Safety of an algorithm for transcatheter aortic valve replacement procedures with or without anesthesiologist

Abstract Background The growing number of transcatheter aortic valve replacements (TAVR) without adequately growing resources is associated with the risk of patient-mortality while on wait list. The presence of anesthesiologists during TAVR may be a limit to reducing wait times. Purpose We aimed to assess the safety of a stratification algorithm to lead patients with high risk of perioperative complications towards anesthesiologist presence and others to non-anesthesiologist managed analgesia. Methods A total of 1369 patients consecutive patients who underwent trans-femoral TAVR in a tertiary university center and prospectively included in the national TAVI registries were include in the analysis. All patients prior to January 2020 (Phase 1, N=502) underwent TAVR in the presence of anesthesiologist. The algorithm led patients to TAVR with anesthesiologist in the presence of at least one of the following: left ventricular ejection fraction 40%, systolic pulmonary artery pressure 60 mmHg, stage ≥ 3 pulmonary disease, body mass index ≥ 35 or difficult femoral access after January 2020 (Phase 2). The primary end point (composite of death, myocardial infarction-MI-, stroke, vascular surgery or tamponade) was compared between phases as well as between groups with and without anesthesiologist during phase 2 using a Chi2 test and an adjusted (variables differently distributed between groups with p0.1) logistic regression model. Results As summarized in the table, the primary event rates were reduced after the introduction of the algorithm with trends towards lower rates of death, MI and vascular surgery, and significantly lower rates of tamponade and 30 days mortality. During phase 2, similar rates of the primary endpoint and other per-TAVR complications were found between groups with and without anesthesiologist, but 30 days mortality rates were lower in those without anesthesiologist. The multivariable model showed a significant reduction of the rates of the primary endpoint between phases (OR 0.52, 95%CI0.3-0.87, p=0.01), but similar results between the groups based on the presence or not of anesthesiologists (OR 0.74 95%CI0.3-1.6, p=0.5). Conclusion Our results show that an algorithm based on pre-defined correlates of complication during TAVR procedures identifies safely patients who may need anesthesiologist support and may participate in reducing wait times and useless medical resource allocation.Table