LETTER TO THE EDITOR |
Vapours and Aerosols of Bitumen: Exposure Data Obtained by the German Bitumen Forum
The recently published paper by Rühl et al. (2006) describes the results of measurements of exposure to bitumen fumes and vapours in a variety of industries where hot bitumen is handled in Germany. Overall, it is to be applauded that results of studies assessing exposure in a variety of industries and workplaces are being published in peer-reviewed literature and do not stay hidden in large inaccessible databases or grey literature. We consider the Annals of Occupational Hygiene to be the world-class forum for publishing such studies and it has done so increasingly over the last two decades. Given that quantitative exposure data are becoming more and more essential to modern epidemiology for establishing exposure–response relations and risk assessment (Stewart et al., 1996; Loomis and Kromhout, 2004) it is of utmost importance that peer review of such studies is done rigorously. (By doing so, the readership and future users of the data are offered the opportunity to fully evaluate and appreciate the presented data.) Having carefully appraised the paper by Rühl et al. (2006) we are of the opinion that in this particular case adequately critical review by the journal has not taken place. We therefore would like to highlight several important issues that will enable readership to appreciate the value of exposure survey described by Rühl et al., as well as the appropriateness of the manner in which it was presented and interpreted in the paper.The manuscript lacks structure common to scientific communications of this nature. Basically, a detailed ‘Methods’ section is missing and is instead replaced by an extremely brief summary of measurement methods without any exposition on the matters of measurement strategy or selection of workers and workplaces for monitoring. It is known that measurement strategy will to a large degree determine the statistical parameters of an exposure concentration distribution. Therefore, failure to describe measurement strategy renders interpretation of measurements impossible and any such attempt meaningless.
It is almost impossible to ensure that in the highly variable environment of asphalt work (outdoor conditions are known to vary enormously depending on meteorology, batch-wise production and application, etc.) a representative sample can be obtained over a 2-h period from a shift that may well exceed 8 h. This contradicts basic principles of survey design and is a fatal flaw, since apparently the aim of the survey was to draw conclusion about typical full-shift exposures. Comments by the authors about how sampling many worksites addressed this problem are not convincing because this does not enable one to separate between worksite variability from within-shift variability. Two-hour time-weighted convenience sampling, as it is endorsed in the German legislation, leads to biased estimates of both variability and mean shift-long exposure. The potential of this bias and its consequences should at least have been discussed and not swept swiftly under the carpet as the authors did by stating: ‘Usually, sampling duration was 2 h. The tasks often lasted for a full 8-h shift, and the sampling period was assumed to be representative of shift-length exposure.’
The validity of the approach taken by the authors to deal with non-detectable values depends on the proportion of non-detectable values in each of the strata in which measurements were classified (Hornung and Reed, 1990). These figures should have been provided in Table 1. The authors should also have described method by which the limit of detection was estimated, as well as other relevant quality control procedures.
The statement by the authors on ‘large’ numbers of worksites/measurements has to be qualified in a sense that it should be made clear why 17 measurements from bitumen production is a large number, while for manual mastic operations 372 measurements is qualified in a similar manner as ‘a large number’ (Fig. 2). The number of worksites monitored in each stratum should have been presented.
In Table 1, standard measures of variability such as geometric standard deviation are lacking. Also, appropriate statistical tests should have been performed to compare exposures in different jobs and industries with exposure limits. The lack of this makes it impossible to judge the validity of conclusions drawn from Table 1 and Figs 2, 6, 7 and 8. There are many speculations in the part about what caused exposure in each particular situation. Judgements of (potential) determinants of exposure should and can nowadays be based on valid statistical models and tests, instead of speculation (Burstyn and Teschke, 1999).
Careful review of the cited paper by Rühl et al. (2002) does not reveal models that can be used to predict exposures. Thus, we think this reference is misleading and overstates the case for there being no longer any need to conduct exposure surveys at particular worksites. If this online resource is to be cited, its relation to data presented in the current manuscript has to be explained. For example, do measurements collected after 2002 (current paper presents data collected till 2005) agree with predictions? This gives an opportunity for model validation that the authors apparently missed. In any case, clear connection with the current manuscript must be shown to justify insertion of a recommendation to use an online resource in lieu of measuring exposures in specific workplaces.
The statement that ‘with the current use of bituminous materials the PAH exposure is several magnitude levels lower than the PAH exposure with the earlier use of tar’ contradicts previously published results from across Europe that indicate that modern exposure to benzo[a]pyrene during mastic work and re-paving can be equal to those that were observed in the past during use of tar-containing materials (Burstyn et al., 2000a, 2002a). In our opinion, Rühl et al. overinterpret their data in this case and fail to place their results in the context of previously acquired knowledge. At the very least, they should have argued why they disagree with or disregard previously reported evidence.
A final point is that references mentioned in the manuscript seem to be highly selective and incomplete. Rühl et al. do not mention papers on exposure of asphalt workers from the Finish Institute of Occupational Health (Heikkila et al., 2002, 2003; Vaananen et al., 2003) and Harvard University (McClean et al., 2004a,b). Furthermore, they cited only one of many relevant manuscripts that emerged from the exposure assessment for a multi-centre epidemiological study of European asphalt workers, omitting five other relevant papers (Burstyn et al., 2000a,b; Burstyn and Kromhout 2000c, 2002a,b). Furthermore, any paper on exposure to bitumen and its health effects in the asphalt industry is also incomplete without at least some reference to the (ongoing) multi-centre study on lung cancer and other diseases among European asphalt workers that was initiated by the International Agency for Research on Cancer in 1995 in close collaboration and with support from the asphalt industry (including the Germany paving industry) (Boffetta et al., 2003a,b; Burstyn et al., 2003, 2005). It should also be noted that three references in the publication by Rühl et al. (2006) are to industry/government reports that are not readily accessible, three references are to materials yet to be published or even yet to be submitted, one reference is a commentary of the authors on their own paper and one reference is to a personal communication. Thus, out of 13 cited references, only five were peer reviewed. This would be seen by most academic journals as clearly inadequate and probably biased towards the authors’ perspective.
We sincerely hope that review standards at our favourite scientific publication in the field of occupational exposure assessment will stay at a level the scientific state of the art requires.
1 Institute for Risk Assessment Sciences, Utrecht University, PO Box 80176, NL-3508 TD Utrecht, The Netherlands
2 Department of Medicine, University of Alberta, Edmonton T6G 2G3, Canada
Received September 4, 2006; in final form October 20, 2006
REFERENCES
Boffetta P, Burstyn I, Partanen T, et al. Cancer mortality among European asphalt workers: an international epidemiological study: I. Results of the analysis based on job titles. Am J Ind Med (2003b) 43:18–27.[CrossRef][Web of Science][Medline]
Boffetta P, Burstyn I, Partanen T, et al. Cancer mortality among European asphalt workers: an international epidemiological study: II. Exposure to bitumen fume and other agents. Am J Ind Med (2003a) 43:28–39.[CrossRef][Web of Science][Medline]
Burstyn I, Boffetta P, Burr GA, et al. Validity of empirical models of exposure in asphalt paving. Occup Environ Med (2002a) 59:620–4.
Burstyn I, Boffetta P, Heederik D, et al. Mortality from obstructive lung diseases and exposure to polycyclic aromatic hydrocarbons among asphalt workers. Am J Epidemiol (2003) 158:468–78.
Burstyn I, Ferrari P, Wegh H, et al. Characterizing worker exposure to bitumen during hot mix paving and asphalt mixing operations. Am Ind Hyg Assoc J (2002b) 63:293–9.
Burstyn I, Kromhout H. Are all the members of a paving crew uniformly exposed to bitumen fume, organic vapour and benzo(a)pyrene? Risk Anal (2000c) 20:653–63.[CrossRef][Web of Science][Medline]
Burstyn I, Kromhout H, Boffetta P. Literature review of levels and determinants of exposure to potential carcinogens and other agents in road construction industry. Am Ind Hyg Assoc J (2000b) 61:715–26.
Burstyn I, Kromhout H, Kauppinen T, et al. Statistical modelling of the determinants of historical exposure to bitumen and polycyclic aromatic hydrocarbons among paving workers. Ann Occup Hyg (2000a) 44:43–56.
Burstyn I, Kromhout H, Partanen T, et al. Polycyclic aromatic hydrocarbons and fatal ischemic heart disease. Epidemiology (2005) 16:744–50.[CrossRef][Web of Science][Medline]
Burstyn I, Teschke K. Studying the determinants of exposure: a review of methods. Am Ind Hyg Assoc J (1999) 60:57–72.[Web of Science][Medline]
Heikkila P, Riala R, Hameila M, et al. Occupational exposure to bitumen during road paving. AIHA J (2002) 63:156–65.[CrossRef]
Heikkila PR, Vaananen V, Hameila M, et al. Mutagenicity of bitumen and asphalt fumes. Toxicol In Vitro (2003) 17:403–12.[CrossRef][Web of Science][Medline]
Hornung RW, Reed LD. Estimation of average concentration in presence of nondetectable values. Appl Occup Environ Hyg (1990) 5:46–51.
Loomis D, Kromhout H. Exposure variability: concepts and applications in occupational epidemiology. Am J Ind Med (2004) 45:113–22.[CrossRef][Web of Science][Medline]
McClean MD, Rinehart RD, Ngo L, et al. Urinary 1-hydroxypyrene and polycyclic aromatic hydrocarbon exposure among asphalt paving workers. Ann Occup Hyg (2004a) 48:565–78.
McClean MD, Rinehart RD, Ngo L, et al. Inhalation and dermal exposure among asphalt paving workers. Ann Occup Hyg (2004b) 48:663–71.
Rühl R, Lechtenberg-Auffarth E, Hamm G. The development of process-specific risk assessment and control in Germany. Ann Occup Hyg (2002) 46:119–25.
Rühl R, Musanke U, Kolmsee K, et al. Vapours and aerosols of bitumen: exposure data obtained by the German bitumen forum. Ann Occup Hyg (2006) 50:459–68.
Stewart PA, Lees PS, Francis M. Quantification of historical exposures in occupational cohort studies. Scand J Work Environ Health (1996) 22:405–14.[Web of Science][Medline]
Vaananen V, Hameila M, Kontsas H, et al. Air concentrations and urinary metabolites of polycyclic aromatic hydrocarbons among paving and remixing workers. J Environ Monit (2003) 5:739–46.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||