Risk of dysphagia after transesophageal echocardiography during cardiac operations John A. Rousou, Dennis A. Tighe, Jane L. Garb, Howard Krasner, Richard M. Engelman, Joseph E. Flack, III and David W. Deaton Ann Thorac Surg 2000;69:486-489
The online version of this article, along with updated information and services, is located on the World Wide Web at: http://ats.ctsnetjournals.org/cgi/content/full/69/2/486
The Annals of Thoracic Surgery is the official journal of The Society of Thoracic Surgeons and the Southern Thoracic Surgical Association. Copyright © 2000 by The Society of Thoracic Surgeons. Print ISSN: 0003-4975; eISSN: 1552-6259.
Downloaded from ats.ctsnetjournals.org by on June 4, 2013
Risk of Dysphagia After Transesophageal Echocardiography During Cardiac Operations John A. Rousou, MD, Dennis A. Tighe, MD, Jane L. Garb, MS, Howard Krasner, MD, Richard M. Engelman, MD, Joseph E. Flack III, MD, and David W. Deaton, MD Divisions of Cardiac Surgery, and Cardiology, and Department of Anesthesiology, Baystate Medical Center, Springfield, Massachusetts
Background. Dysphagia can be a significant complication following cardiac operations. This study evaluates its incidence and relationship to intraoperative transesophageal echocardiography (TEE) for specific indications versus known factors such as stroke or prolonged intubation. Methods. Records of 838 consecutive cardiac surgical patients were reviewed, and categorized into those who received TEE for specific indications versus those who did not (nonTEE). Dysphagia was recorded when symptoms were confirmed by barium cineradiography. Multiple logistic regression identified significant factors causing dysphagia.
Results. TEE was significantly related to the development of postoperative dysphagia by multiple logistic regression (p < 0.001). After controlling for other significant factors (stroke, left ventricular ejection fraction, intubation time, duration of operation), the odds of dysphagia for TEE patients was 7.8 times greater than for nonTEE patients. Conclusions. TEE may be an independent risk factor for dysphagia following cardiac operations.
T
near continuous normothermic (37°C) retrograde blood cardioplegia for myocardial protection. All patients had the same anesthetic treatment according to a fast-track protocol for early extubation, recovery, and discharge from the hospital previously described [11]. Intraoperative TEE was used at the discretion of the surgeon and/or anesthesiologist for various indications such as left ventricular dysfunction (left ventricular ejection fraction [LVEF] ⬍ 30%), the presence of valve dysfunction (especially to evaluate mitral regurgitation in patients with severe coronary artery disease), valve repair operations, intracardiac shunts, cardiac tumors, suspected chamber thrombus or severe aortic disease. Once inserted postintubation, the TEE probe was left in place until the end of the operation. Data collection included documentation of dysphagia when subjective symptoms and signs were confirmed by abnormal barium cineradiography. Patient demographics such as age, LVEF, and gender were recorded. Operative characteristics that could have a bearing on dysphagia, ie duration of cardiopulmonary bypass (CPB) and operation (OR) time, and type of operation were noted. Postoperative outcomes such as duration of intubation and documented neurological dysfunction (stroke) by neurologic consultation and computed tomography scan, percutaneous endoscopic gastrostomy (PEG), postoperative hospital length of stay (HLOS), and mortality were recorded by exhaustive chart review. Univariate analysis with the Student’s t-test and 2 analysis was used to compare baseline preoperative variables between groups. Factors on which TEE groups differed at p less than 0.05 in the univariate analysis, and
ransesophageal echocardiography (TEE) has been used with increasing frequency during cardiac operations both for diagnostic and monitoring purposes [1]. It is a useful tool for monitoring left ventricular filling or detecting regional wall motion abnormalities or valve dysfunction intraoperatively [2– 6]. The risk of complications in ambulatory patients having TEE is very low [7], however the same cannot be said conclusively, of intraoperative TEE. Hogue and colleagues [8] found intraoperative TEE to be an independent predictor of postoperative dysphagia, while at least 2 other studies found no relationship between TEE and dysphagia [9, 10]. In order to evaluate a possible causative relationship of intraoperative TEE and postoperative dysphagia, we reviewed the records of all patients who underwent cardiac operations, with or without intraoperative TEE during a 1-year period.
Patients and Methods The patient population consisted of 838 consecutive patients who underwent cardiac operation (coronary bypass, valve replacement or repair, or combined/other operations) over a 12-month period. Of these, 126 patients had intraoperative TEE (TEE), and 712 patients had no intraoperative TEE (nonTEE). All patients had tepid (32° to 34°C) perfusion on cardiopulmonary bypass and Accepted for publication July 13, 1999. Address reprint requests to Dr Rousou, Division of Cardiac Surgery, Baystate Medical Center, 759 Chestnut St, Suite 4628, Springfield, MA 01107.
(Ann Thorac Surg 2000;69:486 –90) © 2000 by The Society of Thoracic Surgeons
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Downloaded from ats.ctsnetjournals.org by on June 4, 2013
0003-4975/00/$20.00 PII S0003-4975(99)01086-3
Ann Thorac Surg 2000;69:486 –90
20.7%) in the nonTEE group. Conversely, the percentage of coronary bypass patients was 39.7% in the TEE versus 79.3% in the nonTEE group ( p ⬍ 0.0001). Cardiopulmonary bypass (CPB) was significantly longer in the TEE group (159 ⫾ 6.4 versus 136.4 ⫾ 1.9 minutes, p ⬍ 0.001) as was the length of operation (OR time) (378.0 ⫾ 10.5 versus 348.0 ⫾ 3.6 minutes, p ⬍ 0.005). The overall incidence of dysphagia was 10 of 126 (7.9%) in TEE patients, and 13 of 712 (1.8%) in nonTEE patients. Of these TEE patients with dysphagia, 3 patients had a stroke, 3 patients had prolonged intubation and 4 patients had no clear-cut risk factor for developing dysphagia. Similarly, in the nonTEE group, 9 patients had strokes, 3 patients had prolonged intubation and 1 had no other identifiable risk factors for dysphagia. Intubation period was significantly longer for all TEE patients than for nonTEE patients (34.1 ⫾ 10.2 versus 17.5 ⫾ 2.6 hours, p ⬍ 0.050). Patients with dysphagia had much longer intubation periods than patients without dysphagia: 200.9 ⫾ 75 versus 15.3 ⫾ 1.6 minutes. The incidence of stroke was not significantly different between the TEE and nonTEE groups, 7.9% versus 5.1% respectively (not significant). There was a significant difference in the postoperative HLOS between all TEE and nonTEE patients (12.4 ⫾ 1.1 versus 8.4 ⫾ 0.4 days, p ⬍ 0.001). Furthermore, there was a marked and significant difference in HLOS between dysphagic and nondysphagic patients in both groups: (TEE 28.1 ⫾ 6.2 versus 11.0 ⫾ 1.0 days, p ⬍ 0.001); nonTEE 34.5 ⫾ 5.0 versus 7.9 ⫾ 0.4 days, p ⬍ 0.001). Percutaneous endoscopic gastrostomy (PEG) for nutrition was necessary in 12 of 838 patients. The incidence was 6 of 126 (4.8%) in the TEE group and 6 of 712 (0.8%) in the nonTEE group. Operative mortality was higher in the TEE group, 10.3% versus 3.8% in the nonTEE group ( p ⬍ 0.005); however, the difference in mortality between dysphagic and nondysphagic patients was significant only in the nonTEE patients (15.4% in dysphagic versus 3.8% in nondysphagic patients, p ⬍ 0.025). The results of multiple logistic regression are shown in Table 2. The normal statistic (z) for a factor represents the regression coefficient for that factor divided by its standard error, and indicates the relative importance of each factor. For example, a factor with a normal statistic of 4 is twice as important as one with a normal statistic of 2. The significance level of each factor represents the significance of the likelihood ratio test for that factor. An
Table 1. Potential Risk Factors for Dysphagia TEE No-TEE p (126 patients) (712 patients) Value
Variable
Age (y) 66.3 ⫾ 12 Intubation times (h) 34.1 ⫾ 10.2 Stroke (% patients) 7.9 CPB (min) 159.0 ⫾ 6.4 LVEF ⬍ 30% (% patients) 7.8 OR time (min) 378.0 ⫾ 10.5 Type of surgery CABG (% patients) 39.7 NonCABG (% patients) 60.3 a
487
ROUSOU ET AL DYSPHAGIA AFTER TEE
66.9 ⫾ 0.4 17.5 ⫾ 2.6 5.1 136.4 ⫾ 1.9 18.0 348.0 ⫾ 3.6
NS ⬍ 0.050 NS ⬍ 0.001 ⬍ 0.001 ⬍ 0.005
79.3 20.7
⬍ 0.001a
CABG versus non-CABG.
CABG ⫽ coronary artery bypass grafting; CPB ⫽ cardiopulmonary bypass; LVEF ⫽ left ventricular ejection fraction; NS ⫽ not significant; OR ⫽ operating room; TEE ⫽ transesophageal echocardiography.
which could lead to dysphagia suggesting the possibility of confounding, were entered in a multiple logistic regression model in a stepwise fashion [12] to identify independent predictors of dysphagia. A maximum likelihood procedure was used to calculate regression coefficients. The likelihood ratio criterion [13] was used to determine the significance of individual factors in the regression model with p less than 0.1 used as the entry criterion. Factors tested in the model, in addition to TEE, were the duration of CPB, OR time, length of endotracheal intubation, the presence of stroke, LVEF less than 30% and type of operation (CABG versus nonCABG).
Results Potential risk factors for dysphagia following cardiac operations are shown in Table 1. There was no significant difference in age or gender mix (% males, 62.7 versus 67.7) between TEE and nonTEE patients. There was a small but statistically significant difference in preoperative ejection fraction (EF) 42.3 ⫾ 1.3 percent in the TEE group versus 46.2 ⫾ 0.5 percent in the nonTEE group ( p ⬍ 0.005). Abnormal left ventricular ejection fraction (⬍30%) was more prevalent in the nonTEE group (18%) compared to the TEE group (7.8%) ( p ⬍ 0.001). There was a significantly higher percentage of valve (18.3%) and combined/other (42.0%) operations in the TEE group (total 60.3%) versus 7.3% and 13.4% respectively (total
Table 2. Independent Predictors of Postoperative Dysphagia Identified by Multivariate Logistic Regression Coefficient (Beta)
Standard Error
Normal Statistic (z)
Adjusted Odds
Lower 95% CI
Upper 95% CI
Stroke TEE LVEF Intubation time OR time Constant
3.079 2.054 1.709 0.013 0.376 ⫺8.625
0.756 0.744 0.796 0.007 0.214 1.51
4.07 2.76 2.15 1.91 1.76 ⫺5.7
21.7 7.80 5.50 1.01 1.46 0.0002
4.93 1.81 1.16 1.00 0.96 0.00001
95.9 33.6 26.4 1.03 2.22 0.0035
CI ⫽ confidence interval;
LVEF ⫽ left ventricular ejection fraction;
Factor
OR ⫽ operating room;
p Value ⬍ 0.001 ⬍ 0.001 0.023 0.005 0.077
TEE ⫽ transesophageal echocardiography.
Downloaded from ats.ctsnetjournals.org by on June 4, 2013
488
ROUSOU ET AL DYSPHAGIA AFTER TEE
Ann Thorac Surg 2000;69:486 –90
adjusted odds ratio (and 95% confidence interval) are also reported for each factor. The adjusted odds ratio for discrete variables represents the risk of disease for patients with the factor present compared to those with the factor absent, after adjusting for other significant factors in the logistic model. The adjusted odds ratio for continuous variables gives the percent increase in risk for every unit increase of the continuous variable. The most significant factor was the presence of stroke ( z ⫽ 4.07). After adjusting for other significant factors, the odds for developing dysphagia in patients with stroke was 21.2 times greater than patients who did not experience stroke. Next in importance was TEE ( z ⫽ 2.76). After adjusting for other significant factors, patients with an intraoperative TEE were 7.8 times as likely to experience dysphagia than those who did not have TEE. LVEF was next in importance ( z ⫽ 2.15). Patients with impaired LVEF (⬍ 30%) were 5.5 times as likely to have dysphagia as those with normal LVEF. Intubation time was next ( z ⫽ 1.91). No significant cut point could be found for this variable, therefore it was kept in its continuous form in this model. Patients with longer intubation times were more likely to experience dysphagia. The adjusted odds ratio of 1.01, indicates that for each additional minute of intubation, the risk of dysphagia increased by 1% of the previous minute’s risk, albeit the risk was very small with shorter intubation times. Length of operation was least important ( z ⫽ 1.76, and marginally significant, p ⫽ 0.077), as a factor for postoperative dysphagia. The adjusted odds of 1.46 indicate that for each additional hour of operative time, odds of dysphagia increased by 46%. No significant effect could be found on the incidence of dysphagia by patient age, duration of CPB or the type of operation (CABG versus nonCABG). There was no significant interaction between the presence of TEE and any of the other significant variables (stroke, intubation, and OR time).
Comment Intraoperative TEE has gained wide acceptance as a useful diagnostic and monitoring tool [1– 6]. It is invaluable in assessing valve pathology, especially the mitral valve, guiding surgical decisions for repair versus replacement. If repair is undertaken, TEE can also evaluate success or failure of such repair. Other uses of TEE include monitoring ventricular filling, regional and global wall motion in ischemic heart disease, identifying intracavitary thrombus or tumor and evaluating aortic mural or intraluminal disease. Concomitant with its increasing use in the operating room, Hogue and colleagues noted an increase in the incidence of postoperative dysphagia in cardiac surgical patients [8]. In a retrospective review triggered by their observations, they found an independent effect of TEE on the incidence of dysphagia, suggesting a causative association. Age was also a significant factor in their study. Other investigators, however, found no such association in retrospective studies [9, 10].
Like Hogue and colleagues, we also noted an increase in the incidence of dysphagia along with the increasing use of intraoperative TEE. However, dysphagia and/or aspiration can be caused by a variety of factors such as pharyngeal and esophageal motility disorders, neurological disorders (stroke) [14], prolonged intubation and/or tracheostomy [15], or prolonged critical illness among others. In order to identify independent risk factors for dysphagia in our population, we entered stroke, length of intubation, use of TEE, duration of operation, CPB, preoperative ejection fraction (EF) and type of operation in a multivariate logistic regression model. We were also interested in evaluating any possible association of dysphagia with duration of operation, but more specifically the length of time the TEE probe remained in the patient (OR time was taken as the closest estimate of TEE probe indwelling time). In addition, we attempted to identify possible cutpoints in OR time and intubation time above which dysphagia might occur more frequently. As seen in Table 2, stroke had the strongest association with dysphagia (dysphagia was 21.7 times more frequent in patients with stroke than without). Use of TEE increased the risk of dysphagia 7.8 times. Age, duration of CPB, and type of operation had no independent effect. Although TEE was found to be an independent predictor of dysphagia, statistically we could not demonstrate interaction with OR time. It is interesting to note, however, that in operations that lasted less than 4.5 hours there were no patients with dysphagia with or without the use of TEE. This might explain the absence of dysphagia with TEE use in some reported studies [9, 10]. Regarding length of intubation as a cause of dysphagia, no significant cutpoint could be identified, and it was therefore examined as a continuous variable in this model. There is a progressive increase in risk of dysphagia with longer intubation times, albeit this risk is extremely small with shorter intubation times. The mechanisms of dysphagia that could result from TEE use remain unclear. We can only hypothesize that trauma during its insertion or use and/or compression of pharyngoesophageal tissues between the TEE probe and the endotracheal tube, may contribute to this complication. Prolonged intubation could also have similar effects on pharyngeal tissues. The effects of severe and prolonged critical illness on the swallowing mechanisms remain unclear. In conclusion, recognizing the limitations of this study (nonrandomized retrospective review, TEE use in higher risk patient, ie lower LVEF, longer duration of CPB and OR time, and higher percentage of valve or combined operations), we identified TEE to be an independent predictor of dysphagia. Our study does not conclusively prove that intraoperative use of TEE predisposes patients to the development of postoperative dysphagia. It is merely suggestive of an independent effect of TEE based on multiple logistic regression. This should not discourage the use of TEE in cardiac surgical patients when proper indications exist. A large prospective randomized study is necessary to clarify this issue conclusively.
Downloaded from ats.ctsnetjournals.org by on June 4, 2013
Ann Thorac Surg 2000;69:486 –90
ROUSOU ET AL DYSPHAGIA AFTER TEE
References 1. Shintani H, Nakano S, Matsuda H, Sakai K, Taniguchi K, Kawashima Y. Efficacy of transesophageal echocardiography as a perioperative monitor in patients undergoing cardiovascular surgery. J Cardiovasc Surg 1990;31:564–70. 2. Bergquist BD, Leung JM, Bellows WH. Transesophageal echocardiography in myocardial revascularization: I. accuracy of intraoperative real-time interpretation. Anesth Analg 1996;82:1132– 8. 3. Bergquist BD, Bellows WH, Leung JM. Transesophageal echocardiography in myocardial revascularization: II. influence on intraoperative decision making. Anesth Analg 1996; 82:1139– 45. 4. Smith JS, Cahalan MK, Benefiel DJ, et al. Intraoperative detection of myocardial ischemia in high-risk patients: electrocardiography versus two-dimensional transesophageal echocardiography. Circulation 1985;72:1015–21. 5. Sheikh KH, de Bruijn NP, Rankin JS, et al. The utility of transesophageal echocardiography and Doppler color flow imaging in patients undergoing cardiac valve surgery. J Am Coll Cardiol 1990;15:63–72. 6. Freeman WK, Schaff HV, Khandheria BK, et al. Intraoperative evaluation of mitral valve regurgitation and repair by transesophageal echocardiography: incidence and significance of systolic anterior motion. J Am Coll Cardiol 1992;20: 599 – 609.
489
7. Kwan-Leung C, Cohen GI, Sochowski RA, Baird MG. Complications of transesophageal echocardiography in ambulatory adult patients: analysis of 1500 consecutive examinations. J Am Soc Echocardiogr 1991;4:577– 82. 8. Hogue CW, Lappas GD, Creswell LL, et al. Swallowing dysfunction after cardiac operations. J Thorac Cardiovasc Surg 1995;110:517–22. 9. Messina AG, Paranicas M, Fiamengo S, et al. Risk of dysphagia after transesophageal echocardiography. Am J Cardiol 1991;67:313– 4. 10. Hulyalkar AR, Ayd JD. Low risk of gastroesophageal injury associated with transesophageal echocardiography. J Cardiothorac Vasc Anesth 1993;7:175–7. 11. Engelman RM, Rousou JA, Flack JE, et al. Fast-track recovery of the coronary bypass patient. Ann Thorac Surg 1994;58: 1742– 6. 12. Cox DR. Analysis of binary data. London: Methuen, 1970: 76–99. 13. Lee ET. Statistical methods for survival data analysis. New York: John Wiley & Sons, 1992:233– 6. 14. Horner J, Massey EW. Silent aspiration following stroke. Neurology 1988;38:317–9. 15. DeVita MA, Spierer-Rundback L. Swallowing disorders in patients with prolonged orotracheal intubation or tracheostomy tubes. Critical Care Medicine 1990;18:1328–30.
INVITED COMMENTARY The importance of intraoperative transesophageal echocardiography (TEE) in the modern practice of cardiovascular surgery and anesthesia is well established. The information provided by TEE enables the precise definition of cardiac valve lesions, the ability to detect regional myocardial ischemia and dysfunction, and to diagnose dissection, aneurysm, traumatic disruption, or atherosclerosis of the thoracic aorta [1]. Application of intraoperative TEE has contributed to the increased number and success of valve repair procedures [2], the development of minimally invasive cardiac surgical procedures [3], and improved outcome in emergency operations of the thoracic aorta [4]. Despite the rapidly expanding use of intraoperative TEE, complications from its use have been infrequent. Complications were often attributed to unrecognized diseases of the esophagus or difficulty inserting the TEE probe [5]. General anesthesia and cardiopulmonary bypass could further increase risks when using TEE. Anesthetized patients cannot voice discomfort during probe insertion or manipulation and TEE probes are often left in place during conditions of nonpulsatile perfusion and systemic hypothermia. Swallowing dysfunction after cardiac surgery is a serious complication that increases the risk for aspiration pneumonia, the need for tracheostomy, and prolonged hospitalization [6]. In this issue of The Annals, Rousou and colleagues [7] report that the frequency of dysphagia based on barium cineradiographic diagnosis was 7.9% for the 126 cardiac surgical patients who had intraoperative TEE versus 1.8% for 712 patients who did not have TEE. After adjusting for confounding variables including age, duration of intubation and stroke, TEE use was independently associated with dysphagia (odds ratio 7.8, 95%
confidence interval, 1.81 to 33.60, p ⬍ 0.001). These findings were similar to those reported by Hogue and colleagues [6] who found dysphagia in 4% of 869 cardiac surgical patients. In the latter report, intraoperative TEE use was also independently associated with dysphagia even after statistically adjusting for the effects of other risk factors such as age, duration of intubation, stroke and other postoperative complications [6]. In contrast, others have not found an association between intraoperative TEE use and the development of dysphagia [8]. If TEE proves to be causally related to dysphagia after cardiac surgery, the risks of intraoperative TEE may be greater than previously suspected. Despite these reports, it remains difficult to prove that TEE use alone was the cause for postoperative dysphagia. There are multiple problems when linking TEE use to dysphagia as based on the reported series. These include the lack of preoperative assessments for dysphagia, the delay between the TEE examination and the diagnosis of dysphagia, nonuniform documentation of the disorder, and differences in the patient populations and indication for TEE between centers. Further, because in these retrospective studies TEE use was not prospectively randomized, bias for the use of TEE in more complex surgical procedures and for more complicated patients could influence the results. One must acknowledge, though, that multivariate logistic regression analysis used in both the reports by Rousou and associates [8] and Hogue and coworkers [6] adjusted for many of these confounding variables including the type of procedure, duration of cardiopulmonary bypass, and length of the operation. Regardless, these results only show that TEE was statistically associated with dysphagia and its use may or may not be
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Downloaded from ats.ctsnetjournals.org by on June 4, 2013
0003-4975/00/$20.00 PII S0003-4975(99)01086-3
Risk of dysphagia after transesophageal echocardiography during cardiac operations John A. Rousou, Dennis A. Tighe, Jane L. Garb, Howard Krasner, Richard M. Engelman, Joseph E. Flack, III and David W. Deaton Ann Thorac Surg 2000;69:486-489 Updated Information & Services
including high-resolution figures, can be found at: http://ats.ctsnetjournals.org/cgi/content/full/69/2/486
References
This article cites 13 articles, 6 of which you can access for free at: http://ats.ctsnetjournals.org/cgi/content/full/69/2/486#BIBL
Citations
This article has been cited by 10 HighWire-hosted articles: http://ats.ctsnetjournals.org/cgi/content/full/69/2/486#otherarticles
Permissions & Licensing
Requests about reproducing this article in parts (figures, tables) or in its entirety should be submitted to: http://www.us.elsevierhealth.com/Licensing/permissions.jsp or email:
[email protected].
Reprints
For information about ordering reprints, please email:
[email protected]
Downloaded from ats.ctsnetjournals.org by on June 4, 2013
