  See this case in which, after seeing the ECG, I personally asked the patient many times if he had chest discomfort; his only symptom was left hand paresthesia.

Patients can be persuaded to say that they had chest pain.  Here is a scenario that I have seen many times (made a bit humorous):


"Are you sure you didn't have any chest pain?"
"Are you positive?!?"
"You must have had chest pain, no?"
"When did your chest pain stop?"

Patient: "OK, doctor, I did have a little chest pain, I think.  Maybe a little"

Doctor:  "Aha.  You did have chest pain!"

And then it is easy to imagine this doctor who interviewed the patient after seeing the ECG thinks, "Those ER docs, they just can't take a good history!"

But, seriously, in order to appreciate that many patients with MI and STEMI do not have chest pain or SOB, one has to experience these patients PRIOR to the ECG.  Or one must be blinded to the ECG.  This is how we in the emergency department, or medics on the ambulance, experience patients: we get to know their symptoms in a blinded way, in this case before we see an ECG.  

Friday, November 15, 2013

Why we need continuous 12-lead ST segment monitoring in Wellens' syndrome

Also see this incredible case of the use of 12-lead ST Segment monitoring.

This was a male in his 50's with a history of hypertension and possible diabetes mellitus who presented to the emergency department with a history of squeezing chest pain, lasting 5 minutes at a time, with several episodes over the past couple of months.  It was worse on the evening prior to presentation while lying in bed, then recurred and resolved while at rest just prior to arriving in the ED.

Here is the first ED ECG, with no pain:
Sinus rhythm.  Computerized QTc = 419.  There is very subtle Wellens' waves in V2, less subtle in V3, still  less in V4.  This is nearly diagnostic of Wellens' syndrome: pain that resolves, biphasic T-waves with preservation of R-waves.  It extends all the way out to V6, and there is some T flattening in I and aVL, so this is probably a proximal LAD lesion.
The ECG in the chart was read as "no obvious ST changes," (even though no previous ECG was available) and the formal read by the emergency physicians was: "ST deviation and moderated T-wave abnormality, consider lateral ischemia."

Comment: most T-wave inversion is nonspecific, but not these ones!  This T-wave inversion morphology is very specific for Wellens' waves.  You've read in my previous posts that I have a lot of evidence that Wellens' represents spontaneously reperfused STEMI in which the STEMI went unrecorded.  The T-wave inversions are of two types: Pattern A and Pattern B. They occur in the distribution of the LAD and should be in consecutive leads.  (They do also occur with inferior and lateral MI, but are somewhat different and these were not described by Wellens).

Pattern A are biphasic, but with a slightly elevated ST segment and a slight ST upslope.  The  T-wave then drops quickly, as in V2 and V3 here.  This morphology is worth memorizing!  If there is only one wave showing it, it is usually lead V2.

As hours go by, these T inversions always evolve, [unless 1) there is re-occlusion, in which case they go upright and become hyperacute, with or without additional ST elevation, ("pseudonormalize") or 2) no infarction at all (negative troponin, true unstable angina with dynamic T-waves, in which they may normalize).  The evolution is to deeper and more symmetric waves (Wellens pattern B).

Also, as you get further out towards lateral precordial leads, the T-waves are usually more symmetric and more deeply inverted.

The above principles are all well illustrated with this figure from my book, The ECG in Acute MI (2002).  See the Caption:

This figure illustrates the evolution of Wellens' waves from Type A to Type B:

Case Continued:

The physicians considered this ECG nondiagnostic.  Plan was for admission for chest pain workup.  Troponin I level was of course sent.  The first returned at 0.042, which is above the 99% reference of 0.025 ng/mL.  The patient remained asymptomatic, but because of the positive troponin, another ECG was recorded:

Sinus. Computerized QTc = 417.  New ST elevation diagnostic of STEMI [equation value = 25.3 (anterior STEMI) in case you are wondering]
This patient with acute anterior STEMI is without symptoms, even when questioned again!!

It is well documented with continuous 12-lead monitoring that acute re-occlusion is frequently asymptomatic.  Here is the classic article on continuous 12-lead monitoring (in full text) showing that the ECG is a much more reliable indicator of re-occlusion than are symptoms.

Here is another classic article.  It shows that on a 7 day angiogram, only 58% of re-occlusions were symptomatic: Ohman EM, Califf RM, Topol EJ et al.  Consequences of reocclusion after successful reperfusion therapy in acute myocardial infarction.  Circulation 1991;84:1454-1455.

It is important to recognize that coronary thrombosis is dynamic, with spontaneous opening and lysing of the thrombus in the infarct-related artery (we all have endogenous tPA and plasmin to lyse thrombi). The classic article by de Wood (see below) shows that even in completely untreated (not even aspirin) coronary thrombosis, the later the angiogram is done in coronary occlusion, the more likely the artery is to be open. 

Case Continued:

The cath lab was activated.  Before the patient went upstairs (15 minutes later), he remained asymptomatic and another ECG was recorded:
Spontaneous reperfusion again

At cath, there was an 80% mid-LAD active lesion (not proximal, interestingly).  It was stented.  Here is the post PCI ECG:
Minimal change

--Initial and 3 hours troponin I's were (ng/mL):  0.042, 0.054, 0.056, 0.040, 0.039, 0.037

--20 hour echocardiogram showed no wall motion abnormality.  There are those who think that an echo that is done after resolution of ischemia is sensitive for that previous ischemia.  It is not.  When the ischemia is resolved, the wall motion may completely recover, or there may be persistent stunning.


1.  Akkerhuis KM, et al. Recurrent ischaemia during continuous multilead ST-segment monitoring identifies patients with acute coronary syndromes at high risk of adverse cardiac events; meta-analysis of three studies involving 995 patients. Eur Heart J 2001;22:1997–2006.

2.  Gottlieb SO, et al. Silent ischemia as a marker for early unfavorable outcomes in patients with unstable angina. N Engl J Med 1986;314:1214–1219.

3. Jernberg T, et al. The combination of a continuous 12-lead ECG and troponin T; a valuable tool for risk stratification during the first 6 h in patients with chest pain and a non-diagnostic ECG. Eur Heart J 2000;21:1464–1472.

4. Patel DJ, et al. Early continuous ST segment monitoring in unstable angina: prognostic value additional to the clinical characteristics and the admission electrocardiogram. Heart 1996;75:222–228.

5. Alsaab A.  American Journal of Cardiology (Online First, In Press).  Here is the most recent documentation that T-wave inversion does indicate reperfusion:

6.  de Wood et al.  Prevalence of Total Coronary Occlusion during the Early Hours of Transmural Myocardial Infarction.  NEJM 1980;303(16):897-902.

Abstract To define the prevalence of total coronary occlusion in the hours after transmural myocardial infarction, we used coronary arteriography to study the degree of coronary obstruction in 322 patients admitted within 24 hours of infarction. Total coronary occlusion was observed in 110 of 126 patients (87 per cent) who were evaluated within four hours of the onset of symptoms; this proportion decreased significantly, to 37 of 57 (65 per cent), when patients were studied 12 to 24 hours after the onset of symptoms. Among 59 patients with angiographic features of coronary thrombosis, the thrombus was retrieved by Fogarty catheter in 52 (88 per cent) but was absent in seven (12 per cent false positive). Among an additional 20 patients without angiographic features of thrombosis, a thrombus was discovered in five (25 per cent false negative). Thus, total coronary occlusion is frequent during the early hours of transmural infarction and decreases in frequency during the initial 24 hours, suggesting that coronary spasm or thrombus formation with subsequent recanalization or both may be important in the evolution of infarction. (N Engl J Med. 1980; 303:897–902.)

Recommended Resources