PROPOFOL VERSUS REMIFENTANIL SEDATION FOR TRANSCATHETER AORTIC VALVE REPLACEMENT: A SINGLE ACADEMIC CENTER EXPERIENCE

Abstract

Objective:

Comparison of remifentanil versus propofol for sedation during transcatheter aortic valve replacement (TAVR) procedures to analyze the risk of sedation related hypoxemia and hypotension. Secondary outcomes included the rate of conversion to general anesthesia, procedure length, rate of intensive care unit (ICU) admission, ICU and hospital length of stay, and thirty-day mortality.

Design:

Retrospective cohort study.

Setting:

A single tertiary teaching hospital.

Participants:

Two hundred and fifty nine patients who had propofol or remifentanil sedation for TAVR between March 2017 and March 2020.

Intervention:

None.

Measurements and Main Results:

There were 130 patients (50.2%) in the propofol cohort and 129 patients (49.8%) in the remifentanil cohort. The primary outcomes were oxygen saturation nadir values and vasopressor infusion use. Remifentanil was associated with a lower oxygen saturation nadir, as compared to propofol (91.3% vs. 95.4%, p < 0.001). Risk factors associated with hypoxemia (defined as < 92%) were body mass index (p = 0.0004), obstructive sleep apnea (p = 0.004), and remifentanil maintenance (p < 0.001). Vasopressor infusion use was significantly higher with propofol (64.9% vs. 8.5%, p < 0.001). Propofol maintenance and ACEi/ARB use were the only variables identified as risk factors for vasopressor use (p < 0.001 and p = 0.009).

Conclusions:

For patients undergoing TAVR with conscious sedation, remifentanil was associated with more hypoxemia while propofol was associated with a higher rate of vasopressor use.

Introduction

Owing to a similar safety profile and improved secondary outcomes as compared to general anesthesia (GA), conscious sedation has become the preferred type of anesthetic for transfemoral transcatheter aortic valve replacement (TAVR) procedures at many institutions.^1–3 According to data from the Transcatheter Valve Therapy (TVT) registry, the use of conscious sedation for this procedure has increased nationally, from 33% in January 2016 to 64% in March 2019.⁴ Conscious sedation however, is not without risk, and a conversion to GA rate as high as 5.9% has been reported in the TAVR literature, which includes anesthesia as well as procedure related complications.⁵ Anesthetic related complications that necessitate a conversion to GA include hemodynamic instability, hypoxemia, and uncooperative or uncomfortable patients. Respiratory depression secondary to over-medication is a leading cause of injury and liability during monitored anesthesia care for all types of surgeries.⁶

Commonly used medications for conscious sedation include propofol, remifentanil, ketamine, and dexmedetomidine. Among those options, propofol and remifentanil are well known to cause dose-dependent respiratory depression, hypoxemia and hypotension, which can be especially harmful in patients with underlying cardiac disease. The aim of this retrospective review was to analyze the risk of sedation-related hypoxemia and hypotension associated with these propofol and remifentanil for TAVR procedures. Secondary outcomes included the rate of conversion to GA, procedure length, rate of intensive care unit (ICU) admission, ICU and hospital length of stay (LOS), and thirty-day mortality.

Methods

Prior to the study, approval by the Institutional Review Board was obtained (IRB # 18–1883) and the requirement for written informed consent was waived. A retrospective cohort study was performed for patients who underwent TAVR with propofol or remifentanil sedation at the University of North Carolina Hospital over a three-year period (March 14, 2017 to March 14, 2020). The only exclusion criterion was any major procedural complication that resulted in hemodynamic collapse and required surgical intervention. Study data were collected and managed using REDCap® electronic data capture tools hosted at the University of North Carolina.

TAVRs were performed in a hybrid operating room by a cardiologist and a cardiac surgeon. Standard American Society of Anesthesiology monitors were used in all patients. Masimo® (Irvine, California, USA) pulse oximeters were used to monitor SpO₂. Supplemental oxygen was delivered and capnography monitored via a nasal cannula with an end-tidal carbon dioxide (ETCO₂) sampling line or a face mask with an ETCO₂ sampling line placed inside the mask. Femoral arterial and venous lines were placed for invasive blood pressure monitoring and large bore access, respectively. Monitored anesthesia care, henceforth referred to as “sedation,” was provided by qualified anesthesia personnel at all times. Patients were maintained on propofol or remifentanil infusions that were titrated to facilitate patient tolerance of the procedure while maintaining spontaneous ventilation. There was not an objective target concentration rate, as each infusion was titrated to individual patient needs. Vasopressor infusions were initiated to maintain mean arterial pressure (MAP) > 65 mmHg. Post-operatively, patients were either discharged to the post-anesthesia care unit for Phase I recovery, cardiac catheterization lab for Phase II recovery, or ICU.

The primary outcomes measured were intraoperative oxygen saturation nadir, which was identified as the lowest recorded SpO₂, and vasopressor infusion use. Hypoxemia was determined a priori to be defined as SpO₂ < 92%.⁷ The SpO₂ recorded in the medical record is averaged over a minute, therefore, the nadir values are a reflection of sustained oxygen desaturation. Given that dynamic and transient blood pressure swings can occur during a TAVR procedure, particularly during rapid ventricular pacing and balloon valvuloplasty stages, vasopressor infusion use was thought to be a better indicator of the overall trend in hypotension rather than singular blood pressure measurements. The vasopressors used include norepinephrine, phenylephrine, and vasopressin. The outcome was defined as simply yes or no, irrespective of dose and duration of use. The secondary outcomes were rate of conversion to general anesthesia, procedural length, intensive care unit admission rates, intensive care unit length of stay, hospital length of stay, and thirty-day mortality.

The project described was supported by the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health (NIH), through Grant Award Number UL1TR001111. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Statistical Analysis:

Data analysis was carried out using R 3.6.3® (Vienna, Austria). Patient demographics, medical history and outcome variables were summarized and compared by drug group with descriptive statistics. Continuous variables were summarized by mean and standard deviation and categorical variables were summarized by frequencies and percentages. T test or Kruskal—Wallis rank test were used for comparing continuous variables while Chi-Square test or Fisher exact test were used for comparing categorical variables between propofol and remifentanil groups. All tests used a significance level of 0.05. The effect of propofol and remifentanil on hypoxia and vasopressor use were examined using propensity score analyses. Inverse probability weight was estimated via a logistic model, where the response variable was the drug group. The four variables that were most different between the two drug groups, STS score, surgical risk, hypertension, and peripheral vascular disease were adjusted for. Using one-to-one matching with replacement based on fitted propensity score, where estimand is the average treatment (drug group) effect, the effect of the drug on the outcome was estimated. Logistic regression models were used to decide risk factors and calculate odds ratios (OR) for hypoxia and vasopressor use.

Results

Patient Population

Of 262 patients, 259 patients were included in our study. Three patients, one in the propofol cohort and two in the remifentanil cohort, were excluded from our study because they had a procedural complication that necessitated intervention by cardiac surgery. Remifentanil was primarily used during the first half of the data collection period. With a nationwide remifentanil shortage, its use sharply declined and propofol use predominated the latter part of the data collection period (Figure 1). The remifentanil cohort included 129 patients while the propofol cohort included 130 patients. Both cohorts were similarly matched in terms of demographics and medical history (Table 1). The Society of Thoracic Surgeons (STS) risk scores were higher in the remifentanil group at 5.21 ± 3.4 than the propofol group at 4.10 ± 4.3 (p < 0.001).

Figure 1.

Institutional trend in the use of propofol and remifentanil for conscious sedation for TAVRs from March 2017 to February 2020.

Abbreviations: TAVR, transcatheter aortic valve replacement.

Table 1.

Patient characteristics.

	Propofol (n = 130)	Remifentanil (n = 129)	p Values
Demographics
Age	78.2 ± 10.3	79.6 ± 8.4	0.311
Female	54 (41.5%)	65 (50.4%)	0.153
BMI	28.5 ± 6.2	29.7 ± 6.8	0.171
Weight (kg)	81.6 ± 20.7	83.5 ± 21.8	0.437
Medical History
STS Risk Score	4.1 ± 4.3	5.2 ± 3.4	< 0.001
Surgical Risk			0.006
Low Risk	10 (7.7%)	1 (0.8%)
Intermediate Risk	26 (20%)	39 (30.2%)
High Risk	94 (72.3%)	89 (69%)
LV EF (%)	51.7 ± 13.3	50.7 ± 13.4	0.771
NYHA Class			0.232
I	3 (2.4%)	0 (0%)
II	30 (24.0%)	28 (21.7%)
III	87 (69.6%)	92 (71.3%)
IV	5 (4.0%)	9 (7.0%)
Prior Myocardial Infarction	13 (10.0%)	19 (14.7%)	0.683
Cerebrovascular Disease	26 (20.0%)	28 (21.7%)	0.736
Chronic Kidney Disease	48 (36.9%)	45 (34.9%)	0.732
Chronic Lung Disease	30 (23.1%)	32 (32.6%)	0.089
Chronic Pain	27 (20.8%)	25 (27.1%)	0.230
Diabetes Mellitus	52 (40.0%)	63 (48.8%)	0.152
Hearing Loss	36 (27.7%)	45 (34.9%)	0.212
Hypertension	120 (92.3%)	127 (98.4%)	0.019
Obstructive Sleep Apnea	22 (16.9%)	26 (20.2%)	0.503
Peripheral Vascular Disease	16 (12.3%)	31 (24.0%)	0.014
Psychiatric Disorder	42 (32.3%)	34 (26.4%)	0.293
Restless Leg Syndrome	6 (4.6%)	14 (10.9%)	0.060

Abbreviations: BMI, body mass index; LV EF, left ventricular ejection fraction; NYHA, New York Heart Association; STS, Society of Thoracic Surgeons.

Outcomes

Differences in outcomes are outlined in Table 2. The propofol group had a mean SpO₂ nadir of 95.4% (± 4.55). The median was 97% with a range of 77–100%. The remifentanil group had a mean of 91.3% (± 7.41) with a median of 93% and range of 52–100%. The difference between the two groups was statistically significant (p < 0.001). A propensity analysis was performed in which the four variables that were most different between the two cohorts, STS score, surgical risk, hypertension, and peripheral vascular disease, were controlled for. The average effect of drug on hypoxia is 0.191 (p = 0.009) when comparing the remifentanil cohort to the propofol cohort.

Table 2.

Primary and secondary outcomes for propofol and remifentanil cohorts.

	Propofol	Remifentanil	p value
Oxygen Saturation Nadir (%)
Mean	95.4 ± 4.6	91.3 ± 7.5	< 0.001
Median	97	93
Range	77–100	52–100
Vasopressor Infusion Use	59 (45.5%)	9 (7.0%)	< 0.001
Procedural Length (mins)	119.2 ± 30.8	136.8 ± 28.1	< 0.001
Post-Op ICU Admission	52 (40%)	47 (36.4%)	0.740
ICU Length of Stay (hours)	33.8 ± 25.9	66.3 ± 79.0	0.017
Hospital Length of Stay (days)	2.1 ± 2.0	3.1 ± 3.7	0.016
Thirty-Day Mortality	3 (2.3%)	6 (4.7%)	0.303

Abbreviations: ICU, intensive care unit.

Of note, 46% of patients in the propofol cohort received supplemental oxygen via nasal cannula compared to 51% in the remifentanil group who received supplemental oxygen via nasal cannula, with the remainder receiving oxygen via facemask. Average oxygen flow rate for each group was 4.8 liters per minute (LPM) and 4.6 LPM, respectively. In the 30 minutes prior to the oxygen saturation nadir, 39 patients in the propofol cohort and 38 patients in the remifentanil cohort received supplemental bolus doses of propofol (10–20 mg), fentanyl (12.5–25 mcg), or dexmedetomidine (4–8 mcg). For context of the amount of anesthetic maintenance immediately prior to SpO₂ nadir, the average infusion dose of propofol until the point of the nadir was 50.8 mcg · kg⁻¹ · min⁻¹ (± 31.9) and the average dose of remifentanil was 0.054 mcg · kg⁻¹ · min⁻¹ (± 0.030).

Vasopressor infusion use was significantly higher in the propofol cohort, 59 (45.4%) patients compared to 9 (7.0%, p < 0.001). A propensity analysis was performed as above and the average effect of drug on vasopressor infusion use is 0.432 (p = 1.14 × 10⁻⁹) when comparing the propofol cohort to the remifentanil cohort. Of the patients who required vasopressor infusions, the majority required the infusion both before and after valve deployment, 94% in the propofol group, and 78% in the remifentanil group. Three patients in the propofol cohort continued to have a vasopressor requirement at the conclusion of the procedure compared to zero patients in the remifentanil cohort.

Procedural length was 136.8 minutes in the remifentanil group compared to 119.2 minutes in the propofol group. (p < 0.001). Rates for conversion to GA are similar in both groups when accounting for anesthesia-related causes (4.6% for propofol, 5.4% for remifentanil), procedure-related causes (0.8% for propofol, 1.6% for remifentanil), and all causes (5.3% for propofol, 6.9% for remifentanil) (Table 3). There was no difference in ICU admissions (p = 0.740) but remifentanil was associated with longer ICU LOS (p = 0.017) and hospital LOS (p = 0.016). The median ICU LOS was 46 hours and hospital LOS was 2 days for the remifentanil group compared to 27 hours in the ICU and 1 day in the hospital for the propofol group. Thirty-day mortality was 2.3% (n = 3) in the propofol group versus 4.7% (n = 6) in the remifentanil group (p = 0.303).

Table 3.

Reasons for conversion to general anesthesia in the propofol and remifentanil cohorts.

Patient	Reason for Conversion to GA
Propofol
1	Uncooperative patient
2	Uncontrolled, copious bleeding after placement of nasal trumpet; converted to GA for a secured airway
3	Uncooperative patient
4	Uncooperative patient
5	Unknown
6	Patient coughing
7	Valve dislodged and occluded RCA with resultant cardiac arrest, necessitating ACLS and emergent SAVR*
Remifentanil
1	Disinhibition, patient moving and not re-directable
2	Hypoventilation/airway obstruction
3	Became unresponsive and apneic after valve deployment
4	Became unresponsive and apneic after valve deployment
5	Hypoventilation/airway obstruction
6	Anaphylaxis to contrast±
7	Uncooperative patient
8	Uncooperative patient
9	Acute pericardial effusion causing severe hemodynamic instability, altered mental status, and hypoventilation*

^*Procedural complication that was not included in analysis.

^±Procedural complication, however, was included in analysis as it did not result in surgical intervention. Of note, the remifentanil group had another procedural complication, a left ventricular perforation, that did not result in conversion to general anesthesia because it was managed with a pericardial drain, and therefore, it was not included in this table.

Abbreviations: ACLS, advanced cardiac life support; GA, general anesthesia; RCA, right coronary artery; SAVR, surgical aortic valve replacement.

Predictors of Hypoxemia

In the remifentanil group, 38.8% of patients had hypoxemia, defined as SpO₂ < 92%, compared to 14.6% in the propofol group (p < 0.001). The average remifentanil dose prior to the oxygen nadir was 0.061 mcg · kg⁻¹ · min⁻¹ in patients who became hypoxemic compared to 0.049 mcg · kg⁻¹ · min⁻¹ in patients who did not (p = 0.023). The average propofol dose was 52 mcg · kg⁻¹ · min⁻¹ in patients with a SpO₂ nadir < 92% and 43 mcg · kg⁻¹ · min⁻¹ in patients with a higher SpO₂ nadir (p = 0.628). A secondary analysis was performed in patients who had become hypoxemic. Factors associated with hypoxemia included higher drug doses, higher patient weight, BMI, and a history of obstructive sleep apnea (OSA) (Table 4). The OR of hypoxemia with remifentanil use compared to propofol was 3.698 (confidence interval (CI) 2.06–6.88, p = 2.06 × 10⁻⁵). This association still stands when adjusting for BMI (OR 3.56, CI 1.95–6.70, p = 5.33 × 10⁻⁵) and OSA (OR 3.72, CI 2.05–7.02, p = 2.5 × 10⁻⁵). When stratifying patients into obese (BMI > 30), overweight (BMI 25–29.9), normal (BMI 18.5–24.9), and underweight (BMI <18.5), the OR of an obese patient becoming hypoxemic when compared to normal BMI was 3.7 (CI 1.75–8.62, p = 0.001). The OR for patients in the overweight group was 2.1 (CI 0.95–4.90, p = 0.076) compared to the normal group, while the underweight group was 1.059 × 10⁻⁶ (p = 0.989) compared to the normal group.

Table 4.

Univariable predictors for hypoxia and vasopressor use.

Variables	Estimate	Standard Error	z Value	p Value
Hypoxia
Age	0.004	0.015	0.291	0.771
Weight	0.018	0.006	2.77	0.006
BMI	0.077	0.022	3.53	0.0004
Chronic Lung Disease	0.080	0.311	0.257	0.797
Asthma	0.034	0.692	0.048	0.961
COPD	0.051	0.351	0.146	0.884
Pulmonary Fibrosis	1.022	1.42	0.719	0.472
OSA	0.971	0.334	2.91	0.004
CPAP Use	0.214	0.590	0.364	0.716
Smoking History	0.341	0.233	1.46	0.145
Benzodiazepine Use*	0.383	0.450	0.852	0.394
Opioid Use*	0.097	0.435	0.224	0.823
Vasopressor Use
Age	0.005	0.015	0.352	0.725
Preoperative LVEDP	0.025	0.018	1.35	0.176
LV EF	0.006	0.010	0.605	0.545
Loop Diuretic Use	0.183	0.279	0.656	0.512
ACEi/ARB Use	0.767	0.295	2.60	0.009
Use on Day of Procedure	0.243	0.251	0.971	0.332

Abbreviations: ACEi/ARB, angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker; BMI, body mass index; COPD, chronic obstructive pulmonary disease; CPAP, continuous positive airway pressure; DLCO, diffusion capacity for carbon monoxide; FEV1, forced expiratory volume in 1 second; LVEDP, left ventricular end diastolic pressure; LV EF, left ventricular ejection fraction; OSA, obstructive sleep apnea.

^*Benzodiazepine and opioid use are defined as any benzodiazepine or opioid use, regardless of frequency and dose.

Predictors of Vasopressor Infusion Use

Multiple variables were examined in the secondary analysis for patients who required vasopressor use (Table 4). Only angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker (ACEi/ARB) use was associated with increased vasopressor requirement (p = 0.009); however, use of ACEi/ARB on the day of procedure was not (p = 0.332). The OR of vasopressor use with propofol compared to remifentanil was 9.484 (CI 4.84–20.2, p < 1.0 × 10⁻⁵).

Discussion

Propofol and remifentanil are two commonly used medications for conscious sedation with a quick onset and offset and are easily titratable. Both, however, may be associated with side effects including respiratory depression and hypotension. Previous studies have compared propofol and remifentanil sedation for various types of procedures.^8–11 It appears that remifentanil is associated with a higher incidence of respiratory depression with a concomitant decrease in SpO₂, however, this finding was only statistically significant in two of the above mentioned studies. Of the three studies that evaluated hemodynamics, two found that propofol use was associated with lower blood pressure. These studies, however, had small sample sizes and the patient populations were not uniformly affected by advanced cardiopulmonary disease such as in the TAVR population. In addition, with the exception of the study by Krenn, et. al., who studied sedation for carotid endarterectomies, the types of procedures that were investigated are unlikely to cause significant hemodynamic shifts. To date, there is no study that compares the use of propofol and remifentanil for conscious sedation in the TAVR population.

When our program transitioned from general anesthesia to conscious sedation for TAVR procedures, remifentanil was the most frequently used medication for sedation. However, a nationwide remifentanil shortage forced a transition to propofol sedation, which has been the mainstay since (Figure 1). In this retrospective review, we found that the use of remifentanil for sedation was associated with a greater degree of hypoxemia than was seen in the propofol cohort. Our results mirror those from the studies performed by Smith and Mingus.^8,9 A secondary analysis revealed that the predictors of hypoxemia are remifentanil use, BMI, and a history of OSA. Despite a strong association between remifentanil use and hypoxemia, there was no difference in the rates of conversion to GA between the two cohorts. Seven patients in the propofol group and nine in the remifentanil group required a conversion to GA (Table 3).

Propofol sedation was associated with significantly higher vasopressor infusion use than remifentanil. Of the multiple variables that were assessed, propofol and ACEi/ARB use were the only factors associated with increased vasopressor infusion use. The latter finding is consistent with previous studies that demonstrated an increased risk of intraoperative hypotension in patients who take ACEi/ARBs.^12,13 We found no difference in vasopressor use in patients who took their ACEi/ARB on the day of surgery. While this topic has been heavily studied, the decision to continue or withhold ACEi/ARBs on the day of surgery remains controversial.^14,15

This study was primarily limited by its retrospective design. In addition, this was a single center study and the findings may not be generalizable unless further investigated by large, well designed multicenter randomized controlled trials. While patients were sedated with a propofol or remifentanil infusion, patients in both cohorts received intermittent small bolus doses of dexmedetomidine, fentanyl, or propofol. This isolated practice may represent a potential confounder for hypoxemia. The number of patients who received supplemental boluses, however, were similar among both groups, and this, therefore, is unlikely to confound the overall impact. In addition, dexmedetomidine is unlikely to contribute to the risk of hypoxemia. Similarly, small doses of fentanyl and propofol are unlikely to cause hypoxemia but can certainly increase the risk of hypoxemia in a patient that is already on a maintenance infusion of propofol or remifentanil. Markers of ventilation were not accounted for in this analysis. Despite the use of continuous waveform capnography, the accuracy of the documented respiratory rate and ETCO₂ in the setting of spontaneous ventilation with an open airway system is unreliable. The use of supplemental oxygen is a source of confounding for hypoxemia, however, both groups had a similar number of patients receiving supplemental oxygen via nasal cannula vs. face mask with similar oxygen flow rates.

A significant confounder in our data is the time frame in which remifentanil or propofol was used. Remifentanil sedation was the primary technique during the first half of the data collection period, while propofol sedation was primarily used in the latter half. As the cardiologists gained more experience with performing TAVRs, they became faster, which likely explains why propofol sedation was associated with a significantly shorter procedure time than remifentanil. Additionally, as our institution became more comfortable with the postoperative management of TAVR patients, patients were discharged from the ICU and the hospital more quickly. The shorter ICU and hospital length of stay for the propofol cohort compared to the remifentanil cohort can likely be attributed to this change in postoperative management. Lastly, there is the potential for learning effect with titration of medications. However, the use of remifentanil and propofol for sedation reflects common anesthetic practice across a variety of surgical procedures. Therefore, differences in hypoxemic nadir and vasopressor use should not have been affected by learner effect.

This is the first study to compare propofol to remifentanil in the TAVR population, however, studies have compared other medications for conscious sedation for TAVR.^16–18 Mayr et. al. compared propofol-opioid to dexmedetomidine and found that propofol-opioid sedation causes hypercarbia to a greater extent but there was no difference in arterial oxygen saturation or arterial oxygen tension.¹⁶ Chen, et. al. found no difference in hospital LOS, conversion to GA, and postoperative delirium when comparing propofol-only to propofol-dexmedetomidine.¹⁷ Most recently, Kronfli, et. al. compared propofol-only to dexmedetomidine-only sedation and found no differences in both in-hospital and 30-day outcomes and in cost.¹⁸

There is an armamentarium of medications that one can use for conscious sedation in TAVR procedures, but a lack of studies comparing their risks. Future studies are warranted to further expand on this topic in order to identify the optimal anesthetic technique in the TAVR population. Regardless of the preferred medication, familiarity with the multiple available options is advisable in the face of evolving drug shortages, cost variability, and patient related factors. With evolving technology and steep learning curves, the anesthetic management for TAVR procedures has become increasingly safe. However, balancing the risks of sedation with safety in the context of patient-specific co-morbidities and drug side effects are of utmost importance.

Conclusion

For patients undergoing TAVR with conscious sedation, remifentanil was associated with more hypoxemia while propofol was associated with a higher rate of vasopressor use. Propofol is preferred to remifentanil in obese patients or those with OSA because of an increased risk of hypoxemia in those who receive remifentanil. Propofol was, however, associated with more vasopressor use. Nonetheless, both were safe and effective for sedation for TAVRs. Further studies are warranted to identify the optimal anesthetic technique in the TAVR population.

Glossary of Terms

ACEi/ARB

angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker

BMI

body mass index

CI

confidence interval

ETCO₂

end-tidal carbon dioxide

GA

general anesthesia

ICU

intensive care unit

IRB

Institutional Review Board

LOS

length of stay

LPM

liters per minute

MAP

mean arterial pressure

NCATS

National Center for Advancing Translational Sciences

NIH

National Institute of Health

OR

odds ratio

OSA

obstructive sleep apnea

SpO₂

oxygen saturation as measured by pulse oximetry

STS

Society of Thoracic Surgeons

TAVR

transcatheter aortic valve replacement

TVT

Transcatheter Valve Therapy