Double-lumen endotracheal tubes (DLT) (BronchoCath[R]; Mallinckrodt Medical, Inc., St. Louis, MO) and, recently, single-lumen endotracheal tubes with enclosed bronchial blockers, (Univent[R] tube; Vitaid, Lewiston, NY) can be used to achieve one-lung ventilation (OLV) during thoracotomy [1,2]. There is controversy as to which might provide better operative conditions [3,4]. No objective studies have compared these tubes for OLV during thoracotomy. The aim of our study was to compare a left-sided, modified BronchoCath[R] DLT  with a single-lumen endotracheal tube with enclosed bronchial blocker during OLV with thoracotomies that required lung collapse.
After approval by our Human Subjects Review Committee, written, informed consent was obtained from each of 40 patients (aged 21-78 yr; weight 50-103 kg) undergoing either a thoracic or esophageal procedure for which OLV was required. Patients were randomly assigned to one of two groups. Twenty patients were managed with a "left-sided," modified BronchoCath[R] DLT and 20 were managed with a Univent[R] tube. In the latter group, the bronchial blocker was placed in the right main bronchus in 11 patients and in the left main bronchus in 9. There were 10 males and 10 females in the DLT group, and 12 males and 8 females in the Univent[R] group. Surgical procedures performed are listed in Table 1. After standard monitor and arterial catheter placement, patients were anesthetized with intravenous thiopental (3-5 mg/kg) or etomidate (0.1-0.3 mg/kg), fentanyl (5-20 micro g/kg) throughout the procedure, and isoflurane (0.5%-1.5%) in oxygen. A nondepolarizing muscle relaxant was used for paralysis. In 25 patients, elective bronchoscopy was performed by the surgeon via a standard single-lumen endotracheal tube. Surgeons were absent from the operating room during placement of the DLT or Univent[R] tube, by study design, so that their later observations about exposure would be blinded.
In the DLT group, the following method was used for tube placement: The DLT was introduced into the glottis via direct laryngoscopy. After the bronchial cuff had passed the vocal cords, the tube was rotated 90 degrees toward the left and advanced until resistance was encountered. If this failed to place the bronchial lumen in the left main bronchus, then the DLT was with-drawn until the bronchial lumen was above the carina. Then a fiberoptic bronchoscope (FOB) was passed via the bronchial lumen into the left main bronchus, and the tube was guided into position. In the Univent[R] group, once the trachea had been intubated via direct laryngoscopy, the endobronchial blocker was advanced through its anterior channel and directed into the desired main-stem bronchus under direct vision with a FOB. On the left side, the bronchial blocker was inserted deeply into the main-stem bronchus to minimize the chance of dislodgment back into the trachea. On the right side, the blocker was positioned so that cuff inflation would occlude the right upper and main-stem bronchus. For both tube types, accuracy of tube placement was checked exclusively using flexible fiberoptic bronchoscopy . Although auscultation was performed in every patient, this was not considered reliable for satisfactory placement in our study. The following sequence of events was recorded in the two groups of patients: 1) time required to position these tubes; 2) the number of times in each patient that fiberoptic bronchoscopy was required to assure proper position; 3) the frequency of malposition; 4) the effectiveness of lung collapse; 5) the surgical exposure; and 6) cost of each tube per procedure.
The time for initial tube placement, with the patient in supine position, was defined as the time from passage of the tube past the vocal cords until there was satisfactory placement of the bronchial lumen or bronchial blocker. After proper position was achieved, the endotracheal tube (DLT or Univent[R]) was covered with a towel so that surgeons would be unaware of which tube had been placed. Once satisfactory initial placement was achieved, the bronchial cuff of the DLT or the Univent[R] bronchial blocker were deflated, and patients turned to a lateral decubitus position. Once in the lateral position, tube placement was again checked via fiberoptic bronchoscopy on all but one patient.
Criteria used to assess the frequency of tube malposition are shown in Table 2. Once in the lateral position with proper tube placement, tube position and adequacy of lung collapse were evaluated every 30 min. In addition, we recorded the number of times that supplemental flexible fiberoptic bronchoscopy was required to check or correct tube position because of poor collapse or surgical exposure throughout the study. In every instance where malposition was identified it was corrected. Once the pleura was opened, the effectiveness of lung collapse was evaluated by the surgeon who performed the surgery. Complete collapse with perfect surgical exposure was ranked as excellent. If total collapse was obtained but the lung still had residual air, this was ranked as fair; finally, if no collapse was achieved, or partial collapse with interference in surgical exposure, the rank was defined as poor. The time needed for lung collapse began with institution of OLV. When necessary, two types of maneuvers were used to facilitate lung collapse. For the DLT group, a suction catheter was passed into the lumen. For the Univent[R] tube group a 3.0- or 3.5-mm plastic adaptor from a regular endotracheal tube was attached to the suction channel of the bronchial blocker. Lung collapse was recorded as spontaneous, or assisted with suction, or manual collapse. After surgery, the surgeon was asked to comment on the surgical exposure. Three specific categories were used: 1) excellent exposure; 2) fair exposure, and 3) poor exposure. When two-lung ventilation was reestablished and the chest closed, the study was considered complete. Prior to extubation, fiberoptic bronchoscopy was performed in all but one patient to observe whether there was any damage to tracheal or bronchial mucosa.
Unless specified otherwise, values are expressed as mean +/- SD. Positioning time, number of supplement fiberoptic bronchoscopies, and time to lung collapse were analyzed using the Mann-Whitney U-test. The incidence of malposition was analyzed using the Fisher exact test. P values < 0.05 were considered to be significant.
The DLT and Univent[R] groups were equivalent with regard to age, weight, gender, and procedure duration. In the DLT group, 11 patients had right-side procedures (10 thoracotomies, 1 thoracoscopy) and 9 had left-side procedures (6 thoracotomies, 3 thoracoscopies). In the Univent[R] group, 11 had a right-side procedure (8 thoracotomies, 3 thoracoscopies) and 9 had left-side procedures.
Tube sizes used in this study for females were: modified BronchoCath[R] DLTs, one 35 Fr, six 37 Fr, and three 39 Fr; Univent[R] bronchial blockers placed were one 7.0 and eight 7.5 internal diameter. For males the sizes used were: DLTs, one 37 Fr, five 39 Fr, and four 41 Fr; and for Univent[R] tubes two 7.5 and nine 8.0 internal diameter.
The time required for initial tube position was 6.2 +/- 3.1 min for the DLT group versus 5.4 +/- 4.5 min for the Univent[R] group (P > 0.06). The median number of bronchoscopies per patient needed to assure proper position of the tube were two for the DLT group (a minimum of one and a maximum of three) versus three in the Univent[R] group (a minimum of two and a maximum of five) (P > 0.05).
In the DLT group, there were malpositions in 5 of 20 patients during at least one time period versus 11 of 20 in the Univent[R] group (Table 3). The total number of malpositions was higher in the Univent[R] group (P < 0.003). Malpositions in the DLT group within the first 60 min (4 of 20) consisted of both lumens being in the trachea after the patient was turned to the lateral decubitus position (one patient) and bronchial cuff herniation over the tracheal carina (three patients). In all cases, malposition was corrected and collapse achieved. Malposition in the Univent[R] tube group within the first 60 min was detected in 8 of 20 patients. In three patients, the bronchial blockers were too deeply inserted (two right-side and one left-side). With the two right-side bronchial blockers, right upper lobe collapse was not achieved during the malposition. With malposition of the left-side bronchial blocker, only partial collapse was achieved. After repositioning of the bronchial blockers, the respective right or left lungs were collapsed without difficulty. Another left-side bronchial blocker was dislodged back into the trachea after the patient was turned to the lateral decubitus position. In two patients, after initial cuff inflation of the bronchial blocker and proper seal obtained, additional air was required in the cuff to obtain a seal of the desired bronchus. Another two patients in the Univent[R] group (one left and one right side) also needed repositioning. In one of them fiberoptic bronchoscopy was difficult due to a large amount of secretions, and for the other no reason was given for the reposition of the tube. Overall, in the Univent[R] group, among the 15 malpositions, five occurred with right-side bronchial blockers and 10 were with left-side bronchial blockers.
In the DLT group, 10 patients had spontaneous lung collapse, 8 required assisted collapse with suction, and 2 required manual collapse by the surgeon. The time to lung collapse in the DLT group was 7.1 +/- 5.4 min. For the Univent[R] group, 4 had spontaneous collapse, 12 required assisted collapse, and 4 had manual collapse. The time to lung collapse in the Univent[R] group was 12.3 +/- 10.5 min. For the DLT group, exposure was ranked as excellent in 18 cases, fair in 1, and one response was not recorded. In the Univent[R] group, exposure was reported as excellent in 15 patients, fair in 4, and poor in 1.
There were no recognized complications as a result of tube placement or OLV. No abnormal findings were found in the tracheal or bronchial mucosa during the bronchoscopic examination prior to withdrawal of these tubes. Overall acquisition cost for the 20 patients in DLTs was $1360.00; for the 20 patients in the Univent[R] tube the cost was $2600.00.
To facilitate surgical exposure, either the DLT or the Univent[R] tubes can be used to achieve OLV for various thoracic, vascular, and esophageal procedures. Because of potential mismatch between the right upper bronchus and the ventilation slot in the bronchial lumen, right-sided DLTs may have limited use . Therefore, in the large majority of cases where a DLT is used, a left-sided DLT is chosen. Hence in this study a left-sided DLT was chosen for all cases randomized to DLTs irrespective of side of surgery .
This study represents the only objective, prospective, and clinical data comparing the two latest tubes commonly used during anesthesia for OLV. The randomization of our study was aimed to prevent any bias toward any particular tube. This random method allowed us to chose one of these tubes regardless of the type or surgical side.
In this study, there was no difference between groups in time required for initial tube placement. The newest modification of the DLT has eliminated the bevel at the tip of the bronchial lumen. This lack of a bevel had no discernible effect on ease of DLTs placement. Our positioning times are comparable to those in which traditional beveled-tip DLTs have been used . Brodsky and Macario  had questioned whether this newest modification of the DLTs might present difficulties during passage and positioning. We did not find this to be the case in this investigation. In our study, placement was not an issue. Therefore, in patients with normal airways, the ease and time for initial positioning were essentially equivalent. In our study there were no patients with abnormal airways or difficult intubations. The Univent[R] tube might be advantageous, and easier to place in patients with difficult airways [10,11]. Another advantage of the Univent[R] tube is avoiding the need for exchanging a double-lumen tube for a single-lumen tube when postoperative mechanical ventilation is planned. The number of times that supplemental fiberoptic bronchoscopy was needed to check or correct tube position was similar in both groups. Therefore, similar effort was required to assure proper placement while using either tube.
Frequency of malposition was greater for the Univent[R] tube group when compared to the DLT group. Satisfactory initial tube placement was obtained in all patients in the supine position. However, after patients were turned to the lateral decubitus position, head position changed, or bronchial cuff reinflated, we found the Univent[R] tubes to be displaced from the original position more often than the DLTs. In some instances the Univent[R] tube was in too far, and only a segment of the lung could be collapsed. All malpositions were corrected by repositioning the bronchial blocker of the Univent[R] tube or the DLTs. Malpositions have been reported during head movement when DLTs have been used ; however, we are unaware of any report of malpositions during head movements or changes in patient position with the Univent[R] bronchial blocker. Although problems with isolation of the operated lung have been described with the Univent[R] tube , we believe that most malpositions are correctable if a FOB is used and if necessary adjustments in the bronchial blocker are made.
In our study, intraoperative repositioning of the left-sided Univent[R] tubes were performed as follows: the trachea was suctioned to facilitate vision with the FOB and ventilation was stopped. The bronchial blocker cuff was deflated and a FOB passed through the Univent[R] tube. Then the bronchial blocker port was advanced or withdrawn and repositioned under direct vision; the cuff was reinflated to obtain a seal, then ventilation resumed and the cuff checked for any visible air leak or bubbles appearing around the cuff. For a right-sided Univent[R] tube, repositioning similar to that of the left-sided tube was done and the bronchial blocker placed near the right upper bronchus, so that occlusion would occur in the upper and bronchus intermedius. This was correlated with complete collapse of the right lung. Some operations in thoracic surgery require complete and rapid lung isolation. The Univent[R] tube may not be as reliable as the DLT tube in this situation. These problems were more evident after the patient was turned to a lateral decubitus position and bronchoscopic confirmation was made. Benumof  had suggested a routine fiberoptic bronchoscopy while positioning DLTs. We believe this advice also should apply to the Univent[R] bronchial blocker, specifically considering the greater frequency of malposition observed with this device. With respect to effectiveness of lung collapse, the DLT and the Univent[R] group did not differ. Both took the same amount of time to lung collapse, although 12 patients in the Univent[R] group required assisted collapse via suction, in contrast to the DLT where eight required assisted collapse. Surgical exposure evaluated by a surgeon unaware of which tube has been used was better in the DLT group; 18 of 20 were considered excellent exposure. It is important to ask for feedback among surgeons, because in the majority of cases the most common indication for OLV is surgical exposure with facilitation of the surgical condition by deflating the lung.
In our study in the DLT group, there were no left upper lobe obstructions. Perhaps one of the modifications done to this tube, which include a shortened bronchial cuff, had prevented this left upper obstruction. In contrast, with the right-sided Univent[R] tubes, right upper lobe ventilation was experienced in two patients during what was supposed to be right-side total lung collapse. This problem was corrected after repositioning the bronchial blocker on the right-side bronchus. In fact there is a potential complication during right-sided thoracotomy-a risk of transection of the bronchial blocker tip with a stapler if the bronchial blocker is in too far .
In summary, our study found that: 1) The modified BronchoCath[R] DLTs and the Univent[R] tubes are vulnerable to malposition when changing from the supine position to the lateral decubitus position. 2) Fiberoptic bronchoscopic reconfirmation of tube position is mandatory every time the patient's position is changed or unexpected ventilation occurs in the collapsed side. 3) All malpositions were correctable during the lateral decubitus position. 4) Univent[R] bronchial blockers have a greater frequency of malposition when compared to modified BronchoCath[R], DLTs. In the Univent[R] group the incidence of malposition and cost involved were sufficiently greater; therefore, we cannot find cost/efficacy justification for routine use of this device.
The authors wish to express their appreciation and efforts to Marvel Titone, Research Assistant II, for her direct involvement in this investigation; Joyce Jones for her secretarial expertise; and the Thoracic Surgical Team, Drs. Nicholas Rossi, Kemp Kernstine, and Wayne Richenbacher for their participation in this study. Also, Drs. Bradley Hindman and John H. Tinker are acknowledged for reviewing the manuscript.