hydroxocobalamin, since hydroxocobalamin is hardlydistributed in cellular blood components.

Thus, the expected interference of hydroxocobalamin withall the measured parameters is in the range of up to 5%. This iswithin the common variability of the applied analyticalmethodologies. Furthermore, values measured with a particularco-oximeter cannot be reliably extrapolated to other equipment.

Additionally, hydroxocobalamin concentration decreasesrapidly as a result of the initial distribution in the body.Therefore, subsequent measurements are even less subject tointerference.

To summarize, this publication, although scientificallyinteresting, has only low impact on real treatment decisions inemergency situations.

It may be recommended to obtain a blood sample forcarboxyhemoglobin prior to administration ofhydroxocobalamin in order to avoid any potential forinterference.

Stephen W. Borron, MDUniversity of Texas Health Science Center at San AntonioSan Antonio, TX

Wolfgang Uhl, MDArno Nolting, PhDUlrike Hostalek, MDMerck SeronoDarmstadt, Germany

doi:10.1016/j.annemergmed.2007.05.027

1. Lee J, Mukai D, Kreuter K, et al. Potential interference byhydroxocobalamin on co-oximetry hemoglobin measurements duringcyanide and smoke inhalation treatments. Ann Emerg Med. 2007;49:802-805.

2. Feldmeier JJ, Ed. Hyperbaric oxygen 2003: Indications and results:The hyperbaric oxygen therapy committee report. Kensington, MD:Undersea and Hyperbaric Medical Society; 2003:11-18.

3. Hampson NB, Dunford RG, Kramer CC, et al. Selection criteriautilized for hyperbaric oxygen treatment of carbon monoxidepoisoning. J Emerg Med. 1995;13:227-231.

4. Hampson NB, Scott KL, Zmaeff JL. Carboxyhemoglobinmeasurement by hospitals: implications for the diagnosis ofcarbon monoxide poisoning. J Emerg Med. 2006 Jul;31:13-16.

Clarifications Regarding Interference ofHydroxocobalamin With CarboxyhemoglobinMeasurements in Victims of Smoke Inhalation

To the Editor:I read with great interest the study done by Lee et al

regarding interferences of hydroxocobalamin with co-oxymetermeasurements.1 Co-oxymeters are highly valuable tools thathave improved the frequency of diagnosis of carbon monoxidepoisonings.2 However, co-oxymetry cannot be considered as areference method for determination of carboxyhemoglobin.2,3 A

number of interferences have been reported, including lipid-induced turbidity, hyperbilirubinemia, sulfhemoglobin, fetalhemoglobin, and methylene blue.4 The interferences aredependent on the method of measurement, precludinggeneralization of the results.4

The authors limited their study to interference withhydroxocobalamin. However, hydroxocobalamin will beimmediately transformed into cyanocobalamin. In the future,studies aimed at determining interferences of cobalamins mustinclude not only those caused by hydroxocobalamin but alsothose caused by cyanocobalamin.

The authors spiked 5 ml blood sample with 12 mghydroxocobalamin to obtain hydroxocobalamin concentrationsof 1.784, 0.892, and 0.382 mM in whole blood. However,hydroxocobalamin is measured in plasma. Therefore, assuminga hematocrit of 50%, the addition of 12 mg ofhydroxocobalamin to whole blood would result in plasmaconcentrations 2-fold greater than those expected. Data aboutplasma hydroxocobalamin concentrations in humans suggestthat maximal measured concentrations are about 600 �mol/l(range: 267-1011 �mol/l,5 far less than the range studied by Leeet al.1

The use of normal carboxyhemoglobin to assess theinterferences of hydroxocobalamin is questionable. Co-oxymeters are not sufficiently sensitive for very lowcarboxyhemoglobin levels.6 Gourlain et al has shown theinterference of hydroxocobalamin with carboxyhemoglobin.They concluded that the interference of hydroxocobalamin withco-oxymeter measurement, while being moderate, should betaken into account. While being statistically significant, thedifferences are neither of clinical relevancy nor misleading.Indeed, the presence of hydroxocobalamin consistently resultedin slight overestimation of carboxyhemoglobin. The majorconcern in carbon monoxide poisoning is underdiagnosis ratherthan mild overestimation. Carboxyhemoglobin may be used toconsider hyperbaric oxygen. Values of interest are about 15%and 25%. Assuming the same percentage of interference, thefalse values would be 15.9% and 26.6%, respectively. It is ratherunlikely that such limited interferences would significantlyimpair the indications of the attending physician. In clinicalpractice, the 2 major pitfalls we have to face when dealing withthe analytical assessment of carbon monoxide poisoning are: 1)any delay in blood sampling, especially if the patient is treatedwith oxygen, and 2) significant delays between collection ofblood and measurement using co-oxymeters.6 These biaseswould have a far greater impact on carboxyhemoglobinmeasurement than the transient presence of high plasmahydroxocobalamin concentrations.

In cases of suspected carbon monoxide poisoning, emergencycare providers are advised to collect a blood specimen as soon aspossible. Smoke inhalation is a setting where carbon monoxideexposure is highly likely. Consequently, emergency careproviders should collect a blood specimen on anticoagulant forcarboxyhemoglobin measurement while inserting a venous line

Correspondence

Volume , . : November Annals of Emergency Medicine 625

and before hydroxocobalamin administration in fire victimssuspected of smoke-induced cyanide poisoning.

Frederic Baud, MDRéanimation Médicale et Toxicologique.Hôpital LariboisièreParis, France

doi:10.1016/j.annemergmed.2007.05.026

1. Lee J, Mukai D, Kreuter D, et al. Potential interference byhydroxocobalamin on cooximetry hemoglobin measurements duringcyanide and smoke inhalation treatments. Ann Emerg Med. 2007;49:802-805.

2. Duenas-Laita A, Ruiz-Mambrilla M, Gandia F, et al. Epidemiology ofacute carbon monoxide poisoning in a Spanish region. J ToxicolClin Toxicol. 2001; 39:53-57.

3. Allred EN, Bleecker ER, Chaitman BR, et al. Effects of carbonmonoxide on myocardial ischemia. Environ Health Perspect. 1991;91:89-132.

4. Gouget B, Gourlain H. Présentation des caractéristiques des CO-oxymètres. Revue Européenne de Technologie Biomédicale. 1995;17:243-246.

5. Forsyth JC, Mueller PD, Becker CE, et al. Hydroxocobalamin as acyanide antidote: safety, efficacy and pharmacokinetics in heavilysmoking normal volunteers. J Toxicol Clin Toxicol. 1993;31:277-294.

6. Gourlain H, Buneaux F, Levillain P. Mesure du CO et de la COHbdans le sang: interférences de l’hydroxocobalamine et du bleu deméthylène en CO-oxymétrie. Revue Française des Laboratoires.1996;282:144-148.

In reply:We agree that the magnitude of the potential interference in

CO hemoglobin level readings caused by hydroxocobalamin islikely to be small and not be clinically significant under normalcircumstances. The degree of interference should not limitclinical use of the cyanide antidote hydroxocobalamin in

suspected or confirmed cyanide poisoning, or alterdecisionmaking with regard to patient management and we donot want to overestate the possible clinical impact. However,co-oximetry and reporting of CO hemoglobin is relativelyroutine and rapid in many hospitals, particularly burn centerswhere this situation might be expected to be encountered.Therefore, we think it may be important to know about thispotential interference in rare circumstances that might arise suchas hydroxocobalamin overdose or in patients in whom a bloodsample is obtained during a bolus infusion. Furthermore,because of the broad absorption peak of hydroxocobalamin over450 to 600 nm range (and cyanocobalamin as well), thepotential interference would potentially be seen with mostcommercial co-oximeters or any commercial or experimentaloptical devices that employ these wavelength ranges in theirmeasurement.

Jangwoen Lee, PhDBruce Tromberg, PhDLasesr Microbeam and Medical ProgramBeckman Laser InstituteUniversity of California, IrvineIrvine, CA

David Mukai, BSKelly Kreuter, MASari Mahon, PhDMatthew Brenner, MDDivision of Pulmonary and Critical Care MedicineDepartment of MedicineUniversity of California, IrvineIrvine, CA

doi:10.1016/j.annemergmed.2007.06.006

Correspondence

626 Annals of Emergency Medicine Volume , . : November