Friday, January 14, 2011

Cardiac Arrest, Wide Complex, Is it STEMI?

For other cases of cardiac arrest, see these posts

This 200 kg patient was watching TV when he had a witnessed v fib arrest, and had a prolonged resuscitation. In the ED, his pH was 7.20, K of 4.9, and this was his initial ECG:

One might be tempted to diagnose ventricular tachycardia. However, there are p-waves best seen in lead I, with a slightly long PR interval. There are also PVCs which are unlikely to occur during V tach.
Once it is determined to be sinus tach (with aberrancy, in this case a slightly unusual LBBB), one must find the end of the QRS (J-point, the start of the ST segment) in order to find the ST segment and to evaluate for ST elevation or depression. In this case, the end of the QRS is easily found in lead V6. Then, draw a line through that J-point an extend it up and down. I have done this in detail in this previous case:

Here are the lines for this ECG:
Then you can find the end of the QRS in any lead. After doing so, you can see profound concordant ST depression in leads V2-V5, and Concordant ST elevation in I and aVL with reciprocal ST depression in II, III, aVF.

70 minutes after the first ECG, this ECG was recorded:

Now the QRS is narrower. The sinus rhythm is obvious. There is a R-wave in V6 typical of LBBB, but the ST segment is not discordant as it should be. Rather, it is concordant. V1-V3 have concordant ST depression.
The patient's initial troponin was negative. Because of other factors, he was not taken for PCI. Troponin 5.5 hours later was 308. ng/ml. Echo was technically difficult but showed LVH and probable lateral wall motion abnormality. This was presumably a large posterolateral STEMI with LBBB.


  1. Great blog, love it!

    Two questions:

    1) Garcia mentions in his book "12 lead ECG - the art of interpretation" that to determine the width of the QRS it should always be measured in the lead where it is the widest, since some parts of it could be isolelectric in other leads.

    Remembering this "rule", I was a little confused by your method of choosing the lead with the clearest J Point to detect the end of the QRS or the beginning of the ST segment in all other leads. How can I be sure that this lead doesnt involve an isoelectric segment of the QRS?

    2) the P waves in wide complex tachys:
    I know that sometimes in ventricular rhythms, P's can be transmitted retrogradely to the atria. How can I differentiate VTach with retrograde P's from retrograde P's from a low atrial or a junctional pacemaker with aberrency?

    Thanks for the help!

    PS: I may have missed it, but a case with some guidelines on how to differentiate LVH with strain from AMI on an ECG would be very helpful!

  2. 1) that is why you must find the lead where the end of the QRS is clear. That end must also not be inconsistent with other leads. If you look in this case, you can see that not only is the end clear in lead V6, but it also corresponds to a reasonable endpoint in all other leads.

    2) It may be very difficult, but, first, when there are p-waves, a low atrial or junctional pacemaker is comparatively infrequent compared to sinus tach and, second, the complex (preceding or succeeding) that is most proximate to the p-wave is more likely to be the related one, unless there is a very long PR interval.

  3. Dr. Smith,

    This case makes me think of a hypothetically similar one:

    Consider a patient with ischemic symptoms whose ECG shows LBBB, with large ( let’s say 40mm) S waves in the anterior leads. Lets consider these anterior leads only. In previous posts you have described two ways to identify acute coronary occlusion in these anterior leads:

    1) Concordant ST depression (relative to the PR segment)
    2) Excessively discordant ST elevation (>0.2 ST/S ratio)

    Furthermore, you have done a study that found that the normal ratio of ST elevation in anterior leads of LBBB is around 0.10 (ST/S ratio).

    Knowing that this patient should have about 4mm (0.1 x 40 = 4) of appropriate, discordant ST elevation, why would we use the baseline (PR segment) to establish the threshold for “concordant” changes?

    If this patient were to suffer an acute occlusion of the PDA (isolated posterior STEMI) isn’t is possible that the ST segment could deviate negatively (downward, concordantly) below the normal 0.1 ST/S ratio baseline, yet still be above the PR or TP segment? In normal conduction, 3mm of ST depression in the anterior leads could certainly be diagnostic of isolated posterior MI, and would certainly make me get posterior leads. In this theoretical case, however, 3mm of ST depression would result in neither excessive discordance nor concordance (as defined relative to the PR segment). It would result in what I imagine as “insufficient ST elevation” or “proportional concordance.”

    The question I am getting at is this: if LBBB has 0.1 ST/S ratio normally, shouldn’t concordance (ideally, if not practically) be measured with respect to that, instead of the PR/TP segment?

    If true, this would have several implications:

    1) In your study of excessive discordance, were there any isolated posterior occlusions that defied your modified criteria? If so, did they show abnormally small (between 0 and 0.1 ST/S ratio), e.g., “insufficient ST elevation”? If so, could this have accounted for some lost sensitivity?

    2) Just as you have quantified the excessive discordance as 0.2 ST/S ratio, shouldn’t we also be able to quantify concordance as well, instead of assuming that it is simply 0.0 ST/S ratio? What if a study was done that showed the lower limit of the ST/S ratio in the anterior leads was 0.05? Could this identify isolated posterior STEMI’s in the presence of LBBB in which the ST segment didn’t quite make it down to the PR segment?

    3)Of course, this would only be easily visible on LBBB with huge S waves. And of course isolated posterior MI is rare enough, not to mention in the setting of LBBB.

    Do you have any examples that would agree with my thoughts? Have you ever seen an isolated posterior STEMI with LBBB that did not have traditionally concordant (below the PR segment) ST depression?

    Or is it your opinion that, due to the rule of proportionality, an isolated posterior occlusion will almost always deviate the ST/S ratio below 0.0, independent of the magnitude of the S wave? If the proportional cutoff for concordance was truly 0.0 ST/S ratio, that would be very convenient indeed. In reality, it may be so close to 0.0 that it can be approximated to 0.0.

    What are your thoughts on this?


  4. Pendell,
    --Your logic is impeccable.
    --I have never seen a proven isolated posterior MI with "relative" ST depression (relative to baseline, or a normal amount of ST elevation). I have seen patients with normal conduction (early repol) who have posterior MI and their ST segments went from elevated to isoelectric. I also have a case in my book of a patient with LBBB and an inferoposterior STEMI who had relative ST depression in precordial leads (proven by comparison with his baseline ECG). So there is no doubt in my mind that a patient with LBBB and isolated posterior STEMI could present with simple 1 mm, rather than 3 or 4 mm of ST elevation, or zero mm. The problem is finding enough cases to prove it. The other problem is that, even though it has been shown that 8% of all coronary occlusions present only with precordial ST depression (and thus are posterior STEMI), it is hard to get interventionalists to act on isolated posterior STEMI. Now, put LBBB together with an isolated posterior STEMI that is only showing up by relative ST depression, and you'll REALLY have a hard time getting the cardiologist to act.

    --Having said that, it is always worth studying and you should do so if you're interested because every one of your ideas is right on!

    --Lastly, a corollary: if all ST segments should be discordant, then in cases in which the QRS is positive, then wouldn't an isoelectric ST segment, or one with only 0.5 mm of ST elevation, be some degree of concordance? This we studied: I believed that ANY amount of concordant ST elevation, even 0.5 mm, would be more sensitive. It was, but it was less specific: in other words, there were non-occlusions that had this. Finally, I did not measure these ECGs. They were measured by undergrads and med students and there was interrater reliability issues that are very difficult to get into.

    What would the outcome be if experts read the ECGs with these rules in hand?

    Steve Smith

    Steve Smith


DEAR READER: We welcome your Comments! Unfortunately — due to a recent marked increase in SPAM — we have had to restrict commenting to Users with a GOOGLE Account. If you do not yet have a Google account — it should not take long to register. Comments give US feedback on how well Dr. Smith’s ECG Blog is addressing your needs — and they help to clarify concepts of interest to all readers. THANK YOU for your continued support!

Recommended Resources