Carbon monoxide poisoning: pulse oximetry missed carbon dioxide toxicity.

Clinical bottom line (level 4)

  1. Among patients with suspected smoke inhalation or carbon monoxide poisoning, pulse oximetry was inaccurate. The higher the carboxyhaemoglobin level, the greater the inaccuracy.
  2. Consequently, a direct blood gas measurement of the HbCO level will be necessary in all cases.
Buckley et al: Annals of Emergency Medicine 1994; 24: 252-255
Expires July 2003

The study

Setting: emergency department, county hospital, USA

16 patients (aged ?, ?% male) adults with suspected carbon monoxide intoxication

Excluded if
  • <17 years old
  • sustained oximetry signal could not be obtained
  • pulse oximetry and blood gas sampling not performed simultaneously


    • Some patients may have received initial treatment with 100% oxygen.
    Independent unblinded reference standard, applied in all patients from a consecutive ?appropriate spectrum.
    Reference standard:
    • arterial blood gas- oxyhaemoglobin saturation, and HbCO level
    Diagnostic test: finger probe oximetry saturation

    The evidence

    • In all patients, pulse oximetry was 96-100% despite carboxyhaemoglobin ranging from 2.2 to 44% (mean 17%).
    • The difference between the pulse oximetry saturation and the measured arterial oxyhaemoglobin saturation rose linearly with the HbCO, and closely approximated the measured HbCO level (adjusted correlation coefficient= 0.90).

    Comments

    1. The study suggests that pulse oximetry may be inaccurate in smokers as well.

    Citation

    1. Buckley RG, Aks SE, Eshoum JL, et al: The pulse oximetry gap in carbon monoxide intoxication. Annals of Emergency Medicine 1994; 24: 252-255
    Search Terms: carbon monoxide, carbon monoxide poisoning and poisoning in Cochrane and Ovid Medline
    Contributor: Joel Ray and Chris Ball, July 2000
    Reviewer:

    Clinical Question.
    Patient suspected smoke inhalation
    Intervention or Exposure pulse oximetry
    Outcome CO poisoning