Patients

We conducted a retrospective case-control study and reviewed the medical records of 98 consecutive patients hospitalized in the coronary care unit of our institution between 2008 and 2010 with the diagnosis of resuscitated cardiac arrest. The study was approved by the Amiens ethics committee. The Amiens-Picardie University Hospital is a tertiary referral centre for the Picardie administrative region (1.9 million inhabitants); it also directly receives cases of cardiac arrest occurring in its own health territory and in the city of Amiens (total population 520,000 inhabitants) via emergency services.

Patients were included in the study group when cardiac arrest was due to spontaneous primary VF occurring < 48 hours after onset of ACS. ACS was defined as a rise in cardiac troponin concentration with one value above the 99th percentile of the upper reference limit and at least: symptoms of ischaemia; ECG changes indicative of new ischaemia; development of pathological Q waves; or imaging evidence of new loss of viable myocardium or new regional wall motion abnormality . To obtain a homogeneous group of cases with VF at the early phase of ACS, we excluded patients found in asystole without documented previous spontaneous VF ( n = 13) and patients with VF without ACS ( n = 10), VF occurring during or after the revascularization procedure ( n = 8), left bundle branch block on the postcardiac arrest ECG ( n = 8), VF occurring > 48 hours after the onset of ACS ( n = 3) and third-degree atrioventricular block ( n = 3). The study group finally comprised 53 patients (41 men and 12 women) with a mean age of 52 ± 10 years (range 34–76 years; Fig. 1 ).

Flow-chart of patients included in or excluded from the study. ACS: acute coronary syndrome; AV: atrioventricular; CCU: cardiac care unit; LBBB: left bundle branch block; PTCA: percutaneous transluminal coronary angioplasty; VF: ventricular fibrillation.

The control group comprised two age- and sex-matched controls for each patient. Controls were selected from a cohort of patients with ACS but without cardiac arrest due to VF or left bundle branch block, admitted to the coronary care unit during the same period ( n = 106).

The list of patients in the study and control groups was obtained from our institution’s electronic medical records system ( Département d’Informatique Médicale ).

The following data were collected: cardiovascular risk factors; treatment on arrival; previously known coronary artery disease; localization of the ACS; revascularization procedure; thrombolytic therapy; and left ventricular ejection fraction (LVEF) on echocardiography performed during the first 48 hours. The presence of diabetes, hypertension or dyslipidaemia was assessed by patient self-reporting and according to the presence of specific treatment on arrival. The diagnosis of previous coronary artery disease was accepted only when the patient had had a previous anatomical and/or functional evaluation showing coronary artery disease. Localization of the ACS was assessed by the results of coronary angiography and by a 12-lead ECG.

Electrocardiogram analysis

Blinded evaluation of the ECG was performed by two cardiologists (L.A,. S.T.); in the case of disagreement, a consensus was reached after a third blinded analysis (J.-S.H.). PR intervals, QRS durations and QTc intervals (Bazett’s formula) were measured on the ECG recorded on the day of onset of the ACS and VF. ‘J waves’ were defined as in the study by Haïssaguerre et al.: ‘The amplitude of J-point elevation had to be at least 1 mm (0.1 mV) above the baseline level, either as QRS slurring (a smooth transition from the QRS segment to the ST segment) or notching (a positive J deflection inscribed on the S wave) in the inferior lead (II, III, and aVF), lateral lead (I, aVL and V4 to V6) or both. The anterior precordial leads (V1 to V3) were excluded from the analysis to avoid the inclusion of patients with right ventricular dysplasia or the Brugada syndrome.’ ( Fig. 2 ) . As QRS slurring was sometimes difficult to distinguish from ST-segment elevation related to myocardial infarction, we therefore also analysed the ECG after return of the ST segment to the isoelectric line ( Fig. 3 ). Due to the description of ‘J wave’ spontaneous variations in the literature, we considered for analysis all ECGs recorded during hospitalization. Therapeutic hypothermia was performed in all 52 patients in the study group with out-of-hospital cardiac arrest ( Fig. 1 ). ECGs recorded during therapeutic hypothermia were excluded to avoid confusion with Osborne wave.

Ventricular fibrillation (VF) in a patient with ‘J waves’. Presence of QRS notching in leads II, III and aVF during an inferior acute coronary syndrome (ACS) in a 58-year-old man. Small notching is visible before VF but is not enhanced. The 12-lead electrocardiogram was recorded on the first day of the ACS.

Diagnosis of ‘J waves’ during acute coronary syndrome (ACS). (A) Patient who experienced cardiac arrest at the initial phase of an anterior ACS; ‘J waves’ are already visible during the phase of ST-segment elevation. (B) Inferolateral ACS with ST-segment elevation in leads II, III, aVF, V4 and V6 but with no visible notching; detection of ‘J waves’ in leads V5 to V6 after return of the ST segment to the isoelectric line.

Statistical analysis

Differences between cases and controls were assessed using a generalized linear mixed model with the SAS Glimmix procedure. Odds ratios (ORs) and their 95% confidence intervals (95% CIs) were computed for categorical variables and mean differences and 95% CIs were computed for continuous variables.

Differences between patients with or without ‘J waves’ were assessed by a t test for continuous variables and by the Chi ² test for categorical variables. Interobserver agreement was determined by overall proportion of agreement and kappa score for the diagnosis of ‘J waves’. A multivariate logistic regression model (forward conditional model) was used to identify the factors associated with the occurrence of VF and with inhospital mortality. Data that were significantly different ( P < 0.05) between study and control patients in univariate analysis were included in the logistic regression (‘J waves’, RR interval, QTc interval, anterior localization of the ACS, LVEF). An exact logistic model was used for inhospital mortality due to the sparse distribution of the variables. ORs were calculated with 95% CIs. P values were two-sided and values < 0.05 were considered significant. SPSS software (version 11) was used for all analysis.