Ok. I think this is a really important piece of evidence--as increasingly younger lower risk patients have TAVI vs SAVR.
I will cover on TWIC Friday and want to get it right.
I've looked at JACC paper. And I have some questions.
1) The original 2019 trial in NEJM (Popma et al) had a Bayesian noninferiority design, margin ARD 6%. Subsequent reports have switched to frequentist and superiority. I could not find where this was prespecifed or explained. The switch may be important.
2) Why does a 2.8% higher stroke/death outcome in TAVI arm calculate to a p of 0.4? I don't know exactly but in the same JACC issue is a review on dealing with non-proportional hazards
jacc.org/doi/10.1016/j.jacc.… and this data clearly looks non-proportional. So maybe something different like a milestone analysis should be performed. And if this is done, what would the upper bound of the absolute risk increase be? If it was greater than 6%, could we consider TAVI no longer non-inferior? Again, I am not a stats person. Just curious. Why can't NI be assessed at 6 years as well as one year?
2a) For instance, just looking at the reported absolute risk increase of 2.8% [95% CI: −1.9% to 7.6%]; log-rank P = 0.43) the upper bound is > 6% NI margin.
3) If we used the original Bayesian framework what would have been the Pb of harm with TAVI at 6 years?
4) I see these questions as reasonable given the higher re-intervention rate in the TAVI arm (I know, there were technical changes).
5) Also relevant maybe is this meta-analysis in BMJ finding 99% Pb of superiority of SAVR in 5-year mortality
heart.bmj.com/content/early/…
6) UK TAVI presented 5-year results at EuroPCR in May 2025. Higher stroke rates with TAVI. Not published yet??
7) Something seems to change over time. Look at 4-year EVOLUT results vs 7-year
jacc.org/doi/10.1016/j.jacc.…
TAVI mortality goes from 9% to 27.7% = +18.7%
SAVR mortality goes from 12.1% to 23.9% = +11.8%
From the discussion section. "At 7 years, no significant differences were observed between the treatment arms for all-cause mortality (27.7% TAVR vs 23.9% surgery; difference: 3.7% [95% CI: −2.4% to 9.9%]; log-rank P = 0.29) <- LOOK AT THE UPPER BOUND
@kaulcsmc @djc795 @AndrewFoy82
6 yrs after surgery or
#TAVR with CoreValve/Evolut, low-risk AS pts showed no difference in all-cause mortality or disabling
#stroke in the Evolut Low-Risk trial. However, reinterventions were higher after TAVR, largely driven by increased AR.
jacc.org/doi/10.1016/j.jacc.… #JACC