Carboxyhemoglobin measurement by hospitals: implications for the diagnosis of carbon monoxide poisoning.

Most case definitions for carbon monoxide (CO) poisoning include demonstration of an elevated blood carboxyhemoglobin (COHb) concentration. Further, it is generally believed that treatment of CO poisoning is more effective when performed as soon as possible after the exposure. This suggests that a hospital's inability to measure blood COHb could lead to delayed or missed diagnosis or treatment. This study evaluated the ability of hospitals in the Pacific Northwest to measure COHb levels. The clinical laboratory of every acute care hospital in Washington, Idaho, Montana, and Alaska was surveyed regarding the ability to measure COHb levels, the method utilized and the time required. If they could not measure COHb, they were asked whether samples are sent elsewhere, the location of the referral laboratory, and time required. Results were then compared to the list of hospitals referring CO-poisoned patients to a regional center for hyperbaric oxygen therapy from 2003-2004. In the four states, only 44% of acute care hospitals have the capability to measure COHb. The remaining 56% send blood samples to other laboratories. The average time to get a result is 10 +/- 10 min in hospitals with co-oximetry and 904 +/- 1360 min in those without, a difference of 15 h (p < 0.0001). When samples are sent out, the average distance is 121 miles, often bypassing a hospital with CO-oximetry capability. Over 90% of CO-poisoned patients referred for hyperbaric treatment came from hospitals able to measure COHb. Fewer than one-half of acute care hospitals in a four-state region have the capability to measure COHb levels. This has the potential to significantly impact diagnosis or treatment of patients with acute CO poisoning.