Margaret A.WinkerMD, Deputy EditorIndividualAuthorPhil B.FontanarosaMD, Interim CoeditorIndividualAuthor
Copyright 1999 American Medical Association. All Rights Reserved.
Applicable FARS/DFARS Restrictions Apply to Government Use.1999
To the Editor: Dr Shlipak and colleagues1 assessed the value of our electrocardiographic
(ECG) criteria to diagnose acute myocardial infarction (AMI) in patients with
left bundle-branch block (BBB).2 The conclusion,
that our algorithm is worse than a "thrombolysis to all" approach, seems unsubstantiated.
Idiosyncratic characteristics of the studied population may explain the low
sensitivity of our ECG criteria, whereas assumptions in the decision analysis
may have amplified the benefits of thrombolysis. The criterion standard selected
for a study like this is critical. Elevations in creatine kinase-MB isoenzyme
and troponin I (sensitive and specific markers of myocardial damage) cannot
be equated to benefit from thrombolysis. The unusually high prevalence of
AMI (56%) among patients with cardiac arrest in this series is unlikely to
reflect acute coronary thrombosis.3 Unfortunately,
the distribution of clinical presentations was not reported. Although the
sensitivity remained low when only patients with chest pain were considered,
the smaller sample size may have compromised the precision of the estimate.
Sgarbossa EB, Pinski SL, Wagner GS. Left Bundle-Branch Block and the ECG in Diagnosis of Acute Myocardial Infarction. JAMA. 1999;282(13):1224-1225. doi:10-1001/pubs.JAMA-ISSN-0098-7484-282-13-jbk1006