Secondary Logo

Journal Logo

Airway management

Left double-lumen tube with or without a carinal hook

A randomised controlled trial

Dumans-Nizard, Virginie; Parquin, Jean-François; Moyer, Jean-Denis; Dreyfus, Jean François; Fischler, Marc; Le Guen, Morgan

Author Information
European Journal of Anaesthesiology (EJA): June 2015 - Volume 32 - Issue 6 - p 418-424
doi: 10.1097/EJA.0000000000000201
  • Free



Anaesthesiologists have a large choice of double-lumen tubes (DLTs) and endobronchial blockers to provide lung isolation and one-lung ventilation.1 Comparison between these devices gives various results.2–8 The history of DLTs began in 1949 when Carlens9 described a double-cuffed, DLT intended for intubation of the left main bronchus for differential bronchospirometry. Historically, this tube had a hook to engage the carina, helping ‘blind’ placement. The presence of a hook was probably at the origin of some of the cases of tracheal or bronchial rupture reported by Fitzmaurice and Brodsky in 1999.10 Among them, the case reported by Pollack et al.11 was particularly striking, with an obstruction of the tracheal tube orifice of a DLT by the carinal hook. They explained their case by the fact that the hook was bent back, a feature occasionally seen during fibreoptic bronchoscopy (FOB). Rocha et al.12 reported a fracture of the hook during intubation.

Disposable polyvinylchloride DLTs have been in clinical use since the early 1980s with and without a carinal hook, the latter being the more frequently used in Western countries. However, some teams or individuals have continued to use tubes with a hook and a recent study has shown that a retractable carinal hook, that is a modified gum elastic bougie, increases the success rate and shortens the time for correct DLT positioning.13

As commercially available left-sided DLTs provided with or without a carinal hook have never been assessed against each other, we designed a study to compare them, particularly in terms of the time required to the correct placement of the tube, a surrogate for the rate of adequate tube position at the first attempt. Left-sided DLTs were chosen because of their much greater positioning margin of safety, as described by Benumof et al.14

Materials and methods

This single-centre, controlled, randomised, single-blinded study was approved by a regional independent Ethical Committee (Comité de Protection des Personnes Ile de France VIII, No. CPP 09 06 47, 2009), and the French Drug Agency (Agence Française de Sécurité Sanitaire des Produits de Santé). Each patient received detailed oral and written information and gave final written consent during the preanaesthesia visit. The study protocol was registered under Clinical Trial number NCT00969683.

Patients were screened from the operation planning list when the anaesthetic research team was on duty. We enrolled American Society of Anesthesiologists’ physical status 1 to 3 patients, aged 18 years or older, scheduled for lung surgery requiring a left-side DLT for one-lung ventilation. Patients with a Mallampati score of at least 3, a criterion for an expected difficult intubation, with known lesions of the proximal part of the left bronchus or at risk of pulmonary aspiration were not eligible for the study.

Recruited patients were allocated 1 : 1 into two groups (No hook group and Hook group) using a computer-generated randomisation. Codes were kept in numbered envelopes. An independent data safety monitoring board was not instituted because these are low-risk devices. Patients and surgeons were blinded with regard to the device used.

All patients received hydroxyzine orally for premedication. In the operating room, they received standard monitoring and bispectral index monitoring. Patients scheduled for thoracotomy received an epidural anaesthetic at the T6 to T7 level using a mixed solution with levobupivacaine 0.125% and sufentanil 0.25 μg ml−1, started before surgical incision. After 3 min of preoxygenation, patients received propofol 1.5 to 3 mg kg−1, sufentanil 0.2 μg kg−1 and atracurium 0.5 mg kg−1. Anaesthesia was maintained with sevoflurane, sufentanil and atracurium.

Two anaesthetists with more than 5 years experience in thoracic surgery with both types of single-use polyvinylchloride DLT (Broncho-Cath, Mallinckrodt France, Les Ulis, France) and use of an FOB, managed the patients. The size of the device was chosen on the basis of height and sex as follows: women less than 160 cm, 35FG; women more than 160 cm, 37FG; men less than 170 cm, 39FG; men more than 170 cm, 41FG. The stiletted DLT was introduced into the glottis using direct laryngoscopy. The classification by Cormack and Lehane15 was used to quantify the view at laryngoscopy. After the bronchial cuff passed the vocal cords, the stilette was removed and the tube was rotated 90° towards the left. The tube was advanced until slight resistance was encountered. Then, a bronchoscope was inserted through the tracheal tube to verify the position of the DLT. Three abnormalities led to a change in the position of the tube: intubation of the right main bronchus, too proximal a position (bronchial cuff herniation with obstructed view of the tracheal carina) and too distal a position (no visualisation of the proximal edge of the uninflated bronchial cuff below the tracheal carina, obstructed view of the left distal bronchial tree). Once the DLT was in the correct position, the cuffs were inflated and ventilation of the lungs began. After the patient was turned to the lateral position, a systematically performed fibreoptic examination assessed the correct position of the DLT. The lumen of the DLT to the nondependent lung was clamped and its port was opened to atmosphere, leading to passive lung collapse. At the end of the surgical procedure, the surgeon, blinded to the technique, scored the quality of lung collapse [good (complete collapse or some residual air), or poor (residual air interfering with surgical exposure or no collapse)]. At the end of the anaesthetic procedure, the anaesthetist scored the ease of insertion of the DLT and placement in the correct position (easy, difficult or very difficult).

Intraoperative and postoperative management were performed according to our routine protocol; in particular, tracheal extubation was performed in the operating room. Two hours after the end of surgery and on the morning of postoperative day 1, while the patients were still in the postanaesthesia care unit, an investigator not involved in the study and blinded to group assignment asked the patient about sore throat. The evaluation was performed with a verbal numerical scale with 0 being ’no pain’ and 10 being ’the most intense pain imaginable’. Sore throat was noted and considered as clinically significant if the score was more than 3. On day 1, FOB was performed under local anaesthesia, with a target-controlled remifentanil infusion to improve patient tolerance, by a pulmonologist unaware of the device used. The pulmonologist classified the appearance as normal, minor lesions (redness) or major lesions (laceration, rupture).

The primary outcome of the study was the time required to position the tube correctly. It was measured by a nurse who started a stopwatch when the tube was introduced into the mouth and stopped it when correct placement in the supine position was confirmed; this period included the time spent for correction of a malposition detected by bronchoscopy. Secondary outcome measures were the incidence of correct tube position at the first attempt, the types of malposition when the patient was in dorsal decubitus and lateral positions, the number of additional laryngoscopy and FOB examinations required, the quality of lung collapse, the ease of placement of the DLT and the incidences of sore throat and airway lesions.

Statistical analysis

The average time to place a DLT adequately with a carinal hook was 179 ± 118 s in a previous study performed by our team.6 We hypothesised that intubation would be faster when a hook is present, with an expected difference of 60 s. The number of patients required was 83 per group with an α risk of 0.05 and a β risk of 90%. This number was increased to 92 to allow for patients dropping out.

Numbers (%) for categorical variables were compared using Fisher's exact test. Normally distributed continuous variables (Shapiro–Wilk test, P ≥ 0.05), expressed as mean ± SD (range), were compared by means of a t-test. Nonnormally distributed continuous variables (Shapiro–Wilk test, P < 0.05), expressed as median [interquartile range, IQR] (range), were compared by Mann–Whitney rank sum test. Probability values less than 0.05 were considered significant. Data analysis was performed using SPSS version 11.0 (SPSS Science Inc., Chicago, Illinois, USA).


During the period from November 2009 to April 2013, 184 patients were included in the study. A DLT with a hook was used in 92 cases, and a DLT without a hook in the other cases. Postoperative FOB was conducted in 63 patients in the Hook group and 65 patients in the No hook group (Fig. 1). Demographics and operative data of the two groups are summarised in Table 1.

Fig. 1
Fig. 1:
Trial profile. DLT, double-lumen tube; FOB, fibreoptic bronchoscopy.
Table 1
Table 1:
Demographic data and surgical data

Data related to tracheal intubation are summarised in Table 2. The correct placement of a DLT without a hook took a median of 81.0 [50.0 to 146.2] (19 to 600) s and placement of a DLT with a hook took a median of 67.5 [45.0 to 138.7] (15 to 780) s (P = 0.43). The longest times for intubation were with the hooked tubes (Fig. 2). Only 68.5% of DLTs without a hook and 69.6% of DLTs with a hook were placed correctly without the need for repositioning the device after fibreoptic inspection; in the other cases, abnormal locations occurred at similar incidences in both groups. The numbers of laryngeal (P = 0.97) and oesophageal intubations (P = 0.28) were similar (Table 2).

Table 2
Table 2:
Data related to tracheal intubation
Fig. 2
Fig. 2:
Time to correct placement of a double-lumen tube. Data are presented as individual values and box-plot representations. ○, DLT without a carinal hook; ●, DLT with a carinal hook.

After the change to the lateral decubitus position, 78.3% of DLTs without a hook and 85.9% of DLTs with a hook were placed correctly (P = 0.59). The number of supplementary fibreoptic examinations during surgery was similar in the two groups (P = 0.22). The quality of pulmonary collapse was similar (P = 0.68) (Table 2).

Anaesthetists rated the placement of the DLT as very difficult in 5.5% in the No hook group and in 9.9% in the hook group, ratings according to the three classes being similar (P = 0.46) (Table 2).

Similar numbers of patients suffered from sore throat in each group at day 0 (P = 0.80) and day 1 (P = 0.20) (Table 3). No major lesion was found among patients having a fibreoptic examination on day 1. Numerous patients had minor lesions of the vocal cords (39.1% in the No hook group and 38.1% in the Hook group, P = 0.96) and of the trachea (59.4% in the No hook group and 50.8% in the Hook group, P = 0.37); all were minor (erythema) (Table 3).

Table 3
Table 3:
Postoperative data


Our study showed that the use of DLTs with or without a hook gave similar results especially for the primary outcome, time to obtain correct placement, with a difference in median times of 13.5 s between the two groups. An almost identical total of two-thirds of DLTs were correctly placed at the first attempt. Secondary outcomes were also similar, particularly lung collapse, postoperative sore throat and airway complications.

Our study can be compared with a recent one published by Mourisse et al.8 They compared the EZ-Blocker (Teleflex Life Sciences Ltd., Athlone, Ireland) and a left-sided DLT (Broncho-cath; Mallinckrodt, Athlone, Ireland); their methods were quite similar to ours, but they chose the frequency of initial malposition as the primary outcome. They calculated the weighted mean of the reported incidence of malpositioning of left-sided DLTs in several studies to be 53%, each ranging from 10 to 83%.7,16–20 Mourisse et al.8 reported a low rate of success (15%) and explained this result by the ‘large’ number of enrolled patients (50 in their DLT group, although each of our groups contained 92 patients) and by a stringent definition of device malposition.8 This last point is difficult to discuss because, even with exactly the same definition, there is obviously an individual interpretation. Technology now being more user-friendly, we regret that video was not used in our study to permit analysis of this issue by independent assessors.

We decided to use the criterion ‘time to obtain correct placement’ as our primary outcome measure. Time is simple to measure and is related to the number of adequate tube positions. Initial malposition requires repositioning under fibreoptic control and thus a longer time before obtaining an adequate position of the DLT. However, time to obtain correct placement is also difficult to compare between studies because of various definitions: from introduction of the tube into the mouth until correct placement including potential correction of a malposition detected by bronchoscopy in our study, and, for example, from visualisation of the vocal cords until initial positioning without bronchoscopy.8 Finally, we reviewed all charts to look at possible explanations for a prolonged time to obtain correct placement (values greater than the 90th percentile), such as an upper airway or tracheobronchial tree abnormality on preoperative imaging, but found none.

There are three types of malposition: in the right bronchus; too deep; or too shallow in the left bronchus. Initial position in the right bronchus and too deep insertion are probably due to the choice of DLT with an external diameter too small, although shallow insertion is probably due to a DLT with an external diameter too large. Other studies have reported the same problems, with more ‘too deep’ than ‘too shallow’ intubations.18–20 Mourisse et al.8 reported similar results, although they chose more sophisticated methods to determine the size of their DLT (formula using the tracheal width at the level of the clavicles measured on the preoperative anteroposterior chest radiograph21 and direct measurement using computed tomography).

Another important issue is what happened after turning the patient to the lateral position and during surgery. We failed to show a difference between the number of DLTs that dislodge after turning the patient to the lateral position (21.7% in the No hook group vs. 14.1% in the Hook group, P = 0.25) or between the number of fibreoptic bronchoscopies performed in addition to the two in the protocol. These supplementary examinations permitted intraoperative corrections of the position of the DLT and satisfactory surgical conditions, as reported by a very high satisfaction index, which was similar in the two groups.

The relatively surprisingly low rate of complete success at the first attempt, whatever the DLT, and the number of displacements during patient positioning, are two strong arguments to promote FOB in every case. This method of assessment remains a matter for debate. Brodsky22 stated that an experienced operator increases the likelihood of success and consequently that FOB is not always needed for left DLT placement. This was not confirmed in our study despite the presence of highly experienced operators. One major point is tube misplacement especially after turning to the lateral decubitus position and the absence of bronchoscopy could lead to undetected events, such as blood spillage or atelectasis, with an increase in postoperative morbidity. These events could be simply prevented by the routine use of fibreoptic control. We can only reiterate the conclusion of Pennefather and Russell23 in their editorial published in 2000: ‘Issues of resources, training and clinical experience must no longer deflect both the large body of scientific evidence and the simple logic of visually confirming placement of every DLT. For how much longer will it be considered defensible to ignore these facts?’ Such a conclusion has been repeated more recently by Campos.24

We found a very similar incidence of sore throat in both groups, a figure slightly greater than that reported by Mourisse et al.8 in their DLT group (13% moderate and severe postoperative sore throat). In the present study, no serious complications (major glottic or tracheobronchial lesions) were reported in either group. We explain these results by the fact that severe lesions have a very low incidence; our study was not powered to find a difference in terms of severe lesions. However the rate of minor lesions in our study is clinically relevant and similar to that found in the literature.8 The lack of difference between the two groups could be explained by the fact that the lesions induced by DLT depend mostly on the size of the DLT, the role of the hook probably being minimal. We found (6.5% in both groups) some lesions of the right bronchus in our study, which could have been induced by FOB, by the rotation of the DLT or by the tip of a suction catheter.

A limitation of our study is that we had an incomplete evaluation in the postoperative period due to organisational difficulty (failure in reported sore throat rates). In addition, numerous patients refused the postoperative FOB examination on postoperative day 1, despite information about this procedure being presented prior to obtaining consent. Moreover, we did not follow the patients up after they returned to the surgical ward and thus we are unaware of the long-term outcome.

Another point is that we calculated the number of patients to be included from the results of a previous study performed by our team.6 Not surprisingly, the time to correct placement decreased over the years, but the huge difference from a mean value of 179 s (the reference) to 81 s (the observed value in this study) can only be explained by the increased experience of the anaesthetists who participated in the present study. We should also mention that no right pneumonectomy was performed during the study period. In such a case, manual or mechanical suture of the right bronchus may require deflation of the bronchial cuff and a partial withdrawal of the left DLT if it has a hook.

In conclusion, this large study did not show any difference between DLTs with or without a hook. The importance of fibreoptic control of the position of the tube has once more been demonstrated.

Acknowledgements relating to this article

Assistance with the study: none.

Financial support and sponsorship: this work was supported by the Department of Anaesthesiology, Hôpital Foch, 40 rue Worth, Suresnes, France.

Conflicts of interest: none.

Presentation: none.


1. Cohen E. Back to blockers? The continued search for the ideal endobronchial blocker. Anesthesiology 2013; 118:490–493.
2. Bauer C, Winter C, Hentz JG, et al. Bronchial blocker compared to double-lumen tube for one-lung ventilation during thoracoscopy. Acta Anaesthesiol Scand 2001; 45:250–254.
3. Ender J, Bury AM, Raumanns J, et al. The use of a bronchial blocker compared with a double-lumen tube for single-lung ventilation during minimally invasive direct coronary artery bypass surgery. J Cardiothorac Vasc Anesth 2002; 16:452–455.
4. Campos JH, Hallam EA, Van Natta T, Kernstine KH. Devices for lung isolation used by anesthesiologists with limited thoracic experience: comparison of double-lumen endotracheal tube, Univent torque control blocker, and Arndt wire-guided endobronchial blocker. Anesthesiology 2006; 104:261–266.
5. Knoll H, Ziegeler S, Schreiber JU, et al. Airway injuries after one-lung ventilation: a comparison between double-lumen tube and endobronchial blocker: a randomized, prospective, controlled trial. Anesthesiology 2006; 105:471–477.
6. Dumans-Nizard V, Liu N, Laloe PA, Fischler M. A comparison of the deflecting-tip bronchial blocker with a wire-guided blocker or left-sided double-lumen tube. J Cardiothorac Vasc Anesth 2009; 23:501–505.
7. Ruetzler K, Grubhofer G, Schmid W, et al. Randomized clinical trial comparing double-lumen tube and EZ-Blocker for single-lung ventilation. Br J Anaesth 2011; 106:896–902.
8. Mourisse J, Liesveld J, Verhagen A, et al. Efficiency, efficacy, and safety of EZ-blocker compared with left-sided double-lumen tube for one-lung ventilation. Anesthesiology 2013; 118:550–561.
9. Carlens E. A new flexible double-lumen catheter for bronchospirometry. J Thorac Surg 1949; 18:742–746.
10. Fitzmaurice BG, Brodsky JB. Airway rupture from double-lumen tubes. J Cardiothorac Vasc Anesth 1999; 13:322–329.
11. Pollak Y, Kogan A, Grunwald Z. Double-lumen tube malfunction caused by the carinal hook. Anesthesiology 1995; 83:639.
12. Rocha AC, Martins MG, Silva LI, Nunes JM. Detachment of the carinal hook following endobronchial intubation with a double lumen tube. BMC Anesthesiol 2011; 11:20.
13. Al-Metwalli RR, Mowafi HA, Ismail SA. Double-lumen tube placement using a retractable carinal hook: a preliminary report. Anesth Analg 2009; 109:447–450.
14. Benumof JL, Partridge BL, Salvatierra C, Keating J. Margin of safety in positioning modern double-lumen endotracheal tubes. Anesthesiology 1987; 67:729–738.
15. Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anaesthesia 1984; 39:1105–1111.
16. Smith GB, Hirsch NP, Ehrenwerth J. Placement of double-lumen endobronchial tubes. Correlation between clinical impressions and bronchoscopic findings. Br J Anaesth 1986; 58:1317–1320.
17. Alliaume B, Coddens J, Deloof T. Reliability of auscultation in positioning of double-lumen endobronchial tubes. Can J Anaesth 1992; 39:687–690.
18. Lewis JW Jr, Serwin JP, Gabriel FS, et al. The utility of a double-lumen tube for one-lung ventilation in a variety of noncardiac thoracic surgical procedures. J Cardiothorac Vasc Anesth 1992; 6:705–710.
19. Lieberman D, Littleford J, Horan T, Unruh H. Placement of left double-lumen endobronchial tubes with or without a stylet. Can J Anaesth 1996; 43:238–242.
20. Klein U, Karzai W, Bloos F, et al. Role of fiberoptic bronchoscopy in conjunction with the use of double-lumen tubes for thoracic anesthesia: a prospective study. Anesthesiology 1998; 88:346–350.
21. Brodsky JB, Lemmens HJ. Tracheal width and left double-lumen tube size: a formula to estimate left-bronchial width. J Clin Anesth 2005; 17:267–270.
22. Brodsky JB. Fiberoptic bronchoscopy need not be a routine part of double-lumen tube placement. Curr Opin Anaesthesiol 2004; 17:7–11.
23. Pennefather SH, Russell GN. Placement of double lumen tubes – time to shed light on an old problem. Br J Anaesth 2000; 84:308–310.
24. Campos JH. Update on tracheobronchial anatomy and flexible fiberoptic bronchoscopy in thoracic anesthesia. Curr Opin Anaesthesiol 2009; 22:4–10.
© 2015 European Society of Anaesthesiology