The Quality and Safety of Sedation and Monitoring in Adults Undergoing Non-Operative Transesophageal Echocardiography

Abstract

Background:

Sedation is an essential component of the transesophageal echocardiography (TEE) procedure for patient comfort. Utilization and the clinical implications of cardiologist-supervised (CARD-Sed) versus anesthesiologist-supervised sedation (ANES-Sed) are unknown.

Methods:

We reviewed non-operative TEE records from a single academic center over a five-year time period and identified CARD-Sed and ANES-Sed cases. We evaluated the impact of patient comorbidities, cardiac abnormalities on transthoracic echocardiogram (TTE), and the indication for TEE on sedation practice. We analyzed the use of CARD-Sed versus ANES-Sed in light of institutional guidelines, the consistency in the documentation of pre-procedural risk stratification, and the incidence of cardiopulmonary events including hypotension, hypoxia and hypercarbia.

Results:

A total of 914 patients underwent TEE, with 475 (52%) receiving CARD-Sed and 439 (48%) ANES-Sed. The presence of obstructive sleep apnea (p=0.008), a body mass index greater than 45kg/m² (p<0.001), an ejection fraction of less than 30% (p<0.001) and pulmonary artery systolic pressure of more than 40 mm Hg (p=0.015) were all associated with the use of ANES-Sed. Of the 178 (19.5%) patients with at least one caution to non-anesthesiologist-supervised sedation by the institutional screening guideline, 65 (36.5%) underwent CARD-Sed. In the ANES-Sed group, where intraprocedural vital signs and medications were documented in all cases, hypotension (n = 91, 20.7%), vasoactive medication use (n=121, 27.6%), hypoxia (n= 35, 8.0%), and hypercarbia (n= 50, 11.4%) were noted.

Conclusions:

This single-center study revealed that 48% of the non-operative TEE utilized ANES-Sed over five years. Sedation-related hemodynamic changes and respiratory events were not infrequently encountered during ANES-Sed.

Introduction

Transesophageal echocardiography (TEE) is a commonly performed cardiovascular procedure that usually requires sedation.^1,2. However, TEE is an invasive and uncomfortable procedure and as such, at least moderate (“conscious”) sedation is provided for most TEE cases in the United States³. While cardiologist-administered moderate sedation has been studied and found to be safe in the context of other cardiology procedures (such as those performed in the electrophysiology lab), there is minimal existing data on the procedural safety regarding sedation for non-operative TEE.⁴ The American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists guidelines for performing a comprehensive TEE examination include maintaining clinical competency in adult TEE and an understanding of moderate sedation, including the actions, side effects, and risks of a sedative drug.¹ However, there are no specific recommendations on sedation strategy, including the decision on the operator (cardiologist-led vs. anesthesiology-led sedation), choice of sedative agents, and level of monitoring. One cohort study reported rare incidents of cardiac complications (0.08%) during TEE in the form of non-sustained ventricular tachycardia, atrial fibrillation, bradycardia, angina, and cardiac arrest.⁵ These complications could be attributed to both the sedation provided during TEE and the technical aspects of the TEE procedure itself.^6,7 Insertion and manipulation of a TEE probe poses a unique challenge as the probe location in the oropharynx can impede upper airway patency and cause hemodynamic consequences by triggering sympathovagal responses.^8,9 The specific aims of this single-site study were to describe the overall trend in cardiologist-supervised-sedation (CARD-Sed) and anesthesiologist-supervised-sedation (ANES-Sed) for non-operative and non-interventional TEE, examine factors associated with patients undergoing ANES-Sed versus CARD-Sed, and assess the impact of sedation strategy on cardiopulmonary complications such as hypotension, hypoxia, and hypercarbia.

Methods:

The study was undertaken at Harborview Medical Center, a tertiary Level I trauma and comprehensive stroke center affiliated with the University of Washington. TEEs performed from January 1, 2015, to December 31, 2019, were reviewed. Intraoperative TEEs and TEEs during structural heart interventions were excluded. A checklist-based document outlines the institution’s guidelines for non-anesthesiologist-led sedation; however, the decision to involve an anesthesiology service is ultimately left up to the TEE operator (Supplementary Table 1). TEE procedures in which sedation is provided by the cardiologist consists of a nurse responsible for administering and monitoring sedation, the attending cardiologist, and typically, a cardiology fellow. Cardiology attendings and cardiology fellows are certified in the administration of moderate sedation by the institution. The cardiology fellow or attending cardiologist performs a preprocedural sedation examination and reviews the case, including deciding upon the appropriate sedation strategy. An anesthesiology team includes an attending anesthesiologist and a certified nurse anesthetist or resident trainee. Selection of anesthesia monitoring technique, type of sedation regimen, hemodynamic targets, and use of vasoactive agents are not protocolized and left to the discretion of the supervising attending.

Patients undergoing TEE were identified from the TEE case logs. A review of the electronic medical record enabled the extraction of demographic information, medical comorbidities, indications for TEE, and pertinent data from a contemporaneous surface TTE. For ANES-Sed cases, the scanned copies of the paper anesthesia records in the electronic medical record were manually reviewed for anesthetic agents administered and the hemodynamic profile during the procedure. Similarly, we attempted to examine the pre-procedural assessment, procedural medications, and vital sign documentation for CARD-Sed. We assumed that clinicians followed an institution-standard pre-procedural sedation practice checklist documented ASA and Mallampati scores and practiced standardized intra-procedure monitoring and documentation of vital signs and administered anesthetic agents. All CARD-Sed employed moderate sedation, whereas ANES-Sed was listed as either general anesthesia (GA) or monitored anesthesia care (MAC). GA describes a depth of sedation defined by unresponsiveness to painful stimulus that can also include impaired ventilatory function, inability to protect one’s airway, and depressed cardiac function. MAC on the other hand describes a specific anesthesia service performed by a qualified anesthesia provider as opposed to a depth of sedation^10,11.

For the study, we documented all instances of use of CARD-Sed when anesthesiology services were recommended by the institutional guidelines. We defined hypotension as mean arterial pressure < 60 mmHg at any point in the procedure. Hypoxia was defined as any pulse oximetry reading of < 90% (either with or without supplemental oxygen), and hypercarbia was defined as end-tidal CO₂ (ETCO₂) ≥ 40 mmHg if ETCO₂ was monitored at any point in the procedure. Study data were managed using REDCap electronic data capture tools hosted at the University of Washington.¹²

Descriptive analysis summarized sample characteristics. After normality testing (Shapiro-Wilk test), continuous variables such as age, body mass index, and total procedure time were presented as a mean (standard deviation [SD]). Categorical variables were presented as counts (percentages). We categorized procedures as CARD-Sed or ANES-Sed and analyzed the proportion of cases performed by each service during the study period. Group comparisons were performed using the Student’s t-test for continuous variables and the chi-square test for categorical variables. A Bonferroni-corrected p-value of < 0.05 was considered statistically significant. Stata 15¹³ and R studio¹⁴ were used for statistical analysis. A subgroup analysis excluding pre-DCCV cases was performed to assess the impact of the indication for TEE on sedation strategy. We performed a multivariable logistic regression analysis of the indications for TEE (assessment of endocarditis, stroke work-up, valvular assessment) utilizing systolic heart failure, right ventricular dysfunction, pulmonary hypertension, age, body mass index, and gender as covariates.

Results

A total of 914 TEEs were performed during the five-year study period. Baseline characteristics of patients included in the study are shown in Table 1. As a group, the median age was 52 years (SD 15.6). CARD-Sed comprised 475 (52%) procedures and ANES-Sed 439 (48%) procedures. The trends in the proportion of ANES-Sed and CARD-Sed during the entire study period are shown in Figure 1. The number of overall TEE cases steadily increased, but the ratio of the CARD-Sed vs. ANES-Sed remained similar. The most common indication for TEE was endocarditis workup (n= 526, 57.5%) (Table 1).

Table 1.

Characteristics of patients undergoing transesophageal echocardiography by service providing sedation

	Cardiology	Anesthesiology	Odds Ratio Cardiology is reference group	p-value
	N=475	N=439	Empty Cell	Empty Cell
Age in years mean	51.4 (16.2)	52.9 (14.8)	1.01 [1.00;1.01]	0.137
Male sex	319 (67.2%)	329 (74.9%)	1.46 [1.10;1.95]	0.010
BMI mean (kg/m2)	26.9 (6.98)	29.8 (9.29)	1.05 [1.03;1.06]	<0.001
BMI > 45 kg/m2	9 (1.89%)	30 (6.83%)	3.75 [1.82;8.53]	<0.001
ASA Classification	2[2-3]	3[3-3]	11.3[8.03;16]	<0.001
Not documented	167 (35.2%)	7 (1.59%)		.
Mallampati Score
1	101 (21.3%)	61 (13.9%)	Ref.	Ref.
2	130 (27.4%)	186 (42.4%)	2.36 [1.60;3.50]	<0.001
3	65 (13.7%)	72 (16.4%)	1.83 [1.15;2.92]	0.010
4	11 (2.32%)	11 (2.51%)	1.65 [0.66;4.12]	0.279
Not documented	168 (35.4%)	109 (24.8%)	1.07 [0.72;1.60]	0.728
Obstructive sleep apnea	16 (3.37%)	32 (7.29%)	2.24 [1.23;4.26]	0.008
Chronic obstructive pulmonary disease	28 (5.89%)	30 (6.83%)	1.17 [0.68;2.01]	0.564
Liver cirrhosis	14 (2.95%)	23 (5.24%)	1.81 [0.93;3.67]	0.083
Diabetes mellitus	80 (16.8%)	92 (21.0%)	1.31 [0.94;1.83]	0.113
Indication for TEE				<0.001
Endocarditis	300 (63.2%)	226 (51.5%)	Ref.	Ref.
Other	29 (6.11%)	8 (1.82%)	0.4[0.12;1.15]	.
Pre DCCV	0 (0.00%)	137 (31.2%)	*	.
Source of thrombus	111 (23.4%)	49 (11.2%)	0.59[0.4;0.8]	.
Valve assessment	35 (7.37%)	19 (4.33%)	0.7[0.4;1.29]	.
LVEF <30%	10 (2.11%)	37 (8.43%)	4.23 [2.15;9.13]	<0.001
PA pressure greater than 40	39 (8.21%)	58 (13.2%)	1.70 [1.11;2.63]	0.015
Severe aortic stenosis	3 (0.63%)	6 (1.37%)	2.12 [0.54;10.7]	0.287

ASA = American Society of Anesthesiologists; BMI = body mass index; DCCV = direct current cardioversion; LVEF = left ventricular ejection fraction; PA = pulmonary artery; TEE = transesophageal echocardiography.

Continuous variables are expressed by mean (SD) and categorical variables by N (%).

^*Since Pre-DCCV had only anesthesiology service, odds ratio cannot be calculated

Figure 1.

The proportion of transesophageal echocardiograms with procedural sedation provided by cardiology versus anesthesiology was not significantly different year to year.

The comparison of the two groups by patient characteristics, indications for TEE, and pre-procedure TTE findings are presented in Table 1. Of note, all pre-cardioversion TEEs were ANES-Sed (n=137) because ANES-Sed is used for all TEE cardioversion at our institution. Patients with ANES-Sed were more likely to be identified as males (74.9% vs. 67.2%, p=0.01) and had a higher median body mass index (27.3 vs. 25.7 kg/m², p<0.001). Patients of ASA class III-V (n= 379) comprised the majority of cases (86.3%, p < 0.001) performed by anesthesiologists. A chart diagnosis of obstructive sleep apnea (p=0.008), a body mass index greater than 45kg/m² (p<0.001), an ejection fraction of less than 30% (p<0.001) and pulmonary artery systolic pressure of more than 40 mm Hg (p=0.015), both based on a preceding TTE, were all associated with ANES-Sed. In the subgroup analysis excluding pre-DCCV TEE, endocarditis as the indication for TEE was associated with ANES-Sed (OR=1.87, [1.16-3.11], p=0.0002).

The observed frequency of CARD-Sed when institutional guidelines recommended use of anesthesiology services is highlighted in Figure 2. Of the 178 cases with at least one reason for using anesthesiology services per hospital guidelines, 65 (36.5%) underwent CARD-Sed. Of these 65 patients, the most common cautions were pulmonary artery systolic pressure > 40mmHg on TTE (60.0%), ejection fraction < 30% (15.4%), obesity (13.8%), and severe aortic stenosis (10.8%).

Figure 2.

Of the 178 patients, 65 patients (36.5%) who met institutional guidelines for anesthesiology involvement underwent CARDS-TEE-Sed. Of these 65 patients the most common institutional caution met were pulmonary artery systolic pressure >40mmHg TTE (60.0%), an ejection fraction <30% (15.4%), obesity (13.8%), and severe aortic stenosis (10.8%). BMI = body mass index; EF = ejection fraction; PASP = pulmonary artery systolic pressure.

ASA classification was missing in seven (1.2%) ANES-Sed and 167 (35.2%) CARD-Sed. Mallampati score documentation was missing in 109 (24.8%) ANES-Sed and 168 (35.4%) CARD-Sed. Documentation of the sedation agents used and dosing was not always present in the medical record for the CARD-Sed cases. All the CARD-Sed cases with documentation utilized midazolam and fentanyl. However, 24% of the 475 CARD-Sed cases did not document the anesthetic agent used. Furthermore, for the 76% of CARD-Sed patients who were reported as receiving fentanyl and midazolam, only 62% had documentation of the doses administered. Documentation of vital signs could be found in only 4% of patients undergoing CARD-Sed. Thus, the exact incidence of hypotension and hypoxia is not known for these cases. Since CARD-Sed does not routinely include ETCO₂ monitoring, quantifying hypercarbia events was not possible.

Of the 435 ANES-Sed, MAC (n= 301, 68.6%) was more common than GA as listed in the type of anesthetic on the anesthesia records. The use of specific anesthetic agents was as follows: propofol, 72.7% (n= 319), midazolam, 44.6%, (n = 196), fentanyl, 26.2% (n= 115), ketamine, 25.7%, (n = 113). GA was administered with a native airway in 95 patients (68.8%). Forty-five patients (10.2%) were endotracheally intubated for the purpose of the TEE procedure. Most patients received supplemental oxygen via nasal cannula, 313 (71.3%), or via face mask, 74 (16.9%).

Hypotension (n = 91, 20.7%), hypoxia (n= 35, 8.0%) and hypercarbia (n= 50, 11.4%) were observed during ANES-Sed. The frequency of hypotension, the distribution of these frequencies by anesthetic agents used, and the use of vasoactive agents are shown in Supplemental Figure 1. The rates of hypotension were similar between MAC (18.5%) and GA (26.1%). Hypoxia (12.1% vs. 5.1%, p = 0.017) was more common in the GA group than the MAC group but the frequency of hypercarbia (10.1% vs. 8.1%, not significant) was comparable between MAC and GA without endotracheal intubation. This data is shown in Table 2.

Table 2.

Comparison between monitored anesthesia care and general anesthesia without endotracheal intubation in patients undergoing transesophageal echocardiography

Characteristics	Monitored anesthesia care (MAC)	General anesthesia without endotracheal intubation	Odds Ratio MAC as the reference group	p-value
Empty Cell	N=294	N=93	Empty Cell	Empty Cell
Midazolam	136 (46.3%)	42 (45.2%)	0.96 [0.60;1.53]	0.856
Propofol	221 (75.2%)	76 (81.7%)	1.47 [0.83;2.72]	0.193
Fentanyl	64 (21.8%)	23 (24.7%)	1.18 [0.67;2.03]	0.549
Ketamine	79 (26.9%)	26 (28.0%)	1.06 [0.62;1.77]	0.831
Hypotension (MAP < 60 mmHg)	55(18.5%)	24 (25.8%)	1.40[0.97,2.04]	0.077
Lowest mean arterial pressure
≥100-120 mmHg	13 (4.42%)	3 (3.23%)	Ref.	Ref.
≥80<100 mmHg	65 (22.1%)	25 (26.9%)	1.60 [0.46;7.81]	0.481
≥60-80 mmHg	162 (55.1%)	41 (44.1%)	1.06 [0.32;5.00]	0.934
< 60 mmHg	54 (18.4%)	24 (25.8%)	1.85 [0.53;9.07]	0.357
Lowest systolic blood pressure
≥100–120 mmHg	112 (42.7%)	33 (40.7%)	Ref.	Ref.
≥80<100 mmHg	114 (43.5%)	32 (39.5%)	0.95 [0.55;1.66]	0.865
≥60–80 mmHg	34 (13.0%)	15 (18.5%)	1.50 [0.71;3.07]	0.280
< 60 mmHg	2 (0.76%)	1 (1.23%)	1.79 [0.06;22.8]	0.678
Vasoactive medication use	71 (24.1%)	24 (25.8%)	1.10 [0.63;1.86]	0.741
Hypoxia	15 (5.10%)	12 (12.9%)	2.75 [1.21;6.15]	0.017
Hypercarbia	24 (8.16%)	9 (9.68%)	1.22 [0.51;2.65]	0.640

MAP = mean arterial pressure.

Discussion:

This single-center investigation illustrates current practice patterns in procedural sedation for TEE and identifies areas for further investigation. The decision between the two sedation strategies may be driven by the presence of baseline cardiopulmonary comorbidities, which is in line with the institutional guidelines, but the fact that a significant percentage of patients received CARD-Sed instead of recommended ANES-Sed suggests other factors may play a role. Documentation of sedation and complications in CARD-Sed was frequently missing compared to ANES-Sed. In the case of ANES-Sed, hemodynamic changes and respiratory events were not uncommon and were comparable between MAC versus GA.

In this cohort, assessment for endocarditis was the most common indication for TEE and was associated with the choice of ANES-Sed. Besides the TEE indication, the presence and severity of cardiopulmonary comorbidities such as pulmonary hypertension, heart failure with a reduced ejection fraction, obstructive sleep apnea, and obesity, were all associated with the utilization of ANES-Sed, as per institutional guidelines.

The institutional protocols pertain to all procedural sedation, including gastrointestinal and other procedures. Ultimately, the procedure operator determines the final choice of procedural sedation, and cardiologists often manage patients with elevated pulmonary pressures and low ejection fraction in many settings. On the other hand, there may be a perception that ANES-Sed is easier, with faster time to adequate sedation and early start of the procedures, though this must be balanced against the additional time it takes to arrange an extra team’s participation.

Within the ANES-Sed cases, cardiopulmonary complications occurred, including intraprocedural hypotension in 18.5% of cases. Within the ANES-Sed cases, there was no difference in the frequency of intraprocedural hypotension or hypercarbia between the GA group and MAC group, but hypoxia occurred at a higher frequency in the GA group. This indicates that within this high-risk group of patients (ASA 3-4), GA was tolerated similarly to MAC in terms of hemodynamics. We could not compare cardiopulmonary complications between ANES-Sed and CARD-Sed due to a lack of documentation of vital signs in the latter group.

Cardiac comorbidities that were more prevalent in the ANES-Sed group (ejection fraction <30% and pulmonary hypertension) may make these patients more susceptible to hemodynamic instability. In this context, use of propofol, a vasodilatory agent, may put patients at risk for hemodynamic compromise.⁷ Midazolam also causes changes in hemodynamics. The safety of dexmedetomidine, remifentanil, and ketamine have been described in the literature and may have a role in select patients undergoing TEE.^15–17

For other comparable procedures such as gastrointestinal endoscopy there have been 3 to 4-fold increases in anesthesia service involvement in routine procedures from 2003 to 2013 based on insurance claims.^18–23 The gastroenterology literature also demonstrates that anesthesiology administered propofol-based sedation is more costly than endoscopist administered sedation²⁴. Whether such a trend is seen in cardiology-performed TEE procedures on a national level and over a longer timeframe than this study would be worthy of investigation. For our cohort of exclusively nonoperative TEE, sedation for nearly 50% of TEE were performed by the anesthesiology service. We did not observe a year-on-year increase in the proportion of cases performed by anesthesiology, but it is possible that the inflection point where this became common practice pre-dated our study.

To our knowledge, this study is the first to examine differences between CARD-Sed vs. ANES-Sed. Since society guidelines don’t specify a preferred sedation strategy, many TEE operators will rely on institutional guidelines. Based on our single-center experience we recommend tailoring TEE sedation strategies to the individual patient and their risk factors. Patients at risk of respiratory or hemodynamic compromise due to sedation (pulmonary hypertension, low EF, elevated BMI, history of substance use) should have MAC at a minimum.

The strengths of this study are that clinicians individually reviewed a large sample of TEE cases and that this work represents a novel analysis of the area of procedural sedation. Limitations of the study include that data collected during TEE was not specific for research purposes and did not allow for direct comparison between ANES-Sed and CARD-Sed. Additionally, the documentation of intra-procedure vitals for ANES-Sed cases relied on scanned documents, with a paucity of auto-population of beat-to-beat hemodynamic data or area under the curve data. This introduced the risk of under or over-reporting hypotension severity. Furthermore, single-center data may limit the generalization of findings.

Conclusions:

This single-center cohort study explores current practice patterns in procedural sedation for TEE and identifies areas for future investigation. While the lack of intraprocedural vital sign documentation rendered comparison in outcomes between CARD-Sed and ANES-Sed difficult, ANES-Sed data showed that intra-procedure hypotension and other adverse occurrences were not infrequent but did not differ between GA and MAC strategies. Further research into the nationwide trends in TEE sedation, the incidence of adverse cardiopulmonary events with different sedation strategies, and cost analysis between sedation practices are warranted. Such data may inform development of specialty society standardized procedural sedation consensus guidelines.

Abbreviations:

TEE

Transesophageal echocardiography

TTE

Transthoracic echocardiography

CARD-Sed

Cardiologist-supervised-sedation

ANES-Sed

Anesthesiologist-supervised-sedation

ASA

American Society of Anesthesiologists

MAC

Monitored anesthesia care

GA

General Anesthesia

Data Availability:

The data used for this study is available in REDCap, an electronic data capture tool hosted at the University of Washington.