- Email: [email protected]
Incidence and Risk Factors of Postoperative Vocal Cord Paralysis in 987 Patients After Cardiovascular Surgery
Taiga Itagaki, MD, Mutsuhito Kikura, MD, and Shigehito Sato, MD, PhD Department of Anesthesiology, Hamamatsu Medical Center and Department of Anesthesiology and Intensive Care, Hamamatsu University School of Medicine, Hamamatsu, Japan
Background. Vocal cord paralysis (VCP) after cardiovascular surgery can affect the postoperative outcome. The aim of the present study was to clarify the incidence of VCP after cardiovascular surgery and the relationship between the surgery characteristics and the risk of VCP. Methods. A total of 987 consecutive patients who underwent cardiovascular surgery (cardiac, n ⴝ 895; aortic, n ⴝ 92) were enrolled. We retrospectively assessed the incidence and the risk of VCP according to the length and types of surgery and the details of each VCP case and compared them between the aortic and the nonaortic group. Results. Twenty-three patients (2.3%: cardiac, n ⴝ15; aortic, n ⴝ 8) were expertly diagnosed with VCP (left, n ⴝ 19; right, n ⴝ 2; bilateral, n ⴝ 2). In the multivariate analysis, the risk for VCP increased with the duration of
the operation (odds ratio [OR], 4.4, 95% confidence interval [CI] 1.7 to 11.4), and aortic procedures (OR, 5.6, CI 2.3 to 13.5) exhibited higher risk compared with coronary artery bypass grafting. Among the VCP cases, the incidence of poor outcomes (ie, bilateral VCP, repeated airway treatment, death within 6 months) was significantly higher in the aortic group (p ⴝ 0.016). All patients that were intubated for longer than 100 hours exhibited poor outcomes. Conclusions. Aortic procedures and prolonged operation increase the risk of VCP. Severe VCP tended to be associated with aortic surgery and intubation for more than 100 hours.
V
overlooked [11]. The main symptom of VCP is transient hoarseness [1, 12] and it can be treated as a minor complication even when it is noticed. However, in severe cases of VCP, it might be possible to aspirate, due to loss of the laryngeal reflex; in particular, bilateral VCP is potentially fatal [13, 14]. The incidence and risk factors of VCP and the clinical course of these patients have not been fully elucidated in previous reports and those reports did not examine a large number of criteria. The aim of the present retrospective cohort study was to clarify the incidence of VCP after cardiovascular surgery, the relationship between the characteristics of surgery and the risk of VCP, and postoperative outcome in 987 patients treated at our institute between 1991 and 2004.
ocal cord paralysis (VCP) after general anesthesia is considered to be a relatively rare complication [1]. Nevertheless, VCP is one of the most serious complications associated with tracheal intubation [2– 6]. There are numerous claims for airway injury in the American Society of Anesthesiologists Closed Claims database, and the most common site of such injury is the larynx, representing 33% of all claimed airway injuries [2]. In particular, VCP becomes a risk factor for aspiration pneumonia that potentially increases postoperative morbidity and mortality [7]. Moreover, the incidence increases markedly in cardiovascular surgery, with reported rates of 0.67% to 1.9% [8] for cardiac surgery and 8.6% to 32% [9] for aortic surgery. Various mechanisms, such as direct recurrent laryngeal nerve injury at the operating site, in addition to intubation-related injuries, are considered to be implicated in the development of VCP after cardiovascular surgery [7, 8, 10]. Many severe complications can occur after cardiovascular surgery, including hemodynamic instability, respiratory insufficiency, renal failure, cerebral infarction, etc. Therefore, VCP can sometimes be easily
Accepted for publication Feb 5, 2007. Address correspondence to Dr Itagaki, Department of Anesthesiology, Hamamatsu Medical Center, 328 Tomizuka-cho, Hamamatsu, 432-8580, Japan; e-mail: [email protected].
© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2007;83:2147–52) © 2007 by The Society of Thoracic Surgeons
Material and Methods After obtaining approval from the Institutional Review Committee (Hamamatsu Medical Center, Hamamatsu, Japan), we conducted a retrospective cohort study of 1,015 consecutive cases of elective and emergent cardiovascular surgery performed on adults between January 1, 1991 and December 31, 2004 at the Hamamatsu Medical Center. Because our study was performed retrospectively and individual patients were not identified, signed informed consent from each patient was waived as approved by the institutional review committee. 0003-4975/07/$32.00 doi:10.1016/j.athoracsur.2007.02.008
CARDIOVASCULAR
Incidence and Risk Factors of Postoperative Vocal Cord Paralysis in 987 Patients After Cardiovascular Surgery
2148
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
CARDIOVASCULAR
The VCP patients were defined as those who registered “vocal cord paralysis” or “recurrent laryngeal nerve paralysis” or “hoarseness” as the insured disease after undergoing examination of the vocal cord by an otorhinolaryngologist. Ten patients that were preoperatively diagnosed as VCP or had organic lesion in the laryngopharynx were excluded. An additional 18 patients that died within one week after surgery were excluded. Consequently, we studied the data of postoperative VCP in 987 cardiovascular surgery patients. The perioperative management of patients was performed as follows. General anesthesia was standard high-dose fentanyl anesthesia with tracheal intubation in all cases. Tracheal intubation was performed after complete muscle relaxation was achieved. In advance of the operation, radial arterial catheterization was performed. Next, central venous and pulmonary catheters were placed into the right internal jugular vein, and a gastric tube was placed. Transesophageal echocardiography was monitored as necessary. An anesthesiologist performed the tracheal intubation and a standard tracheal tube with a high volume cuff was used. The size of the tube was selected by the attending anesthesiologist according to the generalized concept of approximately 8.0 mm internal diameter (ID) in men and 7.5 mm ID in women. The left-sided double-lumen tube was used in all cases of descending aortic replacement and in some cases of aortic arch replacement, and the size was generally 37Fr in men and 35Fr in women. In this occasion, a doublelumen tube was exchanged for a standard one immediately after the operation. The core hypothermia technique was used in patients requiring total circulatory
Ann Thorac Surg 2007;83:2147–52
arrest. The target core body temperature was about 20°C. A roller-pump cardiopulmonary bypass circuit was routinely used and cold blood cardioplegia was used for myocardial protection. After the operation, all patients were moved to the intensive care unit (ICU) and managed carefully under mechanical ventilation and sedation. Extubation was subsequently performed at the discretion of the cardiovascular surgeons. During the postextubation period, if patients presented with hoarseness or aspiration or any respiratory disturbances, and surgeons suspected vocal cord abnormality as the cause, the patients consulted with the otorhinolaryngologist and underwent examination of the vocal cords by fiberscope. In patients who were diagnosed with postoperative vocal cord paralysis, those patients associated with aspiration or respiratory failure were defined as poor outcome patients. Among them, those patients associated with bilateral VCP, repeated airway treatment, or death within 6 months after surgery were defined as “more critical” patients.
Statistical Analysis The incidences and risks of VCP were calculated. Operation times of the patients who were or were not associated with VCP were compared by unpaired t test. Potential associations of categoric data (patient age, sex, emergency operation, medical history, and surgical procedure) with the occurrence of VCP were analyzed by the 2 test or Fisher exact test for trend as appropriate. In the subgroup of patients who were associated with VCP, the age, height, weight, body mass index, operation time, cardiopulmonary bypass time, total intubation
Table 1. Study Population Characteristics in Patients With or Without Vocal Cord Paralysisa Univariate Analysis Characteristics Age (yr): ⬍39 40–69 70⬍ Sex, male/female Operation time (min) Emergency operation Medical history: Hypertension Diabetes mellitus Surgical procedure: CABG Valvular surgery Ascending aorta and arch Descending aorta Other
Total n ⫽ 987
VCP; Yes (n ⫽ 23)
VCP; No (n ⫽ 964)
p Valueb
37 (3.7) 564 (57.1) 386 (39.1) 674 (68.3)/313 (31.7) 368 ⫾ 153 210 (21.3)
3 (13.0) 13 (56.5) 7 (30.4) 19 (82.6)/4 (17.4) 438 ⫾ 197 3 (13.0)
34 (3.5) 551 (57.2) 379 (39.3) 655 (67.9)/309 (32.1) 366 ⫾ 151 207 (21.5)
0.102 (Trend)
446 (45.2) 433 (43.9)
16 (69.6) 10 (43.5)
431 (44.7) 423 (43.9)
0.031 1
590 (59.8) 217 (22.0) 69 (7.0) 23 (2.3) 88 (8.9)
10 (43.5) 4 (17.4) 5 (21.7) 3 (13.0) 1 (4.3)
580 (60.2) 213 (22.1) 64 (6.6) 20 (2.1) 87 (9.0)
0.107 0.966 0.018 0.014 1
0.175 0.025 0.444
Data are expressed as number (%) unless otherwise noted. Values are the mean ⫾ standard deviation unless otherwise stated.
a
b
p values compare the vocal cord paralysis group versus the nonvocal cord paralysis group.
CABG ⫽ coronary artery bypass grafting; NS ⫽ not significant; septal defect closure; VCP ⫽ vocal cord paralysis.
Other ⫽ resection of left atrial myxoma atrial septal defect closure, ventricular
Ann Thorac Surg 2007;83:2147–52
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
2149
Table 2. Characteristics of Patients With Vocal Cord Paralysis According to the Surgical Proceduresa
Characteristics Age (yr) Sex, male/female Height (cm) Weight (kg) BMI Operation time (min) CPB time (min) Emergency operation Intubation time (hr) ICU stay (day) Side of VCP: Left Right Bilateral Degree of paralysis: Incomplete/complete Aspiration Repeated airway treatmentc Death in 6 months Minimum temperature (°C) Difficult airway Intraoperative TEE Double lumen tube
Nonaortic Surgical Procedures
Total (n ⫽ 23)
Ascending & Arch (n ⫽ 5)
Descending (n ⫽ 3)
CABG (n ⫽ 10)
Valvular (n ⫽ 4)
Other (n ⫽ 1)
p Valueb
59.7 ⫾ 16.0 19/4 160.3 ⫾ 9.0 54.7 ⫾ 8.9 21.3 ⫾ 3.2 438 ⫾ 197 201 ⫾ 92 3 (13.0) 100 ⫾ 146 9.2 ⫾ 8.1
71.8 ⫾ 5.8 4/1 159.0 ⫾ 9.0 58.0 ⫾ 14.2 22.8 ⫾ 4.4 576 ⫾ 209 279 ⫾ 114 1 (20) 236 ⫾ 264 14.2 ⫾ 11.6
45.3 ⫾ 25.5 3/0 172.3 ⫾ 5.1 59.3 ⫾ 4.0 20.1 ⫾ 2.3 647 ⫾ 207 283 ⫾ 87 2 (66.7) 144 ⫾ 42 16.7 ⫾ 5.0
63.7 ⫾ 9.3 8/2 157.1 ⫾ 7.8 54.7 ⫾ 7.8 22.2 ⫾ 2.8 406 ⫾ 125 173 ⫾ 51 0 (0) 57 ⫾ 61 7.1 ⫾ 6.4
51 ⫾ 18.5 3/1 158.3 ⫾ 7.4 47.0 ⫾ 2.2 18.8 ⫾ 0.9 212 ⫾ 34 112 ⫾ 33 0 (0) 13.3 ⫾ 1.2 3.8 ⫾ 1.5
36 1/0 172 54 18.3 408 198 0 (0) 53.8 5
0.637 1 0.160 0.135 0.595 0.001 0.001 0.032 0.010 0.007
19 (82.6) 2 (8.7) 2 (8.7)
4 (80) 0 (0) 1 (20)
2 (66.7) 0 (0) 1 (33.3)
8 (80) 2 (20) 0 (0)
4 (100) 0 (0) 0 (0)
1 (100) 0 (0) 0 (0)
0.589 0.526 0.111
11/12 7 (30.4) 3 (13.0) 2 (8.7) 28.1 ⫾ 6.3 1 (4.3) 3 (13.0) 6 (26.1)
4/1 1 (20) 2 (40) 2 (40) 20 0 (0) 0 (0) 3 (60)
1/2 1 (33.3) 0 (0) 0 (0) 20 0 (0) 0 (0) 3 (100)
4/6 3 (30) 1 (10) 0 (0) 32.2 ⫾ 2.5 1 (10) 1 (10) 0 (0)
2/2 1 (25) 0 (0) 0 (0) 33 0 (0) 2 (50) 0 (0)
0/1 1 (100) 0 (0) 0 (0) 33 0 (0) 0 (0) 0 (0)
0.667 1 0.269 0.348 ⬍0.0001 1 0.526 0.001
Data are expressed as number (%) unless otherwise noted. Values are the mean ⫾ standard deviation unless otherwise stated.
a
p values; compare the aortic surgical procedures (n ⫽ 8) versus nonaortic surgical procedures (n ⫽ 15).
b c
Repeated airway treatment: reintubation and (or) tracheostomy.
BMI ⫽ body mass index; CABG ⫽ coronary artery bypass grafting; CPB ⫽ cardiopulmonary bypass; significant; Other ⫽ ventricular septal defect closure; VCP ⫽ vocal cord paralysis.
time, ICU stay, and minimum core temperature were compared with unpaired t test between the aortic and nonaortic surgery groups. Categoric data, such as side, degree and symptoms due to VCP, recurrence of airway management (ie, reintubation or tracheostomy), the use of double-lumen tube or transesophageal echocardiography (TEE), and difficult tracheal intubation were analyzed by 2 test between the aortic and nonaortic surgery groups. To estimate the odds ratios and 95% confidence intervals by multivariate logistic regression analysis, operation time, history of hypertension, and surgical procedures were treated as independent categoric variables. Vocal cord paralysis was treated as the dependent categoric variable. We initially performed univariate analyses, and all independent variables that were significant (2-tailed nominal p value ⬍ 0.1) in the univariate analyses were subsequently entered into a multivariate logistic analysis by the proportional odds model. Stepwise logistic regression was performed, and variables that were significant (2-tailed nominal p value ⬍ 0.05) were retained. Statistical analyses were performed using Statview 5 (SAS Institute Inc, Cary, NC).
ICU ⫽ intensive care unit;
NS ⫽ not
Results The characteristics of the 987 studied patients (895 patients in cardiac surgery, 92 patients in aortic surgery) are summarized in Table 1. There was no significant difference in the incidence of VCP by age, sex, emergency operation, and history of diabetes mellitus. The incidences were significantly increased in patients who were diagnosed preoperatively with hypertension (p ⫽ 0.031) and in patients who underwent aortic surgeries (p ⫽ 0.004). Characteristics of a total of 23 patients (2.3%) with VCP are summarized according to the surgical procedures in Table 2. All 23 patients developed dysphonia or respiratory disturbance within a day of extubation. Of the 23 patients who suffered VCP, 19 (1.9%) presented with left VCP and both right VCP and bilateral VCP were found in 2 (0.2%) each. The duration of operation, tracheal intubation, cardiopulmonary bypass, and ICU stay were all significantly greater for the aortic surgery group compared with the nonaortic surgery group, and emergency operation, the use of double-lumen tube, bilateral VCP, and death within 6 months after surgery occurred more frequently in the aortic surgery group.
CARDIOVASCULAR
Aortic Surgical Procedures
2150
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
Table 3. Multivariate Logistic Regression Analysis of Risk Factors for Vocal Cord Paralysis in Cardiovascular Surgery Patients (n ⫽ 987) CARDIOVASCULAR
Risk Factors Operation time: Total ⬍10 hr 10 hr ⬍ Aortic surgery ⬍10 hr 10 hr ⬍ Nonaortic surgery ⬍10 hr 10 hr ⬍ Preoperative hypertension Surgical procedures: CABG Valvular surgery Aortic surgery: Ascending & Arch Descending
Odds Ratio
95 % CI
Multivariate Analysis (p Value)
1, as reference 4.4
1.7–11.4
0.003
1, as reference 2.8
0.2–11.4
0.717
1, as reference 1.5 2.6
0.6–12.8 1.1–6.5
0.171 0.035
1, as reference 1.1 5.6 4.5
0.3–3.5 2.3–13.5 1.5–13.7
0.886 ⬍0.0001 0.007
8.7
2.2–34.1
0.002
CABG ⫽ coronary artery bypass grafting;
Ann Thorac Surg 2007;83:2147–52
surgery (p ⬍ 0.0001); in particular, the risk was 8.7 times higher in patients who underwent descending aortic surgery (p ⫽ 0.002), compared with those who underwent coronary artery bypass grafting. The risk was 2.6 times higher in patients with hypertension (p ⫽ 0.035) compared with those patients without that comorbidity. Ten patients with poor outcomes are summarized in Table 4. Of these 10 patients, the outcomes of five patients (Nos. 6 to 10) were “more critical” when associated with bilateral VCP, repeated airway treatment, or death within six months after surgery and the number of such patients was significantly (p ⫽ 0.016) higher in the aortic surgery group. All cases intubated for more than 100 hours were considered the more critical cases.
Comment
CI ⫽ confidence interval.
The results of the multivariate logistic regression analysis of the risk factors for VCP are summarized in Table 3. Among all surgical procedures, the risk of VCP increased with the operation time (for every minute: multivariate odds ratio, 1.003; 95% confidence interval, 1.001 to 1.005; p ⫽ 0.003), and the risk after a long operation of over 10 hours was 4.4 times higher (p ⫽ 0.003) compared with that of an operation of less than 10 hours, although the risk of VCP did not increase statistically with the operation time when the aortic and nonaortic surgery groups were analyzed separately. The risk was 5.6 times higher in patients who underwent aortic
The present study demonstrated that VCP after cardiovascular surgery occurred with a markedly higher incidence (2.3%) than that after general operations in an early investigation. Furthermore, we demonstrated that prolonged operation and aortic procedures increased the risk. Although it is well-known that VCP can result in hoarseness, stridor, poor cough (difficulty in expectoration), dysphagia, or aspiration, VCP is sometimes easily overlooked, particularly when the postoperative condition is critical [11]. Therefore, to recognize the potential risk factors for VCP, an appropriate risk analysis is considered to be very important. The etiologic mechanisms of postoperative VCP generally fall into three types, as follows: the first mechanism is recurrent laryngeal nerve paralysis (RNP), the second one is arytenoid dislocation, and the third is traumatic vocal cord injury. Furthermore, RNP is further subdivided into two patterns of direct injury and indirect injury. The mechanism of RNP due to tracheal intubation is usually indirect injury, neurapraxia without nerve degeneration [15]. The factors of indirect injury are said to include tracheal tube size, location of the cuff, fixing side of the tube, cuff pressure, curvature of tracheal tube, intubation time, etc [13, 16, 17].
Table 4. Clinical and Postoperative Data of 10 Cases of Vocal Cord Paralysis With Poor Outcome Patient 1 2 3 4 5 6 7 8 9 10
Age/Sex
Poor Outcomea
Surgery
Intubation Time (hr)
Side of VCP
Repeated Airway Treatment
29/F 36/M 60/M 69/M 61/F 81/M 77/M 69/M 78/F 20/M
Aspiration Aspiration pneumonia Aspiration Aspiration Aspiration Difficulty in expectoration Choking Difficulty in expectoration Aspiration pneumonia Aspiration
Heart Heart Heart Heart Heart Heart Aorta Aorta Aorta Aorta
13.7 53.8 54.9 29.3 42.7 195.0 624.0 45.5 504.0 160.6
L L L R L L B L L B
— — — — — Tracheostomy Tracheostomy Reintubation — —
a
Poor outcome is defined as the existence of aspiration or respiratory disturbance.
POD ⫽ postoperative day;
VCP ⫽ vocal cord paralysis.
Remarks
Died (48 POD) Died (134 POD) Persistent hoarseness
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
Among our patients, each VCP case used an approximately standard-sized tube, whereas a thicker doublelumen tube was used in six aortic cases (26.1%). Considering the fact that the number of cases of descending aortic replacement was only 2.3% overall, it is suggested that involvement of the double-lumen tube was significant. However, the double-lumen tube was changed to a standard tube just after the operation; therefore, it would not be the sole risk factor, but a risk factor combined with operation time. Furthermore, hyperextension of the neck often occurs during cardiovascular surgery. Therefore, a patient is likely to be in the situation where the compression of the recurrent laryngeal nerve occurs easily, because of displacement of the cuff and esophageal traction [10, 18]. The mechanisms of nonintubation-related indirect RNP include sternal traction, TEE use, and cooling. Excessive lateral sternal traction causes nerve compression by lateral and anterior stretching of the subclavian arteries [7]. The TEE probe is considered to produce nerve compression at the postcricoid area and to generate RNP [19, 20]. In the VCP cases, there were three cases of TEE use. Incidentally, this type of nerve injury is known as a complication of gastric tube placement that is defined as nasogastric tube syndrome [21, 22]. Core hypothermia for total circulatory arrest is also a characteristic of cardiovascular surgery, and includes moderate hypothermia for ordinary cardiopulmonary bypass, cold blood cardioplegia, and topical ice slush for myocardial protection. Cooling causes neurapraxia and contributes to the pathogenesis of RNP [11, 23, 24]. In cardiovascular surgery there is a high possibility that direct RNP may occur, especially as a result of surgical procedures. Specifically, it has been reported that this kind of injury is produced by central venous catheterization [25], sternotomy [7], harvesting of the internal thoracic artery [26], aortic manipulation [9], etc. In general, RNP is mostly found unilaterally, on the left side [13]. This tendency is more prominent in cardiovascular surgery because the left recurrent laryngeal nerve is vulnerable due to its anatomic length [7]. We were not able to find any description of accidental or unavoidable surgical nerve injury in the operation records. Macroscopically, we were easily able to distinguish traumatic vocal cord injury and neurological injury and determine the degree of paralysis. However, it was very difficult to diagnose what type of neurological injury had occurred. Therefore, we were not able to obtain any data about the incidence of each injurious mechanism. Eventually, we were only able to speculate from the clinical course whether the paralysis was curable or not. The symptom of indirect RNP is usually mild and disappears within several months. If symptoms persist for more than this period, it seems more likely that the direct injury was caused by the surgical procedure and that the damage will be permanent [11]. We could not confirm long-term outcomes of all VCP cases, but found that only one aortic patient presented with persistent hoarseness even three month later. This case may have been due to direct recurrent laryngeal nerve injury at the operating site.
2151
The VCP incidence was significantly higher in patients with preoperative hypertension, but not in those with preoperative diabetes mellitus. Several studies have suggested that this may be due to the use of cardiopulmonary bypass related with peripheral nerve injury, and the impaired microcirculation [24, 27]. Both hypertension and diabetes mellitus may generate VCP because they contribute to impaired microcirculation as the main pathophysiologic feature. Further investigation is required to further examine this point. In this study, the incidence and the risk of postoperative VCP increased with prolonged operation. Generally, the prolongation of surgery can increase the chance of direct recurrent laryngeal nerve injury with time. Furthermore, we consider that prolonged surgical cases are intubated longer postoperatively because of the persistence of a critical state. Recurrent laryngeal nerve and laryngeal mucosa injury can occur as a result of the presence of the tube, and insufficiency of the vocal cord will happen [15, 28]. While surgical time was not the sole risk factor in our study, we suppose that total intubation time could potentially be a considerable risk factor for VCP. Our results were well-matched with a previous report that demonstrated that aortic surgery patients had higher incidence and risk of VCP compared with cardiac surgery patients [9]. The duration of surgery, intubation, and cardiopulmonary bypass were significantly longer in the aortic surgery patients compared with the nonaortic surgery patients. Furthermore, there were significant differences in the aortic surgery group among the items of emergency operation, minimum core temperature, and the use of a double-lumen tube. In addition to the time factor, the use of a double-lumen tube and the cooling are all mentioned as the background of VCP, and these factors confirm the high incidence of VCP after aortic surgery. Aspiration and repeated airway treatment did not exhibit statistical significance as the items associated with the severity of VCP between the aortic and nonaortic groups, likely due to the small number of cases. However, two cases of bilateral VCP and two cases that resulted in death within six months after operation were all aortic surgery patients and the ICU stay was significantly longer in the aortic surgery group. Therefore, there is a marked tendency toward an increase in severity of VCP after aortic surgery. Ishimoto and colleagues [9] reported that the incidence of VCP was most common after operation on type I aneurysms (64%, ascending and arch), according to the classification of de Bakey and colleagues, among all the types of thoracic aortic aneurysms. In our study, descending aortic replacement exhibited a higher incidence and doubled the risk of VCP compared with that of the ascending aorta and arch. No significant differences were observed for any items we examined between the two surgical procedures and we consider that the small number of cases of VCP makes it more difficult to compare the degree of risk for each. In theory, because of the anatomic length of the left recurrent laryngeal nerve, it would appear that direct injury could occur more frequently during surgery with manipulation of the aortic arch. Further investigation of this point is warranted.
CARDIOVASCULAR
Ann Thorac Surg 2007;83:2147–52
2152
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
CARDIOVASCULAR
In the present study, the largest severity-related factor of VCP is the duration of intubation. When it is anticipated that the postoperative intubation time will be longer due to a serious condition, tracheostomy should be considered because it would be effective for both respiratory assistance as well as vocal cord protection [11]. As a matter of course, if any abnormalities are recognized after extubation it is important to consider the appropriate therapeutic treatment as well as prompt consultation with an otorhinolaryngologist. Additionally, it can be more difficult to differentiate between RNP and arytenoid dislocation. If arytenoid dislocation exists, surgical treatment is sometimes required and delayed diagnosis leads to a complicated situation where ankylosis of cricoarytenoid articulation makes repositioning more difficult [29]. For this reason, expert diagnosis is also necessary. There are some limitations of this retrospective cohort study that should be mentioned. First, because our investigation focused on patients in whom postoperative VCP after dysphonia was diagnosed, we did not investigate how many patients suffered transient or asymptomatic vocal cord dysfunction without paralysis. We also did not identify patients who had potential preexisting vocal cord pathology, such as smokers or those with gastroesophageal reflux disease. These patients would be at risk for postoperative vocal cord morbidities. It is likely that our study underestimated patients with postoperative vocal cord morbidities, and thus the relationship between preexisting vocal cord pathology and the risk of postoperative VCP was unclear. Second, we were not able to systematically assess some interesting variables, such as body mass index, total intubation time, the difficulty of tracheal intubation, the use of intraoperative TEE, double-lumen tube, or emergent surgery because of the large demographic number. Therefore, the relationship between those factors and postoperative VCP was examined only among the VCP cases. Third, we did not analyze the relation between the incidence of postoperative VCP and the length of ICU stay or perioperative mortality and long-term morbidity as a comprehensive outcome. Therefore, the significance of VCP on the clinical outcome is unclear. Finally, we do not know the incidence of VCP after general surgery in our institute for comparison. In the future, if these problems are resolved, we will obtain more objective and reliable incidence and risk information. In conclusion, this study demonstrated that aortic procedures and prolonged operation increased the risk of VCP, and severe VCP tended to be associated with aortic surgery and intubation for more than 100 hours. Vocal cord paralysis after cardiovascular surgery occasionally presents with life-threatening complications because it may generate aspiration or airway obstruction due to loss of the laryngeal reflex. We should provide a more detailed preoperative explanation to the cardiovascular surgery patients regarding the potential risk for VCP.
References 1. Jones MW, Catling S, Evans E, Green DH, Green JR. Hoarseness after tracheal intubation. Anaesthesia 1992;47:213– 6.
Ann Thorac Surg 2007;83:2147–52
2. Domino KB, Posner KL, Caplan RA, Cheney FW. Airway injury during anesthesia: a closed claims analysis. Anesthesiology 1999;91:1703–11. 3. Mehta S. Safer endotracheal intubation. Lancet 1986;1:1148. 4. Laryngeal paralysis after endotracheal intubation. Lancet 1986;1:536-7. 5. Holley HS, Gildea JE. Vocal cord paralysis after tracheal intubation. JAMA 1971;215:281– 4. 6. David DS, Shah M. Vocal cord paralysis following intubation. JAMA 1971;216:1645– 6. 7. Hamdan AL, Moukarbel RV, Farhat F, Obeid M. Vocal cord paralysis after open-heart surgery. Eur J Cardiothorac Surg 2002;21:671– 4. 8. Dimarakis I, Protopapas AD. Vocal cord palsy as a complication of adult cardiac surgery: surgical correlations and analysis. Eur J Cardiothorac Surg 2004;26:773–5. 9. Ishimoto S, Ito K, Toyama M, et al. Vocal cord paralysis after surgery for thoracic aortic aneurysm. Chest 2002;121:1911–5. 10. Inada T, Fujise K, Shingu K. Hoarseness after cardiac surgery. J Cardiovasc Surg (Torino) 1998;39:455–9. 11. Shafei H, el-Kholy A, Azmy S, Ebrahim M, al-Ebrahim K. Vocal cord dysfunction after cardiac surgery: an overlooked complication. Eur J Cardiothorac Surg 1997;11:564 – 6. 12. Lesser T, Williams G. Laryngographic investigation of postoperative hoarseness. Clin Otolaryngol Allied Sci 1988;13:37– 42. 13. Cavo JW Jr. True vocal cord paralysis following intubation. Laryngoscope 1985;95:1352–9. 14. Gibbin KP, Egginton MJ. Bilateral vocal cord paralysis following endotracheal intubation. Br J Anaesth 1981;53:1091–2. 15. Santos PM, Afrassiabi A, Weymuller EA Jr. Risk factors associated with prolonged intubation and laryngeal injury. Otolaryngol Head Neck Surg 1994;111:453–9. 16. Mandoe H, Nikolajsen L, Lintrup U, Jepsen D, Molgaard J. Sore throat after endotracheal intubation. Anesth Analg 1992;74:897–900. 17. Stout DM, Bishop MJ, Dwersteg JF, Cullen BF. Correlation of endotracheal tube size with sore throat and hoarseness following general anesthesia. Anesthesiology 1987;67:419 –21. 18. Hartrey R, Kestin IG. Movement of oral and nasal tracheal tubes as a result of changes in head and neck position. Anaesthesia 1995;50:682–7. 19. Zwetsch G, Filipovic M, Skarvan K, Todorov A, Seeberger MD. Transient recurrent laryngeal nerve palsy after failed placement of a transesophageal echocardiographic probe in an anesthetized patient. Anesth Analg 2001;92:1422–3. 20. Kawahito S, Kitahata H, Kimura H, Tanaka K, Oshita S. Recurrent laryngeal nerve palsy after cardiovascular surgery: relationship to the placement of a transesophageal echocardiographic probe. J Cardiothorac Vasc Anesth 1999; 13:528 –31. 21. Sofferman RA, Haisch CE, Kirchner JA, Hardin NJ. The nasogastric tube syndrome. Laryngoscope 1990;100:962– 8. 22. Ellis PD, Pallister WK. Recurrent laryngeal nerve palsy and endotracheal intubation. J Laryngol Otol 1975;89:823– 6. 23. Tewari P, Aggarwal SK. Combined left-sided recurrent laryngeal and phrenic nerve palsy after coronary artery operation. Ann Thorac Surg 1996;61:1721–3. 24. Lederman RJ, Breuer AC, Hanson MR, et al. Peripheral nervous system complications of coronary artery bypass graft surgery. Ann Neurol 1982;12:297–301. 25. Sim DW, Robertson MR. Right vocal cord paralysis after internal jugular vein cannulation. J Laryngol Otol 1989;103:424. 26. Phillips TG, Green GE. Left recurrent laryngeal nerve injury following internal mammary artery bypass. Ann Thorac Surg 1987;43:440. 27. Grocott HP, Clark JA, Homi HM, Sharma A. “Other” neurologic complications after cardiac surgery. Semin Cardiothorac Vasc Anesth 2004;8:213–26. 28. Lundy DS, Casiano RR, Shatz D, Reisberg M, Xue JW. Laryngeal injuries after short- versus long-term intubation. J Voice 1998;12:360 –5. 29. Sataloff RT, Bough ID Jr, Spiegel JR. Arytenoid dislocation: diagnosis and treatment. Laryngoscope 1994;104:1353– 61.
Background. Vocal cord paralysis (VCP) after cardiovascular surgery can affect the postoperative outcome. The aim of the present study was to clarify the incidence of VCP after cardiovascular surgery and the relationship between the surgery characteristics and the risk of VCP. Methods. A total of 987 consecutive patients who underwent cardiovascular surgery (cardiac, n ⴝ 895; aortic, n ⴝ 92) were enrolled. We retrospectively assessed the incidence and the risk of VCP according to the length and types of surgery and the details of each VCP case and compared them between the aortic and the nonaortic group. Results. Twenty-three patients (2.3%: cardiac, n ⴝ15; aortic, n ⴝ 8) were expertly diagnosed with VCP (left, n ⴝ 19; right, n ⴝ 2; bilateral, n ⴝ 2). In the multivariate analysis, the risk for VCP increased with the duration of
the operation (odds ratio [OR], 4.4, 95% confidence interval [CI] 1.7 to 11.4), and aortic procedures (OR, 5.6, CI 2.3 to 13.5) exhibited higher risk compared with coronary artery bypass grafting. Among the VCP cases, the incidence of poor outcomes (ie, bilateral VCP, repeated airway treatment, death within 6 months) was significantly higher in the aortic group (p ⴝ 0.016). All patients that were intubated for longer than 100 hours exhibited poor outcomes. Conclusions. Aortic procedures and prolonged operation increase the risk of VCP. Severe VCP tended to be associated with aortic surgery and intubation for more than 100 hours.
V
overlooked [11]. The main symptom of VCP is transient hoarseness [1, 12] and it can be treated as a minor complication even when it is noticed. However, in severe cases of VCP, it might be possible to aspirate, due to loss of the laryngeal reflex; in particular, bilateral VCP is potentially fatal [13, 14]. The incidence and risk factors of VCP and the clinical course of these patients have not been fully elucidated in previous reports and those reports did not examine a large number of criteria. The aim of the present retrospective cohort study was to clarify the incidence of VCP after cardiovascular surgery, the relationship between the characteristics of surgery and the risk of VCP, and postoperative outcome in 987 patients treated at our institute between 1991 and 2004.
ocal cord paralysis (VCP) after general anesthesia is considered to be a relatively rare complication [1]. Nevertheless, VCP is one of the most serious complications associated with tracheal intubation [2– 6]. There are numerous claims for airway injury in the American Society of Anesthesiologists Closed Claims database, and the most common site of such injury is the larynx, representing 33% of all claimed airway injuries [2]. In particular, VCP becomes a risk factor for aspiration pneumonia that potentially increases postoperative morbidity and mortality [7]. Moreover, the incidence increases markedly in cardiovascular surgery, with reported rates of 0.67% to 1.9% [8] for cardiac surgery and 8.6% to 32% [9] for aortic surgery. Various mechanisms, such as direct recurrent laryngeal nerve injury at the operating site, in addition to intubation-related injuries, are considered to be implicated in the development of VCP after cardiovascular surgery [7, 8, 10]. Many severe complications can occur after cardiovascular surgery, including hemodynamic instability, respiratory insufficiency, renal failure, cerebral infarction, etc. Therefore, VCP can sometimes be easily
Accepted for publication Feb 5, 2007. Address correspondence to Dr Itagaki, Department of Anesthesiology, Hamamatsu Medical Center, 328 Tomizuka-cho, Hamamatsu, 432-8580, Japan; e-mail: [email protected].
© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2007;83:2147–52) © 2007 by The Society of Thoracic Surgeons
Material and Methods After obtaining approval from the Institutional Review Committee (Hamamatsu Medical Center, Hamamatsu, Japan), we conducted a retrospective cohort study of 1,015 consecutive cases of elective and emergent cardiovascular surgery performed on adults between January 1, 1991 and December 31, 2004 at the Hamamatsu Medical Center. Because our study was performed retrospectively and individual patients were not identified, signed informed consent from each patient was waived as approved by the institutional review committee. 0003-4975/07/$32.00 doi:10.1016/j.athoracsur.2007.02.008
CARDIOVASCULAR
Incidence and Risk Factors of Postoperative Vocal Cord Paralysis in 987 Patients After Cardiovascular Surgery
2148
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
CARDIOVASCULAR
The VCP patients were defined as those who registered “vocal cord paralysis” or “recurrent laryngeal nerve paralysis” or “hoarseness” as the insured disease after undergoing examination of the vocal cord by an otorhinolaryngologist. Ten patients that were preoperatively diagnosed as VCP or had organic lesion in the laryngopharynx were excluded. An additional 18 patients that died within one week after surgery were excluded. Consequently, we studied the data of postoperative VCP in 987 cardiovascular surgery patients. The perioperative management of patients was performed as follows. General anesthesia was standard high-dose fentanyl anesthesia with tracheal intubation in all cases. Tracheal intubation was performed after complete muscle relaxation was achieved. In advance of the operation, radial arterial catheterization was performed. Next, central venous and pulmonary catheters were placed into the right internal jugular vein, and a gastric tube was placed. Transesophageal echocardiography was monitored as necessary. An anesthesiologist performed the tracheal intubation and a standard tracheal tube with a high volume cuff was used. The size of the tube was selected by the attending anesthesiologist according to the generalized concept of approximately 8.0 mm internal diameter (ID) in men and 7.5 mm ID in women. The left-sided double-lumen tube was used in all cases of descending aortic replacement and in some cases of aortic arch replacement, and the size was generally 37Fr in men and 35Fr in women. In this occasion, a doublelumen tube was exchanged for a standard one immediately after the operation. The core hypothermia technique was used in patients requiring total circulatory
Ann Thorac Surg 2007;83:2147–52
arrest. The target core body temperature was about 20°C. A roller-pump cardiopulmonary bypass circuit was routinely used and cold blood cardioplegia was used for myocardial protection. After the operation, all patients were moved to the intensive care unit (ICU) and managed carefully under mechanical ventilation and sedation. Extubation was subsequently performed at the discretion of the cardiovascular surgeons. During the postextubation period, if patients presented with hoarseness or aspiration or any respiratory disturbances, and surgeons suspected vocal cord abnormality as the cause, the patients consulted with the otorhinolaryngologist and underwent examination of the vocal cords by fiberscope. In patients who were diagnosed with postoperative vocal cord paralysis, those patients associated with aspiration or respiratory failure were defined as poor outcome patients. Among them, those patients associated with bilateral VCP, repeated airway treatment, or death within 6 months after surgery were defined as “more critical” patients.
Statistical Analysis The incidences and risks of VCP were calculated. Operation times of the patients who were or were not associated with VCP were compared by unpaired t test. Potential associations of categoric data (patient age, sex, emergency operation, medical history, and surgical procedure) with the occurrence of VCP were analyzed by the 2 test or Fisher exact test for trend as appropriate. In the subgroup of patients who were associated with VCP, the age, height, weight, body mass index, operation time, cardiopulmonary bypass time, total intubation
Table 1. Study Population Characteristics in Patients With or Without Vocal Cord Paralysisa Univariate Analysis Characteristics Age (yr): ⬍39 40–69 70⬍ Sex, male/female Operation time (min) Emergency operation Medical history: Hypertension Diabetes mellitus Surgical procedure: CABG Valvular surgery Ascending aorta and arch Descending aorta Other
Total n ⫽ 987
VCP; Yes (n ⫽ 23)
VCP; No (n ⫽ 964)
p Valueb
37 (3.7) 564 (57.1) 386 (39.1) 674 (68.3)/313 (31.7) 368 ⫾ 153 210 (21.3)
3 (13.0) 13 (56.5) 7 (30.4) 19 (82.6)/4 (17.4) 438 ⫾ 197 3 (13.0)
34 (3.5) 551 (57.2) 379 (39.3) 655 (67.9)/309 (32.1) 366 ⫾ 151 207 (21.5)
0.102 (Trend)
446 (45.2) 433 (43.9)
16 (69.6) 10 (43.5)
431 (44.7) 423 (43.9)
0.031 1
590 (59.8) 217 (22.0) 69 (7.0) 23 (2.3) 88 (8.9)
10 (43.5) 4 (17.4) 5 (21.7) 3 (13.0) 1 (4.3)
580 (60.2) 213 (22.1) 64 (6.6) 20 (2.1) 87 (9.0)
0.107 0.966 0.018 0.014 1
0.175 0.025 0.444
Data are expressed as number (%) unless otherwise noted. Values are the mean ⫾ standard deviation unless otherwise stated.
a
b
p values compare the vocal cord paralysis group versus the nonvocal cord paralysis group.
CABG ⫽ coronary artery bypass grafting; NS ⫽ not significant; septal defect closure; VCP ⫽ vocal cord paralysis.
Other ⫽ resection of left atrial myxoma atrial septal defect closure, ventricular
Ann Thorac Surg 2007;83:2147–52
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
2149
Table 2. Characteristics of Patients With Vocal Cord Paralysis According to the Surgical Proceduresa
Characteristics Age (yr) Sex, male/female Height (cm) Weight (kg) BMI Operation time (min) CPB time (min) Emergency operation Intubation time (hr) ICU stay (day) Side of VCP: Left Right Bilateral Degree of paralysis: Incomplete/complete Aspiration Repeated airway treatmentc Death in 6 months Minimum temperature (°C) Difficult airway Intraoperative TEE Double lumen tube
Nonaortic Surgical Procedures
Total (n ⫽ 23)
Ascending & Arch (n ⫽ 5)
Descending (n ⫽ 3)
CABG (n ⫽ 10)
Valvular (n ⫽ 4)
Other (n ⫽ 1)
p Valueb
59.7 ⫾ 16.0 19/4 160.3 ⫾ 9.0 54.7 ⫾ 8.9 21.3 ⫾ 3.2 438 ⫾ 197 201 ⫾ 92 3 (13.0) 100 ⫾ 146 9.2 ⫾ 8.1
71.8 ⫾ 5.8 4/1 159.0 ⫾ 9.0 58.0 ⫾ 14.2 22.8 ⫾ 4.4 576 ⫾ 209 279 ⫾ 114 1 (20) 236 ⫾ 264 14.2 ⫾ 11.6
45.3 ⫾ 25.5 3/0 172.3 ⫾ 5.1 59.3 ⫾ 4.0 20.1 ⫾ 2.3 647 ⫾ 207 283 ⫾ 87 2 (66.7) 144 ⫾ 42 16.7 ⫾ 5.0
63.7 ⫾ 9.3 8/2 157.1 ⫾ 7.8 54.7 ⫾ 7.8 22.2 ⫾ 2.8 406 ⫾ 125 173 ⫾ 51 0 (0) 57 ⫾ 61 7.1 ⫾ 6.4
51 ⫾ 18.5 3/1 158.3 ⫾ 7.4 47.0 ⫾ 2.2 18.8 ⫾ 0.9 212 ⫾ 34 112 ⫾ 33 0 (0) 13.3 ⫾ 1.2 3.8 ⫾ 1.5
36 1/0 172 54 18.3 408 198 0 (0) 53.8 5
0.637 1 0.160 0.135 0.595 0.001 0.001 0.032 0.010 0.007
19 (82.6) 2 (8.7) 2 (8.7)
4 (80) 0 (0) 1 (20)
2 (66.7) 0 (0) 1 (33.3)
8 (80) 2 (20) 0 (0)
4 (100) 0 (0) 0 (0)
1 (100) 0 (0) 0 (0)
0.589 0.526 0.111
11/12 7 (30.4) 3 (13.0) 2 (8.7) 28.1 ⫾ 6.3 1 (4.3) 3 (13.0) 6 (26.1)
4/1 1 (20) 2 (40) 2 (40) 20 0 (0) 0 (0) 3 (60)
1/2 1 (33.3) 0 (0) 0 (0) 20 0 (0) 0 (0) 3 (100)
4/6 3 (30) 1 (10) 0 (0) 32.2 ⫾ 2.5 1 (10) 1 (10) 0 (0)
2/2 1 (25) 0 (0) 0 (0) 33 0 (0) 2 (50) 0 (0)
0/1 1 (100) 0 (0) 0 (0) 33 0 (0) 0 (0) 0 (0)
0.667 1 0.269 0.348 ⬍0.0001 1 0.526 0.001
Data are expressed as number (%) unless otherwise noted. Values are the mean ⫾ standard deviation unless otherwise stated.
a
p values; compare the aortic surgical procedures (n ⫽ 8) versus nonaortic surgical procedures (n ⫽ 15).
b c
Repeated airway treatment: reintubation and (or) tracheostomy.
BMI ⫽ body mass index; CABG ⫽ coronary artery bypass grafting; CPB ⫽ cardiopulmonary bypass; significant; Other ⫽ ventricular septal defect closure; VCP ⫽ vocal cord paralysis.
time, ICU stay, and minimum core temperature were compared with unpaired t test between the aortic and nonaortic surgery groups. Categoric data, such as side, degree and symptoms due to VCP, recurrence of airway management (ie, reintubation or tracheostomy), the use of double-lumen tube or transesophageal echocardiography (TEE), and difficult tracheal intubation were analyzed by 2 test between the aortic and nonaortic surgery groups. To estimate the odds ratios and 95% confidence intervals by multivariate logistic regression analysis, operation time, history of hypertension, and surgical procedures were treated as independent categoric variables. Vocal cord paralysis was treated as the dependent categoric variable. We initially performed univariate analyses, and all independent variables that were significant (2-tailed nominal p value ⬍ 0.1) in the univariate analyses were subsequently entered into a multivariate logistic analysis by the proportional odds model. Stepwise logistic regression was performed, and variables that were significant (2-tailed nominal p value ⬍ 0.05) were retained. Statistical analyses were performed using Statview 5 (SAS Institute Inc, Cary, NC).
ICU ⫽ intensive care unit;
NS ⫽ not
Results The characteristics of the 987 studied patients (895 patients in cardiac surgery, 92 patients in aortic surgery) are summarized in Table 1. There was no significant difference in the incidence of VCP by age, sex, emergency operation, and history of diabetes mellitus. The incidences were significantly increased in patients who were diagnosed preoperatively with hypertension (p ⫽ 0.031) and in patients who underwent aortic surgeries (p ⫽ 0.004). Characteristics of a total of 23 patients (2.3%) with VCP are summarized according to the surgical procedures in Table 2. All 23 patients developed dysphonia or respiratory disturbance within a day of extubation. Of the 23 patients who suffered VCP, 19 (1.9%) presented with left VCP and both right VCP and bilateral VCP were found in 2 (0.2%) each. The duration of operation, tracheal intubation, cardiopulmonary bypass, and ICU stay were all significantly greater for the aortic surgery group compared with the nonaortic surgery group, and emergency operation, the use of double-lumen tube, bilateral VCP, and death within 6 months after surgery occurred more frequently in the aortic surgery group.
CARDIOVASCULAR
Aortic Surgical Procedures
2150
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
Table 3. Multivariate Logistic Regression Analysis of Risk Factors for Vocal Cord Paralysis in Cardiovascular Surgery Patients (n ⫽ 987) CARDIOVASCULAR
Risk Factors Operation time: Total ⬍10 hr 10 hr ⬍ Aortic surgery ⬍10 hr 10 hr ⬍ Nonaortic surgery ⬍10 hr 10 hr ⬍ Preoperative hypertension Surgical procedures: CABG Valvular surgery Aortic surgery: Ascending & Arch Descending
Odds Ratio
95 % CI
Multivariate Analysis (p Value)
1, as reference 4.4
1.7–11.4
0.003
1, as reference 2.8
0.2–11.4
0.717
1, as reference 1.5 2.6
0.6–12.8 1.1–6.5
0.171 0.035
1, as reference 1.1 5.6 4.5
0.3–3.5 2.3–13.5 1.5–13.7
0.886 ⬍0.0001 0.007
8.7
2.2–34.1
0.002
CABG ⫽ coronary artery bypass grafting;
Ann Thorac Surg 2007;83:2147–52
surgery (p ⬍ 0.0001); in particular, the risk was 8.7 times higher in patients who underwent descending aortic surgery (p ⫽ 0.002), compared with those who underwent coronary artery bypass grafting. The risk was 2.6 times higher in patients with hypertension (p ⫽ 0.035) compared with those patients without that comorbidity. Ten patients with poor outcomes are summarized in Table 4. Of these 10 patients, the outcomes of five patients (Nos. 6 to 10) were “more critical” when associated with bilateral VCP, repeated airway treatment, or death within six months after surgery and the number of such patients was significantly (p ⫽ 0.016) higher in the aortic surgery group. All cases intubated for more than 100 hours were considered the more critical cases.
Comment
CI ⫽ confidence interval.
The results of the multivariate logistic regression analysis of the risk factors for VCP are summarized in Table 3. Among all surgical procedures, the risk of VCP increased with the operation time (for every minute: multivariate odds ratio, 1.003; 95% confidence interval, 1.001 to 1.005; p ⫽ 0.003), and the risk after a long operation of over 10 hours was 4.4 times higher (p ⫽ 0.003) compared with that of an operation of less than 10 hours, although the risk of VCP did not increase statistically with the operation time when the aortic and nonaortic surgery groups were analyzed separately. The risk was 5.6 times higher in patients who underwent aortic
The present study demonstrated that VCP after cardiovascular surgery occurred with a markedly higher incidence (2.3%) than that after general operations in an early investigation. Furthermore, we demonstrated that prolonged operation and aortic procedures increased the risk. Although it is well-known that VCP can result in hoarseness, stridor, poor cough (difficulty in expectoration), dysphagia, or aspiration, VCP is sometimes easily overlooked, particularly when the postoperative condition is critical [11]. Therefore, to recognize the potential risk factors for VCP, an appropriate risk analysis is considered to be very important. The etiologic mechanisms of postoperative VCP generally fall into three types, as follows: the first mechanism is recurrent laryngeal nerve paralysis (RNP), the second one is arytenoid dislocation, and the third is traumatic vocal cord injury. Furthermore, RNP is further subdivided into two patterns of direct injury and indirect injury. The mechanism of RNP due to tracheal intubation is usually indirect injury, neurapraxia without nerve degeneration [15]. The factors of indirect injury are said to include tracheal tube size, location of the cuff, fixing side of the tube, cuff pressure, curvature of tracheal tube, intubation time, etc [13, 16, 17].
Table 4. Clinical and Postoperative Data of 10 Cases of Vocal Cord Paralysis With Poor Outcome Patient 1 2 3 4 5 6 7 8 9 10
Age/Sex
Poor Outcomea
Surgery
Intubation Time (hr)
Side of VCP
Repeated Airway Treatment
29/F 36/M 60/M 69/M 61/F 81/M 77/M 69/M 78/F 20/M
Aspiration Aspiration pneumonia Aspiration Aspiration Aspiration Difficulty in expectoration Choking Difficulty in expectoration Aspiration pneumonia Aspiration
Heart Heart Heart Heart Heart Heart Aorta Aorta Aorta Aorta
13.7 53.8 54.9 29.3 42.7 195.0 624.0 45.5 504.0 160.6
L L L R L L B L L B
— — — — — Tracheostomy Tracheostomy Reintubation — —
a
Poor outcome is defined as the existence of aspiration or respiratory disturbance.
POD ⫽ postoperative day;
VCP ⫽ vocal cord paralysis.
Remarks
Died (48 POD) Died (134 POD) Persistent hoarseness
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
Among our patients, each VCP case used an approximately standard-sized tube, whereas a thicker doublelumen tube was used in six aortic cases (26.1%). Considering the fact that the number of cases of descending aortic replacement was only 2.3% overall, it is suggested that involvement of the double-lumen tube was significant. However, the double-lumen tube was changed to a standard tube just after the operation; therefore, it would not be the sole risk factor, but a risk factor combined with operation time. Furthermore, hyperextension of the neck often occurs during cardiovascular surgery. Therefore, a patient is likely to be in the situation where the compression of the recurrent laryngeal nerve occurs easily, because of displacement of the cuff and esophageal traction [10, 18]. The mechanisms of nonintubation-related indirect RNP include sternal traction, TEE use, and cooling. Excessive lateral sternal traction causes nerve compression by lateral and anterior stretching of the subclavian arteries [7]. The TEE probe is considered to produce nerve compression at the postcricoid area and to generate RNP [19, 20]. In the VCP cases, there were three cases of TEE use. Incidentally, this type of nerve injury is known as a complication of gastric tube placement that is defined as nasogastric tube syndrome [21, 22]. Core hypothermia for total circulatory arrest is also a characteristic of cardiovascular surgery, and includes moderate hypothermia for ordinary cardiopulmonary bypass, cold blood cardioplegia, and topical ice slush for myocardial protection. Cooling causes neurapraxia and contributes to the pathogenesis of RNP [11, 23, 24]. In cardiovascular surgery there is a high possibility that direct RNP may occur, especially as a result of surgical procedures. Specifically, it has been reported that this kind of injury is produced by central venous catheterization [25], sternotomy [7], harvesting of the internal thoracic artery [26], aortic manipulation [9], etc. In general, RNP is mostly found unilaterally, on the left side [13]. This tendency is more prominent in cardiovascular surgery because the left recurrent laryngeal nerve is vulnerable due to its anatomic length [7]. We were not able to find any description of accidental or unavoidable surgical nerve injury in the operation records. Macroscopically, we were easily able to distinguish traumatic vocal cord injury and neurological injury and determine the degree of paralysis. However, it was very difficult to diagnose what type of neurological injury had occurred. Therefore, we were not able to obtain any data about the incidence of each injurious mechanism. Eventually, we were only able to speculate from the clinical course whether the paralysis was curable or not. The symptom of indirect RNP is usually mild and disappears within several months. If symptoms persist for more than this period, it seems more likely that the direct injury was caused by the surgical procedure and that the damage will be permanent [11]. We could not confirm long-term outcomes of all VCP cases, but found that only one aortic patient presented with persistent hoarseness even three month later. This case may have been due to direct recurrent laryngeal nerve injury at the operating site.
2151
The VCP incidence was significantly higher in patients with preoperative hypertension, but not in those with preoperative diabetes mellitus. Several studies have suggested that this may be due to the use of cardiopulmonary bypass related with peripheral nerve injury, and the impaired microcirculation [24, 27]. Both hypertension and diabetes mellitus may generate VCP because they contribute to impaired microcirculation as the main pathophysiologic feature. Further investigation is required to further examine this point. In this study, the incidence and the risk of postoperative VCP increased with prolonged operation. Generally, the prolongation of surgery can increase the chance of direct recurrent laryngeal nerve injury with time. Furthermore, we consider that prolonged surgical cases are intubated longer postoperatively because of the persistence of a critical state. Recurrent laryngeal nerve and laryngeal mucosa injury can occur as a result of the presence of the tube, and insufficiency of the vocal cord will happen [15, 28]. While surgical time was not the sole risk factor in our study, we suppose that total intubation time could potentially be a considerable risk factor for VCP. Our results were well-matched with a previous report that demonstrated that aortic surgery patients had higher incidence and risk of VCP compared with cardiac surgery patients [9]. The duration of surgery, intubation, and cardiopulmonary bypass were significantly longer in the aortic surgery patients compared with the nonaortic surgery patients. Furthermore, there were significant differences in the aortic surgery group among the items of emergency operation, minimum core temperature, and the use of a double-lumen tube. In addition to the time factor, the use of a double-lumen tube and the cooling are all mentioned as the background of VCP, and these factors confirm the high incidence of VCP after aortic surgery. Aspiration and repeated airway treatment did not exhibit statistical significance as the items associated with the severity of VCP between the aortic and nonaortic groups, likely due to the small number of cases. However, two cases of bilateral VCP and two cases that resulted in death within six months after operation were all aortic surgery patients and the ICU stay was significantly longer in the aortic surgery group. Therefore, there is a marked tendency toward an increase in severity of VCP after aortic surgery. Ishimoto and colleagues [9] reported that the incidence of VCP was most common after operation on type I aneurysms (64%, ascending and arch), according to the classification of de Bakey and colleagues, among all the types of thoracic aortic aneurysms. In our study, descending aortic replacement exhibited a higher incidence and doubled the risk of VCP compared with that of the ascending aorta and arch. No significant differences were observed for any items we examined between the two surgical procedures and we consider that the small number of cases of VCP makes it more difficult to compare the degree of risk for each. In theory, because of the anatomic length of the left recurrent laryngeal nerve, it would appear that direct injury could occur more frequently during surgery with manipulation of the aortic arch. Further investigation of this point is warranted.
CARDIOVASCULAR
Ann Thorac Surg 2007;83:2147–52
2152
ITAGAKI ET AL VOCAL CORD PARALYSIS AFTER CARDIOVASCULAR SURGERY
CARDIOVASCULAR
In the present study, the largest severity-related factor of VCP is the duration of intubation. When it is anticipated that the postoperative intubation time will be longer due to a serious condition, tracheostomy should be considered because it would be effective for both respiratory assistance as well as vocal cord protection [11]. As a matter of course, if any abnormalities are recognized after extubation it is important to consider the appropriate therapeutic treatment as well as prompt consultation with an otorhinolaryngologist. Additionally, it can be more difficult to differentiate between RNP and arytenoid dislocation. If arytenoid dislocation exists, surgical treatment is sometimes required and delayed diagnosis leads to a complicated situation where ankylosis of cricoarytenoid articulation makes repositioning more difficult [29]. For this reason, expert diagnosis is also necessary. There are some limitations of this retrospective cohort study that should be mentioned. First, because our investigation focused on patients in whom postoperative VCP after dysphonia was diagnosed, we did not investigate how many patients suffered transient or asymptomatic vocal cord dysfunction without paralysis. We also did not identify patients who had potential preexisting vocal cord pathology, such as smokers or those with gastroesophageal reflux disease. These patients would be at risk for postoperative vocal cord morbidities. It is likely that our study underestimated patients with postoperative vocal cord morbidities, and thus the relationship between preexisting vocal cord pathology and the risk of postoperative VCP was unclear. Second, we were not able to systematically assess some interesting variables, such as body mass index, total intubation time, the difficulty of tracheal intubation, the use of intraoperative TEE, double-lumen tube, or emergent surgery because of the large demographic number. Therefore, the relationship between those factors and postoperative VCP was examined only among the VCP cases. Third, we did not analyze the relation between the incidence of postoperative VCP and the length of ICU stay or perioperative mortality and long-term morbidity as a comprehensive outcome. Therefore, the significance of VCP on the clinical outcome is unclear. Finally, we do not know the incidence of VCP after general surgery in our institute for comparison. In the future, if these problems are resolved, we will obtain more objective and reliable incidence and risk information. In conclusion, this study demonstrated that aortic procedures and prolonged operation increased the risk of VCP, and severe VCP tended to be associated with aortic surgery and intubation for more than 100 hours. Vocal cord paralysis after cardiovascular surgery occasionally presents with life-threatening complications because it may generate aspiration or airway obstruction due to loss of the laryngeal reflex. We should provide a more detailed preoperative explanation to the cardiovascular surgery patients regarding the potential risk for VCP.
References 1. Jones MW, Catling S, Evans E, Green DH, Green JR. Hoarseness after tracheal intubation. Anaesthesia 1992;47:213– 6.
Ann Thorac Surg 2007;83:2147–52
2. Domino KB, Posner KL, Caplan RA, Cheney FW. Airway injury during anesthesia: a closed claims analysis. Anesthesiology 1999;91:1703–11. 3. Mehta S. Safer endotracheal intubation. Lancet 1986;1:1148. 4. Laryngeal paralysis after endotracheal intubation. Lancet 1986;1:536-7. 5. Holley HS, Gildea JE. Vocal cord paralysis after tracheal intubation. JAMA 1971;215:281– 4. 6. David DS, Shah M. Vocal cord paralysis following intubation. JAMA 1971;216:1645– 6. 7. Hamdan AL, Moukarbel RV, Farhat F, Obeid M. Vocal cord paralysis after open-heart surgery. Eur J Cardiothorac Surg 2002;21:671– 4. 8. Dimarakis I, Protopapas AD. Vocal cord palsy as a complication of adult cardiac surgery: surgical correlations and analysis. Eur J Cardiothorac Surg 2004;26:773–5. 9. Ishimoto S, Ito K, Toyama M, et al. Vocal cord paralysis after surgery for thoracic aortic aneurysm. Chest 2002;121:1911–5. 10. Inada T, Fujise K, Shingu K. Hoarseness after cardiac surgery. J Cardiovasc Surg (Torino) 1998;39:455–9. 11. Shafei H, el-Kholy A, Azmy S, Ebrahim M, al-Ebrahim K. Vocal cord dysfunction after cardiac surgery: an overlooked complication. Eur J Cardiothorac Surg 1997;11:564 – 6. 12. Lesser T, Williams G. Laryngographic investigation of postoperative hoarseness. Clin Otolaryngol Allied Sci 1988;13:37– 42. 13. Cavo JW Jr. True vocal cord paralysis following intubation. Laryngoscope 1985;95:1352–9. 14. Gibbin KP, Egginton MJ. Bilateral vocal cord paralysis following endotracheal intubation. Br J Anaesth 1981;53:1091–2. 15. Santos PM, Afrassiabi A, Weymuller EA Jr. Risk factors associated with prolonged intubation and laryngeal injury. Otolaryngol Head Neck Surg 1994;111:453–9. 16. Mandoe H, Nikolajsen L, Lintrup U, Jepsen D, Molgaard J. Sore throat after endotracheal intubation. Anesth Analg 1992;74:897–900. 17. Stout DM, Bishop MJ, Dwersteg JF, Cullen BF. Correlation of endotracheal tube size with sore throat and hoarseness following general anesthesia. Anesthesiology 1987;67:419 –21. 18. Hartrey R, Kestin IG. Movement of oral and nasal tracheal tubes as a result of changes in head and neck position. Anaesthesia 1995;50:682–7. 19. Zwetsch G, Filipovic M, Skarvan K, Todorov A, Seeberger MD. Transient recurrent laryngeal nerve palsy after failed placement of a transesophageal echocardiographic probe in an anesthetized patient. Anesth Analg 2001;92:1422–3. 20. Kawahito S, Kitahata H, Kimura H, Tanaka K, Oshita S. Recurrent laryngeal nerve palsy after cardiovascular surgery: relationship to the placement of a transesophageal echocardiographic probe. J Cardiothorac Vasc Anesth 1999; 13:528 –31. 21. Sofferman RA, Haisch CE, Kirchner JA, Hardin NJ. The nasogastric tube syndrome. Laryngoscope 1990;100:962– 8. 22. Ellis PD, Pallister WK. Recurrent laryngeal nerve palsy and endotracheal intubation. J Laryngol Otol 1975;89:823– 6. 23. Tewari P, Aggarwal SK. Combined left-sided recurrent laryngeal and phrenic nerve palsy after coronary artery operation. Ann Thorac Surg 1996;61:1721–3. 24. Lederman RJ, Breuer AC, Hanson MR, et al. Peripheral nervous system complications of coronary artery bypass graft surgery. Ann Neurol 1982;12:297–301. 25. Sim DW, Robertson MR. Right vocal cord paralysis after internal jugular vein cannulation. J Laryngol Otol 1989;103:424. 26. Phillips TG, Green GE. Left recurrent laryngeal nerve injury following internal mammary artery bypass. Ann Thorac Surg 1987;43:440. 27. Grocott HP, Clark JA, Homi HM, Sharma A. “Other” neurologic complications after cardiac surgery. Semin Cardiothorac Vasc Anesth 2004;8:213–26. 28. Lundy DS, Casiano RR, Shatz D, Reisberg M, Xue JW. Laryngeal injuries after short- versus long-term intubation. J Voice 1998;12:360 –5. 29. Sataloff RT, Bough ID Jr, Spiegel JR. Arytenoid dislocation: diagnosis and treatment. Laryngoscope 1994;104:1353– 61.