Лекарственная и немедикаментозная терапия нарушений сердечного ритма и проводимости, страница 26

Among 81 pts studied EP studies confirmed the diagnosis of VT in 56 pt. Out of the 56 ECG of VT 51 (89.2%) showed at least two predominantly -negative QRS complex (rs, or QS) among the four chosen lead with inclusion of lead I or V6. In all the 25 pts with SVT (100%) there was no single ECG with predominantly -negative complex in 2 out of the 4 leads. The calculated sensitivity, specificity, positive predictive and negative predictive values using our new ECG lead criteria for diagnosis of VT were 89.2, 100%, 100%, 80.6% respectively. Table 1 shows the prevalence of negative QRS complex in each of the 4 chosen leads in the pt with VT vs pt with SVT.

Multiple logistic regression in a forward manner was conducted using lead I, II, V1 and V6 as independent predictors. Lead I, II proved to be the best predictor with overall predictive accuracy 87% (Table 2).

Comparison between the newly conducted criteria and the conventional criteria of Brugada 1991 and John Camm 1994:

We subjected the 81 ECGs to reanalysis using the criteria of A.John Camm 1993 (21) which diagnose VT by default diagnosis of SVT with aberration based on the typical morphology of BBB and the 4 steps algorithm conducted by Brugada et al., 1991 (8).

Table 3 shows the chi-square and P value of each criteria when applied separately compared to the final EPS diagnosis. Using the discriminative analysis the calculated sensitivity, specificity and predictive value were as follow (Table 4):

In the 12-lead ECG of wide QRS tachycardia shown in Fig. 1 it was correctly diagnosed as SVT by our new criteria (only V1 is negative) and also by Camm group criteria whereas it was misdiagnosed as VT by Brugada. Again in the example shown in Fig. 2 it was correctly diagnosed as VT by our new criteria whereas it was misinterpreted as SVT by both Brugada and Camm group criteria.

We studied the value of 1st and 2nd step in Brugada Algorithm which are absent RS and RS intervals > 100 ms and how much they contribute to diagnosis of VT results was as follow (Table 5):

Out of the 81 cases RS was absent in 28 cases, 24 cases (42.9%) were correctly diagnosed as VT and 4 cases were misdiagnosed with calculated sensitivity, specificity (42.8%, 84% respectively). Out of the 81 pts only 25 pts were having an RS intervals >100 ms, 20 cases (30.9%) were having VT and the remaining 5 were misdiagnosed with calculated sensitivity, and specificity as follow 44% and 74%.

DISCUSSION

This study demonstrated that analysis of only 4 leads out of the 12 lead ECG recorded during wide QRS complex tachycardia could predict the origin of these tachycardia with high specificity and sensitivity. The rationale beyond choosing morphology in these four leads is that divergence of diagnostic criteria of pure pattern of BBB in limb leads (I, II) and pericordial leads (V1, V6) would throw doubt in the diagnosis of pure BBB and would rather point to the possible existence of ventricular depolarization.

The criteria was then made after the observation that VT usually has predominantly -negative QRS in at least 2 out of the four chosen leads including either lead I or V6. This was strongly supported by the morphology criteria described by Wellens et al., (3) that QS morphology in V6 strongly support diagnosis of VT. Moreover, typical BBB morphology described by Marriott’s (23) usually include lead I with V1, and V6 and it is well recognized that lead I and V6 are usually similarly upright. We studied the overall predictive value of the 4 leads when negative (78%) then we study the predictive values of each two leads with negative QRS we found that the combination of lead I, II has a highest predictive value (86.4%) when applying criteria of predominantly negative QRS at least 2 of 4 lead (I, II, V1, V6) provided that lead I or V6 included the overall predictive value increased to 90%. Thus the 4 leads new criteria is easy to interpret particularly in the critical care setting requiring no prices measurement not time consuming and there is no personal variability in application, as we selected 2 electrophysiologists and 2 residents to apply these criteria on all the 81 ECGs included, all of them came to the same result with an average time 10-15 minutes for the 81 ECG together.