Secondary Logo

Journal Logo

Possible Air Embolism During Eye Surgery

Ledowski, Thomas, MD*; Kiese, Felix, MD; Jeglin, Silke, MD*; Scholz, Jens, MD*

doi: 10.1213/01.ANE.0000154304.63828.C0
Ambulatory Anesthesia: Case Report

We report a case of possible air embolism during a three-port pars plana vitrectomy and air-fluid exchange of the vitreous cavity of the eye. After the start of intraocular air flushing, sudden tachycardia, a decrease in oxygen saturation and end-tidal carbon dioxide tension, and a distinct “mill-wheel” murmur were observed. Venous air embolism was suspected but other sources of air entry into the circulation and a thromboembolic event were excluded. Once intraocular air flushing was ceased, clinical variables returned to normal within minutes. In conclusion, during air-fluid exchange of the vitreous cavity, air embolism should be considered as a possible rare complication.

IMPLICATIONS: Although complications of eye surgery are infrequent and normally not severe, this report deals with the potentially life-threatening hazard of air embolism. Because the potential for harm and successful therapy are time-dependent, the possibility of air embolism should be considered, despite its apparent rarity.

*Department of Anaesthesiology and Intensive Care Medicine, University Hospital Kiel; and †Eye Hospital Bellevue, Kiel, Germany

Accepted for publication December 9, 2004.

Address correspondence and reprint requests to Thomas Ledowski, MD, Department of Anaesthesiology and Intensive Care Medicine, University Hospital Schleswig-Holstein, Campus Kiel, Schwanenweg 21, D24105 Kiel, Germany. Address e-mail to

Air embolism is a recognized complication of a variety of surgical procedures other than eye surgery and occurs either when air is entrained into open veins by hydrostatic force or is introduced under high pressure (1–3). Although treatment of retinal detachment by air-fluid exchange (AFE) of the vitreous cavity is a common procedure, systemic air embolism during this procedure has not been reported.

Back to Top | Article Outline

Case Report

A 55-yr-old man (ASA physical status II) with retinal detachment was electively anesthetized for a three-port pars plana vitrectomy and AFE of the vitreous cavity. Anesthesia was induced with propofol, maintained with propofol (6 mg · kg−1 · h−1) and remifentanil (0.25 μg · kg−1 · min−1), and rocuronium was used for muscle relaxation (0.6 mg/kg). Ventilation was performed with an oxygen/air mix at an inspired oxygen fraction (Fio2) of 0.4. Instruments for three-port pars plana vitrectomy were inserted transsclerally (Fig. 1). AFE was started via a 2-mm-long 20-gauge infusion cannula at an air pressure of 40 mm Hg (flushing system: Bausch and Lomb Millenium Microsurgical System, Rochester, NY). Outflow was vented via the vitreous cutter. Within the first few minutes, the surgeon found the choroid to be abnormally swollen and vascularized and was about to terminate the operation because of difficulty visualizing the intraocular contents. At that time, we noted a rapid reduction in peripheral oxygen saturation (Spo2), arterial blood pressure, and end-tidal carbon dioxide tension (Etco2) (Table 1), as well as tachycardia and transient ST-elevation.

Figure 1.

Figure 1.

Table 1

Table 1

Precordial auscultation within a minute of the decrease of Etco2 revealed a distinct mill-wheel murmur. Assuming this to be an air embolus, the anesthetic equipment (infusions, infusion pumps, and IV line) was checked for the possible infusion of air, and an accidental connection between the flushing system for the eye and the infusion lines was excluded. Ventilation was changed to an Fio2 of 1.0, and the propofol infusion rate decreased to 3 mg · kg−1 · h−1. Two doses of norepinephrine 2.5 μg were administered for arterial blood pressure control. Intermittent intraocular air flushing by the surgeon was stopped 4 min after the estimated onset of the event, and the precordial murmur disappeared. During the next 20 min, the Spo2, arterial blood pressure, and heart rate almost normalized (Table 1). Arterial blood gas analysis was performed 10 min after the onset of the event and showed a partial pressure of CO2 of 48 mm Hg (concurrent Etco2 reading of 22 mm Hg) and a partial pressure of O2 of 86 mm Hg (Fio2 1.0; Spo2 98%). IV heparin 5000 IU was given because a thromboembolic event could not be excluded at that time. At the end of the operation, 25 min after the first decrease in Spo2, the patient was admitted to the intensive care unit still ventilated under propofol sedation. Thoracic and brain computed tomography scans (to exclude thromboembolism and cerebral air embolism) were performed within 2 h but showed no pathological alterations except for a tiny bubble of air directly behind the eye bulb. Neither laboratory tests (creatine kinase MB and troponin T iso-enzymes) nor the electrocardiogram indicated an adverse cardiac event. The patient was tracheally extubated 4 h after admission and discharged the next day without any deficit. Two weeks later, the procedure was repeated uneventfully.

Back to Top | Article Outline


Because we eliminated other explanations for the described event, we believe the most likely explanation was air embolism via a choroidal vein. These veins are drained by the cavernous sinus and jugular veins. According to Gildenberg et al. (4), the classical mill-wheel murmur does not occur unless air enters the venous circulation at a rate of 1.96 mL · kg−1 · min−1. The diameter of choroidal veins varies from 300 μm (peripheral veins) to approximately 2 mm (ampulla of vortex veins). The diameter of the infusion cannula was 20-gauge, so direct cannulation of a choroidal vein was feasible, at least for the radially draining vortical veins. According to the Hagen-Poiseuille-Law (V° = πR4Δp/8Lη), at 20°C laminar flow and with a Δp of 20 mm Hg (air inflation pressure 40 mm Hg minus intraocular pressure 20 mm Hg), for a 10-mm tube (cannula plus vortical vein) with a lumen of 0.93 mm (cannula lumen), we calculated a potential air flow of approximately 1600 mL/min. This calculation proves that even lethal air embolism is a possibility. The exact circumstances of air entry in this case remain speculative. The fact that inflation of air during the procedure was not continuous but on demand (activated by a foot switch) might have contributed to the good outcome.

In the case described, the patient had standard monitoring with electrocardiogram, arterial blood pressure, Spo2, and Etco2 assessment. The frequency of uncomplicated procedures and the rarity of air embolism during this type of surgery are such that we are loath to recommend special monitoring. Being alert to the possibility for air embolism might be enough to allow effective management of this complication, and increasing awareness is the main purpose of this report. Nevertheless, we suggest inexpensive, noninvasive monitoring with a precordial Doppler ultrasound and close evaluation of Etco2. In our opinion, routine use of transesophageal echocardiography during eye surgery is too invasive and expensive, despite it being more sensitive (5).

We wish to thank Michael Paech, DM, Associate Professor of Obstetric Anaesthesia, School of Medicine and Pharmacology, University of Western Australia, Perth, WA, for his detailed review.

Back to Top | Article Outline


1. Harrison EA, Mackersie A, McEwan A, Facer E. The sitting position for neurosurgery in children: a review of 16 years experience. Br J Anaesth 2002;88:12–7.
2. Hedge RT, Avatgere RN. Air embolism during anaesthesia for shoulder arthroscopy. Br J Anaesth 2000;85:926–7.
3. Derouin M, Couture P, Boudreault D, et al. Detection of gas embolism by transesophageal echocardiography during laparoscopic cholecystectomy. Anesth Analg 1996;82:119–24.
4. Gildenberg PL, O‘Brien RP, Britt WJ, Frost EA. The efficacy of Doppler monitoring for the detection of venous air embolism. J Neurosurg 1981;54:75–8.
5. Furuya H, Suzuki T, Okumura F, et al. Detection of air embolism by transesophageal echocardiography. Anesthesiology 1983;58:124–9.
© 2005 International Anesthesia Research Society