Document Zv36Vynv79NdJGY8V9232Ggd

fi; i A Polyvinyl Chloride Toxicity in Fires Hydrogen Chloride Toxicity in Fire Fighters Robert F. Dyer, MD, Victor H. Esch, MO Polyvinyl chloride, of all the plastic polymers, has been implicated pri marily in causing the most serious problem in fire fighting today because it releases hydrogen chloride gas when burning. One hundred seventy fire fighters who experienced symptpms from its toxicity have been studied from 1970 to 1975. One died. .{JAMA 235:393-397, 1976) A RECENTLY identified hazard that is of paramount importance to prac ticing physicians is that of chemical injury incurred by persons exposed to fumes from fires. Specifically, such in juries are caused by the inhalation of vapors created by the thermal degra dation of polyvinyl chloride, a plastic polymer. Such substances have now become widely used in the construe* See also p 390. tion of homes, furnishings, office equipment, electric wire, telephone and cable covering, and vehicles. A particular hazard exists in the case of the fire fighter, although it could be equally serious in any circumstance in which an individual might be exposed to such plastic combustion products. THE FIRE At 10:30 ah on Jan 6,.1970, an alarm sounded for a fire reported to toe on the sixth floor of Building 213 at the Washington, DC, Navy Yard. Units of the District of Columbia Fire De partment arriving on the scene found the floor filled with a heavy concen tration of toxic fumes. Although all of the fire fighters carried masks, none were used, since neither the fire nor From the Police and Firemens Clinic, Wash ington. DC. Reprint requests to Chairman of the Board of Surgeons. Suite 1445, Chevy Chase Bldg, 5530 Wisconsin Aye, NW, Washington, DC 20015 (Or Dyer). the smoke was so excessive as to pre vent entry without masks. The fire, which was confined to an office copy ing machine constructed of plastic and Teflon parts, was quickly extin guished. The fire companies entered the building via the elevator, placed sev eral portable fans for ventilation, and left the building about 20 minutes later. On returning to their quarters, many of the fire fighters experienced unusual and unexplainable symp toms, including a constricting tight ness localized to the anterior part of the chest. This chest discomfort was described as a searing, burning sensa tion accompanied by dyspnea as well as a "burning sensation appearing to close off the throat." These symptoms were in addition to the usual head ache, dizziness or vertigo, and nausea, which, in our experience, is fre quently seen in cases of nonchemical smoke inhalation. Some ' symptoms persisted over a course of several days following exposure to the fumes. The dyspnea was noted on mild exer tion and was associated with moder ate apprehension by all of the af fected fire fighters. A number of the fire fighters exposed at this fire expe rienced severe conjunctivitis, lacrimation, and dyspnea persisting for 24 hours after the fire. An intense head ache, localized anteriorly and persist ing about 24 hours, was also common. REPORT OF A CASE Approximately 24 hours after the fire, a 33-year-old fire lighter fainted in the sit ting room of the truck company. It was also reported that he had a muscle spasm similar to an epileptic seizure, but recov ered, was fully conscious, and got up im mediately. His superior officer asked how he felt, and he replied that he was "fine" and walked into the kitchen. Following lunch, the acting sergeant on duty looked for the man to check his condition for driv ing and found him lying on the floor next to his bed. He was cyanotic and, despite all resuscitative efforts by an ambulance crew and the rescue squad, was pronounced dead on. arrival by a fire surgeon at the George Washington University Hospital at 1:10 pm. Postmortem examination by the District of Columbia Department of Public Health Medical Examiner showed severe pulmonary hemorrhage and edema due to chemical pneumonitis secondary to expo- aure to chemical smoke and fire. 1 The I |pathologist also reported coronary [ L atherosclerosis. J --- . EFFECT OF FUMES It has been suspected for several years that the smoke and fumes from electrical fires produced greater in jury now than similar fires in the past.'-* It was strongly suspected that hydrogen chloride released by ther mal degradation of PVC, a common electrical wiring insulation material, was the lethal product causing death in this instance. The medical exam iner's report and the microscopic findings subsequently confirmed our original clinical impression of the cause of death. Hydrogen chloride, inhaled either _ in the gaseous state or combined with water vapor, acts as an irritant to the mucous membranes of the eyes and JAMA, Jan 26. 1976--Vol 235. No 4 Polyvinyl Chloride--Dyer & Esch 393 BFG51223 C00STS9Z i I `''I In the fatalities due to CO, the average concentration of CO was 45%. The burning of cellulose mate ria! resulted in nothing but the re lease of CO. In the mouse study with exposure to burning building mate rials, the ECG abnormalities were recorded before the smoke from the burning materials was evident. In regard to the harmful properties of PVC to mice, as well as those from polyurethane, polysterene, and phe nol, the pathologic effects of the gases, rather than those of the smoke, appeared first, as shown by atypical ECG changes. -It requires less CO in combination with HCl to cause death. In conclusion, HCl and CO have a re ciprocal, potentiating adverse effect. F/g i.--Hydrogen chloride concentration and optical density of polyvinyl chloride electri cal insulation decomposition products.. the respiratory tract. An HCl concen tration of 15 ppm causes localized irritation to the throat after short ex posure. Hydrogen chloride concentra tions of 50 to 100 ppm are tolerable for one hour by humans. More severe exposures result in pulmonary edema and often in laryngeal spasm. Con centrations of 1,000 to 2,000 ppm of HC1 are dangerous even with brief exposure* (Pig 1). Quantitative release of HCl on thermal degradation of PVC has been well documented in the scientific and medical literature. Polyvinyl chloride, of all the synthetic plastic polymers, has been implicated as.causing one of the most insidious, serious problems in fire fighting today, due to its. re lease of HCl gas while burning. Mists of hydrochloric acid are considered less harmful than anhydrous. HCl be cause the droplets have no hydrating action. Animat Experiments The toxicity of PVC has been sub stantiated by Kishitani* in his com prehensive animal experiments per formed at the University of Tokyo. With respect to the toxicity of carbon monoxide, Kishitani states that the injurious nature of PVC is unique. All- mice exposed to harmful gas products died, but the carboxyhemoglobin con centration in the dead animals was low (average, 21.2%). This clearly in dicated that the fatalities were not due directly to CO poisoning. At au topsy, the eyes of the mice were closed, their mucous membranes were injured by chemicals, and the sclerae were discolored. The latter had be come white due to the action of chlo^ rine gas^All of the-mice died in a pe riod oTl& minutes after exposure to PVC fumes. Electrocardiographic ab normalities appeared prior to gener ation of large quantities of smoke. This indicated that a harmful gas produced at an early stage was caus ing the effect. Similar results were found on exposure to polyurethane gas. Kishitani concluded that the in jurious properties of burning mate rials such as wood, fire retardant ply wood, melamine finishing board, and acrylic resin are due to CO. EFFECT OF SMOKE Smoke is a suspension of small par ticles in hot air and gases. It has a particulate fraction and a gaseous fraction. The particles consist of car-; bon and are coated with combustible products such as organic acids and al dehydes. The gaseous fraction has an extremely variable composition. Car bon monoxide and carbon dioxide are always present and constitute the bulk of this fraction. However, a wide variety of other toxic gases are present at the scene of a fire. These may be. formed by the combustion process or leak from com mercial processing or storage equip ment. The particulate and gaseous fractions combine to exert a large space-occupying effect and can fill an enclosed space at the expense of air. The gaseous fraction of smoke is much more dangerous to life than the particulate fraction. The nose, mouth, and throat, which filter, out and trap the particulate fraction, cannot filter out the toxic gases. These gases enter the upper respiratory tract and lungs freely when inhaled. Damage to the lungs may result from exposure to a group of gases and vapors called pul monary irritants (chlorine, phosgene, nitrogen dioxide, sulfur dioxide, and ammonia). On entering the lungs, these agents react chemically with water to produce strong acids or al kalis. A violent inflammatory re sponse occurs, causing destruction of lung tissue. Fire fighters and victims of fires in homes, offices, and industrial com plexes are exposed to smoldering and 394 JAMA, Jan 26. 1976--Vol 235. No 4 Polyvinyl Chloride--Oyer & Esch BFG51224 Y' * * f \ ? > i i i lX I I I t00STS9Z burning plastics and synthetics from furniture, door and wallcoverings, textiles, and building materials. Many plastics produce large volumes of pulmonary-irritant gases when burned. Polyvinyl chloride is used ex tensively in furniture, both in the foamed plastic stuffing as well as the covering. Burning 450 gm of this ma- ' terial'tproduces 320 gm of chlorine, phosgene, and HC1, all of which are strong pulmonary irritants. In all, there are 75 known products of PVC degradation. Practically every struc ture today contains plastics capable of producing lung-damaging gases when burned. Therefore, the fire fighter faces a great, new risk in al most any fire. It became obvious that a particular pattern was emerging in which cer tain fire situations were productive of insidious and unpredictable fire casu alties. The lethal chemical smoke was determined to be HC1 resulting from the decomposition of PVC plastic used in telephone cable and electrical equipment installations. We have found electrical and tele phone cable fires to be especially haz ardous to the fire fighters, due-to the toxic gases produced. We have found that when a short circuit developed in a main switch, usually located in a sub-basement or penthouse of an of fice building or apartment house, the area was always poorly ventilated, and the plastic-covered switches quickly heat to temperatures in ex cess of 93.3 C. The heated conductors cause the plastic insulation to decom pose and emit white, mist-like HCl fumes that resemble steam.When the temperature range of 232 to 321 C is attained, the gas may be invisible. The fire fighter may be fatally de ceived if. he does not know plastic-cov ered cable is involved in the fire he is fighting. OVERHAUL PHASE The period of time just after the fire is extinguished is termed the "overhaul phase." This is when the fire fighters clean up the foreground. Toxic smoke and fumes are still being emitted. This is a serious time for toxic- gas inhalation to occur since many fire fighters remove their self- contained. breathing apparatus dur ing this phase. Another special hazard during this phase is the "heat sink," ie, concrete that retains a great amount of heat and releases fumes throughout the overhaul phase. Our testing has shown highly toxic con centrations of HCl in this concrete for as long as one hour after the fire has been extinguished. FUEL LOADING * In the late 1960s, the type of "fuel loading" in office and high-rise build ings (ie, contents of the buildings) changed significantly and initiated a new awareness. The noncombustible buildings were not burning, but fur niture within them, building insula tion, ceiling tile, wall paneling, and decorations made partly or wholly of plastic were fueling disasters. While a building may be con structed of fire-resistant elements, there is no control over the combus tible material that an individual may elect to bring into the structure. This is the "fuel load" and may be ex pressed as kilograms of combustible material per square meter of floor area. Large differences exist in com bustibility, smoke, and toxic gas re lease among different combustible materials. More than 50 sq m of red oak, all burning at its maximum rate, would be required to produce the same rate of smoke as 2.5 meters of acrylonitrile-butadiene-styrene plas tic pipe.. A fire of plastic materials de velops .`extremely high temperatures because plastics possess a heat of combustion 2V4 times that of other combustibles. Therefore, the fire would magnify the normal problems of radiant heat, structural damage, and exposure. The fire would exhibit a very high burning rate because the flammability of plastics is much higher than that of other combus tibles. The four most common plastics pro duced in the United States in the past few years are polystyrene, polyethyl ene, polypropylene, and PVC. A chlorethylene polymer is a plastic solid marketed under a variety of trade names, and currently being manufac tured at a rate over 8.3 billion kilo grams annually. Polyvinyl chloride (Fig 2) is a plas tic solid widely used as a rubber sub stitute and as electric and telephone wire and cable covering. It is also uti lized frequently in electrical fixtures, plastic drains, as well as waste pipe for interior plumbing, pliable thin sheeting, film finishes for textiles, "noninflammable" upholstery, rain coats, tubing, belting, gaskets, shoe soles, disposable unbreakable plastic bottles, packaging materials, auto mobile and aircraft seatcovers, cor rugated fence toppings, tarpaulins, shower curtains, phonograph records, plastic trays, baby pants, toys, cur tains, and office copying-machines. It is used as an exterior finish in many' vehicles, including buses and subway cars, because of its high resistance to wear and soiling. When PVC is heated to 225 C, noth ing happens. When heated above that, up to 475 C, PVC does not burn. The combustion range or flame point is 475 C, Between 225 C and 475 C, PVC loses over 60% of its weight. During this process, it thermally de grades and HCl is released. Four hun dred fifty-three grams of PVC at deg radation releases 320 gm of chlorine, phosgene, and HCl. The individual toxicity factors of phosgene and the other major decom position products of PVC, ie, benzene, toluene, total xylenes, naphthalene, and vinyl chloride, are very small, compared with the toxicity factor of HCL Further, the total or additive toxicity factors of these minor compo nents are also very small in compari son With.HCl toxicity. Unfortunately, very little is known about the syner gistic effects of toxic gases (ie, the ability of one chemical to increase or decrease the toxicity of another). The main toxic risks from the decomposi tion products of PVC are HCl and CO.1 Stone et al` elucidated the nature of smoke aerosol and its role in the transport of toxic gases in fire situ ations. They confirmed the fact that the toxic gas, HCl, is adsorbed on soot aerosol. Possible inhalation hazards of such soot are positively related to the size of the soot particles and particle agglomerates. It has been shown that Kv ! cn tr 10 o o JAMA, Jan 26. 1976-Vol 235. No 4 Polyvinyl Chloride--Dyer & Esch 395 BFG51225 the amount and location of particle elapses between exposure and onset deposition in the respiratory tract de of severe respiratory and chest symp pends on the particle size.* On inhala toms. Treatment is most effective if tion, soot aerosols would enter the given during this quiescent phase. lower lung region and, to an extent, Once the symptoms of difficult be retained there. Soot particle clus breathing, cyanosis, or other symp ters, with sizes ranging from O.lp to toms, as noted in the reported case, 2.5p in diameter, would be retained in have appeared, it may be too late to the alveolar sacs to an extent of from prevent severe morbidity or even 20% to 40% of those particles inhaled. mortality. The "overhaul phase" after At a breathing rate of 18 liters/min the fire is extinguished is dangerous for an exposure of one hour, a max with respect to fumes also. imum of 0.7 gm of soot bearing 13 mg A Harvard research team headed of loosely bound HC1 would be re by Donald P. Dressier, MD, reported tained in the lower lungs. in 1973 that when burned, plastics and The effect of HC1 in the gas phase modern fibers produced colorless, is largely limited to irritability, odorless fumes that quickly displace mainly affecting the upper respira oxygen and cause a dramatic rise in tory tract, whereas loosely bound HC1 the CO* level of the atmosphere. condensed on soot aerosol gains ac Therefore, a person either dies or is cess to the lower lungs. Experiments rendered unconscious in the early have shown that water droplets of stages of a fire involving these mate respirable size rapidly adsorb HC1 rials. If he is unconscious, he soon, dies from the gas phase and approach of other lethal fume inhalations.* equilibrium concentrations in frac Tests performed in Cambridge, Mass, tions of a second.7 showed that fire fighters were proba SMOKE INHALATION bly affected adversely by the high level of CO, rather than by the initial Respiratory distress may develop release of toxic gases.* in fire fighters or fire victims immedi In establishing the importance of ately or within one to two days of ex detecting the presence of certain posure to combustion products. The toxic gases, the difficulty in detecting best protection against these agents CO has been emphasized. Some hur is the mandatory use of the self-con man response mechanisms are more tained breathing apparatus by every sensitive than others to CO inhala fire fighter. When anyone without a tion.* Fairly high levels of CO affect a breathing apparatus is exposed to human's vision, but even higher levels more than a few breaths of "choking do not affect his sense of timing. A re smoke" characteristic of burning or search team headed by R. A. McFar smoldering plastics, he should receive land, MD, of Harvard University, prompt medical attention. found that when CO replaces 11% to Prior to our investigation, we con 17% of the blood's normal oxygen, the sidered most fireground pulmonary peripheral vision is affected and the casualties to be due to "smoke inhala subject has difficulty viewing stimuli tion" or CO poisoning. Carbon mon at a 20" angle from the line of sight.* oxide had long been considered the The Harvard team also found that at leading cause of death and pulmonary a 17% concentration of COHb, some morbidity in fires.^fhough CO i^notj subjects had a momentary lapse of at _noxious and does not 'actually cause tention and failed to respond to all serious pulmonary injury, the amount the stimuli presented. This is impor of CO in a patient is a good indicator tant in disproving the once-held con of whether more harmful gases are cept that once an individual has de present. tected the sharp odor of HC1 he can Before the introduction of massive walk away. It has been our observa fire loads of combustible and ther tion that the fire fighter cannot es mally decomposable plastic products, cape the HCl fumes. fire surgeons usually considered the Cornish and Abar reported that CO CO victim to be "home safe" in four and hydrochloric acid were the major hours after exposure due to the fast toxic products from the combustion of half-life of CO in the blood. After ex PVC.,# They pyrolized PVC polymers posure to plastic degradation prod in a stream of air by gradually in ucts, a period of one to six hours creasing the temperature from am bient to 600 C. Exposure to this air stream supplemented with oxygen produced pulmonary and interstitial edema in rats. Some of these animals also showed focal bronchial and intraalveolar hemorrhage. In a sealed environment (such as a centrally air-conditioned house, of fice, high-rise unit, or hospital), fumes from plastic degradation could kill, even if the fire is brought under con trol. It has been observed that a sleeping person needs an estimated 12 to 15 minutes to awaken, react to danger, and take appropriate action to save himself. With inhalation of fumes from burning plastic, he will not have that much time. COHb LEVEL AND CIRCULATION As noted by E.P. Radford (writ ten communication), 43.deaths due to. fire demonstrated that a blood COHb level of over 65% is nearly always lethal due to the depression of res piration and circulation. Ventricular fibrillation from acute myocardial hy poxia induced by a rapid rise of the COHb level in individuals with a. restricted coronary circulation was found to be the likely explanation of a lower terminal blood COHb level. These individuals were found to have severe, preexisting coronary vascular disease. We believe that any compromise of the coronary circulation, eg, ath erosclerosis such as found in the fatal' case, would be potentiated by highly irritating chemical fumes. We also believe that other pyrolysis products, especially HCl, have contributed to ventricular fibrillation since its con centration is comparable to that of CO in real fire situations involving plastics. Kishitaui has demonstrated that chlorine affects the myocardium more rapidly and more severely than CO. Carbon monoxide may cause changes in enzyme activities suggestive of myocardial damage. Hydrogen chlo ride has been found to be a more po tent myocardial irritant whose ef fects may be difficult to determine by clinical observation. This is in con junction with the development of pul monary edema over a 24-hour period following exposure to HCl. This fur ther complicates and compromises coronary artery oxygen saturation. This could lead to sudden death syn- 396 JAMA. Jan 26. 1976-Vol 235. No < Polyvinyl Chloride--Dyer & Esch BFG51226 drorni-, similar to that in the reported case. Although in the cases of CO poi soning noted by G. Gordon, (written communication) the ECG changes did appear late if at all and the COHb varied in individuals at the same fire as well as in different fire situations, it has been our experience that all fire fighters exposed to PVCdo mani fest ECG signs of its toxicity. There are variable levels of its toxic effect, on the lungs, mucous membranes (tra cheal peeling effect), eyes', and other parts of the body. CLINICAL STUDY In Washington, DC, a clinical study of fire fighters suspected and later known to have been exposed to toxic products from thermal degradation of PVC was conducted by the DC Board of Police and Fire Surgeons between January 1970 and September 1975. One hundred seventy individuals were exposed from one to four differ ent times to these products and re ported two or alt of the symptoms of acute toxicity, ie, pain in the anterior aspect of the chest, neck and throat pain, dyspnea, severe headache, dizzi ness, and irregular pulse within a few hours after exposure. Initial physical examinations were performed by the fire surgeon on the fireground, in the hospital emergency room, or at the Police and Firemens Clinic. Laboratory studies including complete blood cell count, determina tions of serum glutamic oxaloacetic transaminase, blood urea nitrogen, or creatinine, and a complete urinalysis were done. Electrocardiograms were performed, initially when there was chest pain, extrasystoles, or cardiac arrhythmia. However, ECG studies were not incorporated as part of the physical evaluation until the-latter part of the study when their impor tance was appreciated. About one fifth of the fire fighters had extrasystoles during the first few hours after exposure. Fourteen of the last 26 firemen t.\,.__:u had prema ture ventricular contractions. Some have had recurrence of extrasystoles within the subsequent year. Chest x- ray films were done at the initial examination and have remained nor mal. Serum electrolyte and arterial blood gas determinations were per formed on 12 of the fire fighters who were more acutely ill, or required hos pitalization. Because of the excellent clinic facilities provided by the Dis trict of Columbia government, all fire fighters were observed in a holding area after exposure and examined at the clinic or the hospital emergency room after transportation from the fireground. Twelve men were ad mitted to the hospital for treatment. If symptoms did not worsen over the first 24 hours in the hospital, they were released and followed up on an outpatient basis at the clinic. Specific treatment, including oxy gen given intranasally at 5 liters/ rain, broncbodilators, antihistamines, and decongestants in oral suspension, has been utilized in controlling most of the acute respiratory symptoms. Steroids were given intravenously to three of the fire fighters during hospi talization due to the severity of their signs and symptoms of respiratory distress. Initial and follow-up liver function tests have remained normal in all of these patients. There has been no di rect correlation between severity of symptoms and history of cigarette smoking. None of the fire fighters has had to retire for permanent airway dis orders. Serial studies and follow-up examinations will continue. Since the onset of this study among profes sional fire fighters, the use of the selfcontained breathing apparatus has been required at all fires where t.oxic fumes could be present. This is neces sary during the "overhaul phase" also. We urge all physicians to advo cate restriction of unsupervised use of PVC products, hire reseuren activ ity should be maintained at a na tional level to continue study as well as preventive measures for HC1 tox icity in PVC fires. CONCLUSIONS Hazards of Plastic Fires.--Respira tory distress may develop in fire fighters and fire victims after expo sure to combustion products. The complexity of the problem is empha sized when it is recognized that the thermal degradation of PVC results in the formation of at least 75 identi fiable potentially toxic compounds. Three characteristics of plastic fires are the extremely high temperatures, very high burning rates, and thick toxic smoke. The main danger to ex posed fire fighting personnel at the fireground results from the massive formation of HC1 gas, and in many instances, its adsorption on spherical soot particles that are carried into the victims' respiratory tracts. Prevention.--The mandatory use of the self-contained breathing appa ratus by all who will be vulnerable to toxic smoke will aid in the prevention of plastic fires. Control of hazardous building designs, including equip ment and ventilation is also neces sary. Protection by masks during the "overhaul phase" should be required. Sophisticated air-sampling tech niques are required for definitive study, including gas chromatography. Electrocardiographic studies utilizing base line tracings directly after HC1 exposure are stressed. Areas of Future Study.--All atmo spheres should be suspected by the fire fighter- There should be a search for an effective gas analyzer that can immediately assess gas concentra tions released by burning or smolder ing plastics. Further definitive car-: diologic studies are indicated. The term "smoke inhalation" should be abandoned and replaced by "inhala tion of toxic combustion products." References 1. Lowery R; Injury Control at Heavy Service Electrical Fire9, safety bulletin. City 0/ New York Fire Dept, September 1969. 2. Lee T: Smcke Analysis of Specimens Sub mitted by Fire Marshal Clark of Prince Georges County, Maryland, US Dept of Commerce, Sept 25.1969. 3. Sax NI: Dangerous Properties of Industrial Materials, ed 3. New York, D Van Nostrand Co Inc, 1968, p 883. 4. Kishitant K: Study on injurious properties of combustive products of building materials at the initial stage of fire. J Faculty of Engineering University Tokyo (B) 31:1-3$ (No. 1) 1971. 5. Woolley WD: Decomposition products of PVC for studies of fires. J5r Polymer J 3:186-193, 1971. 6. Stone JP. Hailett RN, Johnson JE, et at: The transport of hydrogen chloride by soot from burning polyvinyl chloride. J Fire Flammability 4:42-51. 1973. 7. Stone JP: Transport of hydrogen chloride by water aerosol in simulated fire. J Fire Flammability 2:127-133, 1975. 8. Fibers, plastic fumes cause smoke deaths, editorial. Irtt Fire Fighter 55:8, 1973. 9. Medical tribune report: Human response mechanism may vary in carbon monoxide sensi tivity. Med Tribune Nov 8,1972, pp 15-16. 10. Cornish H. Abar E: Toxicity of pyrolysis products of vinyl plastics. Arch Environ Health 19:15-21, im JAMA. Jan 26. 1976--Vol 235, No 4 Polyvinyi Chtoride--Dyer & Esch 397 4 1 to Te cc O ov BFG51227