Worse survival after transcatheter aortic valve implantation than surgical aortic valve replacement: A meta-analysis of observational studies with a propensity-score analysis

Long‐term survival after transcatheter versus surgical aortic valve replacement for aortic stenosis: A meta‐analysis of observational comparative studies with a propensity‐score analysis

“…worse ≥3‐year overall survival) than SAVR, which is compelling with data from a total of 4,197 patients in 14 different studies and robust in sensitivity analyses without publication bias. Our previous meta‐analysis of 19 observational studies with a propensity‐score analysis also demonstrated a statistically significant 21% increase in the hazard of mortality with TAVI relative to SAVR (HR, 1.21; 95% CI, 1.05–1.39; P = 0.010). Whereas, similar follow‐up (≥30‐day , 1‐year , mean 1.15‐year , >1‐year , and 2‐year ) mortality after TAVI and SAVR was indicated in recent (published since 2016) meta‐analyses of odds ratios (ORs) and risk ratios (RR) from RCTs only or RCTs and observational studies in overall , intermediate‐to‐high‐risk , intermediate‐risk , and low‐to‐intermediate‐risk patients .…”

“…The far less number of enrolled patients (699 in PARTNER 1 and 750 in CoreValve US Pivotal ) than those (4,197 from 14 studies) included in the present meta‐analysis of the observational studies may generate the underpowered (as reflected in the wide CIs and lack of statistical significance) results in the two RCTs . In our previous meta‐analysis of 4 RCTs and 19 observational studies (with a propensity‐score analysis) reporting ≥6‐month mortality, the mean predicted operative mortality rate in the RCTs was somewhat greater (logistic EuroSCORE I/Society of Thoracic Surgeons Predicted Risk of Mortality of 20.7%/8.2%) than that in the observational studies (16.9%/6.6%). However, the mean observed 30‐day mortality rate after TAVI in the observational studies was slightly greater (5.1%) than that in the RCTs (3.2%), and that after SAVR was similar between the observational studies (4.7%) and the RCTs (5.0%) .…”

“…In our previous meta‐analysis of 4 RCTs and 19 observational studies (with a propensity‐score analysis) reporting ≥6‐month mortality, the mean predicted operative mortality rate in the RCTs was somewhat greater (logistic EuroSCORE I/Society of Thoracic Surgeons Predicted Risk of Mortality of 20.7%/8.2%) than that in the observational studies (16.9%/6.6%). However, the mean observed 30‐day mortality rate after TAVI in the observational studies was slightly greater (5.1%) than that in the RCTs (3.2%), and that after SAVR was similar between the observational studies (4.7%) and the RCTs (5.0%) . In a study by Agarwal et al of the outcomes after TAVI from 11 large representative multicenter registries and the PARTNER 1 trial, pooled 30‐day all‐cause/cardiovascular‐related mortality rates (9.2%/7.0%) from the registries were significantly higher than those from the PARTNER trial (3.8%/3.6%).…”

“…Whereas, similar follow‐up (≥30‐day , 1‐year , mean 1.15‐year , >1‐year , and 2‐year ) mortality after TAVI and SAVR was indicated in recent (published since 2016) meta‐analyses of odds ratios (ORs) and risk ratios (RR) from RCTs only or RCTs and observational studies in overall , intermediate‐to‐high‐risk , intermediate‐risk , and low‐to‐intermediate‐risk patients . In both the previous and the present meta‐analysis, we extracted HRs from each individual study (and then combined them), which is quite different from the aforementioned recent meta‐analyses abstracting ORs or RR . The most appropriate way of summarizing time‐to‐event (survival) data is to use methods of survival analysis and express the intervention effect as a HR, neither an OR nor a RR .…”

“…Patients with previous open cardiac surgery and renal insufficiency (creatinine >3.0 mg/dL or creatinine clearance <20 mL/min ) and/or end stage renal disease requiring renal replacement therapy were excluded in the aforementioned RCTs. Whereas, some observational comparative studies (non‐randomized studies) exclusively included these patients or enrolled “all comers .” In our previous meta‐analysis of observational studies using a propensity‐score analysis, TAVI was associated with an increase in ≥6‐month mortality relative to SAVR. Results of observational studies, however, should be always interpreted with caution when they are included in meta‐analyses because of greater potential biases for non‐randomized studies compared with RCTs .…”