Volume 12, No. 2 • Summer 1997

CRNA Experience with Sevoflurane

Beverly C. Collins, CRNA

CRNA Details Experience with and Analysis of Contaminated Sevoflurane

Editor’s Note: This letter has also been submitted to the American Association of Nurse Anesthetists’ Journal but is reproduced here because of its additional information about this issue of general interest. A follow-up explanatory letter from the manufacturer of sevoflurane was published in the Spring issue of the APSF Newsletter.

To the Editor

In a letter dated November 18, 1996, Abbott Laboratories Hospital Products Division asked directors of pharmacies, directors of anesthesia departments, distributors and wholesalers of sevoflurane, and hospital administrators to return bottles of sevoflurane from a specific lot of sevoflurane (Lot #10-590-DK). Recall was prompted by “… the development in the field of acidic levels outside (Abbott’s) acceptable range. This occurrence may be associated with slight pressure buildup and the presence of a pungent odor and/or cloudiness in the liquid.” Hydrogen fluoride, a highly toxic agent, was suggested to be the contaminant. A second letter issued to Chiefs of Anesthesiology and Directors of Pharmacy on January 22, 1997, repeated the concerns expressed in the letter of November 22, adding that the exposure to 50 ppm of hydrogen fluoride may produce “… respiratory tract irritation (coughing, laryngitis, bronchospasm, etc.) respiratory tract injury (pulmonary edema, respiratory failure, pulmonary fibrosis) tachycardia and hypertension.” A measured concentration of hydrogen fluoride in the affected lot was said to be 863 ppm. Additionally, this letter suggested that silicon tetrafluoride, presumably resulting from the reaction of hydrogen fluoride with the glass bottle, contributed to the pungent odor. It was not clear whether silicon tetrafluoride or hydrogen fluoride produced the odor.

Dr. John Leary wrote of his experience with a contaminated bottle, confirming the pressure buildup and acrid odor.1 This letter reports similar observations and extends the information available in two ways pertinent to AANA members. First, notice of the contamination of sevoflurane and how to recognize it has not been sent to individual members of the AANA or the ASA except in the report by Leary. Second, we supply further information concerning the composition of the contaminants.

On November 11, 1996, while filling a vaporizer with a previously opened bottle of sevoflurane (Lot #10-590-DK), one of our staff noted a strong and foul odor emanating from the bottle. A burning sensation in the nose and mouth persisted for 15-20 minutes after this brief exposure. I confirmed this finding, and all sevoflurane vaporizers were immediately removed from service. Because each of our machines had sevoflurane vaporizers, and because we could not exclude contamination of these and the associated equipment, all surgery was canceled until the machines could be replaced. Examination of other sealed bottles from Lot #10-590-DK revealed a normal odor in some but not all. Of a total of 12 bottles from the affected lot, five were contaminated. We reported our findings to Abbott Laboratories and the FDA, and samples from contaminated bottles were sent to Abbott and Clemson University for analysis. With cooperation from Abbott Laboratories, Ohmeda brought in rental machines for us to use.

At our request, Dr. Melissa Riley, Associate Professor, Department of Plant Pathology and Physiology, Clemson University, analyzed the contents of three of the contaminated bottles. Contaminated bottles had a sharp, pungent, acidic, irritating odor relative to the non-pungent, non-irritating odor of standard sevoflurane. Contaminated bottles contained a thread-like precipitate, and a white crystalline residue developed around the caps of opened bottles. The pH in three bottles (Lot #10-590-DK) with pungent odor ranged from 2.0 to 2.85, compared to a sevoflurane standard (Lot #20-023-DK) with pH 5.8 and the pH1.0 found by Leary for a contaminated bottle.1 Pressure increased in bottles with the pungent odor when they were closed for several days.

Multiple Contaminants

Gas chromatography/mass spectroscopy analysis by Riley of the three bottles (Lot #0-590-DK) and one normal standard (Lot #20-023-DK) gave various results. All analyses showed a peak consistent with pure sevoflurane. The bottle which did not have a pressure buildup or pungent odor yielded one major peak with spectra consistent with standard sevoflurane. However, the contaminated bottles yielded five major peaks and several minor peaks. Some of these peaks were consistent with impurities of sevoflurane such as (CF3)2CHOCH2OCH3 and (CF3)2CHO(CH2O)2CH3. Another contaminant appeared to be (CF3)2CHOCH2OCH(CF3)2. Toluene-2, 4-diisocyanate (also called benzene-2,4-diisocyanate-l-methyl) appeared in several analyses, but was not found consistently. This compound is used in the manufacture of polyurethane foams and other elastomers and has a sharp pungent odor. No hydrogen fluoride was detected in any vapor sample. Definitive identification of the structure and toxicity of all contaminants has not been made. It would appear that “acidic sevoflurane”2 contains more than hydrogen fluoride or silicon tetrafluoride.

The affected lot was manufactured in October 1995 with the last release of product to distributors in March 1996. Abbott believes that all bottles from the affected lot have been accounted for, and that no patient or clinician experienced injury from the contamination. We are not aware of new reports by clinicians of contamination after mid-January 1997. However, we learned that there was a similar event in South America months prior to our situation, and anesthesia providers in the U.S. were not made aware of the potential problem.

No patient at our hospital appears to have been injured, nor did Leary find injury at his hospital. The most likely organ injured by hydrogen fluoride would be the lungs. The only publication associating sevoflurane and pulmonary injury comes from Japan.3 Iwasaki et al. reported five cases of pulmonary edema after sevoflurane anesthesia. They suggested that the injury resulted from a variety of causes, usually radical circulatory changes. They did not invoke the possibility of an effect of contaminated sevoflurane. It does not appear that the contaminants adversely affect vaporizer function. It is not known if the materials leading to contamination could lead to degradation of standard sevoflurane in the vaporizer. Discovery of the cause of the contamination might relieve any further concern. As noted by Leary,1 “presently, it is not known whether the cause might affect sevoflurane in the vaporizer (e.g., produce degradation of sevoflurane and further production of acid) and whether the generation of acid or other by-products would influence vaporizer function.” Abbott recommends that if it is suspected that contaminated sevoflurane has come in contact with vaporizers or anesthetic machines, the equipment should be immediately removed from service and returned to the manufacturer. It appears that recall contained the problem posed by contaminated sevoflurane in Lot #10-590-DK. However, as suggested by Leary, our own experience and the findings of others, the anesthesia practitioner should be aware of the possibility of contamination and how to recognize it. Bottles of contaminated sevoflurane may have an increased pressure that reveals itself as a hiss on opening the bottle. A white precipitate crystal may form on the cap or the lip of the opened bottle. Most obvious is the strong, acrid, irritating odor of the vapor. Such signs should alert the anesthesia practitioner to the possibility of sevoflurane contamination.

Beverly C. Collins, CRNA Lincoln Medical Center Lincolnton, NC

I thank Gil Peterchuck, CRNA, Dr. Richard Symington, Dr. Edmond Eger, and Dr. Melissa Riley for their suggestions regarding the construction of this report.


  1. Leary JP. Sevoflurane containing toxic acid prompts recall. Contaminated sevoflurane use reported from NY State. Anesthesia Patient Safety Foundation Newsletter 1997;11:37-39 (letter).
  2. Callan CM. Sevoflurane containing toxic acid prompts recall. Sevo manufacturer outlines circumstances, response. Anesthesia Patient Safety Foundation Newsletter 1997;11:37- 39.
  3. Iwasaki H, Omote T, Hamada I, Nakamura I, Mamiki A. Acute pulmonary edema in five patients undergoing sevoflurane anesthesia. Masui (Jap J Anesthesiology) 1992;41: 1183-1187.