Needle aspiration in lung cancer.

and injury are well-known complications of CO poisoning, possibly more prevalent in patients with pre-existing atherosclerotic coronary artery disease. As Dr. Marius-Nunez points out, hypoxic injury is multifactorial, probably related both to CO binding to hemoglobin and also to direct cellular effects of CO. All patients with carbon monoxide exposure should undergo routine electrocardiography to evaluate the possibility of myocardial toxicity. It is necessary, however, to clarify the interpretation of laboratory data reported in this case. Quantitative determinations of carboxyhemoglobin saturation are typically performed with a CO-oximeter. CO-oximetry measures total hemoglobin, carboxyhemoglobin (COHb), oxyhemoglobin (HbO2), and methemoglobin by a spectrophotometric method. Arterial hemoglobin saturation with oxygen (Sa02) reported on routine arterial blood gas analysis is calculated from Pa02 and plla. This is relatively reliable for clinical use if there is no significant carboxyor methemoglobin present. Sa02 is an estimate of the fraction of hemoglobin-binding sites occupied by oxygen of those available for oxygen association. In a patient with CO poisoning, the calculated value for SaO3 reported by blood gas analysis must be ignored; instead, the oxyhemoglobin value available from CO-oximetry must be used. In this case, initial simultaneous values for COHb of 52 percent and SaO2 of 69 percent were reported, implying that more hemoglobinbinding sites were occupied than were available. More correctly, 52 percent of the sites were occupied by CO molecules and 69 percent of the remaining 48 percent available were occupied by 0,. I suspect that the values reported for SaO, later in the patient’s course are also incorrect estimates of oxyhemoglobin fraction. Instead of 02 saturations of 98 percent with simultaneous COHb of 23 and 13 percent, the likely correct values for HbO, are 77 and 87 percent. The potential danger of such an interpretation is that the patient will be incorrectly considered to have adequate oxygenation. This patient was extubated from 100 percent oxygen because he was stable and SaO, “appeared” to be 98 percent. If COHb was indeed 13 percent, HbO, could have been no higher than 87 percent. Furthermore, the 87 percent of sites occupied by oxygen would not release the oxygen normally because CO causes the 02 affinity of unoccupied sites to increase (left shift of oxygen dissociation curve). In the face of acute myocardial infarction, longer therapy with oxygen would have been appropriate. Lastly, it should be noted that serious CO poisoning is treated with hyperbaric oxygen when available. Ischemic EKG changes are a standard indication for hyperbaric treatment in the patient with carbon monoxide poisoning, as are metabolic acidosis, chest pain, and significant neurologic impairment.23