The need for the Dutch iVF Registry was stated in our first report, in which we introduced the registry in the Netherlands Heart Journal [7]. With the registry we aim to (1) create a large cohort of patients with an initial iVF diagnosis, (2) focus on additional diagnostic testing (e.g. deformation echocardiography, cardiac magnetic resonance (CMR) and electrocardiographic imaging (ECGI)) to reveal a possible substrate, and (3) collect data on the long-term follow-up, including genetic testing, implantable cardioverter defibrillator (ICD) therapy and family screening.

Eligible patients are those with documented VF, cardiac arrest with a shockable rhythm, or sustained polymorphic ventricular tachycardia (VT), for which known cardiac, respiratory, metabolic and toxicological aetiologies are excluded, as defined by the latest consensus criteria [8]. Preferably, a specific diagnostic approach is followed before diagnosing a patient with iVF. A flowchart indicating which diagnostic tests should be performed at a minimum before diagnosing a patient with iVF is shown in Fig. 2; [1, 4]. If diagnostic test results reveal minor abnormalities, insufficient for a specific diagnosis, this is not an exclusion criterion. Before inclusion in the registry a minimum of 50% of the listed diagnostic tests need to be performed.

Diagnostic approach after a sudden cardiac arrest (SCA). Flowchart indicating all diagnostic tests that need to be performed before diagnosing a patient with idiopathic ventricular fibrillation (iVF). After first investigation with standard testing (electrical testing with an electrocardiogram (ECG) and imaging testing including both echocardiography and coronary imaging) a patient is considered an unexplained cardiac arrest survivor. After additional testing, the diagnosis of iVF can be made. Adapted from Visser et al. [1], with permission from Wolters Kluwer Health, Inc

Data are collected from electronic health records both retrospectively and prospectively, following a standardised form using Castor EDC [9]. We obtain the following data: medical history, physical examination, laboratory testing (including toxicological screening), electrical testing (electrocardiogram (ECG), Holter monitoring, exercise treadmill test (ETT), electrophysiological study), imaging data (echocardiography, coronary imaging, CMR, cardiac computed tomography, positron emission tomography), provocation testing (sodium channel blocker provocation (SCBP), coronary artery spasm provocation), DNA analysis and endomyocardial biopsy results. We consider a diagnostic work-up optimal when an ECG, Holter/telemetry, ETT, echocardiography, CMR, coronary imaging and SCBP are performed. Based on the literature, high-yield diagnostic tests are an ETT, CMR and SCBP [5, 6]. During follow-up, we advise local physicians to complete the diagnostic work-up of iVF patients to reach the most optimal diagnostic work-up in all patients. Outcomes are collected at baseline and during follow-up. Outcomes include appropriate ICD therapy (antitachycardia pacing or shock), inappropriate ICD therapy, detection of a diagnosis, and death. We collect follow-up data every 1–2 years in each participating centre.

The Dutch iVF Registry is incorporated in the VIGILANCE consortium (CVON project). VIGILANCE is an acronym for ‘non-invasive electrocardiographic imaging for individuals at risk for apparently idiopathic ventricular fibrillation’, indicating the fundamental scientific role of ECGI in our research. ECGI enables the non-invasive measurement of the electrical activity at the level of the heart with the use of body surface ECGs and a specific heart-torso geometry [10].

For ECGI measurements and analysis, previously described methods were repeated [11]. In short, approximately184 electrodes were placed on the patient’s torso. Body surface potentials were recorded for 10–60 min. Cardiac computed tomography was used to acquire the heart geometry during end-diastole using intravenous iodine contrast; thoracic computed tomography was used for the positions of the body surface electrodes. This allowed the approximation of the electrostatic torso-heart relationship. Afterwards, for each subject, epicardial electrograms were non-invasively reconstructed for a sinus beat, and each morphologically distinguished premature ventricular contraction (PVC) when applicable. The activation time of each epicardial node was determined by the steepest downslope of the QRS complex; repolarisation time was determined by the steepest upslope of the T‑wave [12].