Strengths and limitations
There were several limitations to this study, the most important one being its retrospective, single-center, non-randomized design.
There was also a significant bias due to the “associated procedures” in the RD-AVR group. In our center, TAVR is indicated in intermediate-risk patients older than 75 years while RD-AVR is actually indicated in patients older than 70 years with more comorbidities needing AVR + CABG. This explains why both subgroups were not similar before matching. However, our aim was to analyze the impact of each heart valve prosthesis on outcomes. To this end, we performed a 1:1 propensity-score matched comparison that allowed us to avoid differences between both groups at the expense of a decrease in the size of the populations being compared. The variables used for matching were the subject of lengthy reflection. Euroscore 2 cannot be used in the propensity score analysis since it includes several variables already used in the model.
We cannot exclude that subclinical leaflet thrombosis(SLT) could have promoted CHF in the TAVR group since CT scans were not routinely performed to confirm the diagnosis [29]. However, all TTE were performed by experienced cardiologists and CT scans were performed if there was any doubt of SLT on TTE.
Finally, our current results reflect only two-year outcomes and do not address the problem of long-term structural valve deterioration (SVD). An extended FU with a larger number of patients would highlight the occurrence and the impact of SVD on a long-term prognosis.