Carboxyhemoglobin half-life in carbon monoxide-poisoned patients treated with 100% oxygen at atmospheric pressure

Study objectives: There are large reported differences for the carboxyhemoglobin (COHb) half-life (COHb t(1/2)) in humans breathing 100% atmospheric O-2 following CO inhalation in tightly-controlled experiments compared to the COHb t(1/2) observed in clinical CO poisoning (range, 36 to 131 min, respectively). Other reports have suggested that the COHb t(1/2) may be affected by gender differences, age, and lung function. We wished to test the hypothesis that the COHb t(1/2) might also be influenced by CO poisoning vs experimental CO exposure, by a history of loss of consciousness (LOC), concurrent tobacco smoking, and by PaO2. The purpose of the present stud,. was to measure the COHb t(1/2) in a cohort of CO-poisoned patients and to determine if those listed factors influenced the COHb t(1/2). Design: Retrospective chart review fr om 1985 to 1995, We calculated the COHb t(1/2) of CO-poisoned patients who were treated with high-flow supplemental atmospheric pressure O-2 delivered by nonrebreather face mask or endotracheal tube. Setting: Hyperbaric medicine department of a tertiary-care teaching hospital. Patients: Of 240 CO-poisoned patients, 93 had at least two COHb measurements > 2% (upper limit of normal) with recorded times of the measurements, permitting calculation of the COHb t(1/2). Results: The COHb t(1/2) was 74 +/- 25 min (mean +/- 1 SD) with a range from 26 to 148 min. By stepwise multiple linear regression analysis, the PaO2 influenced the COHb t(1/2) (R-2 = 0.19; p < 0.001), whereas the COHb t, was not influenced by gender, age, smoke inhalation, history of LOC, concurrent tobacco smoking, degree of initial metabolic acidosis (base excess), or initial COHb level. Conclusions: The COHb t(1/2) of 93 CO-poisoned patients treated with 100% O-2 at atmospheric pressure was 74 +/- 25 min, considerably shorter than the COHb t(1/2) reported in prior clinical reports (approximately 130 +/- 130 min) and mas influenced only bq the patient's PaO2.