Volume 9, No. 2 • Summer 1994

CO Poisoning During Anesthesia Poses Puzzles: New Agent Used in Florida Case

Robert E. Lentz, M.D.

Previous reports regarding the diagnosis of carbon monoxide exposure from an unknown source have suggested that there is an interaction of the halogenated volatile anesthetics and the carbon dioxide absorbent in the breathing circuit, particularly when an anesthesia machine which delivered volatile anesthetic then sat unused for a period of time. As new halogenated anesthetics become available and are used, it remains to be observed whether they too will be associated with this apparent phenomenon. Reported here is an instance where this appears to be so.

Report of a Case

A 46-year-old white female was scheduled as an outpatient for septoplasty, endoscopic bilateral anterior ethmoidal sinus surgery, and excision of a left tonsillar cyst. During her pre-op interview, the patient denied any cardiac or respiratory history. The patient also denied any prior anesthetics and she was not taking any chronic medications. The patient did, however, admit to being a smoker, with a 20 pack/year smoking history.

Routine pre-op labs were within normal limits, and specific values were: Hgb 14.1, Na 141, K 3.8, Cl 108, C02 23, and Ca 9.6.

While in the holding area, an intravenous line was placed with D5LR infusing at 100cc/hr. The patient then received glycopyrolate 0.2mg IV with midazolam 2mg IV for sedation. The patient was taken to the operating room where she was placed on the table, positioned, and monitors including NIBP, pulse oximetry, and EKG were placed. The patient was pre-oxygenated and general anesthesia was induced with alfentanil 500 mg, lidocaine 70 mg, atracurium 5 mg (defasiculating dose), propofol 100 mg, and succinylcholine 100 mg intravenously. The patient was intubated with a 7.0 cuffed endotracheal tube under direct vision. The tube was secured at 20 cm at the lips, breath sounds were equal bilaterally, and EtCO2 was noted on the anesthetic gas monitor. The patient was then placed on the ventilator with the settings of TV 600 n@, Rate 10, PIP 20 cmH20. An additional 10 mg of atracurium was given IV to complete the balanced anesthetic technique. Prior to induction, the patient’s vital signs had been: BP 140/88, Heart Rate 95, SaO2=97% (room air), Temperature 36.5 C. Following induction, the vital signs were: BP 110/60, HR 95, SaO2=100% (FiO2=40%), Temp 36.5 C.

Approximately 40 minutes into the case, the patient’s 02-Hgb saturation decreased to 96% over a period of 2-3 minutes. The pulse oximeter probe was inspected to verify proper placement on the finger. Breath sounds remained equal bilaterally, without wheezes, and there was no change in PIP. The endotracheal tube was also checked for its position, and it was noted to still be secured at 20 cm at the lips. At this point, the patient was placed on 100% 02 and hand ventilated with up to 40 cmH20 pressure. This also failed to bring the patient’s O2 saturation above 96%. The surgeon was made aware of these findings and was asked to complete the procedure as quickly as possible.

Arterial blood gases were sent to check the 02 saturation a COHb level was also requested. The blood gas report read the following values: pH 7.46; PC02=28 mm Hg; P02 467 mm Hg; HCO3 20.3 MEq/l; COHB 31.5%. At this point, the patient’s 02-Hgb saturation remained at 97%. The surgeon was made aware of the new findings, and the procedure was completed over the next 10 minutes. The entire time interval from when the 02Hgb saturation started to decrease to the completion of the surgery was 30 minutes.

It was decided to leave the patient intubated, on 100% 02, and ventilated in the Post Anesthesia Care Unit, until the level of COHB decreased enough to permit safe extubation. In the PACU, the patient was placed on a ventilator with the same settings as were given in the OR with the addition of 5 cmH20 PEEP. The patient arrived in the PACU at 0935 and was noted to be pale and unresponsive to verbal or tactile stimuli. At 0940 the patient became more reactive, requiring midazolam 1 mg for sedation. The 02 saturation by now was 99%. At 1000 another blood gas was drawn, and the results were the following: pH 7.56; pCO2 20 mm Hg; pO2 482.5 mm Hg; HCO3 18.5 MEq/l; COHB 12.4%. The ventilator settings remained unchanged, and at 1030 another blood gas was drawn which showed continued improvement as the COHB level had decreased to 5.3%. At this time, the patient was responsive to verbal stimuli and was able to maintain head lift for five seconds. At 1040, the patient was extubated uneventfully and placed on a venti-mask. The patient maintained her 02-Hgb saturation at 98100% with a respiratory rate of 16 breaths per minute. At 1100 the mask was discontinued and the patient received “blow-by’ 02 only. She continued to maintain her 02-Hgb saturation at 98-1 00%.

At 1200, the patient was transferred to her room and monitors were discontinued. At 1800, the last blood gas was drawn, and the results were as follows: pH 7.44; pCO2 37.4, pO2 93.1 mm Hg; HCO3 24.2 mm Hg; COHB 2.6%. The patient was subsequently discharged, and she suffered no adverse sequelae.


Although carbon monoxide production has been associated with the other halogenated agents, this is the first documented case involving the newest inhalation anesthetic. In all of the other cases reported, the anesthesia machines sat idle for longer than 24 hours, the soda lime was not changed prior to the procedure, and the involved case was usually the first case of the day. Our patient was the first case on a Monday morning, and the anesthesia machine (after use of the volatile anesthetic on Friday) had not been used over the immediately prior weekend. Other similarities included the onset time to the desaturation, about 30-40 minutes after induction in all cases, and rapid recovery without injury to the patient.

The US FDA Center for Disease Control recommendations regarding this subject matter are as follows:

* All soda lime that has been dormant in the anesthesia machine for more than 24 hours should be changed, and dated.

* In addition to changing the soda lime, the anesthesia machine should also be flushed continuously with 100% 02 for at least one minute prior to the first case of the day.

In any patient who develops the type of hemoglobin desaturation described here and who fails to respond to the usual therapeutic measures used to correct this problem, do not hesitate to send either a venous or, preferably, an arterial blood sample for the possibility of COHB “poisoning’ (see article on page 13).

Dr. Lentz is chairman of the Department of Anesthesiology, Palms West Hospital, Loxahatchee, FL.