To the Editor:—

In a previously described method, 1the use of a breath alcohol analyzer combined with the addition of an ethanol marker to irrigation fluid allowed the detection of rapid fluid absorption during operative hysteroscopy. Our usual anesthetic technique for such procedures involves propofol induction and maintenance with periodic end-tidal breath alcohol determinations to detect the presence of ethanol from absorbed irrigation fluid. The device used to detect the ethanol marker is an Alco-Sensor III (Intoximeters, Inc., St. Louis, MO). This device analyzes discrete samples taken as needed from the patient’s expired gas flow and uses fuel-cell technology together with an algorithm to provide a readout. This readout is calibrated to correspond with g/100 ml of blood ethanol and is approved by the U.S. Department of Transportation for evidentiary use. The device is not, however, specific for ethanol, and it will respond to other alcohols as well. Because the algorithm for producing the readout depends on the assumption that the alcohol measured is ethanol, it is not known how other substances relate to the device readout.

With the introduction of an alternative formulation of propofol, pharmacoeconomic considerations led to the use of the preparation manufactured by Gensia-Sicor (Irvine, CA). Shortly thereafter, it was noted that a patient undergoing hysteroscopic surgery showed positive breath alcohol levels before the start of surgery. This was initially attributed to alcohol use by the patient; however, more than 12 subsequent patients administered the Gensia-Sicor formulation were observed, and all were noted to have positive breath alcohol readings only after induction of anesthesia, but before surgery. It was also noted that discontinuing the propofol infusion and switching to sevoflurane eliminated the presence of positive readings and that the reinstitution of the infusion caused a reappearance of the positive readings. Further, positive breath alcohol readings were never present if the brand-name propofol, Diprivan (Astra-Zeneca, Wilmington, DE), was used. The readout on the device after a standard induction dose of the Gensia-Sicor propofol was from 0.010 to 0.020 gm/100 ml, a reading that, if a result of ethanol from fluid absorption, would cause discontinuation of the procedure. The response characteristics of the instrument during these circumstances are also curious. Upon taking a sample of known ethanol, the reading increases and stabilizes within a few seconds, whereas when measuring the sample of a patient administered the Gensia-Secor preparation, the reading increases over the course of as much as 1 min before stabilizing.

These findings have been discussed with the manufacturers of both the generic and the brand-name drugs, but, to date, neither has supplied an explanation. There are differences in the formulation of the two preparations, but no alcohols are added to the Gensia-Sicor formulation. Whatever the cause, further investigation of the source of the phenomenon is warranted because the use of breath alcohol sampling by fuel-cell–based instruments is unreliable in patients administered generic propofol.

O’Connor T: Hyponatremic encephalopathy after endometrial ablation. JAMA 1994; 271:343–5