Background
The initial experience with the miniaturized multiplane micro-transesophageal echocardiographic probe (MTEE) reported high-quality diagnostic imaging in small infants. The aim of this study was to compare the diagnostic accuracy and image quality of the intraoperative MTEE with the pediatric multiplane transesophageal echocardiographic probe (PTEE).
Methods
Infants weighing <5 kg who underwent intraoperative transesophageal echocardiography were identified. Studies using the MTEE were matched 1:1 with those using the PTEE by cardiac diagnosis. The postoperative transesophageal echocardiograms, obtained using either probe, were reviewed for the presence of 11 cardiac abnormalities. Postoperative transesophageal echocardiograms were compared with predischarge transthoracic echocardiograms to assess accuracy. Using receiver operating characteristic curves, the areas under the curve for the MTEE and PTEE were compared. Two pediatric cardiologists scored six image quality metrics on equal numbers of studies obtained with the MTEE and the PTEE. Composite scores from both reviewers were used to compare image quality.
Results
The study included 110 transesophageal echocardiograms per probe type. The mean weight for the MTEE was lower than for the PTEE (3.15 ± 0.58 vs 3.70 ± 0.52 kg, P < .001). There was no significant difference in the diagnostic accuracy of the MTEE and PTEE using receiver operating characteristic curves. The numbers of residual anatomic lesions missed by the MTEE and PTEE were similar (19 vs 22, respectively). The composite image quality score was worse for the MTEE compared with the PTEE (81% vs 92%, respectively, P < .0001).
Conclusions
Although the image quality of the MTEE is inferior compared with the PTEE, its diagnostic accuracy in infants weighing <5 kg is comparable.
Attention ASE Members:
ASE has gone green! Visit www.aseuniversity.org to earn free CME through an online activity related to this article. Certificates are available for immediate access upon successful completion of the activity. Non-members will need to join ASE to access this great member benefit!
Methods
Approval for the study was obtained through the Children’s Healthcare of Atlanta Institutional Review Board. All infants weighing <5 kg who underwent intraoperative TEE at Children’s Healthcare of Atlanta between October 2009 and September 2013 were identified through a review of our institutional electronic echocardiography database. Data collected from the medical records included age and weight at the time of surgery, cardiac diagnosis, type of operation performed, cardiologist performing TEE, and any reported complications related to the transesophageal probe. Patients were divided into two groups on the basis of whether the MTEE or a PTEE was used during intraoperative TEE. The studies performed using the MTEE were matched 1:1 with those performed using a PTEE by cardiac diagnosis. The postoperative transesophageal echocardiograms obtained using either probe were reviewed by a single pediatric cardiologist (B.J.T.) for the presence of 11 defects: residual ventricular septal defects, residual atrial septal defects, tricuspid regurgitation, mitral regurgitation, pulmonary regurgitation, aortic regurgitation, subaortic stenosis, valvar aortic stenosis, subpulmonary stenosis, valvar pulmonary stenosis, and pulmonary venous stenosis. The predischarge complete transthoracic echocardiogram for each study patient was then reviewed by the same pediatric cardiologist (B.J.T.) and compared with the results of postoperative transesophageal echocardiography. To evaluate the influence of the operator’s experience, the cardiologists performing the studies were divided into two groups: (1) those with >5 years’ experience after graduation from cardiology fellowship and (2) those with <5 years’ experience.
Two experienced cardiologists (R.S. and T.C.S.), independently reviewed an equal number of intraoperative transesophageal echocardiogram obtained using the MTEE and a PTEE to assess the image quality of each probe. The studies used to assess image quality were selected from the larger group of studies used to assess diagnostic accuracy. A study-specific image-grading scale was created with the goal of assessing each study for the quality of imaging. The scores were graded on a three-point scale: 1 = poor quality and nondiagnostic, 2 = not high quality but diagnostic, and 3 = high quality and diagnostic. Both cardiologists were asked to independently grade the image quality of six variables: two-dimensional (2D) imaging of the cardiac base and the four-chamber view, color Doppler imaging for mitral and tricuspid regurgitation, and color Doppler imaging of the atrial septum and left ventricular outflow tract (LVOT). Studies deemed to have inadequate imaging of particular variable(s) were excluded from grading by both cardiologists. Composite scores from both reviewers were used to compare quality of the two probes.
Data Analysis
Data were analyzed using SAS version 9.3 (SAS Institute, Inc, Cary, NC) and receiver operating characteristic analysis by MedCalc version 13.1.0 (MedCalc Software, Ostend, Belgium). The results of postoperative TEE and predischarge transthoracic echocardiography for each patient were compared for accuracy in identifying each of the 11 study-specific residual defects. Statistical comparisons between the two groups were performed using the Student t test for continuous variables and χ 2 tests for categorical variables. The accuracy of the two probes was assessed using sensitivity, specificity, positive predictive value, and negative predictive value. Using receiver operating characteristic curves, the areas under the curve for the MTEE and PTEE were calculated and compared. All test results were considered significant at P < .05.
To assess image quality for each probe, the grades from both cardiologists were added to obtain a combined actual score for each study using the six study variables. A probe-specific total score for each variable was calculated by adding the combined scores from each study in the MTEE and PTEE groups. The ratio of combined actual score to maximum possible score was calculated for each of the six variables for the MTEE and PTEE, and these were compared using a population proportion Z test. Agreement between raters was assessed in two ways: (1) using the intraclass correlation coefficient and (2) using a weighted κ coefficient. The intraclass correlation coefficient ranges from 0 to 1, with a score close to 0 indicating high variability among raters’ scores and a score close to 1 indicating little to no variability among raters’ scores. Cohen’s weighted κ is a measure of agreement between raters that takes into account agreement due to chance and the ordering of the data. It has a maximum of 1 when agreement between the two scorers is perfect, whereas a value of 0 indicates no agreement better than chance alone. In this study, the following guidelines were used to interpret measures of agreement: κ = 0.01 to 0.20, slight agreement; κ = 0.21 to 0.40, fair agreement; κ = 0.41 to 0.60, moderate agreement; κ = 0.61 to 0.80, substantial agreement; and κ = 0.81 to 1.00, almost perfect agreement.
Methods
Approval for the study was obtained through the Children’s Healthcare of Atlanta Institutional Review Board. All infants weighing <5 kg who underwent intraoperative TEE at Children’s Healthcare of Atlanta between October 2009 and September 2013 were identified through a review of our institutional electronic echocardiography database. Data collected from the medical records included age and weight at the time of surgery, cardiac diagnosis, type of operation performed, cardiologist performing TEE, and any reported complications related to the transesophageal probe. Patients were divided into two groups on the basis of whether the MTEE or a PTEE was used during intraoperative TEE. The studies performed using the MTEE were matched 1:1 with those performed using a PTEE by cardiac diagnosis. The postoperative transesophageal echocardiograms obtained using either probe were reviewed by a single pediatric cardiologist (B.J.T.) for the presence of 11 defects: residual ventricular septal defects, residual atrial septal defects, tricuspid regurgitation, mitral regurgitation, pulmonary regurgitation, aortic regurgitation, subaortic stenosis, valvar aortic stenosis, subpulmonary stenosis, valvar pulmonary stenosis, and pulmonary venous stenosis. The predischarge complete transthoracic echocardiogram for each study patient was then reviewed by the same pediatric cardiologist (B.J.T.) and compared with the results of postoperative transesophageal echocardiography. To evaluate the influence of the operator’s experience, the cardiologists performing the studies were divided into two groups: (1) those with >5 years’ experience after graduation from cardiology fellowship and (2) those with <5 years’ experience.
Two experienced cardiologists (R.S. and T.C.S.), independently reviewed an equal number of intraoperative transesophageal echocardiogram obtained using the MTEE and a PTEE to assess the image quality of each probe. The studies used to assess image quality were selected from the larger group of studies used to assess diagnostic accuracy. A study-specific image-grading scale was created with the goal of assessing each study for the quality of imaging. The scores were graded on a three-point scale: 1 = poor quality and nondiagnostic, 2 = not high quality but diagnostic, and 3 = high quality and diagnostic. Both cardiologists were asked to independently grade the image quality of six variables: two-dimensional (2D) imaging of the cardiac base and the four-chamber view, color Doppler imaging for mitral and tricuspid regurgitation, and color Doppler imaging of the atrial septum and left ventricular outflow tract (LVOT). Studies deemed to have inadequate imaging of particular variable(s) were excluded from grading by both cardiologists. Composite scores from both reviewers were used to compare quality of the two probes.
Data Analysis
Data were analyzed using SAS version 9.3 (SAS Institute, Inc, Cary, NC) and receiver operating characteristic analysis by MedCalc version 13.1.0 (MedCalc Software, Ostend, Belgium). The results of postoperative TEE and predischarge transthoracic echocardiography for each patient were compared for accuracy in identifying each of the 11 study-specific residual defects. Statistical comparisons between the two groups were performed using the Student t test for continuous variables and χ 2 tests for categorical variables. The accuracy of the two probes was assessed using sensitivity, specificity, positive predictive value, and negative predictive value. Using receiver operating characteristic curves, the areas under the curve for the MTEE and PTEE were calculated and compared. All test results were considered significant at P < .05.
To assess image quality for each probe, the grades from both cardiologists were added to obtain a combined actual score for each study using the six study variables. A probe-specific total score for each variable was calculated by adding the combined scores from each study in the MTEE and PTEE groups. The ratio of combined actual score to maximum possible score was calculated for each of the six variables for the MTEE and PTEE, and these were compared using a population proportion Z test. Agreement between raters was assessed in two ways: (1) using the intraclass correlation coefficient and (2) using a weighted κ coefficient. The intraclass correlation coefficient ranges from 0 to 1, with a score close to 0 indicating high variability among raters’ scores and a score close to 1 indicating little to no variability among raters’ scores. Cohen’s weighted κ is a measure of agreement between raters that takes into account agreement due to chance and the ordering of the data. It has a maximum of 1 when agreement between the two scorers is perfect, whereas a value of 0 indicates no agreement better than chance alone. In this study, the following guidelines were used to interpret measures of agreement: κ = 0.01 to 0.20, slight agreement; κ = 0.21 to 0.40, fair agreement; κ = 0.41 to 0.60, moderate agreement; κ = 0.61 to 0.80, substantial agreement; and κ = 0.81 to 1.00, almost perfect agreement.
Results
The study included 220 transesophageal echocardiograms, 110 each for the MTEE and PTEE. The groups were matched 1:1 for the primary cardiac diagnoses listed in Table 1 . MTEE patients weighed less (3.15 ± 0.58 vs 3.70 ± 0.52 kg, P < .001) and were younger in age (24 ± 38 vs 52 ± 58 days, P < .0001) than PTEE patients. There were no reported complications, such as unintentional extubation, esophageal perforation, or significant hemodynamic compromise, with either probe placement or positioning cited in the intraoperative or postoperative records. However, transient hemodynamic impairment necessitating adjustment of probe position or early withdrawal before the completion of a postoperative transesophageal echocardiographic evaluation were under the control of the attending anesthesiologist in each case, and information regarding such instances was not available in the electronic medical records.
| Cardiac diagnosis | Number of patients |
|---|---|
| Ventricular septal defect | 19 |
| Hypoplastic left heart syndrome | 16 |
| Transposition of the great arteries, intact ventricular septum | 15 |
| Atrioventricular septal defect | 11 |
| Transposition of the great arteries, ventricular septal defect | 9 |
| Tetralogy of Fallot | 9 |
| Total anomalous pulmonary venous return | 7 |
| Interrupted aortic arch | 7 |
| Truncus arteriosus | 4 |
| Other single ventricles | 3 |
| Double-outlet right ventricle | 3 |
| Aorta-pulmonary window | 2 |
| Anomalous left coronary artery from pulmonary artery | 1 |
| Cor triatriatum | 1 |
| Pulmonary atresia, intact ventricular septum | 1 |
| Pulmonary stenosis | 1 |
| Tetralogy of Fallot, absent pulmonary valve | 1 |
| Total | 110 |
Review of transesophageal echocardiograms performed using the MTEE and PTEE revealed similar numbers of missed defects. A total of 19 residual anatomic lesions were missed by the MTEE, including two trivial atrial septal defects, 14 small ventricular septal defect, two aortic stenosis, and one subpulmonary stenosis. With the PTEE, a total of 22 lesions were missed, including two trivial atrial septal defect, 19 small ventricular septal defect, and 1 pulmonary stenosis. None of these patients required reoperation for the residual lesions. There was no difference in the diagnostic accuracy of the two probes. The sensitivity, specificity, positive predictive value, and negative predictive value of each probe’s ability to detect each of the study-specific residual lesions are listed in Table 2 . The comparison of area under the curve for each residual defect betweenthe MTEE and PTEE demonstrated no significant differences between the probes.
| Diagnosis | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | AUC | P | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| MTEE | PTEE | MTEE | PTEE | MTEE | PTEE | MTEE | PTEE | MTEE | PTEE | ||
| Residual ASD | 78 | 79 | 86 | 97 | 87 | 94 | 77 | 89 | 0.82 | 0.88 | .22 |
| Residual VSD | 58 | 56 | 100 | 90 | 100 | 77 | 85 | 76 | 0.79 | 0.73 | .31 |
| Tricuspid regurgitation | 86 | 77 | 71 | 62 | 98 | 94 | 26 | 27 | 0.79 | 0.70 | .45 |
| Mitral regurgitation | 72 | 71 | 87 | 79 | 87 | 83 | 71 | 65 | 0.79 | 0.75 | .47 |
| Pulmonary regurgitation | 74 | 68 | 86 | 73 | 95 | 89 | 50 | 42 | 0.80 | 0.70 | .19 |
| Aortic regurgitation | 38 | 47 | 89 | 85 | 76 | 60 | 61 | 77 | 0.64 | 0.66 | .72 |
| Aortic stenosis | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 1.0 | 1.0 | 1.00 |
| Subaortic stenosis | 100 | 100 | 100 | 99 | 100 | 50 | 100 | 100 | 1.0 | 0.99 | 1.00 |
| Pulmonary stenosis | 78 | 83 | 100 | 99 | 100 | 83 | 98 | 99 | 0.89 | 0.91 | 1.00 |
| Subpulmonary stenosis | 80 | 100 | 100 | 100 | 100 | 100 | 99 | 100 | 0.90 | 1.0 | 1.00 |
| Pulmonary Veins | 100 | 100 | 99 | 100 | 67 | 100 | 100 | 100 | 0.99 | 1.0 | 1.00 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
