Sedated Echocardiograms Better Characterize Branch Pulmonary Arteries following Bidirectional Glenn Palliation with Minimal Risk of Adverse Events

Abstract

Background:

Patients post-bidirectional Glenn (BDG) operation are at risk of left and right pulmonary artery (LPA and RPA) hypoplasia. Transthoracic echocardiograms (TTE) in active children can miss essential elements of anatomy. Procedural sedation improves image quality but increases risk of adverse events. We hypothesized that echocardiograms performed with sedation in patients post-BDG would improve visualization of branch pulmonary arteries with minimal adverse events.

Methods:

Patients post-BDG between 2007–2016 were identified. Exclusion criteria were >12 months of age, absence of complete TTE before discharge, death before discharge, conversion to shunt physiology, and prolonged post-operative course >7 weeks. Of 254 post-BDG patients, 153 met inclusion/exclusion criteria. TTE reports were reviewed for visualization of LPA/RPA and hypoplasia of LPA/RPA. Blinded assessment of image quality was performed (scale of 1[poor] to 5[excellent]). Pertinent clinical data were recorded. Pearson’s chi squared and Wilcoxan Rank Sum tests used for statistical analysis.

Results:

The median age at surgery and hospital stay were 4.8 months and 10 days. Twenty-three patients underwent sedated TTE (15%). Sedated TTE significantly improved visualization of the RPA (100% vs 82%, p=0.029) and LPA, though this did not reach statistical significance (100% vs 91%, p=0.129). Sedated TTEs had significantly better image quality (median of 4 vs 3, p<0.001). There were no serious adverse events due to sedation.

Discussion:

Sedated TTE early post-BDG is safe, improves visualization of the RPA and LPA, and improves overall image quality. Routine sedated TTE in these patients should be considered. Implications for long-term outcome need to be further analyzed.

BACKGROUND

Transthoracic echocardiograms (TTE) are important diagnostic tools for assessing pediatric heart structure and function, especially in congenital heart disease patients. TTEs of children with congenital heart disease have been shown to be at risk for error due to intrinsic factors, such as complexity of heart disease, and situational factors, including performance of a study on a very active infant or child[1,2]. Changing physiology and low patient weight, in multivariate analysis, are also risk factors[1–3]. In addition, TTEs visualize extracardiac structures, like branch pulmonary arteries and cavopulmonary anastamoses, less frequently than intracardiac structures[4]. TTEs that are deemed technically difficult or incomplete result in poor image quality, repeat echocardiograms, and missed information. Controlling for study conditions by using procedural sedation in young, active infants and children, results in a decrease in diagnostic errors and improvement in image quality[5]. There is a risk of adverse events in this population, but in the presence of trained medical professionals, such as sedation nurses under supervision of a pediatric anesthesiologist, these events are minimized and sedation is generally safe[6].

Among the most complex congenital heart disease lesions are patients with single ventricle physiology, such as hypoplastic left heart syndrome (HLHS), who undergo a series of palliative operations, including a superior cavopulmonary anastomosis, also referred to as a bidirectional Glenn (BDG) operation. Patients who undergo BDG are at risk for hypoplastic branch pulmonary arteries and passive blood flow can make visualization of the left and right pulmonary artery (LPA and RPA) difficult. As patients progress from second to third stage palliation, accurate assessment of branch pulmonary artery growth is paramount, as adequately sized branch pulmonary arteries are necessary for optimal outcomes after Fontan palliation[7–9]. Given the difficulty in visualizing branch pulmonary arteries in these patients, our institution created a protocol in late 2011 whereby all patients after BDG would undergo sedated TTE. The objective of this single institution study was to evaluate whether sedated TTE improved visualization and characterization of branch pulmonary arteries and to identify the risks and benefits associated with sedation in patients after BDG. We hypothesized that echocardiograms performed with sedation in patients after BDG would improve image quality, providing better characterization of branch pulmonary arteries with minimal risk of adverse events.

MATERIALS AND METHODS

Enrollment

This retrospective study was approved by the Institutional Review Board. Our cohort was identified through the cardiac surgery database. Between 2007–2016, 254 patients with single ventricle physiology, defined as hypoplastic left heart syndrome (HLHS), tricuspid atresia (TA), double inlet left ventricle (DILV), and single ventricle not otherwise specified (SV-NOS), who underwent BDG or bilateral BDG procedures were retrospectively reviewed. Inclusion criteria were patients who underwent BDG and had a complete TTE prior to discharge, either with or without sedation. Patients were excluded from the study if no complete TTE was performed prior to discharge (n=31), they were greater than 12 months of age at BDG (n=20), they died in the hospital before discharge (n=10), they required conversion to Blalock-Taussig shunt physiology (n=19), or they had a prolonged post-operative course greater than 7 weeks (n=21). Of the 254, 153 patients met inclusion/exclusion criteria.

Data Collection

Basic demographic data, including native single ventricle anatomy, age at operation, age at echocardiogram, and length of hospitalization, were collected. Echocardiogram reports of TTEs with and without sedation, obtained closest to the date of discharge and designated as the complete discharge TTE, were examined for: 1) Visualization of the LPA and RPA. Visualization was defined as visible either by 2-dimensional imaging or color flow mapping and was based on the assessment of the individual reader. 2) Measurement of the LPA and RPA if they were visualized. Measurements are, per our labs protocol, performed on 2-dimensional images only if adequate borders can be delineated. 3) Assessment as stenotic or hypoplastic if the LPA and RPA were visualized. Stenotic/hypoplastic was determined subjectively by the reader based on a combination of color flow mapping, 2-dimensional imaging, and 2-dimensional measurements obtained on the echocardiogram. Pertinent clinical data, including need for intervention as a result of the discharge TTE, were recorded. Post-operative catheterizations performed in addition to the standard, pre-Fontan catheterization were assessed. Results from the standard pre-Fontan catheterization were also analyzed.

One investigator (JS) reviewed every echocardiogram blinded to sedation status and graded the image quality on a scale of 1 (poor) to 5 (excellent). Image quality assessment included overall image quality with a particular focus on the quality of branch pulmonary artery imaging. A second investigator (DP), also blinded to sedation status, reviewed 50 studies to evaluate reproducibility of the image quality assessment.

Sedation data were examined for type and adequacy of sedation and adverse events. Anesthesia records, completed by the procedural sedation team, reported the name of the medication used, adequacy of the medication as defined by the ability to obtain the complete echocardiogram, doses of medication and number of administrations, as well as any adverse events as a result of the medication used. Sedated echocardiograms were performed on the cardiac floor, in a sedation room, or in the cardiac ICU by either our sedation team, cardiac ICU team, pediatric anesthesia, or pediatric cardiac anesthesia. Medications chosen for sedation were at the discretion of the physician, nurse practitioner, or certified registered nurse anesthetist. The most commonly used medications were pentobarbital (n=9) and propofol (n=4), with a few patients receiving some combination of narcotics (n=3), benzodiazepines (n=2), general anesthesia with inhaled anesthetics (n=2), ketamine (n=1), and etomidate (n=1).

Statistical Analysis

Demographic data are reported as mean and standard deviation or median and interquartile range (IQR). Categorical variables were assessed using a Pearson’s chi-squared test or a chi-squared test for trend. Continuous variables were assessed using a Wilcoxan Rank Sum test. Image quality comparison between those with sedated and non-sedated echocardiograms was compared using a Wilcoxan Rank Sum test. Reproducibility of the image quality grading was assessed using an intraclass correlation coefficient and weighted kappa. Analyses were performed with IBM SPSS statistics, version 22.0 (Armonk, NY: IBM Corp). Study data were collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools hosted at Vanderbilt[10].

RESULTS

Demographic Data and TTE Comparison Data

Sixty-three percent of the cohort was male. The most frequent operation performed was the bidirectional Glenn procedure (89%) vs bilateral bidirectional Glenn procedure (11%). The median age at surgery was 4.8 months. The most common underlying diagnosis was HLHS, and the median hospital stay was 10 days (Table 1).

Table 1.

Demographics

	Median (Interquartile Range)
Age at bidirectional Glenn (months)	4.8 (4.1, 6.6)
Age at echocardiograms (months)	5.2 (4.4, 6.9)
Number of echocardiograms performed during hospitalization	1 (1, 2)
Male gender (%)	63%
Hospital length of stay (days)	10 (7, 16)
Total sedated echocardiograms (%)	14% (n=21)
Underlying cardiac anatomy (%)
Hypoplastic left heart syndrome	47%
Single ventricle not otherwise specified	31%
Double inlet left ventricle	12%
Tricuspid atresia	10%
Type of operation (% bidirectional Glenn)	89%

Twenty-three patients underwent sedated TTE (15%). The majority (n=17, 74%) underwent sedated TTE as part of a new protocol instituted at the end of 2011 recommending that all patients after BDG undergo sedated echocardiograms prior to discharge. The remaining sedated studies were ordered to answer specific clinical questions not adequately assessed on a prior unsedated echocardiogram (branch pulmonary arteries n=3, 13%; RV pressure n=2, 9%) or for significant agitation on the prior study (n=1, 4%). There was no difference in age at BDG or age at TTE between patients undergoing sedated TTE and those who underwent unsedated TTE (Table 2). Performance of sedated echocardiograms did not increase length of stay. Patients undergoing sedated TTE underwent fewer echocardiograms during hospitalization but this difference did not reach statistical significance (median of 1 IQR(1,2) vs 2 IQR(1,2), p=0.102). There was no difference in age at surgery, gender, underlying cardiac anatomy, or type of operation (BDG vs bilateral BDG) between those who underwent TTE with sedation vs those who underwent TTE without sedation. Sedated echocardiograms were more likely to have been performed recently (average year of performance 2013 vs 2011, p<0.001).

Table 2.

Comparison of Sedated vs Unsedated Transthoracic Echocardiography (TTE)

	Sedated TTE Median (interquartile range) N=23	Unsedated TTE Median (interquartile range) N=130	p-value
Age at bidirectional Glenn (months)	5.0 (4.0, 7.6)	4.8 (4.1, 6.5)	0.530
Age at echocardiogram (months)	5.4 (4.5, 7.8)	5.2 (4.3, 6.7)	0.248
Number of echocardiograms performed during hospitalization	1 (1, 2)	2 (1, 2)	0.107
Male gender (%)	61 %	64%	0.785
Hospital length of stay (days)	9 (7, 13)	10 (7, 16)	0.923
Underlying cardiac anatomy (%)			0.392
Hypoplastic left heart syndrome	35%	49%
Single ventricle not otherwise specified	30%	31%
Double inlet left ventricle	17%	12%
Tricuspid atresia	17%	9%
Type of operation (% bidirectional Glenn)	87%	89%	0.749

TTE – transthoracic echocardiogram

Sedated TTE was more likely to visualize the RPA (100% vs 82%, p=0.029) (Figure1, Table 3). The LPA was more likely to be seen with sedated TTE, though this did not reach statistical significance (100% vs 91%, p=0.129). There was no difference in incidence of LPA or RPA stenosis between sedated and unsedated echocardiograms. The RPA and LPA were able to be measured more frequently on sedated echocardiograms, though again this did not reach statistical significance for the LPA (RPA: 87% vs 59%, p=0.011; LPA: 91% vs 73%, p=0.060). There was no difference in the maximum measured size of the branch pulmonary arteries on sedated vs unsedated TTEs. The image quality of the sedated echocardiograms was significantly better compared to unsedated echocardiograms, with a median quality score of 4 IQR(4,5) vs 3 IQR(3,4), p<0.001. The intraclass correlation coefficient was 0.81 (p=0.001) and the weighted kappa was 0.50 (p<0.001), suggesting strong and moderate agreement for assessment of image quality (the second reader’s quality scores were also significantly better for sedated vs unsedated echocardiograms with a median of 5 vs 3, p=0.001).

Figure 1.

Patients undergoing sedated transthoracic echocardiograpy (TTE) had improved visualization and measurement of the right pulmonary arteries (RPA) after bidirectional Glenn. There was also improved visualization and measurement of the left pulmonary artery (LPA) though this improvement did not reach statistical significance.

Table 3.

Echocardiographic results for entire cohort and sedated and unsedated patients

	Entire Cohort N=153	Sedated TTE N=23	Unsedated TTE N=130	p-value*
Image quality	3 (3, 4)	4 (4, 5)	3 (3, 4)	<0.001
Right pulmonary artery (RPA) visualized (%)	85%	100%	82%	0.029
RPA measured (%)	63%	87%	59%	0.011
RPA hypoplastic (%)	14%	22%	12%	0.184
Left pulmonary artery (LPA) visualized (%)	92%	100%	91%	0.129
LPA measured (%)	76%	91%	73%	0.060
LPA hypoplastic (%)	41%	48%	40%	0.728

^*p-value for comparison between sedated and unsedated groups

TTE – transthoracic echocardiogram

Comparison with post Glenn intervention

A total of 61 patients underwent 88 catheterizations after BDG, excluding the pre-Fontan catheterization to assess for Fontan candidacy. Twenty-seven patients were noted to have pulmonary artery hypoplasia on their first post-BDG catheterization (20 with LPA hypoplasia, 3 with RPA hypoplasia, and 4 with bilateral hypoplasia; excluding pre-Fontan studies). Including all catheterizations (and specifically, including the cardiac catheterization for pre-Fontan assessment), 50 patients had pulmonary artery hypoplasia (42 LPA, 3 RPA, 5 bilateral). Patients undergoing sedated and unsedated TTE had the same number of cardiac catheterizations. There was no significant difference in rates of pulmonary artery stenosis diagnosed by catheterization between groups. Not surprisingly, patients with stenosis on echocardiogram were more likely to require intervention during cardiac catheterization than those without stenosis (19% vs 6%, p=0.017). There was no difference in time to catheterization between patients who did and did not undergo sedated echocardiograms. A total of 109 patients have progressed to Fontan.

Sedation Data

There were no serious adverse events due to procedural sedation. One patient required a transient increase in oxygen need from 0.5L to 1L via nasal cannula following sedation. One patient was converted from laryngeal mask airway to endotracheal tube during the procedure due to difficulty maintaining an airway, but this did not affect the recovery from sedation or subsequent clinical status.

DISCUSSION

Our data demonstrate that the use of sedation with TTE is safe in the appropriate clinical setting. Sedated TTE improves visualization of the branch pulmonary arteries, particularly the RPA, in patients after BDG palliation and leads to overall improvement in TTE quality scores.

This is the first study, to our knowledge, to focus on the effect of sedation on image quality in patients after BDG. Prior work in patients after Fontan has demonstrated the difficulty in assessing pulmonary arteries in unsedated patients with passive pulmonary blood flow[11]. The safety of multiple different sedation protocols has been demonstrated in pediatric patients undergoing echocardiography[12–16]. Stern, et al. demonstrated that patients from 6–24 months of age were more likely to require an unplanned sedated TTE after an unsedated TTE, presumably due to poor image quality related to agitation[17]. Given the fact that most patients after BDG fall within or just below that age group, it is not surprising that our results demonstrated improved visualization of branch PAs with sedation. The same group demonstrated improved image quality and decreased rate of diagnostic errors in children <3 years of age undergoing sedated echocardiograms[5]. Parthiban, et al., performed a quality improvement initiative that included increased use of sedation in patients with various cardiovascular diagnoses. They also demonstrated improvement in image quality after BDG, though it was a smaller population of patients after BDG and they evaluated multiple interventions[4]. Interestingly, a study of single ventricle patients prior to BDG operation did not demonstrate improved visualization of branch pulmonary arteries with sedated TTE[18]. However, a study by Krupickova, et al. comparing echocardiography and cardiovascular magnetic resonance (CMR) imaging for great vessel measurements prior to BDG reported lower sensitivity and specificity of echocardiography compared to CMR for assessment of branch pulmonary artery stenosis and commented that these echocardiographic measurements were even more unsatisfactory in unsedated patients[19]. In their study, as in ours, echocardiographic assessment of RPA stenosis was worse in non-sedated patients.

The results of our study offer data for heart centers targeting quality improvement in their standard of practice with echocardiography. The addition of sedation in carefully selected populations could improve the quality of the TTE. While this study suggests that sedated TTEs are safe, this is in the setting of robust protocols assessing which practitioners are most appropriate to perform sedation and specifying monitoring parameters and patient recovery. While full sedation protocol details are beyond the scope of this manuscript, the overall health of the patient and the risk of sedation leading to a clinical deterioration must be assessed on an individual basis.

While our cohort had only a trend to fewer total echocardiograms performed, sedated TTE answered critical questions about structural lesions perhaps in a timelier manner. There were no differences in required number of cardiac catheterizations between groups; because of the relatively small number of patients undergoing sedated TTE, and the fact that these were performed more recently so that many patients had not yet undergone Fontan, it is unclear if those undergoing sedated echocardiograms have better long-term outcomes. Moreover, this question would be difficult to answer due to multiple confounders, including underlying anatomy and function, surgical variation, and genetic modifiers. However, given the importance of branch pulmonary artery size to long-term outcome in single ventricles, we feel that consideration of sedated TTE is warranted to better visualize branch pulmonary arteries and potentially identify hypoplasia earlier.

Procedural sedation has several considerations to ensure safety, including availability of designated personnel with the appropriate clinical training to administer the medication, monitor the patient during the echocardiogram, and observe recovery[6,20,21]. The specific choice of medication was not a primary focus of our study results, but moderate sedation is likely sufficient to provide adequate sedation, as was seen in our study. While our data focuses on inpatient use of procedural sedation and analyzes complete TTE after operation, further studies should evaluate safety and improvement in quality regarding outpatient TTEs in patients after BDG.

One of the possible concerns with increased use of inpatient sedation would be the potential to increase length of stay, either due to the requirement for patients to be NPO prior to the study or the need for recovery from sedation. We found no difference in length of stay. It must be noted, however, that outpatient sedated echocardiograms may decrease the productivity of outpatient echocardiographic laboratories and would need to be considered if translating these results to the outpatient setting[22].

Our study is limited by the retrospective design. Specifically, we cannot account for variability in sonographer imaging skill and data reporting. Despite our protocol for obtaining sedated echocardiograms in BDG patients, the majority of TTEs were unsedated, suggesting that selection for sedated echocardiograms was based on the clinical provider, which could lead to heterogeneity between groups. In addition, the majority of sedated echocardiograms were obtained recently, which could lead to bias. Finally, there could be inherent bias introduced by our exclusion criteria, specifically the 35 patients who did not undergo complete TTE prior to discharge.

CONCLUSION

The use of procedural sedation in patients after BDG is safe and improves visualization of branch pulmonary arteries. Routine sedated TTE should be considered in patients after BDG. Implications of earlier diagnosis of branch pulmonary artery hypoplasia in this cohort need to further analyzed.