RESULTS
The initial search yielded 2,808 articles; after electronic and manual deduplication on title review 2,502 studies remained. An abstract review led to the exclusion of the least relevant articles, with 159 selected for full-text review. From that we identified 37 studies with 71,455 patients for which quantitative analysis was possible (figure 1) (6-42). All the studies were observational, and 23 out of 37 included patients who had undergone TAVR at a single institution (62%). The incidence of PPMI after TAVR was 22%. The mean age of our study population ranged between 76–84 years, with 52% female. Logistic EuroSCORE or the STS-PROM score was used to assess the pre-procedural risk in the majority of the studies. By contrast, in seven studies the authors reported no relevant information. Table 1 shows the baseline characteristics of patients in the included studies.
The type of valve implanted and the access routes used in each study are summarized in Table 2. In 17 studies SEV was used exclusively and BEV exclusively in 9; both valve types were used collectively in 11. Of the studies included in the analysis, the incidence of BEV was 64% compared to SEV 35.5%, while other valves were 0.5%. Fifty-four different predictors were eligible for the analysis, and two or more non-overlapping datasets were available for 25 (supplement table 2). Results were divided based on demographic, electrocardiographic, imaging, valvular, and procedural predictors (figure 2).