Log in to MyACC
Conscious Sedation vs. General Anesthesia for TAVR
Study Questions:
What proportion of patients undergoing transcatheter aortic valve replacement (TAVR) receive conscious sedation (CS), as opposed to general anesthesia (GA), and does the chosen sedation technique affect in-hospital mortality?
Methods:
Using the TVT (Transcatheter Valve Therapy) database from the Society of Thoracic Surgeons and American College of Cardiology, temporal and site-specific trends in utilization of CS versus GA in TAVR procedures taking place between January 2016 and March 2019 were determined retrospectively. The proportion of total individuals receiving CS, and each institution’s proportional use of CS for all procedures performed during the study period (site-level preference), were determined. In an effort to adjust for unmeasured confounding, instrumental variable (IV) regression was used, with the site-level sedation preference chosen as the IV. The primary outcome was in-hospital mortality; secondary outcomes were 30-day mortality, hospital length of stay, and percent of patients discharged home. Analysis was by intention-to-treat.
Results:
After exclusion of emergencies, cases requiring valve-in-valve or surgical cut-down access, and cases in which sedation information was missing, 120,080 patients from 559 sites were included in the analysis. During the 3-year study period, the proportion of patients receiving CS grew from 33–64%, and proportion of sites utilizing CS for any of its cases increased from 50–76%.
After IV adjustment, the authors observed a small decrease in hospital mortality among patients receiving CS versus GA (1.1% vs. 1.3%; adjusted risk difference [ARD], -0.2%). Small differences favoring CS were also found in secondary outcomes: 30-day mortality (2.0 vs. 2.5%, ARD, -0.5%), hospital length of stay (3.5 vs. 4.3 days; ARD, -0.7 days), and proportion of patients discharged to home (88.9 vs. 86.1%; ARD, 2.8%) in patients receiving CS versus GA. Vascular complications did not differ by sedation technique in the IV analysis (3.0 vs. 3.1% with CS vs. GA). By comparison, propensity score adjustment produced a larger and statistically significant adjusted outcome difference in vascular complications (2.8 vs. 3.4%; ARD, -0.6%). The relationship between sedation technique and primary outcome was larger in magnitude when the analysis was repeated with propensity matching, as opposed to IV, for adjustment (0.9% vs. 1.5%; ARD, -0.7%).
Conclusions:
Site-specific preference for CS versus GA varies widely between institutions, although an increasing proportion are providing CS to their patients for TAVR. Hospital mortality was modestly lower among patients receiving CS compared to GA for TAVR in this population. Use of IV analysis decreased the apparent advantages of CS, presumably by adjustment for unmeasured confounders.
Perspective:
Decisions on sedation technique for TAVR are likely related to institutional resources including availability of anesthesiologists, operating room block-time, use of transesophageal echocardiography, and experience of the proceduralist. Despite very small differences in outcome in patients receiving CS versus GA, the preferred approach for each institution may have arisen after acknowledgement of these site-specific constraints, and extrapolation of a “preferred” technique overall should be made with caution.
Clinical Topics: Cardiac Surgery, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Valvular Heart Disease, Cardiac Surgery and VHD, Interventions and Imaging, Interventions and Structural Heart Disease, Echocardiography/Ultrasound
Keywords: acc20, ACC Annual Scientific Session, Anesthesia, General, Cardiac Surgical Procedures, Conscious Sedation, Echocardiography, Transesophageal, Heart Valve Diseases, Hospital Mortality, Length of Stay, Patient Discharge, STS/ACC TVT Registry, Transcatheter Aortic Valve Replacement
< Back to Listings
