Before the start of WCD wear, all patients are consented to use their data for quality monitoring, healthcare operations, and research purposes.

The prospective analysis included patients prescribed the WCD at the Sana Klinikum Hof, Germany, from February 2012 to May 2017. Prescription criteria were based upon clinical characteristics at hospital admission which were LVEF ≤ 35% and either ischemic (n = 46, 38%; acute myocardial infarction or ischemic cardiomyopathy), non-ischemic (n = 69, 58%; myocarditis, congestive heart failure, or dilated cardiomyopathy), or others (n = 5, 3%; ICD infection leading to explant or genetic cardiomyopathy). Patients were enrolled if they were prescribed the WCD, presented with LVEF ≤ 35% and cardiomyopathy, and had LVEF measured after 90 days of WCD use. LVEF was measured using Simpson’s biplane method.

Patient characteristics and medical treatment were evaluated at the beginning and after 90 days of WCD use.

Patient use data were obtained from the commercial database. Days of wear was the sum of days in which the WCD was worn for greater than 15 min. Hours of daily use was the ratio of the sum of hours and the sum of days minus 1. The purpose of the 1-day adjustment was to correct for the partial days available on the first and last days of prescribed wear.

Electrocardiogram (ECG) data recorded by the WCD system were remotely transmitted, screened for whether they included information on heart rate and rhythm, and analyzed by the authors, including 2 cardiologists. The WCD recorded the automated detection of and delivered shock in response to VT and VF events. The WCD includes a 2-lead ECG monitoring system and continuously monitors the patient for heart rate and rhythm. In the event of VT or VF, the WCD will record the ECG immediately before and after delivering a shock. Signals other than VT/VF can also initiate an ECG recording. These include irregular and fast heart rhythms that are of supraventricular origin. The lengths of recording typically vary between 45 s and several minutes depending of the nature of the detected heart rhythm and the patient’s interaction with the device’s response buttons. VT or VF was ventricular tachyarrhythmia lasting longer than 30 s. VT was defined as having monomorphic or polymorphic characteristics, and VF as an inconsistent morphology. A single event included all ECG records of VT or VF occurring within 24 h of the index arrhythmia. Appropriate shocks were defined as being delivered in response to an episode of VT or VF. Inappropriate shocks were WCD shocks delivered in the absence of VT or VF.

Data associated with shock events include the ECG record encompassing the event, the number of shocks administered, and whether shock resulted in conversion. All shock events were investigated by the responsible physician. Outcomes were tabulated, with survival defined as alive 24 h after receiving a shock.

Statistical analysis was performed using R [11]. Data were reported as the median and interquartile range (IQR) or as the number of patients and percentage of total patients. Categorical data were analyzed using Fisher’s exact test and continuous data by the Wilcoxon test.

WCD prescription was based upon standard clinical algorithm performed at the Sana Klinikum in Hof, Germany (Fig. 1). Eligible patients diagnosed with acute MI were prescribed the WCD for 3 months when percutaneous coronary intervention (PCI) was performed or for 40 days without revascularization. Prescription of the WCD was for 3 months when the diagnosis was non-ischemic, non-acute ischemic, or other eligible cardiomyopathy. At the end of WCD prescription, LVEF was reevaluated and decision to implant an ICD followed European Society of Cardiology (ESC) guidelines.

Baseline patient characteristics are summarized in Table 1. The 120 patients were primarily male (79%) with median age 66 years. Cardiomyopathy leading to WCD prescription were primarily ischemic and non-ischemic (38% versus 58%) with reduced LVEF ≤ 35%. All patients received optimized medication for heart failure, including beta blocker (92%), ACE inhibitor (80%), and diuretic (86%).

Patient characteristics after 90 days of WCD use, distributed by whether LVEF surpassed 35% or not, are also summarized in Table 1. The distribution of gender (80% versus 77% male), median age (69 versus 64 years), and starting LVEF (26% versus 26%) was similar between both groups. The percentage of patients who achieved LVEF > 35% at 90 days was also similar for ischemic, non-ischemic, and other cardiomyopathy diagnoses. Symptom-based prescription of medication (e.g., beta-blockers, ACE inhibitors, diuretics and digitalis) was similar when compared by LVEF at follow-up. Comparison of WCD use by patients in these two groups indicated that median days of wear (52 versus 45 days) and hours of daily use (23.0 versus 22.7 h, an 18 min difference) were also similar. Overall, the WCD was worn by all patients for at least 3 days, and 112 (93%), 94 (78%), 33 (51%), and 11 (9%) patients wore the WCD for at least 7, 30, 60, and 90 days, respectively. Other than WCD delivered shock preceding an ICD implant, reasons for discontinuing WCD use prior to 90 days were not recorded during the study period.

At the start of WCD wear, all patients had LVEF ≤ 35% with LVEF 25 to 30% being the most frequent, accounting for 49 (41%) patients (Fig. 2, left panel). After 90 days of WCD wear, there were 76 (63%) patients with LVEF ≤ 35% remaining while 44 (37%) responded to guideline-directed medical therapy (GDMT) as indicated by improved LVEF > 35% (Fig. 2, right panel). Analysis of patients by starting LVEF indicated that patients with severely reduced LVEF ≤ 25% were as likely to improve as were patients with LVEF above 25%. Thus, after 90 days of WCD wear, 22 out of 58 patients with LVEF ≤ 25% improved to LVEF > 35%, while 22 out of 62 patients with LVEF > 25% improved to LVEF > 35%, equivalent to one-third of patients.

Three patients (2.5%) ages 60, 78, and 67 years were shocked for VT/VF, 1 non-ischemic, and 2 ischemic (Table 2). In each patient, a single episode of VT/VF was detected on days of WCD wear 36, 6, and 17, and terminated by the first shock. No patient was shocked more than once and no patient was shocked inappropriately by the WCD or died while wearing the WCD. Survival was documented at least 24 h post-shock. The shocked patients were successfully implanted with an ICD for secondary prevention of SCD.

During our study, WCD-based remote monitoring of 2794 ECG recordings were available. Arrhythmia monitoring beyond occurrence of VT/VF was reported via automatic downloads of ECG records when non-sinus rhythm was detected. During the 90 days of prescribed wear, new atrial fibrillation was detected in 1 (1%) LVEF ≤ 35% and 3 (7%) LVEF > 35% patients. A total of 2202 recordings in 96 patients (80%) were artifacts (non-VT/VF, non-atrial fibrillation). In total, 60 patients (50%) initiated 443 recordings manually by pressing the WCD response buttons for 3 s. There was no other supraventricular arrhythmia detected.

In the absence of VT/VF, the decision to implant an ICD correlated with LVEF at follow-up (Table 1). For example, 68 (89%) patients in the LVEF < 35% group versus 3 (7%) among the LVEF > 35% group were implanted with an ICD for prevention of SCD (p < 0.001). There were, however, 5 (4%) patients in the LVEF ≤ 35% group, where the decision to implant was unavailable and corresponded with patient decision to receive follow-up care at another institution.

Subgroup analysis of the reason for WCD prescription along with the decision whether to implant an ICD after 90 days of WCD use indicated that among the LVEF < 35% group, there were 29 out of 29 ICM and 39 out of 44 NICM patients implanted with an ICD. In contrast, among the LVEF > 35% group, there were 15 out of 17 ICM patients and 24 out of 25 NICM patients who were not implanted.

The study has several strengths and limitations. This study is the first to describe a cohort of consecutive patients enrolled from a community-based acute care hospital serving a semirural region. Along with the PROLONG study cohort, this study is one of the first to report LVEF at the beginning and end of WCD use. Limitations include the absence of data on common cardiovascular comorbidities. The definitions and hierarchy used to assign patients to a single cardiac diagnosis likely resulted in overlap among disease etiologies. Additional limitations include the lack of a control group of patients who did not receive the WCD. The possibility that the use of a WCD leads to better care, and subsequent improvement in LVEF, may exist. Inherent to the study design of a non-randomized observational analysis is the possibility of selection bias.