Of course the teaser tittle is misleading.You should ask "better for what".Remember the old auto ad that claimed Fords (or some brand) are better,again better than what and for what.
A better,more focused claim is that ablation is better than drugs to convert atrial fibrillation to normal sinus rhythm.That issue arises only after the decision of rate versus rhythm control has been made.There is convincing evidence to that point that ablation works better. There is also general agreement that patients feel better with a sinus rhythm than when their atria are fibrillating.Atrial fibrillation is a bad method of running a cardiac pump.
What has not been proven with randomized clinical trials is that ablation results in longer lives and fewer strokes.
The recently presented CABANA trial (https://www.acc.org/latest-in-cardiology/clinical-trials/2018/05/10/15/57/cabana) was long awaited and was hoped to answer that question
This was a large (n=2204),multicenter trial with five year followup comparing standard AF ablation procedure with either rhythm or rate control medication.When the data were analyzed by the venerable,preferred, orthodox method of analysis ( intention-to-treat or ITT) there was no difference in the combined end points of death,disabling stroke,or cardiac arrest nor was there for each component of the combined end point.
ITT is also referred to as "once randomized always analyzed". If 1,000 were assigned to medication and another 1000 assigned to ablation, all of the Medication group would be analyzed according to the group to which they were assigned even if they switched over to the ablation group. This method is ,according to standard epidemiologic-statistical dogma, is the only analytic approach which will preserved the "integrity of the randomization process." Randomization is done in the first place to control for the effect of known and unknown variables so that the two groups are balanced in regard to prognostic variables. ITT has been called the de facto standard and it is "conservative", i.e it minimizes Type I error, it is less likely to show a difference when there is no difference. In criticism of ITT one could say it is too conservative and more susceptible to Type II .
Per protocol analysis (PPA) compares treatment groups that include only those who completed the treatment as originally allocated.Whereas ITT makes the two treatments look similar PPA is more able to how differences.
When PPA was applied to the CABANA data there was demonstrated a decrease in mortality in the ablation group.So there are dueling conclusions based on the method of analysis.(1)
My argument is that regardless of there being no difference in mortality (or maybe there is a difference favoring ablation depending of what analysis you prefer) ablation works better to decrease atrial fibrillation and people feel better without AF and the procedure is safe. It is safe according to either way you analyze CABANA and we already knew it was safe. The extensive data from Cleveland Clinic (2) makes that clear. So you do not have to believe that ablation save lives or decreases strokes to favor ablation over rhythm control medication, but of course ablation is not for everyone for everyone.
After the data were presented the predictable flurry of spin emerged touting the results of the PPA and of the "as treated "data.But the EP folks did not really need an alternative analysis to continue with business as usual as ablation is safe,it works better and when successful in decreasing or eliminating the atrial fibrillation "burden" there is better quality of life.
This is certainty not a recommendation to treat everyone with af with ablation but rather an argument in favor of ablation over medication for rhythm control.For many patients for various clinical reasons, rate control may be the better choice and a trial of rhythm control meds before ablation is a reasonable and common approach.
1) "...a man hears what he wants to hear and disregards the rest".The Boxer, Simon and Garfunkel.1982
2)Rehman,KA Life-threatening complications of atrial fibrillation ablation. 16 year experience in a large tertiary care cohort. JACC,March2019, vol 5 no. 3, p 284
(fifteen year period,10,278 patients, no deaths, 100 life threatening complications (mainly pericardial effusions and stroke),no aorto-esophageal fistulas,