Transcatheter aortic valve replacement (TAVR) is conventionally performed under general anesthesia (GA) allowing intraoperative transoesophageal echocardiogram imaging. We present our experience in patients having the procedure under local anesthesia (LA), who were subsequently transferred to a low dependency unit postprocedure, to assess safety and length of hospital stay. We retrospectively assessed all the transfemoral TAVR procedures conducted at our center from January 03, 2011. Of 216 patients, 145 had the procedure under GA and 71 under LA. Both groups were similar with respect to age, co-morbidities, Euro Score, and the severity of the aortic stenosis. The procedure time was significantly shorter in the LA group measured from time in room to skin closure (108 vs 143 minutes, p <0.001). Skin open to skin closure time were the same in both groups (78 vs 79.4 minutes, p = 0.57). There was no difference in 30 days: aortic regurgitation > mild (2.1% in GA and 2.8% in LA, p = 0.67), need for permanent pacing (3.4% in GA and 1.4% in LA, p = 0.32), and disabling cerebrovascular accidents (1.4% and 1.4%, p = 1.0). The 30-day survival was not significantly different (95.9% in GA and 100% in LA, p = 0.17), whereas the median number of days in hospital was shorter in the LA group (4 in GA and 2 in LA, p <0.001). No emergency conversions to GA were performed in the LA group and only 1 patient needed admission to a high dependency (HD) unit. In conclusion, performing a TAVR under LA is at least as safe as GA. In addition, there is a reduced procedural time and length of hospital stay. LA is a safe and cost-effective alternative to GA and patients can be safely transferred to a low dependency unit.
Transcatheter aortic valve replacement (TAVR) is an alternative to surgical aortic valve replacement in patients with a high operative risk. Refinement of TAVR devices, delivery systems, and technique has correlated with a decrease in TAVR-related mortality and complications. As TAVR and vascular closure technologies improve, procedures are increasingly performed in lower risk patients. There is an imperative, therefore, to reduce patient mortality, morbidity, and length of hospital stay even further. When TAVR was first introduced, surgical cut down over the femoral artery to aid sheath insertion and transoesophageal echocardiogram (TOE) to facilitate valve placement were routinely used and patients routinely received a general anesthesia (GA) to reduce discomfort. Patients were transferred to a high dependency (HD) unit with a 1-to-1 patient-to-nurse ratio. Performing TAVR procedure under local anesthesia (LA) is a safer alternative, which can simplify, speed up the procedure, and patients can be recovered in a low dependency (LD) environment with a patient-to-nurse ratio of 2 to 1 or lower. No data have previously been presented in patients transferred directly to an LD unit postprocedure. We, therefore, compared outcomes of patients having the procedure under GA and LA using the valve academic research consortium (VARC) outcomes.
Methods
We retrospectively analyzed all transfemoral TAVR procedures conducted at our center from January 03, 2011, to May 11, 2015. Of 217 patients, 145 had the procedure under GA and 71 under LA. In the GA group, the devices used were Sapien XT 95%, Sapien 3 (S3) 2%, and direct flow medical (DFM) 3%. In the LA group, the devices used were S3 64%, XT 8%, and DFM 28%. The technical aspects of performing a TAVR have previously been explained in the literature. Transoesophageal echocardiography for aortic annular measurements was used in 80% of patients under GA and computed tomography (CT) in the remaining 20%, whereas CT was used in 100% of those under LA. All procedures were performed in a cardiac catheterization laboratory. In the GA group, an anesthetist was required for sedation, endotracheal intubation, and hemodynamic monitoring and patients subsequently recovered in an HD environment. TOE was performed by an experienced imaging consultant cardiologist and was used postvalve deployment to assess valve placement and paravalvular leak.
The GA cases were performed consecutively from January 3, 2011, to January 2, 2014, as GA was the department’s standard practice at that time. We made the transition to routine LA for all patients on January 2, 2014, and no transfemoral TAVR was performed under GA after this date. The main selection criterion for performing a TAVR under LA was the absence of significant peripheral vascular disease on CT that would impede transfemoral access. All patients were discussed in a multidisciplinary team meeting, in which the CT, transthoracic echocardiogram (TTE), coronary angiogram, and clinical state were discussed between members of both the TAVR and surgical team. All LA cases were performed consecutively except in 2 patients who alternatively underwent TAVR through the transapical route because of significant peripheral artery disease. At the beginning, all LA cases were performed with a consultant anesthetist in the catheter laboratory administering analgesia, and from December 12, 2014, most LA cases were performed without an anesthetist present.
In the LA group, 65% of procedures were performed without an anesthetist with sedation and analgesia given by a nurse. TOE was not used, and valve placement was performed under fluoroscopic guidance. Valve performance including paravalvular leak was assessed on the table with aortography, invasive hemodynamics, and transthoracic echocardiography. Patients were transferred to an LD unit requiring only telemetry postprocedure. All patients had a departmental TTE before discharge to assess peak and mean aortic gradients, the presence of aortic regurgitation, and left ventricular function.
Device success and procedural complications were assessed according to the VARC criteria and included major stroke, the need for pacing postprocedure, new paravalvular leak, acute kidney injury, bleeding complications, periprocedural myocardial infarction, success of valve placement, and 30-day all-cause mortality. Statistical analyses were performed using SPSS software package (version 21). Categorical data are expressed as numbers and percentages and compared using the chi-square or Fisher’s exact test with statistical significance at p <0.05. Continuous variables are expressed as mean ± SD and were compared using the Student’s t test. Nonparametric continuous variables are given as median (interquartile range), and the Mann-Whitney 2 independent-sample test was used to compare the data across both groups.
Results
Demographic and clinical details of all patients are presented in Table 1 . Both groups were similar with respect to age, co-morbidities, logistic Euro Score, and the severity of the aortic stenosis. Procedural outcomes are listed in Table 2 . The procedure time was significantly shorter in the LA group measured from time in room to skin closure, whereas skin open to skin closure time were the same in both groups. A prolonged procedure time in the GA group was because of the anesthetic time required to intubate and ventilate the patient. There were no differences in successful valve deployment. Most patients in the GA group had a TOE during the procedure to assess valve positioning and paravalvular leak, whereas no TOE was used in the LA group. Two patients in the LA group were converted to GA and needed intubation during the procedure, one to convert to a transaortic approach because of difficult vascular access and one to perform a femoral artery repair surgically because of failure of percutaneous closure. Surgical cut down for the transfemoral approach was used in 40.6% of the GA group, whereas the percutaneous approach was used in 98.6% of the LA group. The vascular closure technique was surgical closure in 57.9% of patients under GA and percutaneous closure in the remaining patients, whereas percutaneous closure was used in 94.4% of the LA group.
