Data analysis
The main endpoint of the study was any episode of ventricular
tachycardia (VT) or ventricular fibrillation (VF) during the study
period (see Supplementary Material).
Programming of detection zones, counters, and discrimination algorithms
were set in any patient according to indication, routine practice and
patient’s history. Device detected ventricular episodes labelled as VT
or VF were adjudicated by two expert cardiac electrophysiologists; in
case of discordance a third electrophysiologist was asked to classify
the event and the episode was labelled according to majority vote.
Episode data were categorized into binary variables indicating the
presence or absence of a VT, VF, treatment with ATP or with shock during
the study period. Data analyses were conducted separately for VA
episodes self-terminating and/or treated with ATP or with at least one
shock.
Descriptive statistics for continuous variables were reported as mean
(standard deviation) and binary variables as percentages. The event rate
was calculated as the number of patients having an event in the study
period divided by the total number of patients. Event rate in the study
period was compared to that of the reference period using chi-squared
test. The mean number of events in the study period was calculated as
total number of events divided by the total sample size; comparison
between periods was carried out by t t-test. Univariate and multivariate
logistic regressions were conducted in order to examine the associations
between variables of interest and the occurrence of an arrhythmic event.
For all tests, a significance level of 0.05 was used. All analyses were
performed using the Stata software version 13.0 (StataCorp, Texas, US).