Collins, James J. PhD; Molenaar, Donald M. MD, MPH; Bowler, Larry O. PE; Harbourt, Tom J. CIH; Carson, Michael DO; Avashia, Bipin MD; Calhoun, Teresa MSN, FNP; Vitrano, Craig MD; Ameis, Paul BS; Chalfant, Richard MS, CIH; Howard, Pete BS
Phosgene gas (carbonyl chloride, CAS 75-44-5, COCl2) is a lower respiratory tract irritant and choking gas first synthesized in 1812 by exposing a mixture of carbon monoxide and chlorine to sunlight.1 Phosgene became important in the chemical industry in the 19th century, particularly in dye manufacturing. It is probably best known due to its use as a weapon in World War I. Today, phosgene is an important intermediate industrial chemical used in organic synthesis, including the manufacture of isocyanates, pesticides, pharmaceuticals, and polycarbonates. Approximately 80% of the world's production of phosgene is used to make isocyanates.1 Nearly 1 million tons of phosgene is produced each year in the United States, and the National Institute of Occupational Safety and Health estimates that there are 10,000 workers in the United States who have potential exposure to phosgene in the workplace.2,3 Currently in the United States, the American Conference of Government and Industrial Hygienists sets a threshold limit value at 0.1 ppm.4,5 Phosgene is typically handled under strict safety precautions, including in-time production, enclosure, redundant controls, with extensive area and personal monitoring.
The extent of respiratory injury from phosgene exposure is mainly determined by exposure dose rather than concentration.6,7 Thus, dose is usually expressed as the product of the atmospheric concentration in the breathing zone and the exposure duration or ppm-minutes. For example, as shown in Table 1, doses less than 60 ppm-minutes are not thought to be associated with adverse health effects. A lethal dose of phosgene is estimated to be approximately 300 ppm-minutes or higher.6,8–11 Exposure for several hours at low levels could in theory result in tissue damage and possibly death. Phosgene has a distinctive odor often described as moldy hay with an odor threshold somewhere between 0.1 and 1 ppm.12 However, olfactory inhibition occurs rapidly, and phosgene odor can be masked by other smells.6 This odor threshold is generally deemed an insufficient level for early warning of exposure.6,13 Table 1 summarizes various effects from phosgene exposure by levels and doses in ppm-minutes.7,12,14,15
The most serious health effects from phosgene exposure are due to its reactivity within the lower respiratory tract. Concentrations greater than 3 ppm can cause discomfort and irritation of the eye, nasopharynx, and upper respiratory tract because of the hydrolysis on moist membranes to hydrochloric acid.5,6 Lung tissue damage is dose dependent and occurs from penetration of the phosgene gas into the alveolar structures, resulting in chemical reactions with biologic macromolecules. Pulmonary edema may develop rapidly after a high-level exposure or slowly from a long low-level exposure after a latent period. Three distinct pathologic phases are associated with overexposure.6,8,16 The initial phase is characterized soon after overexposure by airway and mucous membrane irritation, leading to various combinations of coughing, choking, tearing, and tightness in the chest. This may be due to the partial hydrolysis of phosgene to hydrochloric acid within the aqueous microenvironment of the mucous membranes and upper respiratory tract. These initial symptoms may be followed by nonspecific secondary symptoms, such as headache, nausea, and vomiting. These secondary symptoms are thought to be due to psychological factors and are unlikely a direct effect of phosgene, which targets primarily the lungs.6 The second phase is often clinically asymptomatic and can last from 1 to 24 hours. This latency phase is inversely proportional to the dose. In the third phase, lung damage from high doses may result in pulmonary edema and respiratory failure may develop.6,8,16 A majority of persons with significant phosgene exposure have good prognoses.8,17 However, some persons complain of dyspnea and reduced physical fitness for several months after the exposure.14 Persons with certain chronic lung conditions like chronic bronchitis appear to be more susceptible to these longer-term effects. In some cases, complete recovery after phosgene exposure may take several years.14
In 2004, the American Chemistry Council phosgene panel consisting of industries producing phosgene in the United States established a phosgene registry. This registry collects health outcome information regarding acute phosgene exposure. Medical treatment practices are normally escalated as exposure dose increases.3,6 While medical treatment is generally considered unnecessary below 60 ppm-minute, more uncertainty surrounds interventions to prevent and/or treat pulmonary edema, which can be anticipated at higher exposure doses.7 This registry was named the Diller Registry in honor of Dr Werner F. Diller, a pioneering researcher who did much of the early work on phosgene badge programs and developed the rationale for treatment for overexposure while serving as an occupational physician at Bayer AG in Germany.6,8,14,18
We describe this Registry covering approximately 1455 workers at 15 sites in 2009 and provide a summary of the data collected to date. To advance understanding of the prognosis of exposed workers, we examined the relationship between phosgene exposure and selected outcome measures, primarily self-reported symptoms.
We included all phosgene-exposed workers entered into the Registry. The exposure incidents occurred any time between the start-up of the Registry in the second quarter of 2004 until the end of 2009. During this period, there were 338 workers reporting exposures at the Registry sites. An exposure incident is defined as any known or suspected unprotected phosgene exposure. Exposure to phosgene is usually determined by a personal monitoring device, phosgene detector system, but could also be based on a perceived phosgene exposure from odor or irritation. The Registry records these exposure incidents for each worker at the participating sites with a known or suspected exposure. The exposure dose to phosgene in the Registry is usually documented by a personal monitoring device, otherwise known as a phosgene badge. The badges are commercially available from multiple manufacturers, but all function by changing color on exposure to phosgene. Color change is from white to red or white to blue, depending on the type of badge that is utilized. Dose is estimated by matching the intensity of the color on a badge reader or color wheel comparator with graduated color intensities, which correspond to dose in ppm-minutes. The badges are typically very sensitive to phosgene and may start showing a color change at dose levels less than 1.0 ppm-minutes of phosgene.19 The color-change chemistry utilized by badge manufacturers is specific to phosgene, which greatly reduces the effects of cross-sensitivity to other chemicals. However, some badges are affected by ultraviolet light and moisture and are susceptible to color intensity fading when exposed to hydrochloric acid vapor. The phosgene badge is usually placed in the breathing zone by attaching to the shirt collar or adhering under the brim of the hard hat of a worker with potential for exposure. After an exposure, the duration, concentration, and dose are all estimated from a combination of worker recollection, badge readings, area monitoring, and informed industrial hygiene expertise. Thus, while in many instances, the estimated dose is the same as the badge reading, this is not always the case.
After an exposure, a questionnaire is filled out by the exposed worker and an incident assessment and a medical evaluation are completed. This questionnaire of the exposed workers records demographics (age and sex), previous lung impairments, smoking habits, and the symptoms present at time of the exposure or soon thereafter. Many of the questions in the questionnaire were taken from standardized questionnaires; however, no standardized questionnaire was fully used in this Registry.20,21 Health care professionals record the results of any diagnostics and provide a summary of the medical evaluation. Medical diagnostic measures may include baseline and periodic pulse oximetry, respiratory rate, blood pressure, chest auscultation, and chest radiograph results. Pulmonary function tests may also be administered. Pulmonary function measurements are compared to routine surveillance spirometry from before the exposure when available. Thirty days after the initial exposure, another questionnaire is completed, which records subjective symptoms; results of medical testing, including a diagnosis, if one is rendered; and the type of treatment and medications prescribed for the worker. Treatment approaches are not standardized for dose or symptoms but are left to the discretion of the treating physician. One purpose of the second questionnaire is to record any persisting respiratory tract symptoms and other effects, such as eye irritation or anxiety, after the initial exposure.
We examine the relationships between exposures and symptoms soon after exposure (up to 48 hours) and 30 days after exposure, taking potential confounding factors, such as age and sex, into account. We record whether the incident case has a preexisting lung condition (asthma, chronic bronchitis, or other long-standing lung problem), currently a smoker (“currently smoking”), or ever smoked (“smoked more than 100 cigarettes in lifetime”). We use a logistic regression model, with the presence or absence of symptoms treated as the dependent variable. Smoking habits, previous lung impairment, dose level in ppm-minutes, and phosgene exposure duration in minutes are treated as independent variables with the potential confounders of age and sex. We use a statistical software package, SAS (SAS, Inc, Cary, NC) for all calculations.22
Figure 1 shows the number of workers exposed by quarter for each year of the Registry. Some of these exposure incidents included multiple workers. One incident that occurred in December 2008 included 35 workers, and another incident in November 2008 included 13 workers. However, most incidents included only 1 (28% or 95 of 338) or 2 (22% or 73 of 338) workers. The three largest numbers of phosgene exposures occurred in the fourth quarter of the years 2004 (60 cases), 2008 (57 cases), and 2009 (40 cases).
The characteristics of the workers exposed to phosgene over the history of the Registry are presented in Table 2. The exposed workers had an average age of 39.7 years and were mostly men (98.4%). A total of 26.8% were currently smokers, 39.6% smoked more than 100 cigarettes in their lifetime, and 13.1% had a preexisting lung condition before exposure. The dose among the workers averaged 8.3 ppm-minutes and ranged from less than the limit of detection to 159 ppm-minutes. The average duration of exposure was 0.1 minutes, although two exposures were reported to last 90 minutes. We also list the percentage and the number of missing information for each of the key characteristics in Table 2. Twenty-four percent (80 of 338) of the exposure incidents had no estimated dose of phosgene exposure recorded and 52% (117 of 338) had no estimate of exposure duration. The distribution of the doses in ppm-minutes is shown in Figure 2. If we included only workers with known or estimated exposures, most workers had exposures less than 10 ppm-minutes (90.7% or 234 of 258). Only three of these workers had exposures greater than 60 ppm-minutes (1.2% or 3 of 258).
The smell of phosgene was by far the most frequently cited reason for warning of an exposure, as shown in Table 3. Almost half of the exposed workers (49.1% or 166 of 338) detected an “odor,” 18.0% were alerted by an alarm, 14.2% were alerted by a color change of the exposure badge, 8.6% were warned by other employees, and 7.4% were alert by other factors such as eye irritation, or “saw something.” Detection of an “odor” also had the highest mean exposure, 8.1 ppm-minutes, indicating that odor was effective in detecting these higher exposures. However, the color badges and the alarms were also effective in the warning of exposure, especially for lower doses, as shown in Table 3.
As reported in Table 4, almost a quarter, 24.6%, of the exposed workers reported one or more symptoms within 48 hours of exposure. The most prevalent symptoms among these workers are shown in Figure 3. The most common symptoms within the first 3 hours of exposure were coughing; taste in mouth; irritation of eyes, throat, and nose; followed by nausea, headache, and chest tightness. Most of these symptoms as well as other symptoms disappeared after the 3-hour period. Only anxiety and “Other” symptoms persisted in a few cases after 48 hours. Thirty days after exposure, however, the percentage of workers with symptoms dropped to 1.2%, as shown in Table 4. One worker had signs of pulmonary edema within 48 hours after exposure, but no workers had signs of persistent lung impairment 30 days after exposure. The worker with signs of edema was the only patient who had an abnormal result of chest radiograph among the 22 workers who underwent a radiography within 48 hours. The worker with pulmonary edema underwent a second radiography 30 days after exposure, and it was normal. This worker had an initial colorimetric badge reading of 30 ppm-minutes, which on further investigation of the incident was deemed to be unrepresentative of this worker's actual exposure. According to a written summary, he was exposed to a “mist that hit the back of his leg.” The exposure to the mist containing phosgene lasted just 5 seconds, and his badge, which was located on his collar, registered 30 ppm-minutes. Subsequent investigation determined that the badge reading was not an accurate estimate of dose because of the presence of moisture. Thus, the exposure to phosgene likely produced a much higher dose than that registered by the badge.
Eight workers (2.4%) received some medical treatment within 48 hours of phosgene exposure, and all 8 workers were given work restrictions. The average length of these restrictions was 2.9 days. One of these workers (0.3%) was also prescribed medication. Thirty days after exposure, all restrictions had been lifted on these workers and the medication for the one worker had been discontinued. Three hundred twenty-nine of 336 workers (97.9%) stated that they had completely recovered from the exposure, as shown in Table 4. The seven workers, who had reported that they were not fully recovered, did not report specific symptoms. Physicians examined 52 workers and stated that all had completely recovered.
Table 5 presents the nominal logistic regression results for the presence of symptoms within 48 hours and 30 days, considering dose in ppm-minutes, duration of exposure in minutes, and potential confounding factors. Examining dose in ppm-minutes within 48 hours of exposure, both age (P = 0.007) and the level of exposure (P = 0.001) were significant predictors of the presence of symptoms. The other potential confounding factors, including sex, ever smoked, and previous pulmonary condition, were not associated with symptoms. We also substituted the actual badge exposure level for the estimated exposure levels and current smoker for ever smoked and obtained similar results (data not shown). When duration of exposure in minutes was substituted for concentration in ppm-minutes, no association with any factor was observed.
We examine predictors of symptoms 30 days after exposure in the final two columns of Table 5. Exposure measured either as dose in ppm-minutes or duration of exposure in minutes is not predictive of the presence of symptoms 30 days after exposure. We were unable to include either sex or smoking behavior into the model because the coefficients were too unstable. We also substituted the actual badge exposure level for the estimated exposure levels and had similar results (data not shown).
The Diller Registry has several important strengths. First, the collection of exposure incidents among workers since 2004 is systematic. Since exposure to phosgene is uncommon in any single producing or using facility, the collection of more than 300 exposures to phosgene requires compiling of multiple companies' experiences. This Registry represents one of the larger groups of exposed cases analyzed to determine whether effects from exposure are occurring and the largest collection known to us specific to phosgene. Second, the Registry includes assessments of dose levels in ppm-minutes and duration of exposure for a large number of the cases. Most previous studies of phosgene-exposed workers failed to either report dose levels or durations of exposure.2 Third, the Registry collects medical outcome data at two points in time, within the hours of the exposure and 30 days after the exposure, to better assess the latent period for evidence of persistent lung impairment. Finally, this Registry is a classic example of occupational exposure and health surveillance combining exposure and outcome data, which should provide data for better understanding the risks of accidental exposures to phosgene.23 These elements are the keys to an effective occupational exposure and health surveillance program protecting and informing the worker and assuring the company and the worker of the safety of the operation.24 For low dose exposures, the Registry provides two important pieces of information: the medical signs and symptoms at two points in time and the dose as estimated in ppm-minutes by using several pieces of information, including the colorimetric badge. The absence of persistent symptoms related to dose 30 days after the low exposures and the absence of physician diagnoses of long-term impairment should be reassuring for workers, the clinicians caring for them, and the companies in the Registry.
There are also some important limitations in this study. First, the exposures to phosgene are estimates. While badges are used extensively in the phosgene areas, the dose is estimated from a combination of worker recollection, badge readings, area monitoring, and informed industrial hygiene expertise. Thus, these exposure estimates are not as precise as the controlled exposures of experimental animals in inhalation toxicology studies. Experimental studies may indeed provide a better estimate of the no-effect level of phosgene exposures than the current study or from epidemiology studies in general. Nevertheless, the results from experimental studies are useful in conjunction with our study and other epidemiology studies to provide more confidence in defining safe exposure levels in humans.
Another limitation of the study is that the conclusions about the lack of long-lasting effects rely on workers' reporting of symptoms and physician's diagnoses. The biometric data recorded in the Registry are too sparse to allow analyses for these latent effects with more objective measures. While a formal study might require all workers exposed and unexposed to have several biometric tests, such as pulmonary function, chest radiography, and arterial blood gas testing, an exposure Registry would conduct these tests only if the patient's condition determined by a physician warranted them. Thus, the data collected in this Registry, while extensive, does not have completeness typically found in a formal study. A third limitation is that the study is observational and, as such, is subject to selection bias. For example, it is possible that some workers entered the Registry on the basis of their perception of exposure rather than a real exposure. However, there was documentation of exposure for most cases in the Registry. Also, since we relied entirely on the Registry for data, we had no group of unexposed workers to serve as a comparison.
Most of the exposures in this study were low with phosgene doses less than 60 ppm-minutes and thus were expected not to require treatment according to previous studies.6,14 The few workers who had exposures to greater than 60 ppm-minutes exhibited no persistent symptomatic lung impairment or an increase in reported symptoms. One worker did develop pulmonary edema, which resolved to an asymptomatic stage. Unfortunately, we were unable to reliably estimate the dose for this worker because of the presence of moisture on the badge.
The indication of exposure to phosgene was, in almost half of the cases, discovered by odor. While the use of the personnel exposure-monitoring indicator badges allowed us to more accurately access exposure, the badges by themselves were not the most-effective method for early warning, particularly for high exposures. However, the badges and the alarm were able to warn of phosgene exposure in a number of cases, especially for lower exposures, when odor would not be a useful indicator. The badges can assist in making medical decisions by providing an estimated level of dose. These dose estimates are particularly helpful when deciding how long to observe the worker within a medical facility. While the detection of odor will likely remain a major warning device for phosgene exposure, the alarms and badges will still provide additional early warning.
The Registry establishes a database for examining exposure–response relationships for phosgene exposure and persistence of symptoms. While there is information on how much high levels of phosgene may cause death, the relationship between much-lower total doses and the persistence of symptoms have not been well studied in humans. This Registry provides a research vehicle to examine this relationship. While the current data are limited by being heavily skewed to low doses, our study still represents the largest database known to systematically examine phosgene exposures in humans. We determined that higher phosgene concentration was related to the presence of symptoms within 48 hours of exposure. However, we found no indication of symptoms being related to exposure after 30 days, and physicians diagnoses found no persistent lung impact among the exposure incidents investigated. These observations are consistent with the acute symptoms being due to concentration-dependant, transient, irritant effects on mucous membranes of the face and upper respiratory tract. These transient effects are likely a separate mechanism resulting from hydrolysis to hydrochloric acid and have separate prognostic significance from the potentially life-threatening damage deep within the lungs from higher phosgene exposures.8 These findings lend credence to the theory that prolonged respiratory effects do not occur with doses less than 150 ppm-minutes. In addition, the algorithm for the three distinct pathological phases, proposed by Diller many years ago, is also supported by our study.8,14
The number of workers with exposures greater than 60 ppm-minutes is small, and the number with lower respiratory tract symptoms is even smaller. This does limit the Registry's utility in addressing efficacy of treatment options and prognosis of phosgene-induced lower respiratory tract disease. However, the Registry already provides a unique resource for refining understanding of the outcomes of low concentration and low-dose exposures.
This research was funded by American Chemistry Council. This study conduct was pursuant to review and oversight by a Human Subjects review board in Midland, Michigan.
1. International Program on Chemical Safety. Environmental Health Criteria 193: Phosgene. Geneva, Switzerland: International Program on Chemical Safety; 1997.
2. National Institute for Occupational Safety and Health. Criteria for a Recommend Standard: Occupational Exposure to Phosgene. HEW Publication No. (NIOSH) 76-137. Washington, DC: National Institute for Occupational Safety and Health; 1976.
3. Sciuto AM, Hurt HH. Therapeutic treatments for phosgene-induced lung injury. Inhal Toxicol. 2004;16:565–580.
4. National Institute for Occupational Safety and Health. NIOSH Pocket Guide to Chemical Hazards. DHHS (NIOSH) Publication No. 2005-149. Third printing ed. Washington, DC: Department of Health and Human Services; 2007.
5. American Conference of Governmental Industrial Hygienists. 2010 Guide to Occupational Exposure Values. Cincinnati, OH: American Conference of Governmental Industrial Hygienists; 2010.
6. Borak J, Diller WF. Phosgene exposures: mechanisms of injury and treatment strategies. J Occup Environ Med. 2001;43:110–119.
7. Pauluhn J, Carson A, Costa DL, et al.. Workshop summary: phosgene induced pulmonary toxicity revisited: appraisal of early and late markers of pulmonary injury from animal models with emphasis on human significance. Inhal Toxicol. 2007;19:789–810.
8. Diller WF. Pathogenesis of phosgene poisoning. Toxicol Ind Health. 1985;1:7–15.
9. Pauluhn J. Acute nose-only exposure of rats to phosgene. Part I: concentration × time dependence of LC50s, nonlethal-threshold concentrations, and analysis of breathing patterns. Inhal Toxicol. 2006;18:423–435.
10. Pauluhn J. Acute nose-only exposure of rats to phosgene. Part II. Concentration × time dependence of changes in bronchoalveolar lavage during a follow-up period of 3 months. Inhal Toxicol. 2006;18:595–607.
11. Pauluhn J. Acute head-only exposure of dogs to phosgene. Part III. Comparison of indicators of lung injury in dogs and rats. Inhal Toxicol. 2006;18:609–621.
12. Henschler D. Gesundsheitschadliche Arbeitstoffe, Toxikologisch-Arbeitmedizinische Bergrundung von MAK-Werten, Loseblattsammlung, Phosgen, 1. Weinheim, Germany: Lieferung 1972, VCH-Verkagsgesellschaft; 1972.
13. International Program on Chemical Safety. Phosgene: Health and Safety Guide 106. Geneva, Switzerland: International Program on Chemical Safety; 1998.
14. Diller WF. Late sequelae after phosgene poisoning: a literature review. Toxicol Ind Health. 1985;1:129–136.
15. Cucinell SA, Arsenal EA. Review of the toxicity of long-term phosgene exposure. Arch Environ Health. 1974;28:272–275.
16. Russell D, Blaine PG, Rice P. Clinical management of causalities exposed to lung damaging agents: a critical review. Emerg Med J. 2006;23:421–424.
17. Galdston M, Leuetscher JA, Longcope WT, et al.. Residual effects of phosgene poisoning in human subjects-II: after chronic exposure. J Clin Invest. 1947;26:169–181.
18. Ott MG, Diller WF, Jolly AT. Respiratory effects of toluene di-isocyanate in the workplace: a discussion of exposure-response relationship. Crit Rev Toxicol. 2003;33:1–59.
19. Dunlap KL. Phosgene. Kirk-Othmer Encyclopedia of Chemical Technology. New York, NY: John Wiley & Sons Inc; 2001.
20. Venables KM, Farrer N, Sharp L, Graneek BJ, Newman-Taylor AJ. Respiratory symptoms questionnaire for asthma epidemiology: validity and reproducibi-lity. Thorax. 1993;48:214–219.
21. Guyatt GH, Berman LB, Townsend M, Puglsey SO, Chambers LW. A measure of quality of life for clinical trials in chronic lung disease. Thorax. 1987;42:773–778.
22. SAS. SAS System (for Windows). Release 8.0. Cary, NC: SAS Institute; 2003.
23. Schulte PA, Trout D, Zumwalde RD, et al.. Options for occupational health surveillance of workers potentially exposed to engineered nanoparticles: state of the science. J Occup Environ Med. 2008;50:517–526.
24. Halperin WE. The role of surveillance in the hierarchy of prevention. Am J Ind Med. 1996;29:321–323.