"Flu-Like" Illness and CO Poisoning
                ===================================

þ A pertinent reference is the study by Baker, et al. entitled
  Carboxyhemoglobin levels in children with nonspecific flu-like symptoms.
  (Journal of Pediatrics, September, 1988, pp. 501-504). They studied
  sublethal CO toxicity in a pediatric population presenting with headache,
  nausea, weakness, dizziness or altered mental status who were thought to
  have a flu syndrome. They excluded patients who had a fever, watery stools,
  or a viral rash and included patients who lived in homes with combustible
  fuel heat (oil, kerosene, wood stoves, etc.). In patients presenting between
  January 5 and March 30, 24 out of 46 had COHb levels above 2%. Six had
  levels over 10% (up to 27.6%), seven had levels between 5% and 10%, and 11
  had levels between 2% and 5%. They cited studies which showed that smokers
  may have levels between 5% and 9%. Since none of the children were smokers,
  and since the average COHb level in controls was 1.3% (range 0.6% - 1.9%),
  levels above 2% were considered abnormal and were treated with 100% oxygen.
  The number of smokers in the household did not differ between study groups.
  They state "It might be argued that children with COHb levels between 2% and
  5% have minimal symptoms and that their ill effects would likely be
  self-limited. We contend that the crucial importance of accurate diagnosis
  in this group is...the prompt recognition of a hazardous environment".

þ One study in adults looked at 55 patients who presented to the ED with
  flu-like symptoms (1).  Thirteen (23.6%) of them had CO levels greater than
  10%.  These levels were drawn in triage and no physician correctly diagnosed
  these patients as having subacute carbon moxoxide poisoning. COHb levels
  ranged from 0 to 21%.   No statistically significant difference was noted in
  vital signs at presentation between the groups with COHb levels > or < 10%.
  Smokers and non-smokers had similar levels.  The use of wood heating in the
  home did correlate with elevated levels.  It is not clear to me from a
  listing of their final diagnosis  whether or not the elevated COHb levels
  were coincindental to their symptoms.  The authors seem to think that they
  were not.

þ A study done on children (2) looked at a group of kids with both flu-like
  symtoms and a history of possible exposure. Levels above 2% were considered
  to be significant. A control group of 10 patients had an average level of
  1.3%.  Of 46 study patients 22 had levels less than 2%. The remaining 24 had
  the following distribution:

N       COHb

11      2-5%
7       5-10%
6       >10%

They concluded that  without the routine checking os COHb levels 54%
would not have been diagnosed. They noted that all patients with levels
above 2% noted symptomatic improvement with oxygen therapy. The number
and extent of household smoking did not corelate with the children's COHb.

Neither of these studies is beyond criticism.  Nevertheless using
stricter criteria for likely toxicity (15% in adults and 5% in children)
would still leave 4/55 = 7% (adults) and 13/46 = 28% of children with
unequivically elevated levels.


H. Louzon MD


(1)

Dolan et. al.  Carboxyhemoglobin Levels in Patients with Flu-Like
Symptoms.  Ann Emer Med 1987;16:782-786

(2)

Baker et. al.  Carboxyhemoglobin Levels in Children With Nonspecific
Flu-like Symptoms.  J Peds  1988;113(3):501-504