Carbon Monoxide Exposures After Hurricane Ike --- Texas, September 2008

During power outages after hurricanes, survivors can be at risk for carbon monoxide (CO) poisoning if they use portable generators improperly (1). On September 13, 2008, Hurricane Ike struck the coast of Texas, leaving approximately 2.3 million households in the southeastern portion of the state without electricity (2). Six days later, 1.3 million homes were still without electrical power (2). To assess the impact of storm-related CO exposures and to enhance prevention efforts, CDC analyzed data from five disparate surveillance sources on CO exposures reported during September 13--26 in counties of southeast Texas that were declared disaster areas by the federal government. This report describes the results of that analysis, which indicated that one data source, Texas poison centers, received reports of 54 persons with storm-related CO exposures during the surveillance period. Another data source, the Undersea and Hyperbaric Medical Society (UHMS) hyperbaric oxygen treatment database, reported that 15 persons received hyperbaric oxygen treatment for storm-related CO poisoning. Medical examiners, public health officials, and hospitals in Texas reported that seven persons died from storm-related CO poisoning. Among the data sources, the percentage of reported storm-related CO exposures caused by improper generator use ranged from 82% to 87%. These findings underscore the need for effective prevention messages during storm preparation, warnings, and response periods regarding the correct use of generators and the installation and maintenance of battery-powered CO detectors.

For this analysis, a storm-related CO exposure was defined as evidence of inhalation of CO (e.g., self-reported activation of a CO detector) that was related to the storm. Storm-related CO poisoning was defined as storm-related inhalation of CO that resulted in symptoms of CO poisoning. Only poison center calls and deaths associated with CO exposures deemed to be unintentional were included in this analysis. CDC obtained surveillance data from five different sources: 1) the National Poison Data System (NPDS); 2) the Texas Poison Center Network (TPCN), operated by the Texas Department of State Health Services (TDSHS); 3) the TDSHS disaster mortality surveillance system; 4) the UHMS hyperbaric oxygen treatment database*; and 5) CDC's BioSense system.†

NPDS and TDSHS provided CDC with information on all storm-related CO-related calls to poison centers during the surveillance period originating from Texas counties that were declared federal disaster areas. All 61 poison centers in the United States upload call data, including demographic and symptom information, to NPDS; the six poison centers in Texas simultaneously submit similar data to TDSHS via TPCN (3). For each call, poison center staff members determined the number of CO-exposed or CO-poisoned persons described by the caller. Data on CO-related calls provided by NPDS and TPCN were matched to avoid duplication. However, cases reported by poison centers could not be reconciled with cases from the other data sources because some poison center calls lacked identifying information.

In addition to providing poison center call data, TDSHS also provided CDC with information on storm-related deaths with CO exposure listed as the cause of death. Mortality data came from medical examiners, public health officials, and hospitals for deaths in persons residing in counties in the disaster area.

UHMS maintains a national online reporting system in collaboration with CDC. This system collects information on hyperbaric oxygen treatments administered for severe CO poisonings. Participating UHMS physicians enter patient demographic and clinical data for those receiving treatment for CO poisoning in their hyperbaric facilities.

Reports from BioSense, a CDC-sponsored and maintained automated surveillance system that receives data from approximately 590 civilian hospitals in the United States (4), included free-text, patient-reported chief complaint data and International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes from emergency departments of the seven participating hospitals in southeast Texas. Visits were included if "carbon monoxide exposure" or "carbon monoxide poisoning" was listed as the chief complaint and/or ICD-9-CM code 986 was listed as either a working or a final diagnosis. For UHMS and Biosense, because no data were provided on the address of the patient, residency was determined using the location of the health-care facility.§

Poison Center Calls

Calls to poison centers regarding 54 storm-related CO exposures were reported to CDC during September 13--26 (Figure). The median age of the exposed persons was 24 years (range: 1 month--71 years), and most (64.8%) were women (Table). Headache (63%), nausea (44%), and vomiting (28%) were the most commonly reported symptoms. Twenty-seven patients were treated in a health-care facility, of whom 25 (93%) were treated and released; two (7%) were hospitalized. Generators were the source of CO exposure in 82% of cases. Most (93%) of the exposures occurred in a residential setting; four (7%) exposures occurred at a workplace.

Hyperbaric Oxygen Treatments

UHMS data indicated that 15 persons were treated for severe, storm-related CO poisoning at one hyperbaric oxygen treatment facility in the disaster area during September 13--26 (Figure). Additional cases were identified by those persons undergoing treatment but are not described in this report. Thirteen CO poisonings were caused by gasoline-powered generators, and two were caused by house fires. The median age of patients was 49 years (range: 1--86 years), and eight were women (Table). The mean carboxyhemoglobin measurement was 18% (range: 7%--40%). Three of the persons treated with hyperbaric oxygen were treated and released, seven were hospitalized, and five had an unknown outcome. Generators were the source of CO exposure in 13 cases. Thirteen patients were exposed to CO in a residential setting.

Emergency Department Visits

Among the seven participating hospitals in the disaster area, BioSense reported that eight persons made emergency department visits to four health-care facilities with a chief complaint or final diagnosis of CO poisoning (Figure). The median age of the patients was 57 years (range: 17--72 years), and five of them were female. The symptoms most frequently reported were headache (four patients) and nausea (four patients). Three of the eight patients were admitted to the hospital.

Deaths

During the surveillance period, TDSHS received reports from medical examiners, public health officials, and hospitals of seven deaths in the disaster area caused by storm-related CO poisoning (Figure). Among those seven decedents, the median age was 32 years (range: 4--76 years). Six of the decedents were male, and five were of Hispanic ethnicity (Table). The source of exposure for six of the deaths was a generator placed inside the home or garage. All of the poisonings occurred in residential settings, and all occurred within 4 days after Hurricane Ike.

Reported by: N Hampson, MD, S Dunn, Center for Hyperbaric Medicine, Virginia Mason Medical Center, Seattle, Washington. A Bronstein, MD, American Assoc of Poison Control Centers, Alexandria, Virginia. C Fife, MD, Memorial Hermann-Texas Medical Center, Houston; J Villanacci, PhD, D Zane, MS, M Forrester, J Hellsten, PhD, R Beal, MPH, C Beasley, MS, Texas Dept of State Health Svcs. J Clower, MPH, F Yip, PhD, S Iqbal, PhD, C Martin, MSPH, C Cooper, A Wolkin, MSPH, Div of Environmental Hazards and Health Effects, National Center for Environmental Health; M Podgornik, MPH, J Tokars, MD, S Benoit, MD, Div of Emergency Preparedness and Response, National Center for Public Health Informatics, CDC.

Editorial Note:

Even when placed outdoors or in well-ventilated areas, generators can be a dangerous source of CO. After Hurricane Katrina in 2005, 50 (98%) of the 51 reported CO-poisoning cases involved generator use (5). The analysis in this report indicates that CO exposure also was an important source of morbidity and mortality after Hurricane Ike. Most of the exposures occurred within 2 days after the storm, likely because of widespread power outages and increased generator usage (1). Improper generator use, including placement inside residential settings, was the primary cause for these reported CO exposures. These findings emphasize the need for effective, storm-related prevention messages concerning proper generator use, and underscore the need for ongoing prevention messages regarding the installation and maintenance of battery-powered CO detectors in homes.

In post-disaster situations, prevalence estimates of household generator usage have ranged from 18% to 31% (6,7), indicating that a substantial number of persons affected by a storm could be at risk for CO exposure. Previous studies have shown that, in nondisaster situations, children are disproportionately affected by CO poisonings that result in emergency department visits. Women often make up the majority of persons exposed to CO, whereas men make up the majority of deaths (8,9). Increased rates of CO toxicity in men have been attributed to engagement in high-risk activities, such as fuel-burning tool and equipment use (9). The results described in this report indicate that, of storm-related CO exposures reported to poison centers in counties of southeast Texas that were declared disaster areas, approximately one third were in persons aged <1--17 years. Additionally, most CO exposures reported by poison centers, BioSense, and UHMS occurred among women (65%, 63%, and 53%, respectively).

These CO exposures occurred despite efforts to warn the public of CO-related hazards. TDSHS issued statewide press releases on CO poisoning and prevention within 1 day of hurricane landfall and again on day 3, and public health workers distributed CO-poisoning prevention materials at ice and water distribution locations. Public health officials in Houston distributed prevention materials to residents and evacuees returning to their homes, and during door-to-door community health assessments and, along with the Harris County Medical Examiner Office, produced a press release in both English and Spanish to warn residents about indoor generator placement. In addition, public health officials in Galveston distributed approximately 6,000 flyers containing CO-poisoning and prevention information.

The findings in this report are subject to at least three limitations. First, although CDC was able to match cases from mortality, emergency department, and hyperbaric oxygen treatment facility data sources, cases reported by poison centers could not be matched with those from other data sources because some poison center data were missing identifying information. This might have resulted in some duplication of cases. Second, estimates of nonfatal CO exposures in this report are likely underestimations of the overall number of CO exposures after Hurricane Ike; presumably, not all exposed persons contacted poison centers or sought treatment. Finally, data could be obtained for fewer than half of the patients receiving hyperbaric oxygen treatment.

CO exposure is preventable, yet it continues to pose a substantial public health problem in the wake of hurricanes. The public, especially those in the path of an impending storm, should be reminded that 1) installation of a battery-operated CO detector outside each sleeping area in the home and routine battery changes can save lives, and 2) generators should never be operated in a basement or garage and should be placed as far away from the home as possible. These surveillance results, in addition to results from previous post-disaster situations, can help in the development of public health interventions during storm preparation, warnings, and response periods.

Acknowledgments

The findings in this report are based, in part, on contributions by J Wiersema, PhD, S Derrick, PhD, J Coston, B Begay, Harris County Medical Examiner Office; S Pustilnik, MD, Galveston County Medical Examiner Office; D Wiltz-Beckham, DVM, Galveston County Health District; T Brown, MD, P Moore, MD, S Jones, Southeast Texas Forensic Center; participating hospitals; V Harris, K Shofner, T Haywood, R Taylor, PhD, A Valadez, MD, members of the Disaster Surveillance Workgroup, Texas Dept of State Health Svcs; and T Bayleyegn MD, National Center for Environmental Health, CDC.

References

1. Hampson NB, Stock AL. Storm-related carbon monoxide poisoning: lessons learned from recent epidemics. Undersea Hyperb Med 2006;33:257--63.
2. CenterPoint Energy. CenterPoint Energy concludes emergency operations. October 2, 2008. Available at http://www.centerpointenergy.com/newsroom/newsreleases/1e107fbef1bbc110VgnVCM1000005a1a0d0aRCRD.
3. CDC. Using the National Poisoning Data System for public health surveillance. Atlanta, GA: US Department of Health and Human Services, CDC; 2005. Available at http://www.cdc.gov/nceh/hsb/chemicals/pdfs/npds.pdf.
4. CDC. BioSense for public health departments. Atlanta, GA: US Department of Health and Human Services, CDC; 2008. Available at http://www.cdc.gov/biosense/publichealth.htm.
5. CDC. Carbon monoxide poisoning after Hurricane Katrina---Alabama, Louisiana, and Mississippi, August--September 2005. MMWR 2005;54:996--8.
6. Daley WR, Smith A, Paz-Argandona E, Malilay J, McGeehin M. An outbreak of carbon monoxide poisoning after a major ice storm in Maine. J Emerg Med 2000;18:87--93.
7. CDC. Epidemiologic assessment of the impact of four hurricanes---Florida, 2004. MMWR 2005;54:693--7.
8. CDC. Nonfatal, unintentional, non-fire related, carbon monoxide exposures---United States, 2004--2006. MMWR 2008;57:896--9.
9. CDC. Carbon monoxide related deaths---United States, 1999--2004. MMWR 2007;56:1309--12.

Use of trade names and commercial sources is for identification only and does not imply endorsement by the U.S. Department of Health and Human Services. References to non-CDC sites on the Internet are provided as a service to MMWR readers and do not constitute or imply endorsement of these organizations or their programs by CDC or the U.S. Department of Health and Human Services. CDC is not responsible for the content of pages found at these sites. URL addresses listed in MMWR were current as of the date of publication.

All MMWR HTML versions of articles are electronic conversions from typeset documents. This conversion might result in character translation or format errors in the HTML version. Users are referred to the electronic PDF version (http://www.cdc.gov/mmwr) and/or the original MMWR paper copy for printable versions of official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

**Questions or messages regarding errors in formatting should be addressed to mmwrq@cdc.gov.

Date last reviewed: 8/13/2009