Epidemiology of severe carbon monoxide poisoning in children.

OBJECTIVE To describe the characteristics of severe pediatric CO poisoning cases treated with hyperbaric oxygen (HBO2) to determine key features that might aid prevention efforts. DESIGN Using data on 250 consecutive children treated with HBO2 for CO poisoning at Virginia Mason Medical Center in Seattle, we calculated descriptive statistics and used Fisher's Exact test to determine differences in case frequency of CO sources with respect to race/ethnicity. RESULTS There were 236 (94%) unintentional and 14 (6%) intentional poisonings. Loss of consciousness occurred in 50%. The average initial carboxyhemoglobin level was 19.0 +/- 10.3% (mean +/- SD; range 0.3% to 57.6%). Months with the highest proportion of incidents were November (15%), December (12%) and January (14%). Children 0-2 years old were most often poisoned by charcoal briquettes (40%), in contrast to older children who were most often poisoned by motor vehicle. Females were most often poisoned by charcoal briquettes (28%), in contrast to males in which motor vehicle exhaust predominated (41%). Minorities were disproportionately affected by CO poisoning compared with non-Latino whites (P < 0.001). Overall, motor vehicle exhaust predominated for Non-Latino whites (33%), while Non-Latino blacks (55%) and Latinos (71%) had significantly higher proportions of poisonings by charcoal briquettes (P < 0.001). All intentional cases occurred in adolescents aged 15 to 17 years, 71% of these in males, mostly by motor vehicle exhaust (79%). CONCLUSIONS Severe pediatric CO poisoning demonstrates a seasonal pattern with key sources of exposure varying by gender, age, and race/ethnicity. Minorities are at increased risk for severe CO poisoning. Intentional CO poisoning occurs primarily among adolescent males and by motor vehicle exhaust. Prevention efforts should take these factors into consideration.