In contrast, the potential bias in
crossover strengthen our findings as only the patients with the best
response to AAD therapy should have been included in the pure AAD
group whereas some patients with no objective evidence of treatment
success may have been included in the RFA group.