This observational cohort study included all patients receiving a WCD at our tertiary care University Center between 2012 and September 2016. All patients were fitted with a ZOLL Life Vest™ system (Pittsburgh, USA).

All patients received OMT. Each subject provided consent for the de-identified analysis of standard clinical data. The study was approved by the local ethics committee and conforms to the 1975 Declaration of Helsinki.

The WCD ZOLL Life Vest™ system (Pittsburgh, USA) has been described previously [8]. The WCD programming was individually adapted to the patient’s underlying heart disease and electrocardiographic patterns. In general, for older patients the ventricular tachycardia (VT) zone was programmed at a heart rate of 150 bpm with a VT response time of 60 s and the ventricular fibrillation (VF) zone at a rate of 200 bpm with a VF response time of 25 s. For younger and more active patients the VT zone was programmed at a heart rate of 180 bpm. First shock energy was set at maximum output (150 J) in all patients. Any arrhythmia episode was considered as a separate episode when occurring with a minimum delay of 3 min from the previous one. Each individual episode was reviewed and classified into the following categories: sustained VT (lasting 30 s or longer) or VF with WCD shock therapy, non-sustained VT (lasting less than 30 s), bradycardia of 30 beats per minute or less or asystole. Inappropriate WCD therapy was classified as a non-VT/− VF episode treated by WCD shock.

Data were prospectively collected from the time of the initial hospitalization with WCD implementation. Baseline data included the indication for WCD, co-morbidities, baseline medications, ECG data, and echocardiographic results. LVEF was calculated using Simpsons’s method. WCD was generally prescribed for 3 months irrespective of the underlying WCD indication. Data on arrhythmias during follow-up were prospectively collected clinically and simultaneously retrieved from the ZOLL LifeVest Network™.

In patients with primary preventive indication follow-up visits were scheduled 2 months after diagnosis. If LVEF had increased beyond 35% after 2 months, WCD therapy was terminated early. If LVEF had not increased above the 35% threshold, echocardiography was repeated 1 month later. If LVEF remained below 35% on OMT patients received primary prophylactic ICD implantation.

If a patient with primary preventive WCD indication was incompliant and returned the device earlier than planned while LVEF was still below 35%, the patient received normal clinical and echocardiographic controls. An ICD was implanted if LVEF remained below 35% after 3 months on OMT.

If patients had an appropriate WCD shock, a secondary prophylactic ICD was implanted within the following few days. In patients with an indication for cardiac resynchronization therapy (CRT), a cardiac resynchronization therapy defibrillator (CRT-D) was implanted according to current guidelines [4]. Patients with previously explanted ICDs were followed according to clinical indications.

Patients suspected of or demonstrated to have sinus bradycardia, asystole or intermittent AV-block (Mobitz II) or third-degree AV-block received a 2-chamber ICD. Patients with no suspected or reported bradycardia received an S-ICD or a 1-chamber ICD.

Long-term follow-up was counted from the day on which the patient returned the WCD. During long-term follow-up, all study patients irrespective of the duration of WCD therapy or subsequent ICD implantation received clinical and echocardiographic assessments every 6 months and when clinically indicated.

Heart failure medication consisted of angiotensin converting-enzyme inhibitor/angiotensin receptor blocker (ACE-I/ARB), betablocker and mineralocorticoid receptor antagonist (MRA) according to current heart failure guidelines [4]. Procoralan was prescribed in selected cases. Medication doses were adjusted during each follow-up visit. Since January 2016 selected patients received angiotensin receptor-neprilysin inhibitor instead of ACE-I or ARB. Furthermore, all patients were screened for iron deficiency in heart failure. Intravenous iron substitution was performed if required [4].

For missing data, particularly in cases of missed follow-up visits, the patient or other treating physicians were contacted.

Between April 2012 and September 2016 114 patients were prescribed a WCD. Eight patients returned their WCD during the first hours after initiation because of unwillingness or inability to handle it. One patient was lost to follow-up. 105 patients had complete data sets of baseline and follow-up during and after WCD use and were included in the data analysis.

Patients were classified according to WCD indication: The most common indication for WCD (in 84 of 105 patients) was primary preventive therapy in patients with heart failure symptoms and newly diagnosed ICM or NICM with baseline LVEF ≤35% (Fig. 1, Table 1). These patients received the WCD to bridge the duration of implementing OMT to monitor whether LVEF rose above 35%. Further indications are shown in Fig. 1.

Median patient age of the whole patient population was 60 years with a male predominance (82%). Further details of the patients’ baseline data are presented in Table 1.

Comparing baseline data of patients with newly diagnosed ICM and NICM, patients with ICM were older (ICM versus NICM, median (range): 62 (43–78) versus 54 (30–78)) (p = 0.0008) and had a lower prevalence of left bundle branch block (p = 0.03) than patients with NICM. They were more likely to suffer from arterial hypertension (p < 0.0001) and hyperlipidemia (p = 0.005) than patients with NICM and were less often treated with mineralocorticoid-receptor antagonists (p = 0.008). LVEF was significantly higher in ICM patients than in NICM patients (p = 0.0002).

The mean WCD wear time of all 105 patients was 68.8 ± 50.4 days with a mean daily use of 21.5 ± 3.5 h (Table 2).

Of the 84 patients with primary preventive indication, 37 returned their WCD earlier than the prescribed 3 months due to early LVEF improvement (21 of 37 patients), non-compliance (12 of 37 patients), early ICD-implantation due to an appropriate WCD shock (3 of 37 patients) and non-cardiac death (1 of 37 patients).

Patients with a previous explanted trans-venous ICD had a mean WCD wear time of 92.5 ± 87.0 days.

Patients with previous explanted ICDs had a higher mean daily use (23.1 ± 1.1 h/day) than did patients in the other patient groups (Table 2). This was caused by the higher risk awareness in this patient group.

Five of the 105 patients (4.8%) received appropriate WCD shocks. Characteristics of these patients are presented in Table 3. Four patients had WCD shock because of VF, one patient experienced WCD shock because of VT. In all five patients VT or VF was successfully terminated with the first WCD shock. In each of the five patients WCD shocks occurred after discharge from hospital between 3 to 154 days after initiation of WCD.

The single patient who received an appropriate WCD shock after 154 days of WCD wear was a 71-year-old man with ICM. He received prolonged WCD therapy due to persistent bacterial infection triggered by his knee prosthesis. The WCD shock was delivered because of sustained VT with ventricular heart rate of 180/min. Then, an S-ICD was successfully implanted within the following days.

One patient had asymptomatic non-sustained VT that was detected via the ZOLL LifeVest Network™. Furthermore, one patient had an asystole of 10s during sleep that was detected via the ZOLL LifeVest Network™.

One 74-year-old female patient with mild cognitive defects and newly diagnosed ICM received an inappropriate WCD shock on day 3. This was triggered by artifactual voltage fluctuations misinterpreted by the WCD as ventricular arrhythmia. The patient had ignored both tactile and audible alarms and failed to press the response button of her WCD.

None of the 21 patients with early LVEF improvement and none of the 12 patients who returned their WCD early due to non-compliance showed signs of ventricular tachyarrhythmias during the missed WCD time period.

After end of WCD use all 103 patients alive were followed for a mean period of 18.6 ± 12.3 months- irrespective if they had received an ICD or not.

Several observational studies have shown WCD to successfully identify and interrupt VT and VF with shock efficacy rates of 99–100% [8‐10, 12]. The present study confirmed effectiveness and safety of the WCD therapy. 4.8% of the patients had sustained VT/VF, and all episodes were successfully terminated by WCD shock.

The percentage of VT/VF was slightly higher than in multi-center registries reporting appropriate WCD shocks in 1.1 to 2.1% of the patients [8‐10]. Other single center data showed similar results to our patients with appropriate shocks in 3.9 to 7% of patients [13‐16].

Several publications have reported higher WCD shock rates in patients with newly diagnosed ICM than in patients with newly diagnosed NICM [9, 16, 17]. Sing et al. even questioned the utility of the WCD in patients with NICM [17]. Nevertheless, Duncker et al. found an incidence of VT in the early stage of NICM of 38.7 of 100 person-years, reflecting the relevant arrhythmogenic risk in patients with NICM and non-optimized heart failure medication [18]. High rates of appropriate WCD shocks have been described in several publications evaluating patients with explanted ICDs [8, 10]. As a result, WCD prescription is now generally accepted in this patient group [11].

In our study, the majority of rescued patients had new onset ICM. However, one patient with NICM and LVEF < 35% whose primary prophylactic ICD had to be explanted and another patient with myocarditis also received lifesaving WCD shocks because of VF. The numbers of patients in the four groups in our study (especially in the patient group with explanted ICDs and myocarditis) were too small to permit interesting conclusions about predictive factors for arrhythmias. Nevertheless, our data confirm the usefulness of WCD in both patients with ICM and NICM and support its application in patients with a wide range of indications.

One of the problems associated with WCD therapy is incorrect use [19] which caused inappropriate shock in one patient. However, rates of inappropriate WCD shocks were low (0.4–3.0%) [8‐10, 12]. Importantly, no death has been attributed to WCD technical failure since its introduction [20].

Another common problem [19] we faced in the study was that several patients were reluctant to wear the WCD or returned the device earlier than planned. Poor compliance or inappropriate use of the WCD may have catastrophic consequences [19]. Patient education on how to properly wear the device, change the battery and disable shock delivery is crucial. Furthermore, patients should understand their cardiac disease and the potential benefits associated with the use of the WCD. Patients should be selected carefully and the device should not be deployed in patients unfit or unwilling to properly manage it. Because WCD use may be monitored online, detection of incorrect use of the device is possible and should be used to provide prompt feedback and motivation to a non-compliant patient. In non-compliant patients who are at high risk for ventricular tachyarrhythmias, other forms of monitoring or therapy should be discussed (patient monitoring in-hospital, early ICD implantation, etc.)

ICD implantation rate in the whole patient cohort was 54%. This parallels other studies with similar patient collectives that reported implantation rates of 34 to 57% [9, 14, 16, 21, 22]. Naturally, ICD implantation rate was highest in the patient group with a prior explanted trans-venous ICD. Following WCD, ICD implantation could be avoided in almost half of the patients with ICM and NICM. LVEF improvement was the most common reason not to implant an ICD.

Current heart failure guidelines recommend primary preventive ICD implantation in both patients with ICM and NICM with LVEF ≤35% despite at least 3 months of OMT [4]. OMT is currently defined as combination of high dose betablocker and high dose angiotensin converting-enzyme inhibitor, with the addition of a mineralocorticoid receptor antagonist in patients with persistent symptoms of heart failure [4]. As a result, ICD implantation 3 months after diagnosis might be too early in many patients because patients would be assumed to be on OMT already at the moment of diagnosis.

A previous study including patients with acute myocardial infarction who were treated with modern therapies including early revascularization, showed greatest LVEF improvement within the first month. Nevertheless, a smaller part of patients showed LVEF improvement beyond the initial 40 days after acute myocardial infarction [23].

On the other hand, recent studies including patients with NICM reported late LVEF improvement after implantation of a primary preventive ICD [24, 25]. Verma et al. reported LVEF improvement to > 35% in 12% of patients [24]. Grimm et al. found LVEF improvement in 24% of patients with NICM after ICD implantation reducing incidence of appropriate ICD therapies to ∼1% per year in patients with improved LVEF [25]. In both studies, multivariate analysis identified no other significant predictor for LVEF improvement after ICD implantation than a short time from diagnosis to device implant. Consistent with the data of Varma and Grimm, 22.0% of our patients with NICM showed late LVEF improvement beyond 35%. It should be noted that approximately half of the NICM patients in our study who had late LVEF improvement had received a CRT-D device (12.2%). The CRT is known to have an important effect on LVEF recovery.

In patients with ICM we observed a trend towards late LVEF improvement. 9.3% of ICM patients skipped the ICD threshold during long-term follow-up and developed LVEF > 35% while 7.0% had received CRT-D.

Patients in our study received optimization of medical therapy not only during WCD period but were seen every 6 months in our heart failure outpatient clinic where doses of medication were adjusted. Our study overlapped the debut of newer heart failure therapies such as sacubitril/valsartan which may have played a role in improving LVEF. Sacubitril/valsartan has been shown to improve morbidity and mortality in heart failure [1]. Furthermore, this medication has also been shown to reduce the incidence of VT and arrhythmogenic deaths which might contribute to the low rate of shocks delivered to patients who received an ICD [26].

A recently published study by Duncker et al. which included 156 patients with newly diagnosed NICM or ICM (with a majority of NICM) showed 11 ± 11% improvement in LVEF during 3 months of OMT and WCD use [15]. Even in patients without CRT-D implant they demonstrated improvement of LVEF to > 35% in 33% of their patients during a prolonged period of OMT beyond 3 months [15]. Nevertheless, patients had a relevant risk for life-threatening VT as long as LVEF was ≤35%. Therefore, they proposed a prolonged regimen of WCD use beyond 3 months in patients with one of the three indications: improvement of LVEF to 30–35% or Δ LVEF ≥5% during the first 3 months of WCD wear or insufficient optimization of medical dosages (especially MRA).

Duncker et al. did not stratify subjects by etiology of heart failure (ICM vs NICM) which might have improved their ability to personalize indications for WCD use. Conversely the smaller numbers of subjects in the present study preclude further subanalysis of specific patient subgroups who might benefit from prolonged use of WCD or ICD implantation reduction.

The recently published DANISH trial confirmed the risk of life-threatening ventricular arrhythmias in the chronic phase of NICM, during which primary preventive ICD implantation did not reduce overall mortality [27]. Given that primary preventive ICD implantation in NICM has become debatable, a prolonged WCD regimen might offer prevention from SCD during careful optimization of heart failure therapy and may avoid too early or even not mortality reducing ICD implantation in patients with NICM.

In general, the potential for LVEF improvement indicates a more favorable outcome in both patients with ICM and NICM [28]. Nevertheless, existing guidelines, especially for patients with dilated cardiomyopathy, lack sensitivity and specificity in the selection of patients who need primary- prevention ICD implantation. There is a need for a more precise risk stratification algorithm that considers new markers other than LVEF to provide a more comprehensive system for disease phenotyping. Such new promising markers include the extent or pattern of myocardial replacement fibrosis detected by late gadolinium enhancement cardiac magnetic resonance imaging, cardiac autonomic dysfunction detected by ¹²³I-meta-iodobenzylguanidine myocardial scintigraphy, several ECG-derived markers and genetic testing [29, 30]. Future ICD studies should also consider the antiarrhythmic effects of modern heart failure therapies such as sacubitril/valsartan.

Next to its function to detect and terminate VT/VF the WCD also acts as an external monitoring system to identify bradycardias and asystole. Asymptomatic asystole was identified in one patient who subsequently received a 2-chamber ICD. Currently the WCD only records bradycardias with heart rates less than 10/min. Recording of less severe bradycardia might help to better identify patients at risk for symptomatic bradycardia.

In our study 30 patients were transitioned to S-ICD after wearing the WCD. The S-ICD, like the WCD, is an innovative device without the need to implant trans-venous ICD leads avoiding lead related complications. The S-ICD, similar to the WCD, has no permanent pacing functions and is therefore not indicated in patients who need bradycardia support, cardiac resynchronization or antitachycardiac pacing [11]. During post WCD long-term follow-up, both trans-venous ICDs and S-ICDs successfully terminated VT/VF. Importantly, none of the S-ICD patients needed implantation of a pacemaker or device change to a trans-venous ICD system. In the literature, changes from S-ICD to trans-venous ICD due to the need for ATP have been described in about 0.5% to 0.8% of patients [31, 32]. Changes from S-ICD to resynchronization therapy have been described in 0.4% of patients [32]. Furthermore, several cases have been described where a pacemaker was added to S-ICD for ventricular pacing [33, 34]. In our patients, the monitoring function of the WCD helped to optimize patient selection for S-ICD implantation.

Five patients died, primarily because of fatal infection, and no patient died from SCD during the study period. Four of the five patients treated with the WCD had VF as a first arrhythmic event and would likely have died without the WCD therapy. Patients who had received appropriate WCD shocks and patients whose LVEF did not improve beyond 35% were implanted with a permanent ICD. In the other patients, the vulnerable phase until recovery of LVEF was safely bridged with the WCD. None of the patients with recovered LVEF and no ventricular tachyarrhythmia event during WCD wear showed late signs of tachyarrhythmia. Therefore, with the use of the WCD all study patients could be provided a comprehensive safety net avoiding SCD.