2.3. Statistical analysis

The primary endpoint will be estimated annually and it will be expressed as absolute and relative numbers. Survival curves free from study endpoint, eventually adjusted according to variables of interest such as gender, age, underlying heart disease, comorbidities and programming mode will be calculated according to the Kaplan-Meier method and compared by Log-rank test for primary and secondary endpoints. Primary endpoint subgroup analyses for each of the aforementioned variables will be performed too. According to the number of centres involved, we expect to recruit at least 500 patients. Considering an annual incidence of the primary endpoint (death from all causes) of 11% according to our experience [7], we expect about 55 deaths at 1-year follow-up and about 165 deaths at the end of the 3-year follow-up. The statistical analysis will be conducted using the SPSS 21.0 for Windows software (SPSS Inc. Chicago, IL, USA). Every test will be two-tailed and a value of p > 0.05 will be considered significant. Multiplicity between the different endpoints will be controlled by the stepwise Holm procedure.