• 
• 

Three Part Question

In [patients presenting to the ED with possible acute coronary syndrome and who have ventricular paced rhythm on the ECG], can [Sgarbossa criteria or Smith-modified Sgarbossa criteria] be used to [rule in AMI and/or acute coronary occlusion (ACO)]?

Clinical Scenario

A 70-year-old man presents to the ED with acute chest pain. His ECG shows ventricular paced rhythm with left bundle branch block morphology. You are aware that the Sgarbossa criteria could be used to diagnose acute myocardial infarction (AMI) in this context1 and that the modification to those criteria proposed by Smith et al 2 could improve diagnostic accuracy. You wonder if the same criteria can be applied to diagnose AMI in patients with ventricular paced rhythm.

Search Strategy

We searched the MEDLINE (1946 to May Week 2, 2016) and Embase (1974 to 2016 Week 21) databases, together with the Cochrane Database of Systematic Reviews (2005 to May 18, 2016) using the Ovid interface. We used the following search terms:

(exp Cardiac Pacing, Artificial/OR exp Pacemaker, Artificial/OR (pacemaker OR pacing OR paced).mp) AND (sgarbossa.mp OR concordan$.mp OR discordan$.mp) limit to human and English language.

Search Outcome

We identified a total of 750 papers (including 688 in Embase, 58 in MEDLINE and 4 in the Cochrane Database of Systematic Reviews), of which 8 were relevant to the three-part question. Four were excluded as they were case studies. One study was excluded as the analysis combined both ventricular paced rhythms and left bundle branch block with no differentiated results available.3 The three remaining papers were included

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Caldera et al, 2002 USA	Patients with 100% VPR admitted for a coronary angiography for an acute chest pain and had acute myocardial infarction (AMI) confirmed enzymatically between 1994 and 2002. Patients were excluded if thrombolytic therapy was given during the current clinical event and/or >1 week delay of angiography post presentation. 13 patients were included.	Retrospective cohort study level 2b	AMI	Discordant ST elevation >5 mm = Five cases (38%)	The type of cardiac marker used was not specified and at what level. It would have been useful to note the threshold which was used in practice at the time. The question only addressed one element of the Sgarbossa criteria without reporting any data for the remaining two.
Fordyce et al, 2014 Canada	Patients with pacemaker admitted for a coronary catheterisation due to ACS from May 1999 to January 2012 at two hospitals (528 patients). The patients needed to have had at least one 12-lead ECG with ventricular pacing to be included. 157 patients were included. Sgarbossa criteria were applied to each paced ECG by adjudicators blinded to the angiogram results. ACS (all cases had ACS by virtue of the occlusion criteria)	Retrospective cohort study level 2b	Proportion of cases (with ACS) that had discordant ST elevation >5 mm	6.4% (n=10) patients met this criterion	Only a conference abstract was available at the time of the literature review so limited appraisal can be done. Only cases with ACS were included so it is not possible to evaluate specificity or PPV. The modified Sgarbossa criteria, which were already shown to be more accurate, could have been used for this study. Criteria for adjudicating the diagnosis of ACS are unclear.
			Proportion of cases with ST-segment depression >1 mm in leads V1–V3	3.8% (n=6)
			Proportion of cases with any Sgarbossa criteria	10.2% (n=16)
Freitas et al, 2015 Portugal	Patients with pacemaker admitted for a coronary angiography due to suspected ACS. Biventricular or atrial (AAI) pacing modes were excluded. 51 patients with ventricular pacing of 5072 coronary angiography exams performed due to ACS were identified between 2010 and 2014. Eight patients were excluded due to atrial (AAI) pacing. ST elevation myocardial infarction (STEMI) was diagnosed in 26 patients (60%).	Retrospective cohort study level 2b	STEMI, which was defined in this study as acute coronary occlusion or stenosis (with angiographic evidence and troponin I ≥10 ng/mL at 24 hours) Sgarbossa	Sensitivity: 16% Specificity: 100% PPV: 100% NPV: 46%	Only a conference abstract was available at the time of the literature review so limited appraisal could be done. Furthermore the ECGs were analysed by two cardiologists but the level/experience/training was not available/provided.
			Smith-modified Sgarbossa	Sensitivity: 40% Specificity: 83% PPV: 77% NPV: 50%
			Selvester criteria	Sensitivity: 40% Specificity: 72% PPV: 67% NPV: 46%

Comment(s)

The ECG is the investigation of first choice in patients with suspected acute coronary syndromes to immediately ‘rule in’ ST elevation myocardial infarction (STEMI), which requires urgent revascularisation. Arguably, the condition that the ECG must diagnose is ACO. It is not possible to interpret ST elevation in patients with left bundle branch block (LBBB) or ventricular paced rhythm. In such patients, waiting for biomarker concentrations can introduce important delays to coronary intervention or fibrinolysis. Sgarbossa et al proposed three criteria that predict the diagnosis of AMI in patients with LBBB as follows: (1) concordant ST-segment elevation of 1 mm (0.1 mV) in at least one lead (sensitivity 18%, specificity 94%); (2) concordant ST-segment depression of at least 1 mm in leads V1–V3 (three points) or (3) excessively discordant ST-segment elevation, defined as ≥5 mm of ST-segment elevation when the QRS complex is negative (two points).1 In a case-control study of 34 patients with ventricular paced rhythm, Sgarbossa confirmed that the same criteria appeared predictive of AMI.4 Smith et al further improved on this with a ‘modified Sgarbossa rule’ in which Sgarbossa's third criterion (excessively discordant ST elevation as defined by 5 mm) was substituted to interpret discordant ST elevation in relation to the depth of the preceding S wave (defined as >25% of the previous S wave). As ventricular paced rhythm causes similar ECG changes to LBBB, we aimed to determine whether the original or modified Sgarbossa criteria could be used to ‘rule in’ ACO or AMI. We identified only three relevant papers. Caldera et al 5 only evaluated one of the three Sgarbossa criteria. Fordyce et al 6 only included patients with confirmed acute coronary syndrome (ACS) precluding calculation of specificity and positive predictive value. In the only work to have directly answered this question, Freitas et al 7 included 43 patients with ventricular paced rhythm (VPR) of whom 60% had STEMI. This study demonstrated that the original Sgarbossa criteria had 100% specificity and positive predictive value, and are therefore useful to ‘rule in’ ACO. The Smith-modified Sgarbossa criteria had superior sensitivity but specificity and positive predictive value fell (to 83% and 77%, respectively), introducing more uncertainty as a ‘rule in’ test. Two of the included studies noted that patients meeting Sgarbossa criteria had a greater number of coronary occlusions and a greater extent of coronary disease.5 ,6

Editor Comment

NPV, negative predictive value; PPV, positive predictive value

Clinical Bottom Line

The clinical bottom line is that there is little current evidence to determine the value of the original and modified Sgarbossa criteria for diagnosing AMI or acute coronary occlusion. While further research is clearly needed, the existing evidence suggests that the original Sgarbossa criteria can ‘rule in’ acute coronary occlusion with high specificity.

References

1. Sgarbossa EB , Pinski SL , Barbagelata A , et al Electrocardiographic diagnosis of evolving acute myocardial infarction in the presence of left bundle-branch block (Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries) Investigators. N Engl J Med 1996;334:481–7.
2. Smith SW , Dodd KW , Henry TD , et al . 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. Ann Emerg Med 2012;60:766–76.
3. Aguinaga-Meza M , Barboza J , Cabrera R , et al . Abstract 14643: Clinical and electrocardiographic characteristics are insensitive indicators of acute myocardial infarction in patients who present with chest pain and left bundle branch block or paced rhythm. Circulation 2012;126:A14643.
4. Sgarbossa EB , Pinski SL , Gates KB , et al Early electrocardiographic diagnosis of acute myocardial infarction in the presence of ventricular paced rhythm. GUSTO-I Investigators. Am J Cardiol 1996;77:423–4.
5. Caldera AE , Bryce M , Kotler M , et al Angiographic significance of a discordant ST-segment elevation of > or =5 millimeters in patients with ventricular-paced rhythm and acute myocardial infarction. Am J Cardiol 2002;90:1240–3.
6. Fordyce CB , Alipour S , Pu A , et al . Electrocardiographic diagnosis of acute myocardial infarction in paced rhythms: utility of the Sgarbossa Criteria. Can J Cardiol 2014;30:S61–2.
7. Freitas P , Santos M , Castro M , et al . ECG evaluation in patients with pacemaker and suspected ACS: which score to apply? Eur Heart J Acute Cardiovasc Care 2015;4(Suppl 5):348.