One-Lung Ventilation

Indications for One-Lung Ventilation


  1. Protective Isolation
    1. Massive Hemorrhage
    2. Infection
  2. Control of Ventilation Distribution
    1. Bronchopleural or bronchopleural cutaneous fistula
    2. Giant cyst or bullae (risk of rupture with PPV)
    3. Major bronchial disruption or trauma
  3. Unilateral Lung Lavage
  4. VATS

Relative (Strong) – Surgical Exposure

  1. Thoracic aortic aneurysm
  2. Pneumonectomy
  3. Upper lobectomy

Relative (Weak) – Surgical Exposure

  1. Esophageal surgery
  2. Middle and lower lobectomy
  3. Thoracoscopy under general anesthesia

Lung Isolation

Bronchial Blockers

Major disadvantages are that they can become dislodged (life-threatening in some cases), may be difficult (or impossible) to place, and cost more than double lumen tubes. Advantages are that they allow a single lumen tube to be placed (no tube exchange post-operatively, ideal for long cases with expected post-operative mechanical ventilation), are safer (with SLTs) in difficult airway situations, can be used to isolate individual lobes, and can be used in smaller patients (smaller tube, can be placed next to tube, not necessarily through it)

Common varieties include the Univent (fixed hockey-stick end), Arndt (wire-loop), and the Cohen (wheel). The Arndt is particularly useful in that it comes in 5F, 7F, and 9F sizes, thus can be used in a 4.5 ETT.

Double Lumen Tubes

Most common is the Robertshaw tube, available as both a L and R-sided tube. Sizes include a 28F (pediatric) as well as 35F, 37F, 39F, and 41F. All are PVC with D-shaped lumens, disposable. Blue represents the endobronchial lumen/cuff. Note that R-sided tubes have a donut-shaped endobronchial cuff, which allows separate access to the RUL.

Traditional positioning of the DLT is accomplished as follows:

Positioning of the DLT

  1. Because DLTs are so much more difficult to place, always check both cuffs, lubricate the inside and outside of the tube, and place a stylet prior to insertion.
  2. Attempt your DL with a Mac-3 (leaves the largest amount of physical space with which to work in the oropharynx), and remove the stylet as soon as the tube is past the glottis.
  3. Rotate the tube appropriately, then inflate the tracheal cuff and verify bilateral inflation / equal breath sounds.
  4. Next, clamp the tracheal lumen (clear) and slowly inflate the endobronchial cuff until no air leak occurs (so as not to overinflate).
  5. Unclamp the tracheal lumen (clear) and verify that both lungs are inflated with both cuffs up (i.e. to ensure that the endobronchial cuff is not obstructing the opposite brochus)
  6. Lastly, clamp each lumen individually and verify unilateral chest movement

The aforementioned technique, when studied in 23 patients, showed a 48% failure rate [Smith et al. Br J Anaesth 58: 1317, 1987], although most of these malpositioning incidences were clinically insignificant. While FOB is necessary for positioning, it is not necessarily needed for initial placement – one study comparing blind to FOB-based placement in 59 showed a higher success rate and 93 seconds of saved time when initial placement relied on the blind technique [Boucek LD et al. J Clin Anesth 10: 557, 1998]

The major malpositioning errors are 1) entrance into opposite bronchus [opposite lung will collapse] 2) too deep into the bronchus [diminished breath sounds contralaterally] 3) underinsertion [no breath sounds when tracheal lumen used, as endobronchial cuff is still in the trachea 4) right-sided DLT may occlude the RUL [mean distance to RUL 2.2 cm] 5) LUL may be obstructed by the L endobronchial tube. Key FOB landmarks in DLT placement include [tracheal lumen initially] 1) lack of herniated bronchial cuff 2) visualization of three orifices in the RUL (only lobe to have three orifices), [bronchial lumen] 3) LUL and RUL orifices

Placement of a left-sided tube is significantly easier, however there are instances in which a right-sided tube is necessary – anytime that the L main bronchus has been altered significantly (ex. compression by tumor or thoracic aortic aneurysm), or anytime an operation involves the L main bronchus (ex. lung transplantation)

Approach to One-Lung Ventilation

Avoidance of hypoxemia is the primary goal, and while there are no evidence-based recommendations regarding the lower limit of acceptable SpO2, most practitioners try to maintain 90% or higher (PaO2 60 mm Hg), adjusting as needed based on other comorbidities. The incidence of hypoxemia during OLV has fallen from 20% from the 1950’s to 1980’s, to 10% in the 1990’s [Hurford WE et al. J Cardiothorac Vasc Anesth 7: 517, 1993], to 1% most recently [Brodsky J and Lemmens HJ. J Cardiothorac Vasc Anesth 17: 289, 2003]

Critical to avoiding hypoxemia is an understanding of the basic goal of physiologic management in OLV – maximizing PVR in the operative lung, and minimizing PVR in the dependent lung. Of note, PVR is generally lowest at FRC, and increases hyperbolically as volumes deviate in either direction from FRC. During OLV, FRC occurs at slightly lower-than-normal volumes due to paralysis, lateral positioning, the open operative hemithorax, and the weight of mediastinal structures

Immediately after placing the patient in the lateral position, the DLT position should be verified with an FOB. TLV should be continued as long as possible, and when OLV is finally required, start with an FiO2 of 1.0, keeping plateau pressures < 25 cm H2O and PaCO2 at 35 mm Hg. Give a recruitment maneuver to immediately address atelectasis in the dependent lung – increasing peak inspiratory pressure to 40 cm H2O combined with a peak end-expiratory pressure level of 20 cm H2O for 10 consecutive breaths increased PaO2 to 244 mm Hg, as compared to 144 mm Hg in patients who did not get the recruitment maneuver (p < 0.001) [Tusman G et al. Anesth Analg 98: 1604, 2004; FREE Full-text at Anesthesia & Analgesia]. PaO2 may fall for up to 45 minutes, thus frequent ABGs are important in this initial period

Hypoxemia during OLV should prompt FOB examination. If the DLT is properly placed, add CPAP10 to the operative lung (except in VATS cases, in which it disruptive – in these instances, PEEP10 to the dependent lung may be tried). As a last resort, intermittent two lung ventilation (which requires cooperation from the surgeon) may be tried

Management of One Lung Ventilation

Most practitioners will tolerate an SpO2 of 90%, although this threshold may change based on the presence of coronary artery disease, cerebrovascluar disease, or anemia. Keep in mind, however, that there are no evidence-based recommendations regarding the lower limit of acceptable SpO2

It is useful to at least try to predict who will desaturate during one lung ventilation – preoperative risk factors include 1) very high or low V/Q [Hurford WE et al. Anesthesiology 64: 841, 1987] 2) normal spirometry 3) restrictive lung disease and/or severity of COPD, and intraoperative factors include 1) supine positioning 2) R-sided thoracotomy [Lewis JW et al. J Cariothor Vasc Anesth 6: 705, 1992] 3) hypoxemia with two lung ventilation (the most predictive of all) [Slinger P et al. Can J Anaesth 39: 1030, 1992]


An FiO2 of 1.0 (leading to PaO2 150-200 mm Hg) is usually used during OLV, because it a) provides a margin of safety and b) dilates the pulmonary vessels (especially in the ventilated, dependent lung). The major disadvantage is the potential for absorption atelectasis, however this can be counteracted with PEEP

FiO2 as low as 0.25 to 0.50 have been used, generating a PaO2 of 62-87 mm Hg. Lower FiO2 should be considered in any patient on bleomycin. Lower FiO2 may also reduce the concentration of volatile anesthetic requirements, which is potentially beneficial in particularly sick patients [Barash]

Tidal Volme and Respiratory Rate

While every patient has optimal ventilatory settings for OLV, it takes time and effort to discover these, and may not be practical in the context of an anesthetic. Thus, a reliable starting point is desired. 5-6 cc/kg IBW with 5 cm H2O PEEP is recommended by Miller as a good starting point in patients without COPD. According to Barash, however, a TVOLV < 8 mL/kg can result in dependent-lung atelectasis as well as a decrease in FRC, thus the recommended TVOLV is 10-12 mL/kg. By contrast, a TVOLV > 15 mL/kg can increase PVR in the dependent lung, thus shunting blood to the operative lung. Barash also advocates a PaCO2 of 35 mm Hg, as significant hypocapnia may inhibit the hypoxic pulmonary vasoconstriction response

PEEP in the Dependent Lung

PEEP10 has the theoretical advantages of increasing FRC in the dependent lung, thus improving the V/Q ratio and preventing atelectasis, and does not require cessation of surgery. That said, early studies of PEEP during FiO2 of 1.0 failed to find significant improvements in oxygenation [Tarhan et al. Can Anaesth Soc J 17: 4, 1970; Capan LM et al. Anesth Analg 59: 847, 1980]. A subsequent study of PEEP10 in diseased lungs (with PaO2 < 80 mm Hg) found some benefit [Cohen E et al. Anesth Analg 64: 200, 1985]. More recent studies have shown mixed results – Mascotto et al. randomized 50 patients to ZEEP versus PEEP, and found that PEEP lowered the PaO2/FiO2 ratio, with no differences in the requirement for 100% oxygen, reinflation of the operative lung, or PACU time [Mascotto et al. Eur J Anaesthesiol 20: 704, 2003]. Valenza et al. showed that 10 cm of PEEP in OLV was able to improve oxygenation in patients with an FEV1 of > 72%, but not in those with poor pulmonary function (< 72%) [Valenza et al. Eur J Anaesthesiol 21: 938, 2004, 48 patients]. Senturk et al. showed that the addition of 4 cm PEEP to PCVOLV, PEEP lowered plateau pressures, peak pressures, and Qs/Qt, and increased PaO2 in [Senturk et al. J Cardiothorac Vasc Anesth 19: 71, 2005, 25 patients]


Tends to occur in emphysematous or elderly patients, and is worsened by the small DLT lumen. Can be beneficial if it move the patient towards FRC on the compliance curve (rare), but more often than not it moves that patient away from FRC. Most standard anesthesia ventilators cannot detect auto-PEEP (this requires end-expiratory flow interruption). Auto-PEEP is proportional to tidal volume and inversely proportional to expiratory time, thus, if it is suspected, consider lowering TV and increasing the expiratory time

CPAP in the Operative Lung

The MOST effective method by which PaO2 can be increased is application of CPAP to the operative lung [Capan LM et al. Anesth Analg 59: 847, 1980; Cohen E et al. J Cardiothorac Vasc Anesth 2: 34, 1988; Hogue et al. Anesth Analg 79: 364, 1994], but it must be preceded by a recruitment maneuver, as the opening pressure of atelectatic lung regions is > 20 cm H2O [Rothen HU et al. Br J Anaesth 71: 788, 1993]. In fact, if one fully inflates the lungs, only 1-2 cm H2O CPAP may be needed [Hogue CW. Anesth Analg 79: 364, 1994], which will improve intraoperative conditions

By contrast, some authors have suggested simply giving 5-10 cm H2O immediately after an inspiratory breath [Capan LM et al. Anesth Analg 59: 847, 1980], although it has been shown that insufflation only can increase PAO2 eventually (~ 45 minutes [Rees and Wansbrough. Anesth Analg 61: 507, 1982], as does intermittent reinflation of the lung [Malmkvist. Anesth Analg 68: 763, 1989]. CPAP10 has no hemodynamic consequences [Van Keer et al. J Clin Anesth 1: 284, 1989]. According to Barash, anything greater than 10 cm H2O may cause overdistention and/or hemodynamic consequences

Volume Control versus Pressure Control Ventilation

Pressure control has several theoretical advantages, ex. lower peak airway pressures, a decreased incidence of sudden airway pressure changes due to surgical manipulation, however none of these have resulted in improved outcomes as compared to volume control ventilation, with the possible exceptions of COPD patients [Tugrul M et al. Br J Anesth 79: 306, 1997], or those undergoing pneumonectomies or lung transplantations [Slinger P. Anesth Analg 103: 268, 2006; FREE Full-text at Anesthesia & Analgesia]. Tugrul et al. randomized 48 patients to VCV (10 mL/kg) → PCV (peak pressures titrated to 10 mL/kg) versus PCV → VCV and found that PCV led to decreases in plateau airway pressure (17.8 vs. 18.5, p < 0.05), peak airway pressure (23.65 vs. 28.3, p = 0.000001), and pulmonary shunt (36.2 vs 40.2%, p = 0.03), with no effect on cardiac output or oxygen saturation [Tugrul et al. Br J Anaesth 79: 306, 1997]. A repeat study of 58 patients (this time with a goal of 9 mL/kg in both groups) found no difference in plateau pressures (19.51 vs. 19.81 mm Hg), but did find a decrease in peak pressures (24.43 vs. 34.16 mm Hg, p < 0.001), with no difference in arterial saturation [Unzueta MC et al. Anesth Analg 104: 1029, 2007; FREE Full-text at Anesthesia & Analgesia.]. In summary, it can be assumed that PCVOLV reduces peak airway pressures and may have a small effect on plateau pressures.