Using radiological measurements of the airway may be a useful and objective method of choosing the correct size of double-lumen endobronchial tubes (DLTs) . Objective selection criteria based on the measured tracheal width from a chest radiograph have been successfully used . This selection method, however, seems to be more accurate when selecting larger sizes of DLT for an individual patient . The correct size of the DLT, however, was not reliably predicted in some Asian patients, who may require a smaller DLT . A computer tomographic (CT) scan can measure the width of the left mainstem bronchus before selecting the DLT size . Hence, we investigated this method of selecting the DLT in predicting the correct DLT sizes in our Asian patients.
Fifty-two consecutive Asian-born adults aged 17-80 yr who required a DLT during anesthesia for elective thoracic surgery were entered into the study, which was approved by our hospital ethics committee. Informed consent was obtained from all participants. The inclusion criteria included patients who already had a preoperative chest CT scan that was performed by using a spiral scanner in our institution. The axial scans of the thorax were obtained with a slice thickness of 7-10 mm with a pitch of 1 or 1.2.
The patients were identified the day before surgery. Their chest CT scans were reviewed again by the radiologist. Measurements of the left main bronchus were made at the plane one slice (7-10 mm) below the slice that showed the carina and where the left main bronchus could be seen as a singular structure. The internal diameter of the left main bronchus was measured perpendicular to the portion of the bronchus that was most parallel. The measurements were similar to those made by Hannallah et al. . However, our measurements were made on a selected slice displayed in soft tissue windows and magnified 2 times. It was also performed with the electronic calipers of the spiral scanner to the nearest millimeter. These measurements were then made known to the anesthesiologist on the day of the operation.
The size of the left-sided BronchoCath[registered sign] DLT (Mallinckrodt, St. Louis, MO) used for the patients was selected based on the internal diameters of the patients' left main bronchi. Our initial protocol did not include the 32F DLT, which was unavailable to us. However, soon after our study began, we encountered two patients whose bronchial diameter measured 9 mm, into which the bronchial end of the selected 35F DLTs could not pass. In both patients, a 6.5-mm Univent[registered sign] bronchial-blocker tube (Fuji System Corp., Tokyo, Japan) was used successfully. These two patients were not included in the data. Since then, we have acquired the 32F DLT and modified our protocol to include the use of a 32F DLT for patients with bronchial diameters <10 mm. Table 1 summarizes the estimated diameters of the bronchial end of various left-sided BronchoCath[registered sign] DLT sizes and our protocol for selection of the DLT sizes.
The tracheas of all patients were intubated by anesthesiologists who had at least 5 yr experience using a DLT. If the bronchial end of the DLT failed to enter the left bronchus after two attempts, the next smaller size DLT would be chosen. All DLT placements were confirmed and positioned by using fiberoptic bronchoscopy. The proximal end of the bronchial cuff of the DLT was positioned just beyond the carina; the orifice of the left upper lobe bronchus should not be obstructed in this position. Lung isolation was confirmed using a positive pressure technique similar to that used by Hannallah et al. , in which the bronchial side of the DLT was pressured to 30 cm H2 O while the tracheal side was connected to a 1-cm underwater seal. The bronchial cuff was then inflated in increments of 0.5 mL until no more air escaped from the underwater seal. The DLT size was deemed appropriate if the volume of air in the bronchial cuff required for isolation was >0 but less than the recommended resting volume . DLTs that had no air leak with a deflated bronchial cuff or that required a bronchial cuff volume more than the recommended resting volume were considered oversized or undersized, respectively.
Descriptive statistics was used. An unpaired t-test was used to compare the means of the bronchial diameters of the male and female patients. One-way analysis of variance, followed by Student-Newman-Keuls test for post hoc comparison, was used to compare means of the height, weight, and age among the groups of correctly selected DLT sizes. P < 0.05 was considered significant.
Thirty-two male and 18 female patients participated in the study. Of the 50 patients, 34 (68%) were predicted to require smaller DLT (37F or smaller). Based on the CT scan measurements, the mean bronchial diameter of our female patients was significantly smaller (P < 0.01) than that of our male patients (9.6 +/- 1.0 vs 11.6 +/- 1.4 mm).
The tracheas of all patients were successfully intubated with the predicted DLT sizes. However, there were 12 patients (24%) who received an oversized DLT (Table 2). Six of these patients were predicted to receive a 35F DLT. Among these were two patients whose 35F DLTs were intubated into the wrong bronchus in the first attempt. Their tracheas were successfully reintubated with the same size DLT in the second attempt with their heads turned to the right side after the bronchial cuff had passed the vocal cords.
Six female patients were correctly predicted to receive a 32F DLT. The diameter of their left main bronchi was 9 mm. All six patients were significantly shorter and weighed less than most of the other patients (Table 3). In one patient whose left main bronchus diameter was 8 mm, even a 32F DLT was oversized, as good lung isolation was achieved with a deflated bronchial cuff.
A major difficulty encountered when inserting the left DLT is the inability to place the bronchial end of the DLT correctly into the left main bronchus . This task would be made easier if the correct size DLT was selected, as these tubes must fit appropriately into an orifice of a given size .
To select the appropriate size DLT, the external diameter of its bronchial end must be just smaller than the measured diameter of the left main bronchus of the patient. Although the diameter of the left main bronchus can be directly measured in the chest radiograph, it can only be seen 50% of the time . However, the tracheal diameter, which bears a correlation to the diameter of the left main bronchus, is more easily visualized from the chest radiograph and has been used to select the size of the DLT . However, this method of selecting the DLT is not very accurate in selecting smaller DLT sizes .
The CT scan of the thorax, however, allowed us to clearly visualize and measure the diameter of the left main bronchus. Hannallah et al.  used fine calipers to measure the diameter of the left main bronchus from a CT film, but this requires some training and may not have been accurate enough, especially when the choice of the DLT size was based on a difference of approximately 1 mm. However, the measurements made by our radiology colleagues were performed with electronic calipers of the spiral scanner on an enlarged CT scan and thus were more accurate.
Our overall positive predictive values of 61.1% for the female patients seems better than the 45.5% reported in our earlier work, in which the DLT sizes were chosen based on measurements of the chest radiographs . This was because we were able to correctly predict the small 32F DLT size in all patients who required them (and who were all female). Indeed, we found that some of our smaller Asian patients who were <160 cm tall required a DLT size smaller than 35F, which is normally the smallest DLT recommended for adults.
There were, however, a few problems that we faced in our study. First, an appropriate calculated allowance (difference between the internal diameter of the left main bronchus and the corresponding external diameter of the bronchial end of the deflated DLT) is important when selecting a DLT. One that is too small will increase the chances of selecting an oversized DLT, which may result in failure to insert the DLT into the bronchus or rupture of the bronchus . However, a calculated allowance that is too large (>2 mm) may result in an inappropriately small DLT being used, which carries a risk of misplacement  and injury to the airway . Unfortunately, it was difficult to ensure that the desired calculated allowance was present in each case in our study because not all the bronchial ends of the DLT of the same size had exactly the same external diameter (they may vary by 0.4 mm), and our CT scan measurements were accurate only to the closest millimeter. This may explain why some patients who were predicted to receive the 35F or 37F DLT, in which the calculated allowance was the smallest, received an oversized DLT (Table 1).
Second, a limitation of our method is that it required the patients to have a preoperative CT thorax scan. Fortunately, most of the patients who presented for thoracic surgery in our institution (except for patients undergoing thoracoscopic sympathectomies for palmar hyperhidrosis and thoracoscopic pleurodesis for recurrent spontaneous pneumothoraces) had already had a preoperative CT thorax scan performed as part of their preoperative evaluation.
Finally, our method of DLT selection requires the assistance of the radiology department. Although it did not incur much additional time and work for our colleagues from radiology, these routine sophisticated measurements may not be practical for all patients undergoing thoracotomy. Perhaps only smaller patients or those with smaller airway anatomy, identified through chest radiograph, may benefit most from these accurate measurements of the chest CT scans.
We thank Mr. T. Agastian and radiographers from the Departments of Radiology in Singapore General Hospital and Tan Tock Seng Hospital, Singapore.
1. Slinger P. Choosing the appropriate double-lumen tube: a glimmer of science comes to a dark art. J Cardiothorac Vas Anesth 1995;9:117-8.
2. Brodsky JB, Macario A, Mark JBD. Tracheal diameter predicts double lumen tube size: a method for selecting left double lumen tubes. Anesth Analg 1996;82:861-4.
3. Brodsky JB, Mackey S, Cannon WB. Selecting the correct size double-lumen tube. J Cardiothorac Vasc Anesth 1998;11:924-5.
4. Chow MYH, Liam BL, Lew TWK, et al. Predicting the size of double-lumen endobronchial tube based on tracheal diameter. Anesth Analg 1998;87:158-60.
5. Hannallah M, Benumof JL, Silverman PM, et al. Evaluation of an approach to choosing a left double-lumen tube size based on chest computer tomographic scan measurement of left mainstem bronchial diameter. J Cardiothorac Vasc Anesth 1997;11:168-71.
6. Hannallah MS, Benumof JL, McCarthy PO, Liang M. Comparison of three technique to inflate the bronchial cuff of left polyvinyl chloride double-lumen tubes. Anesth Analg 1993;77:990-4.
7. Hannallah MS, Benumof JL, Bachenheimer LC, Mundt DJ. The resting volume and compliance characteristics of the bronchial cuff of left polyvinyl chloride double-lumen endobronchial tubes. Anesth Analg 1993;77:1222-6.
8. Klein U, Karzai W, Bloos F, et al. Role of fiberoptic bronchoscopy in conjunction with the use of double-lumen tubes for thoracic anesthesia. Anesthesiology 1998;88:346-50.
9. Benumof JL. Difficult tubes and difficult airways. J Cardiothorac Vasc Anesth 1998;12:131-2.
10. Hannallah MS, Benumof JL, Ruttimann UE. The relationship between left mainstem bronchial diameter and patient size. J Cardiothorac Vasc Anesth 1995;9:119-21.
11. Hannallah M, Gomes M. Bronchial rupture associated with the use of a double-lumen tube in a small adult. Anesthesiology 1989;71:457-9.
12. Brodsky JB, Shulman MS, Mark JBD. Malposition of left-sided double lumen endobronchial tubes. Anesthesiology 1985;62:667-9.
13. Sakuragi T, Kumano K, Yasumoto M, Dan K. Rupture of the left main-stem bronchus by tracheal portion of a double-lumen endobronchial tube. Acta Anaesthesiol Scand 1997;41:1218-20.