Tuesday, September 4, 2012

Modified Sgarbossa Rule Published Online: Annals of Emergency Medicine

Link: Diagnosis of ST-Elevation Myocardial Infarction in the Presence of Left Bundle Branch Block With the ST-Elevation to S-Wave Ratio in a Modified Sgarbossa Rule

There are differences between previous posts and the findings in this paper.  

Previously, the best ratio was 0.20.  Due to slight differences in methodology, the final rule uses 0.25.  It is important to realized that the use of 0.20 will result in slightly higher sensitivity and lower specificity for STEMI. 

Also, I did not use the absolute value of the ratio.  Thus, whereas, before, excessive discordance was greater than 0.20, it is now less than -0.25 (less than a negative number).  This may be confusing, but was more accurate in terms of simple arithmetic (dividing a positive number by a negative one).

Thus, for the revised rule, the third component of the rule [greater than or equal to 5 mm discordant ST elevation in leads with a negative QRS (S-wave)] is replaced by a ratio of ST elevation at the J-point, relative to the PR interval (a positive number), divided by the preceding S-wave (a negative number, so the result is a negative number) that is less than or equal to -0.25, was far more sensitive and was more accurate than the Sgarbossa rule at diagnosing coronary occlusion.  Additionally, the discordant ST elevation must be at least 1 mm.  The criteria need to be met in only one lead to be positive.

Furthermore, we found that a simple rule using only any excessive discordance (excessively discordant ST elevation or ST depression in just one lead, without paying attention to concordance), with a ratio less than or equal to -0.30, was the most sensitive (100%), with excellent specificity (88%) and the best accuracy.

Both rules need validation in another study.  We are working on that.

The full text is not free now; I'm not sure if it will be when published in print.

Here are some example cases.

17 comments:

  1. Can you clarify the simple rule that you mention in the last paragraph. What is the ratio you are using? Is it the ST/S ratio?

    ReplyDelete
    Replies
    1. Yes, it is the ST/S ration whether the ST segment is elevated or depressed, but always discordant to the predominant part of the QRS. That is, excessively discordant ST depression OR excessively discordant ST elevation.

      Delete
  2. I have read this full text,
    I have a question ,How to calculate prevalence for relate pretest and post test probability in this study?
    Are you have likelihood ratio nomogram?
    Thank you

    ReplyDelete
    Replies
    1. Good question. The exact pretest prob of coronary occlusion in patients presenting to the ED with chest pain and/or SOB is uncertain, but is probably about 2-3%. Therefore, a positive modified Sgarbossa criteria would still have a low post test probability, but the consequences of missing the diagnosis are severe, so a false + is not so bad. The post test prob of a negative mod Sgarbossa would be incredibly low and all but rule out occlusion
      Steve Smith

      Delete
    2. thank you very much for you answer
      but I'm sorry to bother you again , I can't find a 2 by 2 table in this full text for calculate sensitivity,specitivity,LR+,LR- by myself

      could you please tell me about how to computation 2 by 2 table from data in this journal.Or are you have more data for calculate 2 by 2 table?

      thank in advance.

      Delete
    3. With out using a nomogram , the ST/S,R ratio is directly infered from the previous Sgarbossa ECG criteria for LBBB and STEME .

      Delete
    4. not sure what you mean here, either

      Delete
  3. There is not a 2 x 2 table, but all the necessary data can be found in Table 2 if you dig a little.
    Steve Smith

    ReplyDelete
  4. Hello Dr. Smith,
    When you say LBBB does this include pacemakers? The paper did not mention this patient population.

    Salim

    Salim

    ReplyDelete
    Replies
    1. Salim, we did not study patients with paced rhythm. There are two previous papers on use of sgarbossa criteria in paced rhythm, but both with very few cases. I nevertheless think it is safe to treat paced rhythm like LBBB and have few cases here on my blog which illustrate STEMI in paced rhythm.
      Steve

      Delete
    2. I my own personal opinion , the paced rhythm is similarly to the LBBB , moreover ,the chest pain and biomarkers are additional diagnostic information .

      Delete
  5. Does the modified rule has 2 point also ? If so its only total less than 3 points if we use Sgarbossa Criteria ? so what the next step if either modified or Sgarbosa criteria give only 2 points ?

    ReplyDelete
    Replies
    1. Modified Sgarbossa does NOT use points. It is "UNWEIGHTED".

      Delete
  6. Better to use the pure percentage rather than negative 0.25 , better to say more than 0.25 or equal or more than 0.30 .

    ReplyDelete
  7. Sir, I don't understand why the ratio 0.3 is more sensitive (100%) than 0.25, while the ratio 0.2 is more sensitive than 0.25, too. Is it a mistake?

    ReplyDelete
    Replies
    1. it was more sensitive because it included discordant ST depression as well, not just ST elevation. In our validation study, that criterion did not turn out to be as sensitive, though it was very specific.

      Delete

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.