記住我
Abstract
Background: General anesthesia has traditionally been used in transcatheter aortic valve replacement; however, there has been increasing interest and momentum in alternative anesthetic techniques.
Aims: To perform a descriptive study of anesthetic management options in transcatheter aortic valve replacements in the United States, comparing trends in use of monitored anesthesia care versus general anesthesia.
Settings and Design: Data evaluated from the American Society of Anesthesiologists' (ASA) Anesthesia Quality Institute's National Anesthesia Clinical Outcomes Registry.
Materials and Methods: Multivariable logistic regression was used to identify predictors associated with use of monitored anesthesia care compared to general anesthesia.
Results: The use of monitored anesthesia care has increased from 1.8% of cases in 2013 to 25.2% in 2017 (p = 0.0001). Patients were more likely ages 80+ (66% vs. 61%; p = 0.0001), male (54% vs. 52%; p = 0.0001), ASA physical status > III (86% vs. 80%; p = 0.0001), cared for in the Northeast (38% vs. 22%; p = 0.0001), and residents in zip codes with higher median income ($63,382 vs. $55,311; p = 0.0001). Multivariable analysis revealed each one-year increase in age, every 50 procedures performed annually at a practice, and being male were associated with 3% (p = 0.0001), 33% (p = 0.012), and 16% (p = 0.026) increased odds of monitored anesthesia care, respectively. Centers in the Northeast were more likely to use monitored anesthesia care (all p < 0.005). Patients who underwent approaches other than percutaneous femoral arterial were less likely to receive monitored anesthesia care (adjusted odds ratios all < 0.51; all p = 0.0001).
Conclusion: Anesthetic type for transcatheter aortic valve replacements in the United States varies with age, sex, geography, volume of cases performed at a center, and procedural approach.
Keywords: Anesthesia type, general anesthesia, monitored anesthesia care, transcatheter aortic valve replacement
How to cite this article: Introduction 
Transcatheter aortic valve replacement is a procedure that has been historically popularized in view of its specific utility in patients with prohibitive surgical risk.[1] As such, over the past 2 decades, transcatheter aortic valve replacement has demonstrated utility in the management of severe symptomatic aortic stenosis with ever increasing application in patients considered to be high-, intermediate-, and more recently, low-risk.[1],[2] This growing trend is supported by a recent meta-analysis that confirmed reduction in all-cause procedural mortality and stroke, regardless of baseline survival risk when comparing transcatheter aortic valve replacement to standard surgical aortic valve replacement.[2] As such, transcatheter approaches to aortic valve replacement are fast becoming a reasonable alternative for most patients. Nevertheless, despite the increasing popularity for this minimally invasive option, no consensus currently exists with respect to the optimal anesthetic approach.
Anesthetic management in transcatheter aortic valve replacement has traditionally been general anesthesia. Indeed, this is true of the initial reports of high-risk patients.[3] Nevertheless, there has been increasing interest and momentum in the use of alternative anesthetic techniques in the management of this often-frail cohort of patients. Efforts have shifted toward establishing a safe anesthesia alternative without compromising procedural or clinical outcome.[4],[5] To this end, monitored anesthesia care and sedation techniques have been reported to decrease procedural time, simplify monitoring, decrease intraprocedural vasopressor requirements, decrease risk of hemodynamic instability, decrease inotropic requirement, and decrease intensive care unit length of stay.[1],[6],[7],[8],[9] In theory, there may be an opportunity to defray costs through a less intensive approach with a reduced need for hemodynamic infusions and associated intensive monitoring.[4] A minimal approach to both the procedure and the delivery of anesthesia may further decrease overall length of stay as well as in-hospital and 30-day mortality.[10]
In a bid to better delineate contemporary patterns of practice, we sought to evaluate trends in transcatheter aortic valve replacement anesthetic management using national data from the American Society of Anesthesiologists' (ASA) Anesthesia Quality Institute's (AQI) National Anesthesia Clinical Outcomes Registry (NACOR). We hypothesized that the use of monitored anesthesia care has increased over time, but that general anesthesia remains the anesthetic used most frequently in transcatheter aortic valve replacements.
Materials and Methods We performed a retrospective analysis of data pertaining to all adult patients 18 years of age or older who had undergone transcatheter aortic valve replacements between 2013 and 2017. NACOR is a nationwide registry in the United States that collects and organizes electronic reports on anesthesia care. The NACOR database incorporates approximately 25% of all anesthesia case volume in the United States, including patient, procedural, facility, and provider characteristics.[11] All United States anesthesia practices can partake in submission to this database and are expected to contribute at minimum 20 elements per case but without strict requirements as to what data is required.[12],[13] The database comprises 9% small (1–25), 23% medium (26–75), 7% large (76–250), and 1% mega (251+) sized practices.[14] Patients who had undergone emergent procedures were excluded. Additionally, those in whom greater than one procedural code was listed were excluded as this suggested that conversion from an initial approach to an alternative approach to transcatheter aortic valve replacement had occurred. This manuscript adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.[15] The West Virginia University institutional review board approved the study and considered it exempt from the informed consent requirement.
Statistical analysis
The variable of interest was primary anesthetic management type, and we specifically compared general anesthesia to monitored anesthesia care. We combined the variable “conscious sedation”, which was a less precise “catchall” data field used prior to 2016 when NACOR data was compiled largely based on local definitions, and “monitored anesthesia care”, which became the standard terminology adopted in 2017 when the AQI released standardized data definitions. The AQI's current definition of monitored anesthesia care is, “A specific type of anesthesia service in which a qualified anesthesia provider has been requested to participate in the care of a patient undergoing a diagnostic or therapeutic procedure. Indications for monitored anesthesia care depend on the nature of the procedure, the patient's clinical condition, and/or the potential need to convert to a general or regional anesthetic. Deep sedation/analgesia is also included in monitored anesthesia care.”[16] Thus, according to the ASA's Standards and Guidelines, monitored anesthesia care does not necessarily denote the continuum of depth of sedation; rather, monitored anesthesia care may include minimal sedation anxiolysis, moderate sedation/analgesia (”conscious sedation”), or deep sedation/analgesia in which qualified anesthesia personnel have been requested to participate in the care of a patient undergoing a diagnostic or therapeutic procedure.
We evaluated the frequency of use of each of the two anesthesia management types over time. Cochran-Armitage testing was used to determine whether or not the trends were statistically significant. We used multivariable logistic regression model to evaluate patient, hospital, and procedural characteristics to identify variables associated with the primary outcome. We did not use variables that had rates of missing data >10%. Other variables were either imputed to the mode, median, or complete case analysis was used. Investigators selected factors based on clinical and theoretical reasoning. The models were adjusted for age, sex, race, ASA physical status, US census region in which a center is located, procedural approach, median income by patient zip code, and center volume of transcatheter aortic valve replacements. Volume of transcatheter aortic valve replacements was calculated as the total number of all transcatheter aortic valve replacements for each practice during each year for our study period. In order to account for practice level variation, we used three-level hierarchical modeling and nested patients within practices within each year to properly generate our variance estimates. Model fit was assessed using C-statistic, Hosmer-Lemeshow test, as well as a model calibration plot. Continuous variables were presented as medians (interquartile range, 25th and 75th percentile) and differences across anesthesia type were assessed using Mann-Whitney U tests. Categorical variables were presented as counts (proportions) and differences across anesthesia type were assessed using the Chi-square test.
The secondary outcome was case duration in minutes, defined as duration in minutes from the recorded anesthesia start to anesthesia finish. Based on the distribution of this outcome, we excluded cases with duration less than or equal to 60 min as this was considered to be the lower 1% and deemed unlikely to be realistic. Additionally, in order to account for the skewed nature of the data, we log-transformed the outcome. We fit a multivariable linear model to assess the association between anesthesia management type and case duration. We also utilized the same three-level hierarchical structure as described above, also adjusting for the same covariates. All statistics were performed using SAS v 9.4 (SAS, Cary, NC).
Results Study population
From 2013 to 2017, the NACOR database includes 44,808,877 observations from 836 practices. After excluding cases based on inclusion criteria, the final data set pertained to 19,909 transcatheter aortic valve replacements from 160 practices [Figure 1]. A total of 15.99% were performed at academic hospitals, 47.54% at community hospitals (29.64% large [>500 beds], 17.88% medium [100-500 beds], and 0.02% small [<100 beds]), 2.99% at specialty hospitals, 0.02% at an attached surgery center, 0.26% at a freestanding surgery center, 0.01% at a surgeon's office, and 33.22% had unknown/unavailable facility type. The geographical distribution was 24.73% in the Northwest, 26.96% in the Midwest, 32.04% in the South, and 16.27% in the West with 13% occurring in rural areas, 66.38% in urban areas, and 20.62% unknown.
Figure 1: Flow diagram for information retrieved from the National Anesthesia Clinical Outcomes Registry (NACOR) DatabaseUnivariable analysis
Monitored anesthesia care increased over time from 1.8% of transcatheter aortic valve replacements in 2013 to 25.2% of cases in 2017 (p = 0.0001) [Figure 2]. Performance of 25.9% of transcatheter aortic valve replacements managed with monitored anesthesia care occurred at teaching facilities while only 14% of cases managed with general anesthesia occurred at teaching facilities (p = 0.0001). Patients receiving monitored anesthesia care were more likely to be 80 + years of age (66% vs. 61%; P = 0.0001), male (54% vs. 52%; P = 0.0001), have ASA physical status > III (86% vs. 80%; P = 0.0001), have received care in the Northeast (38% vs. 22%; P = 0.0001), and reside in zip codes with higher median income ($63,382 vs. $55,311; P = 0.0001) [Table 1].
Figure 2: Trends of anesthesia management type for Transcatheter Aortic Valve Replacements (TAVR) over time
Table 1: Descriptive table of patient and center characteristics by anesthesia typeMultivariable analysis
Each one-year increase in age was associated with a 3% increase in the odds of the patient undergoing monitored anesthesia care for transcatheter aortic valve replacement (p = 0.0001). Male patients had 16% higher odds of receiving monitored anesthesia care than females (p = 0.026). Centers in the Midwest, South, and West were less likely to use monitored anesthesia care than those in the Northeast (adjusted odds ratio 0.28, 0.27, and 0.1, respectively; all P < 0.005). A 33% increased odds of use of monitored anesthesia care was observed for every 50 transcatheter aortic valve replacements performed annually at any given practice (p = 0.012). Furthermore, patients who underwent open femoral arterial, open axillary arterial, transaortic, or transapical approaches were less likely to receive monitored anesthesia care compared to those who underwent percutaneous femoral arterial approach (adjusted odds ratios all < 0.51; all P = 0.0001) [Table 2]. Model fit was good with a c-statistic of 0.93.
Table 2: Multivariable adjusted odds ratios for monitored anesthesia care in transcatheter aortic valve replacements (TAVR)Cases using monitored anesthesia care were associated with shorter median case durations than those using general anesthesia (174 [IQR 141-223] and 182 [IQR 149-231] min, respectively; P = 0.0001). Additionally, adjusting for covariates revealed that cases with monitored anesthesia care had 11% shorter median case durations compared to those undergoing general anesthesia (Model estimate 0.89 [0.87–0.9]; P = 0.0001).
Discussion Our results suggest a steady increase in the use of alternatives to general anesthesia in the management of patients undergoing transcatheter aortic valve replacement. The increase in use of monitored anesthesia care was greater than 20-fold over the 4 year duration of the study period. Interestingly, we did not observe a difference in the odds of receiving monitored anesthesia care vs. general anesthesia by ASA physical status in multivariable analysis. However, monitored anesthesia care and sedation have each been associated with satisfactory, even superior outcomes with respect to procedural time, decreased intraoperative vasopressor support, optimized hemodynamic profiles, decreased inotropic requirement, and shorter intensive care unit stays.[1],[6],[7],[8],[9],[17],[18] This improvement in outcomes has further fueled the popularity in the use of monitored anesthesia care.
In previous reports, 5–17% of patients are reportedly transitioned from conscious sedation to general anesthesia, the most common reason for which is the unanticipated occurrence of vascular and pulmonary complications.[10] Conscious sedation has been described as a drug-induced depression of consciousness during which the patient may purposefully respond to verbal commands or light tactile stimulation.[10] Sudden and emergent intubation of these patients can inadvertently result in hemodynamic instability.[10] This aside, the growing majority of outcomes favor the use of monitored anesthesia care in patients deemed appropriate for its use. Nevertheless, patients may not necessarily be candidates for monitored anesthesia care for myriad reasons, and this includes those unable to undergo preoperative computed tomography angiography for valve sizing. This cohort may benefit instead from intraoperative three-dimensional transesophageal echocardiography sizing under general anesthesia as an alternative to non-invasive imaging.
The use of monitored anesthesia care, nevertheless, does not necessarily proffer a panacea. This is most apparent in the restriction in transesophageal echocardiography for intraoperative imaging, though, there is growing interest in its use in patients receiving monitored anesthesia care.[19] Echocardiographic imaging is more sensitive than fluoroscopy and contrast injection to detect paravalvular leak in both the balloon expandable and self-expandable valves. For some patients, however, transthoracic echocardiography may pose challenges. Indeed, the management and detection of the complications such as paravalvular leaks, aortic root disruption, or the imaging of complex vascular access under monitored anesthesia care may be difficult. Paravalvular leaks, for example are associated with poor long-term outcomes. However, a retrospective review of 111 patients by Sengupta et al.[20] has shown statistically similar outcomes with the use of transthoracic echocardiography guidance for transcatheter aortic valve replacement under monitored anesthesia care in patients with regard to procedural success, paravalvular regurgitation degree, the need for additional valve implantation, and periprocedural stroke and mortality. As such, though our data may have been restricted largely to epidemiological evaluation of the use of monitored anesthesia care, more evidence may still be necessary to best establish whether potential short-term gain such as decreased procedural duration and decreased inotropic requirement translates into long-term quality improvement.
Our study adds to the growing body of evidence that has identified monitored anesthesia care as an increasingly popular option for transcatheter aortic valve replacement. Clinical outcomes are increasingly more favorable with lower length of stay and without compromising other outcomes.[19] The ethos of lower depths of sedation than general anesthesia on the continuum of depth of sedation is certainly in keeping with the intent to maintain a minimally invasive approach in valve implantation as a whole, particularly true in a clinically debilitated cohort with prohibitive risk. The evolution of surgical techniques has elevated the safety and efficacy of many high-risk procedures in contemporary cardiovascular clinical practice, such as extracorporeal support, for example, and transcatheter aortic valve replacement appears to be on the same trajectory. This has duly been coupled with improvements in contemporary clinical and intensive care management, each of which has played a crucial role in optimizing these outcomes as a whole. Our results suggest that transcatheter aortic valve replacement has benefitted in the same way and may explain how the outcomes have prompted its extrapolation onto patients with lower surgical risk.
The geographical variation in use of monitored anesthesia care may suggest an ongoing regional disparity in access to care. The use of monitored anesthesia care versus general anesthesia varies with region and this, nevertheless, is likely further proof of pervasive disparities that tend to affect vulnerable populations from specific regions of the country (south and Midwest). Many of these areas have fewer providers and less favorable patterns of insurance penetration.[21],[22],[23] The utilization of monitored anesthesia care may arguably lower the costs of provision of anesthesia in these cases, that may in turn have a significant effect on lowering costs overall, broadening access at multiple levels of provision and management both locally and regionally.
Our study has several limitations. First, the use of retrospective registry data is fraught with bias and confounding factors. Second, the lack of granular detail precludes the ability to evaluate nuances within clinical practice and postoperative outcomes. Third, because the data were drawn from a North American registry, generalizability is restricted in geographical context to the United States. Fourth, the absence of follow-up data precludes the ability to estimate long-term outcomes. Fifth, the comparison of monitored anesthesia care with general anesthesia introduces the confounding phenomenon of the simultaneous interplay of technological and surgical advances over the temporal course of the duration of the study. Sixth, the lack of consensus in nomenclature regarding monitored anesthesia care and general anesthesia introduces a wide scope of heterogeneity in the discussion of the continuum of depth of sedation that is open to misinterpretation. Seventh, the dataset does not have a unique identifier to determine repeat procedures on the same patient. Moreover, the data is coded for the primary anesthetic; thus, conversion from monitored anesthesia care to general anesthesia would have only been categorized as general anesthesia. The dataset also lacks long-term follow-up or detail pertaining to morbidity and mortality. Finally, NACOR is a database specifically for anesthetic management; therefore, sedation performed by the proceduralist without the involvement of anesthesia personnel would not be captured. Many of these limitations lend nicely to areas of future research that will strengthen our current results.
Conclusion This study draws upon a large national denominator that proffers a legitimate comparison between two different techniques of anesthesia performed in contemporary practice. The patterns confirm the increased popularity of monitored anesthesia care for transcatheter aortic valve replacements in the United States over time. Anesthetic type, however, varies with age, sex, geography, volume of cases performed at a center, and procedural approach.
Authorship statement and justification
All authors listed satisfy requirements for authorship. All authors have read and approve of this manuscript and believe that it represents honest work. Data acquisition involved a core group of researchers, and this, along with those involved in statistical analysis, data interpretation, and manuscript writing necessitate inclusion of more than 6 authors as each provided meaningful contribution.
Financial support and sponsorship
Support was provided solely from institutional and/or departmental sources.
Conflicts of interest
There are no conflicts of interest.
References 
Correspondence Address:
Heather K Hayanga
Division of Cardiovascular and Thoracic Anesthesiology, Heart and Vascular Institute, West Virginia University, 1 Medical Center Drive, Morgantown, WV 26505
United States
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/aca.aca_311_20
留言 (0)