- Email: [email protected]
Carbon monoxide detectors and emergency physicians
90
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
TABLE1.
Treatment Time Effect on Mortality: "What If" Projections
Prehospital delay (hr) Door to needle time (hr) Total time (hr) Mortality
TABLE3.
GUSTO IIAcute Coronary Syndromes ST-elevation acute MIs
4
3
2
2
1
1/2"
2 6
1 4
1 3
1/2 21/2
1/2 11/2
1/2" 1*
12% to 15%
10%
--
--
--
1%
* Seattle Standard.
Non-ST-elevation, non-Q-wave acute MIs Unstable angina Severe prolonged pain onset Prodromal chest discomfort (not pain)
presentation comparable with those experienced already in "door to needle" times. RAYMONDD. BAHR,MD Medical Director The Paul Dudley White Coronary Care System Saint Agnes Hospital
Professor of Medicine and Public Health at Boston University: "The day has to come when we consider heart attack in our patients as not the first indication for treatment, but as a medical failure." Acute heart attacks can be prevented through appropriate treatment at the prodromal unstable angina stage in patients with coronary disease. 6 Realization of this potential, however, awaits reductions in time to TABLE2.
Frequently Used Terms
Thrombolytic Index (TI) = Thrombolytic MIs/Total MIs Optimal Thrombolytic Index (OTI) = 1st Hour Thrombolytic MIs/Total Thrombolytic MIs Effective Thrombolytic Index (ETI) = 1st Hour Thrombolytic MIs/Total MIs "Early Index" or Acute Prevention Index = Unstable Angina Patients/Thrombolytic MIs 4- Non Thrombolytic MIs + Unstable Angina Patients "Crashing Index" or Myocardial Infarction Index or Failure to Detect Index = Thrombolytic MIs + Non Thrombolytic MIs/Thrombolytic MIs + Non Thrombolytic MIs + Unstable Angina Patients Better Defined Early Index = 1st Hour Thrombolytic MIs + Unstable Angina/-rhrombolytic MIs less 1st Hour + Non Thrombolytic MIs + Unstable Angina Patients
JAMESTONASCIA,PHD Department of Biostatistics Johns Hopkins School of Hygiene and Public Health Baltimore, MD REFERENCES 1. Vogel RA: HCFA's Cooperative Cardiovascular Project: A nationwide quality assessment of acute myocardial infarction. Clin Cardiol 1994; 17:354-356 2. National Heart Attack Alert Program Coordinating Committee, 60 Minutes to Treatment Working Group: Emergency Department: Rapid identification and treatment of patients with acute myocardial infarction. Ann Emerg Med 1994;23:311-329 3. Bahr RD: Access to early cardiac care: Chest pain as a risk factor for heart attacks, and the emergence of early cardiac care centers. MMJ February 1992;133-137 4. Balagtas R, Herrada J, Bahr R: Thrombolytic therapy in MI: A retrospective study. MMJ June 1990;561-564 5. Bahr RD: Reducing time to therapy in AMI patients: The new paradigm. Am J Emerg Med 1994;12:501-503 6. Bahr RD: Thechanging paradigm of acute heart attack prevention in the emergency department: A futuristic viewpoint? Ann Emerg Med 1995;25:95-96
Carbon Monoxide Detectors and EmergencyPhysicians During the winter of 1991, 10 members of one Chicago family succumbed to carbon monoxide (CO) while sleeping in their home. A defective furnace appeared to be the cause of this tragedy. With this episode (along with other intermittent clusters of CO related deaths and illnesses) as a trigger, the Chicago City Council on March 2, 1994 (after more than 2 years of committee hearings) passed the nation's first carbon monoxide detector ordinance, essentially mandating carbon monoxide detectors in single-family and multi-unit dwellings. 1 The ordinance went into effect on October 1, 1994. The initial phases of ordinance implementation were certainly rocky. From October 1st through December 31, 1994, the Chicago Fire Department received over 12,000 calls of CO detector alarming; in approximately 85% of
these alarms, less than 9 ppm (parts per million) of CO was reportedly measured by Fire Department personnel. Questions arose regarding the causes of these "nuisance" alarms: Did ventilation of the house play a role in the low readings by the Fire Department? Did a gradual accumulation of CO set off some detectors? What was the effect of room placement of these detectors on alarm activation? The magnitude and focus of these alarm calls were highlighted on December 21-22, 1994, when an unusual thermal air inversion (warmer in the upper atmosphere than in the lower atmosphere) caused elevated ambient CO levels of 13 to 20 ppm in the Chicago area (about a fivefold increase), thus setting off multiple CO alarms (whose sensitivity was 15 ppm over an 8-hour period). To further complicate matters, there was no easy reset mechanism for
EDITORIALS
the majority of these detectors, often requiring that the device be dismantled and the sensor be aired out for many hours. Thus, during this 2-day period, the Chicago Fire Department responded to 3,464 CO investigations. Ironically, at a meeting at Underwriter's Laboratory (U.L.) 2 weeks earlier, steps were devised to minimize these occurrences. Although the high-end threshold was not affected (400 ppm within 15 minutes, 200 ppm within 35 minutes, or 100 ppm within 90 minutes, all equivalent to a carboxyhemoglobin level of 10% in an individual with exertion), the low-end CO stability and sensitivity threshold was changed from 15 ppm over an 8-hour period to 15 ppm over 30 days. In this way, detector sensing units will not alarm to accumulation of low levels of CO (about 10 ppm) over several hours and will essentially reset when the detector is exposed to clean air. 2 Additionally, a reset button mechanism was included in the revised U.L. standards. 2 Despite these changes, criticism of the ordinance was ferocious and concentrated from a few media outlets. One daily city newspaper printed at least seven editorials or columns within 7 days (more than 12 within a period of 80 days) attacking the ordinance. Every aspect of this ordinance was demonized; city building inspectors, aldermen, the CO detector industry, and even Underwriter's Laboratory were characterized as either corrupt or incompetent. Ironically, the only participant to escape criticism was carbon monoxide--its toxicity was minimized, and the toll on morbidity and mortality was mischaracterized. Essentially, these editorials coincided with those opinions of the Gas Research Institute (GRI, Chicago, IL), a research, development, and demonstration management organization of the natural gas industry. Despite the intense and focused criticism of this ordinance, no witnesses spoke against the ordinance during a 6hour open city council meeting on January 4, 1995. Additionally, a newspaper poll printed on December 2, 1994 (though hardly scientific) noted that 77% of respondents supported the ordinance. Obscured by the rhetoric was the fact that 68 individuals were transported by Chicago paramedics because of CO detection alarming and treated for CO poisoning in area emergency departments during the first 3 months of this ordinance. What can emergency physicians learn from this experience? Public education should precede implementation. CO detectors resemble smoke detectors in appearance, but public familiarity should not be assumed. Although it had been recommended that these detectors be placed near to sleeping quarters, a significant number of alarming detectors were located near the furnace or kitchen (whereupon cooking can set off the alarms). Furthermore, response to a CO alarm should be two-tiered. If any cohabitant has symptoms, they should seek medical help as soon as possible. Asymptomatic CO alarms do not require emergency attention. Fire department response to a domestic CO alarm may not be required on an immediate basis. Furthermore, maximum fire department response need not be dispatched to all alarms. Certainly, affected symptomatic individuals require urgent attention and transport to a health care facility. Additionally, an adequate supply of carbon monoxide moni-
91
tors should be issued to fire department personnel prior to ordinance implementation. Training of personnel in use of these meters along with coordination with local natural gas companies should occur well in advance of ordinance implementation. Project Medical Directors of local Emergency Medical Services should be involved in protocol arrangement for each individual urban setting (see Appendix). Increased sensitivity is not a desirable characteristic of carbon monoxide detectors when marketed as an acute lifesaving device. Unlike smoke detectors, carbon monoxide detectors do not lend themselves well to low thresholds for alarm. Although atmospheric conditions in the 15 to 30 ppm range can have detrimental effects on patients with coronary artery disease, most individuals will panic at a CO alarm going off at 3 AM, believing that an acute fife-threatening situation exists. 3,4 Consumer awareness of the product is essential but an alarming CO detector can override a significant amount of education, thus causing a panic atmosphere. This can be especially true if it is difficult to reset the device immediately. It should be noted that the revised U.L. Standard for CO detectors addresses both of these concerns. 2 Prior to implementation, multiple brands of CO detectors should be available and plentiful. In Chicago, one brand (First Alert, B.R.K. Industries, Aurora, IL) had approximately 85% market share of alarms sold. Since First Alert was the most sensitive product on the market with the largest market share, most of the nuisance alarms were ascribed to this brand. As a result, First Alert offered a refund to any individual wishing to return the device by January 31, 1995. Any individual problems (whether real or perceived) with this one detector brand contaminated the entire CO detector industry, the CO ordinance, and Underwriter's Laboratory. Consumer choice should precede implementation. Criticism of this concept will be fanatical and will come from essentially few sources. (One media outlet even appointed itself the unofficial "Journal of Carbon Monoxide Poisoning.") The Chicago experience will be characterized as a negative one and these lessons along with detector adjustments by U.L. will not be recognized. Medical statistics may be used and quoted (and misused and misquoted) but medical opinions of the worthiness of the ordinance will not be sought or published. Local medical societies and regional poison control centers will need to be leaders in this effort. It should be noted that both the American Medical Association as well as the Illinois College of Emergency Physicians have endorsed the use of CO detectors. The most important lesson is that preventive medicine will have its skeptics. Whether one deals with antihypertensive medication, seat belt use, motorcycle helmet use, lead screening, or vaccinations, preventive medicine techniques in an asymptomatic individual unaware of risks will not be popular. Carbon monoxide does deserve a preventive medicine approach on a population basis. Carbon monoxide is the leading cause of death from acute poisoning in the US, with unintentional nonvehicular deaths accounting for about (as a minimum estimate) 500 deaths yearly. (Illinois accounts for about 8.7% of all unintentional carbon monoxide deaths nationally, s) CO poisoning in
92
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
France (17 million population) caused by gas water heaters accounted for 800 cases of illness and 32 deaths in 1979. 6 In 1994 there were 14,926 (6.9 calls/100,000 population/year) human exposures to CO called into 65 regional poison control centers. Of these, 14,556 (97.5%) were considered unintentional. 7 Of 90 cases of CO exposure reported to a regional poison control center in one year, 53.4% was from a home setting. 8 The prevalence of CO-induced headache and dizziness in adults presenting to EDs in the winter was noted to be from 3% to 5%. 9 It has been estimated that nationally 10,000 individuals annually seek medical attention or lose at least one day of normal activity due to carbon monoxide. 10 Although the exact mechanism of carbon monoxide poisoning has not been established, it is known that CO binds to red blood cells with 250-fold greater affinity than oxygen, thus reducing oxygen delivery and shifting the oxyhemoglobin dissociation curve to the left. CO also directly hinders cellular respiration due to cytochrome oxidase blockade. 1~ Blood carboxyhemoglobin levels do not correlate well with clinical effects. 11 Residual nervous system effects (ranging from chronic headaches, memory impairment, and neuropsychiatric damage) from acute CO poisoning may occur in from 14% to 40% of severely poisoned individuals. Brain lipid peroxidation can lead to neuronal dysfuncdon.1031 An increasing amount of cardiopulmonary complaints presenting to EDs has been associated with increased amounts of ambient levels of carbon monoxide. 4 Patients with heart disease will suffer from more frequent exacerbations of angina at CO concentrations ranging from 50 to 100 ppm. 12 It has been estimated that an increase of ambient carbon monoxide concentrations from 7 to 20 ppm can result in 11 excess deaths daily in Los Angeles. 3 Recently, ambient carbon monoxide levels were correlated with hospital admissions for congestive heart failure with the relative risk for hospitalization ranging from 1.I to 1.37 when associated with an increase of 10 ppm of carbon monoxide concentrations in the seven U.S. cities studied.13 CO crosses the placenta, with the fetus concentrating CO at 10% to 15% greater amount than the mother. Additionally, CO is eliminated over a much longer period of time in the fetus as opposed to an adult. 13,14 Increased teratogenic risk (ie, mental retardation and limb abnormalities) and fetal death have been correlated with CO poisoning. Animal studies have shown that chronic in utero exposure to 150 ppm CO results in permanent memory deficits. 13,14 Patients in the lower socioeconomic strata will be particularly adversely affected by CO poisoning because of their reliance on ancillary (stoves, ovens, space heaters) sources for heat. 8,9 It has been estimated that as many as 30% of patients with significant carbon monoxide poisoning will carry an erroneous initial diagnosis. (Food poisoning is usually the most common diagnosis given. 6) Because of the variability of clinical presentations, occult carbon monoxide poisoning can be one of the most difficult conditions to diagnose. However, one of the most remarkable (and, to some extent, the most satisfying) moments in emergency medicine is when one treats a patient with carbon monoxide poisoning, who, invariably, will say he or she is "lucky to be alive." With adequate preparation, consumer choice, education, and
appropriate emergency response, CO detectors can be used to give domestic dwelling a safer, healthier environment, and luck will then be taken out of the equation. JERROLDB. LEIKIN,MD Medical Director Rush Poison Control Center Rush Presbyterian St. Luke's Medical Center Chicago, IL
REFERENCES 1. City Council of the City of Chicago: Amendment of Title 13, Chapter 64 of Municipal Code of Chicago by Addition of New Sections 190 Through 300 Requiring Carbon Monoxide Detectors in Various Building. Meeting of March 2, 1994 2. Underwriters Laboratory: Proposed First Edition: The standard for single and multiple station carbon monoxide detectors. UL 2034, Northbrook, IL. 1991. Revised 1994 (AppendixA) 3. Hexter AC, Goldsmith JR: Carbon monoxide association of community air pollution with mortality. Science 1971 ;12:265- 267 4. Kurt TL, Mogielnicki P, Chandler JE, Hirst K: Ambient carbon monoxide levels and acute cardiorespiratory complaints: An exploratory study. Am J Public Health 1979;69:360-363 5. Cobb N, Etzel RA: Unintentional carbon monoxide- related deaths in the United States. JAMA 1991;266:659-663 6. Barret L, Danel V, Faure J: Carbon monoxide poisoning, a diagnosis frequently overlooked. J Toxicol Clin Toxicol 1985;23: 309-313 7. Litovitz TL, Feldberg MA, Soloway RA, et al: 1994 Annual Report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1995;13:551-597 8. Spiller HA: Carbon monoxide exposure in the home: Source and epidemiology. Vet Hum Toxicol 1987;29:383-386 9. Herkerling PS, Leikin JB, Maturen A: Occult carbon monoxide poisoning: Validation of a prediction model. Am J Med 1988;84:251255 10. Goldfrank L (ed): Goldfrank's Toxicologic Emergencies (ed 5). Norwalk, CT, Appleton and Lange, 1994, pp 1199-1214 11. Hardy KR, Thorn SR: Pathophysiology and treatment of carbon monoxide poisoning. J Toxicol Clin Toxico11994;32:613- 629 12. Anderson EW, Andelman RJ, Strauch JM, et al: Effect of low-level carbon monoxide exposure on onset and duration of angina pectoris: A study in ten patients with ischemic heart disease. Ann Intern Med 1973;79:46-50 13. Morris RD, Naumova EN, Munaisinghe RL: Ambient air pollution and hospitalization for congestive heart failure among elderly people in seven large U.S. cities. Am J Public Health 1995;85:13611365 14. Longo LD: The biologic effects of carbon monoxide on the pregnant woman, fetus and newborn infant. Am J Obstet Gynecol 1977;129:69-103 15. Mangold KJ, Silverstein S, McGuigan MA, et al: Carbon Monoxide. In Rumack BH, Hess AJ, Gelman CR (eds): PoisindexR System. Engiewood, CO, Micromedex, Inc, 1995, vo186 16. Bizovi K, Leikin J, Hryhorczuk D, Frateschi L: In-home CO detectors: Suburban Chicago experience. J Toxicol Clin Toxicol 1995;33:524 (abstr) 17. Krenzelok EP, Full R, Roth R: Carbon monoxide... The silent killer with an audible solution. J Toxicol Clin Toxicol 1995;33:536 (abstr) 18. Unintentional Carbon Monoxide Poisonings in Residential Settings--Connecticut November 1993-March 1994. MMWR Morbid Mortal Wkly Rep 1995;44(41):765-767
APPENDIX: CHICAGOFIRE DEPARTMENTPROTOCOL FOR EVALUATIONOF CARBON MONOXIDEDETECTORALARMS PURPOSE
This Order: A. establishes a procedure for the Chicago Fire Department's response to reports of carbon monoxide (CO) incidents.
EDITORIALS
II.
HI.
B. rescinds General Order 94-006 and supercedes all memos and directives pertaining to carbon monoxide incident response. C. becomes effective immediately. GENERAL Carbon monoxide is an odorless, tasteless, colorless gas that is deadly. It is a byproduct of a fuel burning process. Many appliances such as furnaces, kitchen stoves, hot water heaters, automobiles, etc, can produce carbon monoxide. When a faulty or unusual condition exists, carbon monoxide may be vented into areas where people are present. Carbon monoxide poisoning may be difficult to diagnose. Its symptoms are similar to the flu, which may include headache, nausea, fatigue, and dizzy spells. The Occupation Safety and Health Administration has established a maximum safe working level for carbon monoxide at 35 parts per million (ppm) over an 8-hour period, in the general workplace. The US Environmental Protection Agency has established that residential levels are not to exceed 9 ppm over an 8-hour average. The gas company will no longer respond automatically to all carbon monoxide (CO) investigations. They will respond only if the initial call received by them indicates that someone is ill or if the Fire Department requests that they respond to the scene. PROCEDURES A. Issuance and Maintenance of Meters: 1. The Department will issue a carbon monoxide meter to each company designated as a meter company. 2. Company commanders shall enter the meter onto their company inventory as per current Department procedure. 3. This meter shall be used to monitor the atmosphere during any suspected carbon monoxide investigation. 4. Each morning the CO gas meter shall be examined, tested, and the results entered in the company journal. 5. Should any problem be noted, the company officer shall immediately contact the Hazardous Incident Team (ML 9222) who will take steps to correct the problem or issue a replacement meter. The Hazardous Incident Team will also provide periodic recalibration and maintenance of all CO gas meters, as per the manufacturer's recommendations. 6. Company officers shall see that their assigned meter is stored in a secured location on their apparatus. B. Fire Alarm Office Shall Attempt to Verify if Mechanism that is Sounding the Alarm is a Smoke Detector or a Carbon Monoxide Detector. 1. If it is a smoke detector, the Fire Alarm Office will determine what, if any, fire response is necessary. 2. If it is a carbon monoxide detector: a. Determine if any persons at the scene are exhibiting symptoms of carbon monoxide poisoning. (1) If yes, dispatch the closest meter company and ambulance; (2) If no, dispatch the closest meter company to conduct an investigation. C. The first arriving Company Shall Establish Scene Control--As Per C.ED. Incident Command Procedure. 1. Verification shall be made if the alarm is coming from a smoke detector or carbon monoxide detector. The cause of the alarm shall be determined, ie, true alarm, low battery indication, poor location of device, etc. a. If it is a smoke detector alarm they shall: (1) investigate the cause of the alarm;
93
(2) take the necessary action to mitigate the situation; (3) advise the fire alarm office of the situation. b. If it is CO detector they shall: (l) determine if anyone is exhibiting any symptoms of possible carbon monoxide poisoning; if so, immediately evacuate and ventilate the premises; (2) request necessary E.M.S. response; (3) begin the investigative procedure. c. If no one exhibits any symptoms of carbon monoxide poisoning, it will not be necessary to evacuate or ventilate the premises unless a level of over 9 ppm is detected by a meter as described below. d. The incident commander shall request that the gas company respond to the scene through the appropriate Fire Alarm Office if: (1) a CO level of over 9 ppm is indicated on their meters, (2) or the responding company shuts off gas appliance, (3) or someone is showing signs of being ill due to carbon monoxide, (4) or the incident commander feels a response by the gas company is needed. D. Carbon Monoxide Investigations (Procedure) 1. Zero their meter in fresh air and comply with all other start up procedures as recommended by the manufacturer of the metering equipment. 2. Initiate a survey of the premises to determine if there are any amounts above 9 ppm of carbon monoxide present. 3. All members shall make complete use of the Self Contained Breathing Apparatus (S.C.B.A,) in any atmosphere that is in excess of 35 ppm of CO. 4. Reading of 9 ppm or less: (1) Inform the occupants that our instrument did not detect an elevated level of CO at this time. (2) Recommend occupants check their CO detector per manufacturer recommendations. (3) Attempt to reset detector. (4) Inform occupants that if it activates again, call 9-1-1. 5. Readings of more than 9 ppm but less than 100 ppm: (1) Any reading above 9 ppm shall be considered above normal reading. (2) Occupants shall be informed that we have detected a potentially dangerous level of CO. (3) Recommend that all persons leave the premises and begin ventilation. (4) If it is determined that an appliance is malfunctioning and thereby producing CO, it shall be shut down. (5) Once the premises have been reduced to a safe level of CO, the premises may be occupied--at the discretion of the occupants. (6) Attempt shall be made to reset the detector. (7) Inform occupants that if it activates again, call 9-1-1. (8) The occupants shall be informed of the action that has taken place and that Peoples Gas Company has been requested to respond by the Fire Department. 6. Readings of 100 ppm or greater: (1) Any reading of 100 ppm or greater--inform the occupants that we have detected a potentially lethal level of CO.
94
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
(2) Order the occupants to leave the premises immediately. (3) ff it is determined that an appliance is malfunctioning and thereby producing CO, it shall be shut down. (4) Once the premises have been reduced to a safe level of CO, the premises may be occupied at the
discretion of the occupant. (5) Attempt shall be made to reset the detector. (6) Inform occupants that if it activates again, call 9-1-1. (7) The occupants shall be informed of the action that has taken place and that Peoples Gas Company has been requested to respond by the Fire Department.
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
TABLE1.
Treatment Time Effect on Mortality: "What If" Projections
Prehospital delay (hr) Door to needle time (hr) Total time (hr) Mortality
TABLE3.
GUSTO IIAcute Coronary Syndromes ST-elevation acute MIs
4
3
2
2
1
1/2"
2 6
1 4
1 3
1/2 21/2
1/2 11/2
1/2" 1*
12% to 15%
10%
--
--
--
1%
* Seattle Standard.
Non-ST-elevation, non-Q-wave acute MIs Unstable angina Severe prolonged pain onset Prodromal chest discomfort (not pain)
presentation comparable with those experienced already in "door to needle" times. RAYMONDD. BAHR,MD Medical Director The Paul Dudley White Coronary Care System Saint Agnes Hospital
Professor of Medicine and Public Health at Boston University: "The day has to come when we consider heart attack in our patients as not the first indication for treatment, but as a medical failure." Acute heart attacks can be prevented through appropriate treatment at the prodromal unstable angina stage in patients with coronary disease. 6 Realization of this potential, however, awaits reductions in time to TABLE2.
Frequently Used Terms
Thrombolytic Index (TI) = Thrombolytic MIs/Total MIs Optimal Thrombolytic Index (OTI) = 1st Hour Thrombolytic MIs/Total Thrombolytic MIs Effective Thrombolytic Index (ETI) = 1st Hour Thrombolytic MIs/Total MIs "Early Index" or Acute Prevention Index = Unstable Angina Patients/Thrombolytic MIs 4- Non Thrombolytic MIs + Unstable Angina Patients "Crashing Index" or Myocardial Infarction Index or Failure to Detect Index = Thrombolytic MIs + Non Thrombolytic MIs/Thrombolytic MIs + Non Thrombolytic MIs + Unstable Angina Patients Better Defined Early Index = 1st Hour Thrombolytic MIs + Unstable Angina/-rhrombolytic MIs less 1st Hour + Non Thrombolytic MIs + Unstable Angina Patients
JAMESTONASCIA,PHD Department of Biostatistics Johns Hopkins School of Hygiene and Public Health Baltimore, MD REFERENCES 1. Vogel RA: HCFA's Cooperative Cardiovascular Project: A nationwide quality assessment of acute myocardial infarction. Clin Cardiol 1994; 17:354-356 2. National Heart Attack Alert Program Coordinating Committee, 60 Minutes to Treatment Working Group: Emergency Department: Rapid identification and treatment of patients with acute myocardial infarction. Ann Emerg Med 1994;23:311-329 3. Bahr RD: Access to early cardiac care: Chest pain as a risk factor for heart attacks, and the emergence of early cardiac care centers. MMJ February 1992;133-137 4. Balagtas R, Herrada J, Bahr R: Thrombolytic therapy in MI: A retrospective study. MMJ June 1990;561-564 5. Bahr RD: Reducing time to therapy in AMI patients: The new paradigm. Am J Emerg Med 1994;12:501-503 6. Bahr RD: Thechanging paradigm of acute heart attack prevention in the emergency department: A futuristic viewpoint? Ann Emerg Med 1995;25:95-96
Carbon Monoxide Detectors and EmergencyPhysicians During the winter of 1991, 10 members of one Chicago family succumbed to carbon monoxide (CO) while sleeping in their home. A defective furnace appeared to be the cause of this tragedy. With this episode (along with other intermittent clusters of CO related deaths and illnesses) as a trigger, the Chicago City Council on March 2, 1994 (after more than 2 years of committee hearings) passed the nation's first carbon monoxide detector ordinance, essentially mandating carbon monoxide detectors in single-family and multi-unit dwellings. 1 The ordinance went into effect on October 1, 1994. The initial phases of ordinance implementation were certainly rocky. From October 1st through December 31, 1994, the Chicago Fire Department received over 12,000 calls of CO detector alarming; in approximately 85% of
these alarms, less than 9 ppm (parts per million) of CO was reportedly measured by Fire Department personnel. Questions arose regarding the causes of these "nuisance" alarms: Did ventilation of the house play a role in the low readings by the Fire Department? Did a gradual accumulation of CO set off some detectors? What was the effect of room placement of these detectors on alarm activation? The magnitude and focus of these alarm calls were highlighted on December 21-22, 1994, when an unusual thermal air inversion (warmer in the upper atmosphere than in the lower atmosphere) caused elevated ambient CO levels of 13 to 20 ppm in the Chicago area (about a fivefold increase), thus setting off multiple CO alarms (whose sensitivity was 15 ppm over an 8-hour period). To further complicate matters, there was no easy reset mechanism for
EDITORIALS
the majority of these detectors, often requiring that the device be dismantled and the sensor be aired out for many hours. Thus, during this 2-day period, the Chicago Fire Department responded to 3,464 CO investigations. Ironically, at a meeting at Underwriter's Laboratory (U.L.) 2 weeks earlier, steps were devised to minimize these occurrences. Although the high-end threshold was not affected (400 ppm within 15 minutes, 200 ppm within 35 minutes, or 100 ppm within 90 minutes, all equivalent to a carboxyhemoglobin level of 10% in an individual with exertion), the low-end CO stability and sensitivity threshold was changed from 15 ppm over an 8-hour period to 15 ppm over 30 days. In this way, detector sensing units will not alarm to accumulation of low levels of CO (about 10 ppm) over several hours and will essentially reset when the detector is exposed to clean air. 2 Additionally, a reset button mechanism was included in the revised U.L. standards. 2 Despite these changes, criticism of the ordinance was ferocious and concentrated from a few media outlets. One daily city newspaper printed at least seven editorials or columns within 7 days (more than 12 within a period of 80 days) attacking the ordinance. Every aspect of this ordinance was demonized; city building inspectors, aldermen, the CO detector industry, and even Underwriter's Laboratory were characterized as either corrupt or incompetent. Ironically, the only participant to escape criticism was carbon monoxide--its toxicity was minimized, and the toll on morbidity and mortality was mischaracterized. Essentially, these editorials coincided with those opinions of the Gas Research Institute (GRI, Chicago, IL), a research, development, and demonstration management organization of the natural gas industry. Despite the intense and focused criticism of this ordinance, no witnesses spoke against the ordinance during a 6hour open city council meeting on January 4, 1995. Additionally, a newspaper poll printed on December 2, 1994 (though hardly scientific) noted that 77% of respondents supported the ordinance. Obscured by the rhetoric was the fact that 68 individuals were transported by Chicago paramedics because of CO detection alarming and treated for CO poisoning in area emergency departments during the first 3 months of this ordinance. What can emergency physicians learn from this experience? Public education should precede implementation. CO detectors resemble smoke detectors in appearance, but public familiarity should not be assumed. Although it had been recommended that these detectors be placed near to sleeping quarters, a significant number of alarming detectors were located near the furnace or kitchen (whereupon cooking can set off the alarms). Furthermore, response to a CO alarm should be two-tiered. If any cohabitant has symptoms, they should seek medical help as soon as possible. Asymptomatic CO alarms do not require emergency attention. Fire department response to a domestic CO alarm may not be required on an immediate basis. Furthermore, maximum fire department response need not be dispatched to all alarms. Certainly, affected symptomatic individuals require urgent attention and transport to a health care facility. Additionally, an adequate supply of carbon monoxide moni-
91
tors should be issued to fire department personnel prior to ordinance implementation. Training of personnel in use of these meters along with coordination with local natural gas companies should occur well in advance of ordinance implementation. Project Medical Directors of local Emergency Medical Services should be involved in protocol arrangement for each individual urban setting (see Appendix). Increased sensitivity is not a desirable characteristic of carbon monoxide detectors when marketed as an acute lifesaving device. Unlike smoke detectors, carbon monoxide detectors do not lend themselves well to low thresholds for alarm. Although atmospheric conditions in the 15 to 30 ppm range can have detrimental effects on patients with coronary artery disease, most individuals will panic at a CO alarm going off at 3 AM, believing that an acute fife-threatening situation exists. 3,4 Consumer awareness of the product is essential but an alarming CO detector can override a significant amount of education, thus causing a panic atmosphere. This can be especially true if it is difficult to reset the device immediately. It should be noted that the revised U.L. Standard for CO detectors addresses both of these concerns. 2 Prior to implementation, multiple brands of CO detectors should be available and plentiful. In Chicago, one brand (First Alert, B.R.K. Industries, Aurora, IL) had approximately 85% market share of alarms sold. Since First Alert was the most sensitive product on the market with the largest market share, most of the nuisance alarms were ascribed to this brand. As a result, First Alert offered a refund to any individual wishing to return the device by January 31, 1995. Any individual problems (whether real or perceived) with this one detector brand contaminated the entire CO detector industry, the CO ordinance, and Underwriter's Laboratory. Consumer choice should precede implementation. Criticism of this concept will be fanatical and will come from essentially few sources. (One media outlet even appointed itself the unofficial "Journal of Carbon Monoxide Poisoning.") The Chicago experience will be characterized as a negative one and these lessons along with detector adjustments by U.L. will not be recognized. Medical statistics may be used and quoted (and misused and misquoted) but medical opinions of the worthiness of the ordinance will not be sought or published. Local medical societies and regional poison control centers will need to be leaders in this effort. It should be noted that both the American Medical Association as well as the Illinois College of Emergency Physicians have endorsed the use of CO detectors. The most important lesson is that preventive medicine will have its skeptics. Whether one deals with antihypertensive medication, seat belt use, motorcycle helmet use, lead screening, or vaccinations, preventive medicine techniques in an asymptomatic individual unaware of risks will not be popular. Carbon monoxide does deserve a preventive medicine approach on a population basis. Carbon monoxide is the leading cause of death from acute poisoning in the US, with unintentional nonvehicular deaths accounting for about (as a minimum estimate) 500 deaths yearly. (Illinois accounts for about 8.7% of all unintentional carbon monoxide deaths nationally, s) CO poisoning in
92
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
France (17 million population) caused by gas water heaters accounted for 800 cases of illness and 32 deaths in 1979. 6 In 1994 there were 14,926 (6.9 calls/100,000 population/year) human exposures to CO called into 65 regional poison control centers. Of these, 14,556 (97.5%) were considered unintentional. 7 Of 90 cases of CO exposure reported to a regional poison control center in one year, 53.4% was from a home setting. 8 The prevalence of CO-induced headache and dizziness in adults presenting to EDs in the winter was noted to be from 3% to 5%. 9 It has been estimated that nationally 10,000 individuals annually seek medical attention or lose at least one day of normal activity due to carbon monoxide. 10 Although the exact mechanism of carbon monoxide poisoning has not been established, it is known that CO binds to red blood cells with 250-fold greater affinity than oxygen, thus reducing oxygen delivery and shifting the oxyhemoglobin dissociation curve to the left. CO also directly hinders cellular respiration due to cytochrome oxidase blockade. 1~ Blood carboxyhemoglobin levels do not correlate well with clinical effects. 11 Residual nervous system effects (ranging from chronic headaches, memory impairment, and neuropsychiatric damage) from acute CO poisoning may occur in from 14% to 40% of severely poisoned individuals. Brain lipid peroxidation can lead to neuronal dysfuncdon.1031 An increasing amount of cardiopulmonary complaints presenting to EDs has been associated with increased amounts of ambient levels of carbon monoxide. 4 Patients with heart disease will suffer from more frequent exacerbations of angina at CO concentrations ranging from 50 to 100 ppm. 12 It has been estimated that an increase of ambient carbon monoxide concentrations from 7 to 20 ppm can result in 11 excess deaths daily in Los Angeles. 3 Recently, ambient carbon monoxide levels were correlated with hospital admissions for congestive heart failure with the relative risk for hospitalization ranging from 1.I to 1.37 when associated with an increase of 10 ppm of carbon monoxide concentrations in the seven U.S. cities studied.13 CO crosses the placenta, with the fetus concentrating CO at 10% to 15% greater amount than the mother. Additionally, CO is eliminated over a much longer period of time in the fetus as opposed to an adult. 13,14 Increased teratogenic risk (ie, mental retardation and limb abnormalities) and fetal death have been correlated with CO poisoning. Animal studies have shown that chronic in utero exposure to 150 ppm CO results in permanent memory deficits. 13,14 Patients in the lower socioeconomic strata will be particularly adversely affected by CO poisoning because of their reliance on ancillary (stoves, ovens, space heaters) sources for heat. 8,9 It has been estimated that as many as 30% of patients with significant carbon monoxide poisoning will carry an erroneous initial diagnosis. (Food poisoning is usually the most common diagnosis given. 6) Because of the variability of clinical presentations, occult carbon monoxide poisoning can be one of the most difficult conditions to diagnose. However, one of the most remarkable (and, to some extent, the most satisfying) moments in emergency medicine is when one treats a patient with carbon monoxide poisoning, who, invariably, will say he or she is "lucky to be alive." With adequate preparation, consumer choice, education, and
appropriate emergency response, CO detectors can be used to give domestic dwelling a safer, healthier environment, and luck will then be taken out of the equation. JERROLDB. LEIKIN,MD Medical Director Rush Poison Control Center Rush Presbyterian St. Luke's Medical Center Chicago, IL
REFERENCES 1. City Council of the City of Chicago: Amendment of Title 13, Chapter 64 of Municipal Code of Chicago by Addition of New Sections 190 Through 300 Requiring Carbon Monoxide Detectors in Various Building. Meeting of March 2, 1994 2. Underwriters Laboratory: Proposed First Edition: The standard for single and multiple station carbon monoxide detectors. UL 2034, Northbrook, IL. 1991. Revised 1994 (AppendixA) 3. Hexter AC, Goldsmith JR: Carbon monoxide association of community air pollution with mortality. Science 1971 ;12:265- 267 4. Kurt TL, Mogielnicki P, Chandler JE, Hirst K: Ambient carbon monoxide levels and acute cardiorespiratory complaints: An exploratory study. Am J Public Health 1979;69:360-363 5. Cobb N, Etzel RA: Unintentional carbon monoxide- related deaths in the United States. JAMA 1991;266:659-663 6. Barret L, Danel V, Faure J: Carbon monoxide poisoning, a diagnosis frequently overlooked. J Toxicol Clin Toxicol 1985;23: 309-313 7. Litovitz TL, Feldberg MA, Soloway RA, et al: 1994 Annual Report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 1995;13:551-597 8. Spiller HA: Carbon monoxide exposure in the home: Source and epidemiology. Vet Hum Toxicol 1987;29:383-386 9. Herkerling PS, Leikin JB, Maturen A: Occult carbon monoxide poisoning: Validation of a prediction model. Am J Med 1988;84:251255 10. Goldfrank L (ed): Goldfrank's Toxicologic Emergencies (ed 5). Norwalk, CT, Appleton and Lange, 1994, pp 1199-1214 11. Hardy KR, Thorn SR: Pathophysiology and treatment of carbon monoxide poisoning. J Toxicol Clin Toxico11994;32:613- 629 12. Anderson EW, Andelman RJ, Strauch JM, et al: Effect of low-level carbon monoxide exposure on onset and duration of angina pectoris: A study in ten patients with ischemic heart disease. Ann Intern Med 1973;79:46-50 13. Morris RD, Naumova EN, Munaisinghe RL: Ambient air pollution and hospitalization for congestive heart failure among elderly people in seven large U.S. cities. Am J Public Health 1995;85:13611365 14. Longo LD: The biologic effects of carbon monoxide on the pregnant woman, fetus and newborn infant. Am J Obstet Gynecol 1977;129:69-103 15. Mangold KJ, Silverstein S, McGuigan MA, et al: Carbon Monoxide. In Rumack BH, Hess AJ, Gelman CR (eds): PoisindexR System. Engiewood, CO, Micromedex, Inc, 1995, vo186 16. Bizovi K, Leikin J, Hryhorczuk D, Frateschi L: In-home CO detectors: Suburban Chicago experience. J Toxicol Clin Toxicol 1995;33:524 (abstr) 17. Krenzelok EP, Full R, Roth R: Carbon monoxide... The silent killer with an audible solution. J Toxicol Clin Toxicol 1995;33:536 (abstr) 18. Unintentional Carbon Monoxide Poisonings in Residential Settings--Connecticut November 1993-March 1994. MMWR Morbid Mortal Wkly Rep 1995;44(41):765-767
APPENDIX: CHICAGOFIRE DEPARTMENTPROTOCOL FOR EVALUATIONOF CARBON MONOXIDEDETECTORALARMS PURPOSE
This Order: A. establishes a procedure for the Chicago Fire Department's response to reports of carbon monoxide (CO) incidents.
EDITORIALS
II.
HI.
B. rescinds General Order 94-006 and supercedes all memos and directives pertaining to carbon monoxide incident response. C. becomes effective immediately. GENERAL Carbon monoxide is an odorless, tasteless, colorless gas that is deadly. It is a byproduct of a fuel burning process. Many appliances such as furnaces, kitchen stoves, hot water heaters, automobiles, etc, can produce carbon monoxide. When a faulty or unusual condition exists, carbon monoxide may be vented into areas where people are present. Carbon monoxide poisoning may be difficult to diagnose. Its symptoms are similar to the flu, which may include headache, nausea, fatigue, and dizzy spells. The Occupation Safety and Health Administration has established a maximum safe working level for carbon monoxide at 35 parts per million (ppm) over an 8-hour period, in the general workplace. The US Environmental Protection Agency has established that residential levels are not to exceed 9 ppm over an 8-hour average. The gas company will no longer respond automatically to all carbon monoxide (CO) investigations. They will respond only if the initial call received by them indicates that someone is ill or if the Fire Department requests that they respond to the scene. PROCEDURES A. Issuance and Maintenance of Meters: 1. The Department will issue a carbon monoxide meter to each company designated as a meter company. 2. Company commanders shall enter the meter onto their company inventory as per current Department procedure. 3. This meter shall be used to monitor the atmosphere during any suspected carbon monoxide investigation. 4. Each morning the CO gas meter shall be examined, tested, and the results entered in the company journal. 5. Should any problem be noted, the company officer shall immediately contact the Hazardous Incident Team (ML 9222) who will take steps to correct the problem or issue a replacement meter. The Hazardous Incident Team will also provide periodic recalibration and maintenance of all CO gas meters, as per the manufacturer's recommendations. 6. Company officers shall see that their assigned meter is stored in a secured location on their apparatus. B. Fire Alarm Office Shall Attempt to Verify if Mechanism that is Sounding the Alarm is a Smoke Detector or a Carbon Monoxide Detector. 1. If it is a smoke detector, the Fire Alarm Office will determine what, if any, fire response is necessary. 2. If it is a carbon monoxide detector: a. Determine if any persons at the scene are exhibiting symptoms of carbon monoxide poisoning. (1) If yes, dispatch the closest meter company and ambulance; (2) If no, dispatch the closest meter company to conduct an investigation. C. The first arriving Company Shall Establish Scene Control--As Per C.ED. Incident Command Procedure. 1. Verification shall be made if the alarm is coming from a smoke detector or carbon monoxide detector. The cause of the alarm shall be determined, ie, true alarm, low battery indication, poor location of device, etc. a. If it is a smoke detector alarm they shall: (1) investigate the cause of the alarm;
93
(2) take the necessary action to mitigate the situation; (3) advise the fire alarm office of the situation. b. If it is CO detector they shall: (l) determine if anyone is exhibiting any symptoms of possible carbon monoxide poisoning; if so, immediately evacuate and ventilate the premises; (2) request necessary E.M.S. response; (3) begin the investigative procedure. c. If no one exhibits any symptoms of carbon monoxide poisoning, it will not be necessary to evacuate or ventilate the premises unless a level of over 9 ppm is detected by a meter as described below. d. The incident commander shall request that the gas company respond to the scene through the appropriate Fire Alarm Office if: (1) a CO level of over 9 ppm is indicated on their meters, (2) or the responding company shuts off gas appliance, (3) or someone is showing signs of being ill due to carbon monoxide, (4) or the incident commander feels a response by the gas company is needed. D. Carbon Monoxide Investigations (Procedure) 1. Zero their meter in fresh air and comply with all other start up procedures as recommended by the manufacturer of the metering equipment. 2. Initiate a survey of the premises to determine if there are any amounts above 9 ppm of carbon monoxide present. 3. All members shall make complete use of the Self Contained Breathing Apparatus (S.C.B.A,) in any atmosphere that is in excess of 35 ppm of CO. 4. Reading of 9 ppm or less: (1) Inform the occupants that our instrument did not detect an elevated level of CO at this time. (2) Recommend occupants check their CO detector per manufacturer recommendations. (3) Attempt to reset detector. (4) Inform occupants that if it activates again, call 9-1-1. 5. Readings of more than 9 ppm but less than 100 ppm: (1) Any reading above 9 ppm shall be considered above normal reading. (2) Occupants shall be informed that we have detected a potentially dangerous level of CO. (3) Recommend that all persons leave the premises and begin ventilation. (4) If it is determined that an appliance is malfunctioning and thereby producing CO, it shall be shut down. (5) Once the premises have been reduced to a safe level of CO, the premises may be occupied--at the discretion of the occupants. (6) Attempt shall be made to reset the detector. (7) Inform occupants that if it activates again, call 9-1-1. (8) The occupants shall be informed of the action that has taken place and that Peoples Gas Company has been requested to respond by the Fire Department. 6. Readings of 100 ppm or greater: (1) Any reading of 100 ppm or greater--inform the occupants that we have detected a potentially lethal level of CO.
94
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
(2) Order the occupants to leave the premises immediately. (3) ff it is determined that an appliance is malfunctioning and thereby producing CO, it shall be shut down. (4) Once the premises have been reduced to a safe level of CO, the premises may be occupied at the
discretion of the occupant. (5) Attempt shall be made to reset the detector. (6) Inform occupants that if it activates again, call 9-1-1. (7) The occupants shall be informed of the action that has taken place and that Peoples Gas Company has been requested to respond by the Fire Department.