False Carboxyhemoglobin
False carboxyhemoglobin readings occur through two primary mechanisms: falsely elevated readings from standard pulse oximeters that cannot differentiate COHb from oxyhemoglobin, and falsely elevated laboratory measurements in infants due to fetal hemoglobin interference with spectrophotometry. 1, 2
Understanding the Problem
False carboxyhemoglobin measurements create dangerous clinical scenarios where:
- Standard pulse oximeters (2-wavelength) will show falsely normal oxygen saturations (SpO2 >90%) even when COHb levels exceed 25%, because COHb and oxyhemoglobin have similar light absorbance at 660 nm 1
- Older blood gas machines without CO-oximeters calculate oxygen saturation from PaO2 and pH alone, reporting normal SaO2 (97-98%) even when 40% of hemoglobin is bound to CO 1
- In infants under 6 months, fetal hemoglobin interferes with spectrophotometric COHb analysis, producing falsely elevated readings in clinically well-appearing babies 2
Critical Pitfalls to Avoid
The Pulse Oximetry Trap
A patient with 40% COHb and PaO2 of 100 mmHg will show:
- Pulse oximeter reading: SpO2 ~90-95% (falsely reassuring)
- Calculated SaO2 from old blood gas machines: 97-98% (falsely reassuring)
- True oxygen-carrying capacity: only 60% 1
This creates a false sense of security. Never rely on pulse oximetry or calculated oxygen saturation to rule out CO poisoning.
The Infant Exception
In a 3-month-old exposed to smoke, persistently elevated COHb levels (despite clinical wellness and oxygen therapy) likely represent fetal hemoglobin interference rather than true CO poisoning 2. This can lead to unnecessary prolonged ED monitoring and oxygen therapy.
How to Evaluate Properly
Always confirm CO poisoning with laboratory spectrophotometry (CO-oximetry) that directly measures COHb, not calculated values 1:
- Obtain arterial or venous blood (levels are similar when equilibrated) 1
- Send to laboratory CO-oximeter that measures multiple hemoglobin species by spectrophotometry at multiple wavelengths
- Diagnostic thresholds:
When Laboratory Values Don't Match Clinical Picture
Recent case reports demonstrate both false negatives and false positives can be clinically significant 3:
- False negative: Patient with clear CO exposure showing COHb of 1.8% on initial ED measurement, but 20.1% on later analysis of the same blood sample (machine calibration was within specification) 3
- False positive: Non-smoking patient in all-electric home with COHb 7.1%, later confirmed as 1.9% by gas chromatography/mass spectrometry 3
When COHb levels conflict with clinical presentation, trust your clinical judgment and repeat the measurement 3. Consider:
- Time elapsed since exposure (COHb may have already declined)
- Oxygen therapy already administered
- Presence of other exposed individuals with elevated levels
- Environmental CO measurements
Management Approach
Regardless of COHb measurement delays or uncertainties, immediately administer 100% oxygen to any patient with suspected CO poisoning 1. The clinical diagnosis drives initial treatment, not the laboratory value.
For Fingertip Pulse CO-Oximetry
The newer pulse CO-oximeters (available since 2005) have questioned accuracy 1. If using pulse CO-oximetry for screening, always confirm with laboratory spectrophotometry before making decisions about hyperbaric oxygen therapy 1, given the inconvenience, cost, and risks of transfer.
Special Consideration: High COHb Levels
Only when COHb exceeds 40% does the difference between pulse oximeter and calculated SaO2 become apparent (SaO2 approximately 5% higher than SpO2) 1. At lower toxic levels, both remain falsely reassuring.
Key Clinical Points
- Never use pulse oximetry alone to exclude CO poisoning - it will appear normal even with lethal COHb levels 1
- In infants <6 months with suspected CO exposure, interpret elevated COHb cautiously if the infant appears clinically well, as fetal hemoglobin causes false elevation 2
- When laboratory COHb conflicts with strong clinical suspicion, repeat the measurement and consider alternative testing methods 3
- COHb levels serve only to confirm diagnosis, not predict symptoms or outcomes - treat based on clinical presentation 1