Cluster of unintentional carbon monoxide poisonings presenting to the emergency departments in Kingston, Ontario during 'ice storm 98'.

Abstract

The purpose of this report is to describea cluster of unintentional CO poisoningsthat presented to the two emergencydepartments (EDs) in Kingston, Ontario.These poisonings occurred following thesevere ice storm that swept through west-ern Quebec and eastern Ontario in January1998. This analysis demonstrates: theimportance of CO poisoning as a publichealth issue during power outages; and,how an injury surveillance program can beused to monitor sentinel patterns of injuryat a community level. METHODSThe Canadian Hospitals InjuryReporting and Prevention Program(CHIRPP) is an emergency room-basedcomputerized injury surveillance programoperated in sentinel Canadian hospitals.Both the Kingston General and HotelDieu Hospitals (Kingston) participate inthis program. Records are maintained forall persons presenting for the treatment ofan injury or poisoning.Cases of carbon monoxide poisoningpresenting for treatment at the two emer-gency departments were identified throughthe CHIRPP system. The individual chartswere then abstracted for more detailedinformation. The study period spanned thetime from January 7, 1998 at 12 noon(start of the ice storm) through Monday,January 19, 1998 at 2 pm (lifting of thestate of emergency in Kingston). All caseswere described by: age and sex; source ofpoisoning; carboxyhemoglobin (COHgb)level; treatment and disposition; and day ofpresentation. Univariate (frequencies,medians, inter-quartile ranges) and bivari-ate statistics (cross-tabulations, t-tests) wereemployed.On several occasions, the same incidentcaused multiple cases of CO poisoning.Incidents and cases were described sepa-rately by source of CO poisoning, and dateof presentation.RESULTSDuring the 13-day ice storm, 22 cases ofCO poisoning presented to the KingstonEDs, representing 11 incidents of COexposure (Table I). All cases presented withsymptoms suggestive of CO poisoning andhad measurable COHgb levels. Eight indi-viduals (36%) arrived by ambulance. The most common sources of CO poi-soning were gas generators (n=13; 59%)followed by charcoal barbecues (n=5;23%). The source of CO poisoning wasbased on the patient’s self-report; therewere no cases of patients reporting morethan one source.There were equal numbers of males andfemales affected, and this did not vary bysource of exposure. The median age ofthose seen was 47 years (inter-quartilerange: 28-67). Age varied by source ofexposure with older individuals (>65)being poisoned by charcoal barbecues andkerosene heaters, and those less than 65 bygenerators and propane barbecues(p<0.001).The median COHgb level for all casescombined was 15% (inter-quartile range:2-19 %). COHgb levels varied by source ofexposure (Table I) with charcoal briquettescausing the highest levels.Most persons (n=19; 86%) were advisedand/or treated with high-flow oxygen inthe ED and discharged home (Table I).One individual was admitted due to amyocardial infarction, a known complica-tion of carbon monoxide poisoning. Twowere transferred to a hyperbaric chamberin Toronto for high-pressure oxygen thera-