A sudden increase in end-tidal nitrogen (ETN2) concentration during craniotomy is usually attributed to venous air embolism (VAE), and requires urgent measures to prevent serious cardiovascular and neurologic sequelae [1-3]. This report describes a more benign reason for the occurrence of this phenomenon, which arose due to changing volatile drugs during a neurosurgical procedure. If increases in ETN2 are to remain diagnostic of VAE and not result in deleterious alterations in patient management, all false-positive sources of nitrogen within the anesthesia system should be controlled or eliminated.
A 67-yr-old female weighing 73 kg was undergoing a frontotemporal craniotomy to clip an anterior communicating artery aneurysm which had bled 3 days earlier. She was anesthetized with 1% isoflurane in 66% nitrous oxide and 33% oxygen using an Ohmeda Modulus II anesthetic machine and an Ohmeda 7810 ventilator (Ohmeda, Madison, WI). Fentanyl 250 micro gram, pancuronium 10 mg, and a sodium nitroprusside infusion of 1 micro gram centered dot kg-1 centered dot min-1 were also administered. She was positioned supine with 5 degrees of head-up tilt. Using a circle anesthesia system with soda-lime absorber, ventilatory settings had remained unchanged since the beginning of the case in order to obtain moderate hyperventilation (PaCO2 25-30 mm Hg) with tidal volumes of 800 mL and a respiratory frequency of 12 breaths/min. Fresh gas flow from the anesthetic machine had been unchanged for 2 h at 3 L/min. Monitoring included a 20-gauge left radial artery catheter, a 7.5-Fr right heart catheter inserted via the right internal jugular vein, precordial Doppler probe, electroencephalogram, and an Ohmeda Rascal Registered Trademark II Anesthesia Gas Monitor (Ohmeda, Salt Lake City, UT) to record and display inspired and expired anesthetic gas concentrations, including nitrogen.
After an uneventful clipping of the aneurysm, the surgeon was obtaining final hemostasis before preparing to close the dura. At this point isoflurane was discontinued and 4% desflurane substituted in the belief that emergence at the end of the procedure, still approximately 1 h away, would be more rapid and allow more prompt neurologic evaluation. The fresh gas flow remained unchanged at 3 L/min. Within 20 s of turning on the fully filled Tec 6 desflurane vaporizer, the ETN2 concentration, which had remained stable at 0.4% for 2 h, suddenly increased to 1.3% Figure 1. ETN2 variations during steady-state anesthesia are very small, and a change of 0.1% is considered clinically important . No disturbances characteristic of VAE were heard from the precordial Doppler. No bubbles could be aspirated from the right heart catheter. No change in end-tidal CO2 was detected, and all hemodynamic variables remained at their previous levels. A peripheral nerve stimulator revealed complete paralysis of the adductor pollicis muscle. The capnograph did not show any waveform change suggestive of diaphragmatic movement. Close inspection of the N2 waveforms revealed an increase in inspired N2 immediately prior to the increase in ETN2. Within 5 min, ETN2 concentration had returned to its baseline value of 0.4% Figure 1. The remainder of the operation was uneventful, and the patient was tracheally extubated in good condition at its conclusion.
A subsequent machine check did not reveal any vaporizer fault or malfunction, and no leak was discovered at the Selectatec Registered Trademark (Ohmeda Inc., Madison, WI) mountings on the backbar. The anesthetic gas scavenging system was also found to be functioning appropriately.
Subsequently, 10 adults undergoing general surgical procedures were briefly studied in an attempt to repeat this incident. Nitrous oxide, oxygen, and isoflurane were administered to all patients for the initial phase of the anesthetic. Fresh gas flow was maintained at 3 L/min, a level identical to that in the case described above. When a steady state had been reached and ETN2 was 0.4%-0.6%, isoflurane was discontinued and the desflurane vaporizer set to deliver 4% drug while N2 levels in the circuit were monitored. During the preoperative machine check, the filled desflurane vaporizer had been flushed with air for 30 s before switching it off so as to "contaminate" with N2 any available space within the vaporizer above the desflurane sump. All cases transiently demonstrated a 1.5%-2.0% increase in both inspired and ETN2 levels within 20 s of switching on the desflurane vaporizer.
The sudden, transient increase in ETN2 concentration coincided with switching on the Tec 6 desflurane vaporizer, which must have contained air. This would explain the increase in inspired N2 immediately prior to the increase in ETN2Figure 1, suggesting that the N2 in end-tidal gas originated from the anesthetic machine and not some extraneous source. Exactly where within the Tec 6 vaporizer this air was trapped is unclear, since in this novel device no fresh gas enters the vaporizing chamber [4,5]. The N2 "contaminating" the desflurane vapor presumably arose from a previous anesthetic in which air and desflurane had been administered. When desflurane was discontinued at the end of that earlier anesthetic, air remained "trapped" in the vaporizer until it was turned on during this craniotomy after the aneurysm had been clipped.
Furthermore, filling the desflurane vaporizer during an anesthetic did not result in any detectable increase in inspired or ETN2, suggesting that air does not enter the vaporizing chamber during this process. A full Tec 6 vaporizer contains 425 mL of liquid desflurane [4,5].
VAE is a well recognized hazard of several procedures, particularly those involving craniotomy. As little as 1 mL of venous air can have serious consequences should paradoxical passage occur to the arterial side of the circulation via a patent foramen ovale. Therefore, several measures are taken to detect small volumes of air in the circulation [6,7]. These include the use of transesophageal echocardiography, precordial Doppler, and analysis of end-tidal gas for N2 and CO2. An increase in ETN2 is a more sensitive and specific diagnostic indicator of a large VAE than a decrease in end-tidal CO2 or a change in pulmonary artery pressure [1,8]. Its early detection during craniotomy must be taken seriously and treated promptly before volumes of air large enough to impair right ventricular ejection or cause paradoxical air embolism are entrained. Intraoperative VAE may occur in the absence of changes in precordial Doppler sounds, making continuous measurement of ETN2 vitally important [2,9]. Suspicion of VAE often requires surgical intervention, such as flooding the operative field and repositioning the patient, both of which interfere with surgery and could complicate some delicate intracranial procedures. However, the priority in cases of suspected VAE is to prevent further air entrainment and treat any hemodynamic instability before searching for a "false-positive" source of ETN2.
The case described above shows that an increase in ETN2 may occasionally be benign and result from switching vaporizers during anesthesia. Nevertheless, the increase in ETN2 is large enough to implicate VAE and warrants that protective maneuvers be taken by both surgeon and anesthesiologist unless the benign source of N2 is realized. It can be differentiated from true VAE by its temporal relation to switching on a vaporizer, and by the simultaneous increase in inspired N2. It may be avoided altogether by preoperatively flushing all vaporizers with N2 O/O2 to ensure removal of any air that might remain within the vaporizer after a prior anesthetic that included air as a carrier gas. Other causes of an abrupt increase in ETN2 include a disconnection or leak in the anesthesia circuit or endotracheal tube cuff, resulting in intraoperative air entrainment into the breathing system , particularly if any inspiratory effort is made . A sudden reduction in fresh gas flow into a circle anesthesia system will also produce an apparent increase in ETN2[8,11]. However, none of these differential causes were present in the case reported, strongly implicating the vaporizer as the source of nitrogen.
1. Matjasko J, Petrozza P, Mackenzie CF. Sensitivity of end-tidal nitrogen in venous air embolism detection in dogs. Anesthesiology 1985;63:418-23.
2. Cucchiara RF, Nugent M, Seward JB, Messick JM. Air embolism in upright neurosurgical patients: detection and localization by two-dimensional transesophageal echocardiography. Anesthesiology 1984;60:353-5.
3. Cucchiara RF, Black S, Steinkeler JA. Anesthesia for intracranial procedures. In: Barash PG, Cullen BF, Stoelting RK, eds. Clinical anesthesia. Philadelphia: JB Lippincott, 1989:849-76.
4. Dorsch JA, Dorsch SE. Understanding anesthesia equipment. 3rd ed. Baltimore: Williams & Wilkins, 1994:116-22.
5. Graham SG. The desflurane Tec 6 vaporizer. Br J Anaesth 1994;72:470-3.
6. Gronert GA, Messick JM Jr, Cucchiara RF, Michenfelder JD. Paradoxical air embolism from a patent foramen ovale. Anesthesiology 1979;50:548-9.
7. Bedell EA, Berge KH, Losasso TJ. Paradoxic air embolism during venous air embolism: transesophageal echocardiographic evidence of transpulmonary air passage. Anesthesiology 1994;80:947-50.
8. Matjasko MJ, Hellman J, Mackenzie CF. Venous air embolism, hypotension, and end-tidal nitrogen. Neurosurgery 1987;21:378-82.
9. Michenfelder JD, Miller RH, Gronert GA. Evaluation of an ultrasonic device (Doppler) for the diagnosis of venous air embolism. Anesthesiology 1972;36:164-7.
10. Lanier WL. Intraoperative air entrainment with Ohio Registered Trademark Modulus Trademark anesthesia machine. Anesthesiology 1986;64:266-8.
11. Matjasko J, Gunselman J, Delaney J, Mackenzie CF. Sources of nitrogen in the anesthesia circuit. Anesthesiology 1986;65:229.