Although considered relatively low-risk procedures, vitrectomies can be associated with venous air embolism, a life-threatening complication. As with many common procedures associated with rare, yet devastating complications, limited data exist to fully explain the mechanisms. To prevent the complications, we must first better understand how they occur.

Modern pars plana vitrectomy, developed by Robert Machemer in 1970, facilitated the first controlled access to the posterior segment of the eye. Nearly a half-century after Machemer’s report, Gayer and associates documented that the suprachoroidal space is a wealth of arteries and veins that may become the portal to fatal venous air emboli (VAE) during routine vitrectomy. Should air embolism occur, hypotension, decreased cardiac output, and cardiovascular collapse are likely to soon follow. However, the number of reports of this dreaded complication in both the anesthesiology and ophthalmology literature is limited, perhaps validating the impression that this is a very uncommon event. Nonetheless, the current study by Gayer serves an important role not only by highlighting this important subject but also by presenting data to support a mechanism of action .

How can air embolism occur as a result of pars plana vitrectomy? These authors describe, investigate, and illustrate just how a breach in the integrity of the large suprachoroidal veins could provide the entry pathway for a large volume of air to quickly enter microcirculation, followed by its inevitable transit to the heart, thereby disrupting the central circulation. Although we must be cautious in over-interpretation owing to the limited data available from Gayer and associates, this experimental model and the inferred mechanism are very plausible.

VAE is a well-known complication during a host of surgical operations, including neurosurgical procedures (especially those in the sitting position), Cesarean delivery, scoliosis surgery, and, more recently, even retrograde endoscopic procedures of the pancreas ( Table ). In addition, venous air embolism can also be associated with medical procedures such as central venous catheterization, hemodialysis, vacuum gynecologic procedures, and radiocontrast injection for computed tomography and other imaging modalities. Lastly, VAE may even occur outside the hospital setting, such as when scuba divers are exposed to rapid pressure changes while breathing a gas mixture that includes nitrogen.

The pathophysiology of VAE classically occurs when 2 conditions are met simultaneously:

• (1)
  

  Procedures where the surgical site is above the plane of the heart (eg, sitting craniotomy), creating a gravitational gradient for air, and

  
  
• (2)
  

  Surgical incisions that invade noncompressible venous channels (such as dural veins), creating an open portal for air to enter the circulation.

However, medical and surgical procedures have created a second, more insidious pathway for fluids (air and other gases) to enter the circulation. Procedures such as imaging contrast injections, laparoscopic abdominal procedures, and intra-aortic balloon rupture inject air/gases/fluids into various body cavities under pressure. Although the pressure of injections is typically closely monitored, the volume of gas injected is seldom measured or known. Thus, large amounts of air or gas can rapidly be forced into small, nearly invisible vascular and tissue rents.

So, what can physicians caring for a patient about to undergo a pars plana vitrectomy do to mitigate the risk of air embolism? Prevention is the best management for VAE. We believe the key steps include:

• (1)
  

  Improved communication/preoperative brief: A preoperative brief is the ideal opportunity to engage all caregivers to develop a “shared mental model” regarding the patient, the operation, and any potential concerns or possible complications, potentially including VAE. Although checklists have certain limitations, the value of a well-done preoperative briefing with a checklist including VAE should be well received by physicians, nurses, and even patients. For patients likely to be at increased risk of VAE (eg, those with a traumatic intraocular foreign body or other perforating injury, endoresection of a choroidal melanoma, etc), the ophthalmologist/attending surgeon leads the brief and, among other elements, reminds personnel of the remote but still real possibility of VAE during the air insufflation phase of the vitrectomy. In addition, operating room personnel should be aware of one recommendation that the air infusion pressures not exceed 30 mm Hg—at least until air is seen entering the vitreous cavity. Lastly, the team may wish to review the necessary and immediate steps in the event of suspected VAE compromising the cardiovascular status (see Diagnosis and Treatment steps below). Early intervention saves lives. It’s hard to imagine a team working optimally when they are not comfortable working with each other and have never experienced such a disaster. Thus, some centers may explore the use of high-fidelity simulation so that teams can practice how to respond to such an event.

  
  
• (2)
  

  Diagnosis. The paper in this issue by Gayer and associates highlights the extreme sensitivity of echocardiography for the presence of emboli, especially air emboli (capable of detecting air volumes as small as 0.02 mL/kg), within the central circulation. However, this modality is costly, semi-invasive, and impractical for the vast majority of eye procedures, as the presence of a transesophageal probe would at best be cumbersome near the surgical field itself. In addition, specialized education is necessary for safe and appropriate use of echocardiography. Similarly, other recognized but costly monitors, such as the pulmonary artery catheter or central venous catheter, are feasible but rather unrealistic options for “routine care” during ophthalmologic surgery. Thus, in practical terms, clinicians will need to rely on 4 core monitors looking for the prodrome of signs, such as:

  
  
  
  
  • (a)
    

    Sharp drop in end-tidal carbon dioxide (owing to dramatic decrease in cardiac output and increased ventilatory dead space).

    
    
  • (b)
    

    Sharp rise in end-tidal nitrogen concentration—assuming your standard operating room gas monitor detects the presence of nitrogen (which is by no means a standard modality nationwide). Though neither end-tidal carbon dioxide nor nitrogen monitoring is as sensitive as transesophageal echocardiography for detection, either should alarm with air volumes of 0.5 mL/kg.

    
    
  • (c)
    

    Otherwise unheralded hypotension, tachycardia, and perhaps oxygen desaturation. We acknowledge that these signs are nonspecific, but in the setting of “at-risk” surgery, the sudden appearance of this constellation of findings should rapidly lead the team to consider the diagnosis of VAE. Unfortunately, consistent with the experimental findings of Gayer and associates, oxygen desaturation is a relatively late and nonspecific sign.

    
    
  • (d)
    

    An observation during scoliosis surgery suggests that the intraoperative bubbling of visible air may be the first sign of VAE. Whether such a prodrome might occur during pas plana vitrectomy is unreported and unknown at this time.

  
  
  
  
• (3)
  

  Treatment. In the event that prevention strategies fail, early diagnosis may provide the opportunity for rescue prior to cardiac arrest. Rescue steps should include:

  
  
  
  
  • (a)
    

    Immediate termination of the insufflation source of any air or gas.

    
    
  • (b)
    

    Placing the patient on high-flow, 100% oxygen.

    
    
  • (c)
    

    Consideration of patient position. In order to reduce the air lock within the right atrium and right ventricle, it was traditional to place the patient in left lateral decubitus position and/or steep Trendenlenburg position. However, experimental animal studies have found no reliable improvement in cardiac output or blood flow subsequent to such maneuvers, and human data are totally lacking.

    
    
  • (d)
    

    Aspiration of air from the central venous catheter. This maneuver, while theoretically appealing and potentially effective, presumes the presence of a catheter at the time of air entrainment, a routine recommendation that exists only for sitting craniotomy patients.

    
    
  • (e)
    

    Full cardiac resuscitation with epinephrine, vasopressin, and chest compressions. In selected cases, extracorporeal hemodynamic support should be considered, included modalities such as percutaneous extracorporeal membrane oxygenation.

    
    
  • (f)
    

    Because of the complexity of therapy in advanced VAE, and the rarity of such events, simulation-based team training may be considered.

  
  

In summary, ophthalmologists and anesthesiologists must remain vigilant for very uncommon but potentially devastating complications such as VAE. At the first signs, it is vital to call early for help. An orderly team-based response will maximize the opportunity for a successful resuscitation from hemodynamically disruptive VAE. Because of the complexity and infrequency of these events, simulation training should always be considered an option for a well-prepared team.