Limitations of infrequent holter monitoring in assessing antiarrhythmic efficacy