Document jgYVpyGX9YGD25oM78pE55On9

94178996 ' Authors Belletti I. Troschel L. Pisani P. Berrino F. Title Retrospective assessment of exposure by experts: the Bexample of formaldehyde, solvents and mineral oils among m4um mounw6 textile and metal workers. coql Source # L m . -1 International Journal of Epidemiology. 22 Suppl 2:S127-33, 1993. I tal exposures in Aogical survey is '.echerche et de la Travail, Technale d'Assurance en supported by a del'Enseignement per J A. Discovering ,ivironrnent: a novel IInst 1981; 66:217-2S. D T. Aci occupational ly of occupationalcar2-26, m of occupationalexitrol studies. General posure matrices. h r J ning retrospective ex' Hyg 1991; 6: 421-27. of long term exposure re data: a tool for Wed 1987; 4Q: 133-41. ig historicalexposures tality study. Appl Ind ined by questionnaire. a: 284-88. ilidity of self reported `9-32. :rger R J, Flares G H. rorn interview%Am J JI. Validity of work demiologic purpose\. . Blurn S. Ai1 analysis onal histories using a J Ind ,Wed 1987; 44. nnairc reliability and 7,. Scan J Work En- z-ilitv of self-reported jrn patients in EEC -79. r W R. Development occupational health k 26: 50-54. ). Obtaining occupa. case-control studie\. i i International Journal of Epidemiology InternationalEpidemiological Association 1993 Vol. 27. No. 6 ISuppl. 2) Printed in Great Britain Retrospective Assessment of Exposure by Experts: The Example of Formaldehyde, Solvents and Mineral Oils among Textile and Metal Workers ISABELLA BELLElTI, LOREDANA TROSCHEL PAOLA PlSANl AND FRANC0 BERRINO Belletti I (EpidemiologyDivision, lstituto Nazionaledei Tumori. Via Venezian, 1, 1-20133 Milan, Italy),Troschel L, Pisani P and Berrino F. Retrospectiveassessment of exposure by experts: The example of formaldehyde, solvents and mineral oils among textile and metal workers. InternationalJournal o f Qidemiology 1993; 22 (Suppl. 2): S127-5133. The validity of retrospective assessment of occupational exposure greatly depends on the amount of detail in the available information, on the knowledge of the specific industrial process by the experts, and on the criteria adopted to define relevant exposure. These criteria are difficult to standardize and are rarely made explicit in published reports, which makes it difficult to interpret inconsistencies among different studies. In two ongoing case-control studies on kidney cancer and, respectively. malignant lymphomas, a detailed occupational history was obtained and supplemented by 19 additional questionnaires, specifically addressing industrial activities where the knowledgeof job title alone would have been insufficient for reliable exposure assessment. One further questionnaire was used to collect details of task and environment for all the other activities. These data are used to establish probability, intensity and frequency of exposure to 30 substances known or suspected to be carcinogenic from previous studies. There are two basic steps in the exposure assessment procedure: firstly, general rules are defined for each job within each activity covered by specific questionnaires; secondly the judgement is modulated according to the detailed tasks, working conditions and environment. To illustrate the process and to facilitate comparison with other studies, examples are given for a few common exposures in the textile and metal industries-thetwo most frequent economic activities in the study area-namely exposure to mineral oils, formaldehyde, aromatic solvents, chlorinated solvents and other organic solvents. In two population-based case-control studies on kidney cancer and malignant lymphomas, which are ongoing in the Varese province (Northern Italy). complete j o b histories were collected, and supplementary questionnaires for specific industrial activities were also used. These specific questionnaires (SQ)were designed to improve the expert assessment of exposure to a number of nuisances whose presence in the work environment cannot easily be determined from the job title alone and the economic activity of the establishment.) They have been developed in order to obtain a detailed description of any single task and work en- vironment. * Beyond the qualitative assessment of the exposure, this detail helps to estimate its intensity and ~ Epidemiology Division, lstituto Nazionale del Tumon, Via Venezian. I.1-20133 Milaii, Ital). frequency. Even with SQ the validity of assessment greatly depends on the knowledge of industrial processes by the experts, on the information actually elicited with the questionnaires, and on the concept of what is individual exposure, namely the physical conditions and minimal dose of a compound regarded as a relevant exposure.3 They also depend on the ability of the expert to translate anecdotal information into a quantitative assessment. All these determinants of classification may vary from study to study and are not usually described in detail when presenting results. In order to facilitate comparison with other studies this paper attempts to report explicitly.-with the help of a few examples, the criteria followed and uncertainties persisting in the process of retrospective assessment of exposure. Suggestions for improving the process are also given. SI28 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY (SUPPLESlENT 2) MATERIALS AND METHODS Substances and Industrial Activities Considered Occupational exposure to all the substances belonging to IARC Monographs' groups 1, 2A, 2B (carcinogenic, probably and possibly carcinogenic substances) and related to malignant lymphomas or kidney carcinomas have been considered. J5 Some other substances have been added because of their mention in epidemiological studies or to validate a job exposure matrix constructed in a previous study.? In the end 30 substances or groups of substances have been included: acetamide, acrylonitrile, p-aminoazobenzene, aromatic solvents, arsenic, asbestos, atrazine, benzene, bitumens, 1,3-butadiene, cadmium, chlorinated pesticides, chlorinated solvents, chlorophenols and chorophenoxy compounds, chromium, coal tars and derived products, p-dichlorobenzene, ethylene oxide, formaldehyde, lead, mineral oils, nickel, nitrilotriacetic acid, N-nitrosodiethanolamine, other organic solvents (OOS), polychlorinated biphenyls, polynuclear aromatic hydrocarbons, styrene, tris (2.3-dibromopropyl) phosphate, vinyl chloride. In order to exemplify the criteria used to assess exposure, this paper focuses on the textile and metal industries, because of their prevalence in the study area. Relevant substances present in both industries have been selected: mineral oils, formaldehyde, aromatic solvents, chlorinated solvents and 0 0 s . Together with industrial hygiene reports from local industries a number of references have been Questionnaires Specific questionnaires on the following 19 industries/operations have been used whenever the corresponding industry or job was mentioned in the occupational history: agriculture, breeding, slaughterhouses and meat processing, foundries, metal industry, welding, plating and coating metal products, body shops, chemical industry, paint manufacture, rubber industry, plastic manufacture, textile industry, wood industry, leather and footwear industry, building, transport, electricians, laundries. One further general questionnaire has been used for all the other activities. The questionnaires have been designed following a detailed analysis of the relevant working areas, jobs and operations within an industry. l 3 For instance, the metal industry questionnaire asks about which operations are performed in hot or cold metal rolling, in quenching and with different machine tools. It also enquires about the use of mineral oils as lubricants, as cutting or quenching fluids and for metal forming; about the presence of furnaces in hot metal rolling; about the use of specified or non-specified solvents; about the presence of different kinds of painting; about insulating operations and materials. All SQ enquire about company's activities, shift work, working environment (presence of fumes, dusts or vapours, use of exhaust fans), individual protection, job type (including temporary and occasional work), raw materials and final products. Even if some direct questions about particular chemicals are present (formaldehyde in the textile industry, mineral oils and paints in the metal industry, solvents in different activities, etc.), the principal aim is to collect a detailed and complete description of each job, leaving to the experts the task of establishing exposure to specific chemicals. It has been shown, in fact, that a subject is more likely to remember operations and equipment than chemical names. 2*iJ.Is Exposure Categories The exposure categories chosen for this study have evolved from those used in previous studies. Pastorino ef a/. l6 assessed only the probability of exposure and classified subjects' exposure as certain o r highly prob- able (+), uncertain (?), absent (-). N o judgement of intensity and frequency of exposure is present in this classification. Ferrario et 01. developed a job exposure matrix in which the exposure was categorized with regard to both probability and intensity: 1) job-related exposure is not higher than for the general population; 2) job may entail a cumulative exposure slightly higher than for the general population; 3) job may entail exposure to levels definitely higher than for the general population, but available information is not sufficient to discriminate between exposed and not exposed; 4) job entails exposure to the specific agent; 5 ) j o b entails exposure to the specific agent and exposure is known to be particularly high. Category 3) (uncertain exposure) was further divided into three levels according to the probability of exposure (<1/3. 113-2/3, >2/3) but did not give information about intensity. In the present study, in order to retain most of the available detail, probability. intensity and frequency of exposure have been separately assigned for each job title and for each substance. Categories of exposure Probability: 1-not exposed; 2-exposure probability <33%; 3-exposure probability 33-66%; 4-exposure probability >66%; 5-exposed. Intensify: 0-not exposed; 2-very low intensity: groups of the general population can be exposed with this intensity for non-occupational causes; I I ! i I 3-mediui general r - Frequen: posed; 2 4-freque Exposure As: The procedur a) For each Si are decided. working area. (e.g. in the t t systematicall? in the spinnir different dep: ing); jobs wir present in t chemical deli\ similar charL transport, go( b) Exposure decided on modified by ment and task Further rei posure assess: present in all - Descripti specify workers . direct ex For exan cleaning the near to a sol, such as Labor 0. exposure - The pres1 of an e?. have heir - A possik units, for resulted posure. 1 be consic: RESULTS Exposure A SJ Mineral oils. . used as: ot metal rolling; 3ecified solvents; nds of painting; :rials. A11 SQ enft work, working USIS or vapours, tection, job type nal work), raw about particular in the textile in:metal industry, the principal aim :scription of each f establishing exbeen shown, in remember operaames. 2.14.15 this study have tudies. Pastorino of exposure and 1 or highly prob\lo judgement of is present in this td a job exposure categorized with ty: 1) job-related ieral population; re slightly higher > bmay entail ex7 for the general I is not sufficient not exposed; 4) nt; 5 ) job entails posure is known ) (uncertain exlevels according 113-2/3, >2/3) tensity. .tain most of the r` and frequency :ned for each job ries of exposure fxposure probaability 33-66%; -exposed. y low intensity: can be exposed iational causes; f ASSESSMENT OF EXPOSURE TO FORMALDEHYDE. SOLVENTS AND MINERAL OILS S129 3-medium intensity: definitely higher than in the general population; 4-high intensity. - Frequency: (within a working year) 0-not ex- posed; 2-frequency <540; 3-frequency 5-50%; &frequency >50% of the working time. - lubricants for all types of natural fibre spinning machines. - fibre lubricants in the spinning of wool and jute (generally an aqueous solution of mineral oils is - sprayed on the fibres). loom lubricants. In particular Sulzer looms which Exposure Assessment The procedure is the following: a) For each SQ general criteria of exposure assessment are decided. Criteria are mainly based on the principal work at high speed and have `projectiles' needing continuous lubrication; often the weaver has to operate in front of the loom where a box containing hot oil is situated. working areas on which the questionnaire is structured Therefore, a high probability and a medium inten- (e.g. in the textile industry formaldehyde exposure is systematically considered in the finishing area and not in the spinning area), but might also includc: jobs in different departments (e.g. maintenance, room clean- sity of exposure to mineral oils have been estimated for the spinning and weaving of natural fibres; a higher intensity of exposure has been estimated for Sulzer weavers and for jute and wool spinners. Specific ex- ing); jobs with possible exposure to all the chemicals amples of exposure assessment are given in Table 1. present in the specific industry (e.9. warehouse, chemical deliveries to various working units); jobs with similar characteristics in different industries (e+. transport, goods loading and unloading, etc.). TABLE 1 Examples of exposure as~essnren~-iMineraoI ils in the textile industry b) Exposure assessment of any single job period is decided on the basis of the general criteria and Prob. Int. ILOa Taskb modified by the answers about the specific environ- ment and task. Further relevant information to modify the ex- posure assessment derives from the general questions present in all questionnaires. For instance: - Description of the working unit has helped to specify the operations performed by other workers near the subject and thus any possible in- direct exposure from the working environment. For example, in the metal industry, painting and cleaning operations have often been reported in the near surroundings of the work place. This led to a solvent exposure assessment for workers, such as some machine assemblers (International Labor Office-ILO code 841 lo).'' for whom this - exposure was not always expected. The presence of fumes, dusts or vapours, the use of an exhaust fan or of individual protection, - have helped in estimating exposure intensity. A possible temporary move to other working units, for example to substitute other workers, has resulted in an evaluation of low frequency ex- posure. The issue of frequency, however, will not be considered here. 2 2 97145 Warehouse porter: carries and stacks goods in warehousc; no indication of handling chemicals 2 3 75440 Weaver lubricating looms: possible use of graphite as lubricant instead of mineral oil\ 3 2 75430 Hand-operated machine weaver; other nonspecified looms near !he work place 4 2 79450 Cutting and sewing machine operator and ironer; spinning operations near the work place 4 3 70070 Supervisor and fibre preparing machine setter 4 3 75440 Weaving machine operixor 4 4 75220 Wool spinner and non-specified loom weaver 5 2 75430 Hand-operated machine weater: loom lubrication wi!h mineral oils 5 3 85560 Maintenance electrician in the textile industry 5 4 75440 Sulzer loom operator ~~ Prob. = Categories for probability of exposure. h i . = Categories for intensity of exposure. a The complete description of the working activity has bren used for each exposure assessment; the 1LO code" was not used for this purpose. It is present in the various Tables IO facilitate comparisons (e.g. with job exposure matrices). Synthesis of information from questionnaires which led to exposure assessment is reported. RESULTS Exposure Assessment in the Textile Industry Mineral oils. In the textile industry mineral oils can be used as: Solvents. A direct question concerning the use of solvents is present for printing and finishing, but solvent exposure is possible also for other operations. In SI30 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY (SUPPLEMENT 2) rayon acetate spinning, for example, some exposure to benzene. toluene, or dichloromethane is likely. In additives) and in sizing operations (in crease-resistant agents such as formaldehyde resins). Exposure to for- TABLE 3 EXU rexrile indusrrs weaving, loom lubrication may involve exposure to solvents in loom cleaning and for hand cleaning. In maldehyde is also possible for all operations related to direct contact-with finished fabrics that could have Prob. Int. sizing operations chlorinated solvents, such as per- been previously treated with formaldehyde releasing chloroethylene, instead of warer, can be used. In dye- agents, for example: cutting, sewing, ironing, packag- i 22 ing. chlorinated solvents, such as perchloroethylene, ing, textile quality control and so on. 23 trichlorocthylene and 1.1.1 -trichloroethane, can be used instead of aqueous solutions which are more frequently used. In colour mixing, solvent exposure can TABLE 2 Examples of exposure assessment-OOS solvents in the textile indumy 32 be due, in particular, to printing paste preparation because of the use of organic solvents as coadjutants. The use of this paste is then responsible for solvent exposure in printing. In finishing, solvents can be used in cleaning and sizing operations (in particular chlorinated solvents for wool and 0 0 s solvents for all other fibres). A certain probability of exposure to solvents is also possible for operations on finished fabrics that could have been just cleaned by solvents, such as cutting, sewing, ironing, textile quality control and so on. When no direct information on the class of solvents Prob. Int. ILO Task 2 2 97145 Warehouse poner: opens packages. loads trolleys for dyeing operations 2 2 70070 Supervisor; occasionally machinery maintenance 2 3 75470 Ironing and textile quality control; no direct 56060 use of chemicals 2 3 92950 Textile printer; says he has never used solvents (prob. 2. in . 3 for chlorinated artd aromatic solvents) 1 33 34 42 42 43 43 53 was available, a higher probability of exposure has been assigned for chlorinated and 0 0 s solvents than for aromatic ones; the latter, in fact, were less fre- quently mentioned in the industrial hygiene reports from the study area that were available to us. High probability and intensity of exposure have been frequently assigned for chlorinated and 0 0 s solvents in printing, colour mixing and finishing. Medium probability and usually low intensity of ex- posure have been estimated for chlorinated and 0 0 s solvents in cutting, sewing, dyeing, and unspecified artificial fibre spinning; medium probability and usually high intensity exposure to aromatic solvents have been estimated for printing, colour mixing and artificial fibre spinning. Low probability and medium-low intensity of ex- posure have been estimated for chlorinated and 0 0 s 3 2 79450 Cutting and sewing machine operator (prob. 4, in!. 2 for chlorinated solvents) 3 4 7S690 Textile cleaner (prob. 4. int. 4 for chlorinated solvcnts) 3 4 75220 Spinner of unspecified fibres; presence in the factory of acrylic, polyester and rayon fibres (prob. 3. int. 4 for chlorinated and aromatic solvents) 4 4 92950 Textile printer: use of unspecified solvcnlS (prob. 4, in!. 4 for chlorinated solvent-; prob. 3, int. 4 for aromatic solvents) 5 4 92950 Textile printer; use of turpentine (prob. 3. in1 . 3 for chlorinated and aromatic tolvefllS) N.B. See footnotes to Table I . Medium or high values of probability and intensity of exposure to formaldehyde have been estimated for printing, sizing, finishing (particularly for high tem- N.B. See fool ing, machi Mineral oi rosive coal long time. Exposur the metal affirmativ not answe been estin the subjec High ir operation quenchini machine 1 solvents in sizing; for aromatic solvents in dyeing; for perature operations such as ironing and calendering). A lowe all solvents in warehouse-related activities. Low probability of exposure has also been estimated for workers Low probability and low intensity of exposure has been estimated for warehouse porters and goods lubricatic exposure in printing and finishing who claim never to have used distribution operators. Specifi solvent<;it has been assumed, in fac ,that the operator may have been unaware of the presence of solvents in Specific examples of exposure assessment are reported in Table 3. reported printing pastes or auxiliarie\, which might involve medium intensity exposure ai least. Specific examples of exposure assessment are reported in Table 2; the Table refers to 00s solvents, but it also comments on aromatic and chlorinated solvents. Formaldehyde. In the textile industry formaldehyde may be present in corrosion printing (in printing paste Ekposure Assessment in the Metal Industry Mineral oils. Mineral oils have different functions in the metal industry. They can be used, for example, for machine lubrication, friction reductioii, metal cooling, corrosion and rust prevention and chip removal. With emulsifying agents they can also be used as aqueous solutions. Operations related to mineral oil exposures are, for example. rolling, pressing, drawing. quench- Solvents. possible i the quest of solver specific I jects, dir is availa followin; n crease-resistant Exposure to forrations related to that could have dehyde releasing ironing, packag- -00ssolvenrs in rhe \ packages, loads itions machinery iy control; no direct ias never used for chlorinated and bine operator (prob. solvents) int. 4 for abres; presence in the rter and rayon fibres mated and aromatic ispccified solvents mated solvent-; itic solvents) rpentine (prob. 3. 1 aromatic solvents) ity and intensity cn estimated for y for high temnd calendering). of exposure has .ers and goods assessment are /UStQJ mt functions in .cor example, for metal cooling, 3 removal. With sed as aqueous a1 oil exposures awing. quench- ~ ASSESSMENT OF EXPOSURE TO FORMALDEHYDE, SOLVENTS AND MINERAL OILS SI31 TABLE 3 Exomples of exposure wessmenl-Formoldehyde in rhe rexrile indusrry TABLE 4 Exompies of e.rposure ossesmenr-Minerol orls in rhe merol industrv prob. Int. ILO Task Prob. Int. ILO Task 2 2 97145 Warehouse porter: carries and stacks goods 2 3 75670 Calender operator who declared that fabrics were coming from bleaching (and so probably formaldehyde free) 3 2 39140 Storeroom clerk; yarn distribution and final product deliver); fabric cutting near the work place; vapour presence deriving from other rooms 3 3 92950 Textile printer 3 4 75690 Sizer; the worker said he used chemicals vapour presence 4 2 79550 Cutting and sewing machine operator 4 2 795IO Embroidered table cloth repairer 4 3 56060 Pressing machine operator: cutting and packaging operations near the work place 4 3 75670 Calender operator 5 3 75440 Weaving, cutting and finishing operator; use of formaldehyde and exhaust ventilation system presence N.B.See footnotes to Table I . ing, machine tool working and machine lubricating. Mineral oils can also be used as antirust and anticorrosive coatings for parts which have to be stored for a long time. Exposure to mineral oils has often been assessed in the metal industry because many subjects answered affirmatively to the direct question. If the question was not answered, a medium probability of exposure has been estimated for machine tool operators and for all the subjects working close to them. High intensity of exposure has been estimated for operations using high temperatures such as forging or quenching and for the presence of many automatic machine tools in the working environment. A lower intensity has been assessed for machine tool lubrication when this was the only possible source of exposure. Specific examples of exposure assessment are reported in Table 4. Solvents. Solvent exposure in the metal industry is possible in different working areas. For this reason in the questionnaire a direct question concerning the use of solvents is included, but without reference to any specific operation or department. So, for most subjects, direct information is available. When no answer is available, solvent exposure is assumed for the following operations: painting and surfacing; metal 2 2 84910 Engine maintenance wlth hand tools; similar operations near the work plac ; no prexnce of vapours and fumes 2 2 85920 Security inspector for whole factory 3 2 99910 Scrap man also involved in cleaning Of production departments 3 3 84175 Textile machinery assembler w t h hand tools and installer in place of use . 3 3 99910 Metal-sawing machine operator; no direct use of mineral oil.; presence of different machine tools near the work place 3 3 99910 Apprentice: cleaning of mechanical equipment 4 3 70050 Machine assembler with hand tools; automatic lathes near the work place 4 3 97145 Warehouse porter: lathes, millers and other machine tools near the work place 4 4 84910 Machine tool operator: the interviewee (next of kin) could not respond about the use Of mineral oils 5 3 83420 Lathe operator without cutting oils; machine lubrication; other machine tools near the work place 5 4 83420 Lathe and grinder operator; use of aqueous solutions of mineral oils and machine lubrication 5 4 84910 Oil quenching N.B. See footnotes to Table i . workparts degreasing; materials, machines and work place maintenance; any job entailing solvent manipulation (fuels included); hand cleaning for workers using lubricants or paints. Solvents can also be present in exhaust fumes from fuel engines or furnaces. Exposure to chlorinated solvents has been assumed for painting and for workers involved in machine tool lubrication, the latter because of cleaning of machines, metal work parts, work place and hands. Aromatic and 0 0 s solvents have been associated with the presence of exhaust fumes, with painting and with the use, for any purpose, of gasoline, naphtha, diesel oil, kerosene and white spirif. Exposure graduation has been estimated as follows: - 00s solvents. High probability has been assigned for any use of gasoline, diesel oil, kerosene, naphtha, white spirit; for thepresence of exhaust fumes from gasoline or diesel oil; for painting operations. Medium probability of exposure has been assessed for machinery lubrication and fp_r operators of non-specified furnaces. . SI32 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY (SUPPLEMENT 2) - Aromatic solvents. Graduation of probability and - intensity was very similar to those for 00s. Chlorinated solvents. High probability and low TABLE 6 Examples of exposure arressment-Formaldehyde in the metal industry intensity of exposure havi' been assessed for gasoline use; medium probability and high intensity for degreasing and painting. Specific examples of exposure assessment are reported in Table 5 ; Table 5 refers to 00s solvents, but it also contains comments on aromatic and chlorinated solvents. TABLE 5 Examples of exposuw assmment-OOS solvents in the mtd indust~ Prob. hi. ILO Task 2 2 84175 Textile machine assembler and installer in place of use 2 2 99910 Apprentice: cleaning operations in a mechanic unit 3 2 83130 Oil quenching 3 2 83130 Pressing wirh nontmulsified mineral oils 4 2 84985 Work pans controller; machine tools with emulsified mineral oils in the work place 4 3 83420 Lathe and grinder operator; ux of aqueous solutions of mineral oils and machine lubrication ~~ ~ ~~~ Prob. Int. ILO Task N.B. See footnotes to Table I 3 3 83130 Pressing and forging; non-specified furnaces; f u m e in the environment (prob. 3. int. 3 for aromatic solvents) 3 4 83970 Bending machine operaror: usc of unspecified solvents for unspecified operations (prob. 3. int. 4 for aromatic and chlorinated solvents) 4 2 99940 Cleaning operations in painting unit (prob. 4. in'. 2 for aromatic solvent.; prob. 3, ini. 2 for chlorinated solvents) 4 2 84175 Machine assembler; painting near the work place but in an isolated room (prob. 4. int. 2 for aromatic solvents; prob. 3. int. 2 for chlorinated solvents) 4 2 85920 Safety control of machina; presence of painting operations (prob. 4. ini. 2 for aromatic solvents; prob. 3, int. 2 for chlorinated soivents) 4 3 87370 Car body repairer; use of nitrocellulose stopping (prob. 4. in!. 3 for aromatic solvcnts) 4 4 84110 Compressor assembler; spray painting; no information about solvent u x (prob. 4. in!. 4 for aromatic solvents; prob. 3, int. 4 for chlorinated solvents) 5 3 841 IO Machine tool assembler; painting near the work plac ; nitrocellulose paint vapours in the environmeni (prob. 5. int. 3 for aromatic solvents) 5 3 83110 Fuel furnace operator (prob. 5. int. 3 for aromatic solvents) 5 4 84910 Machine tool operator; u x of nitroceliulose solvents to clean machine tools and spray painting (prob. 5. int. 4 for aromatic solvents) J 4 87350 Lathe operator; use of naphtha and trichloroethylene to clean both the machine and the work parts (prob. 5. int. 4 for chlorinated and aromatic solvents) N.B. Sce footnotes to Table I . Formaldehyde. Mineral oils, in particular aqueous solutions, can release formaldehyde because of the addition of bactericides. Formaldehyde exposure has been assessed for workers exposed to mineral oils, with a higher probability for those using emulsified ones. Specific examples of exposure assessment are reported in Table 6. DISCUSSION The use of specific questionnaires permits different exposure judgements to be given to subjects with the same job title and to solve many uncertainties that arise when only the job title is available. Sometimesthe answer to a direct question is sufficient for a straightforward assessment: this occurred, for example. for mineral oils in the textile and metal industries. Exposure assessment from a direct question is possible. however, only for substances which are very common in a given activity; which are generally used with the awareness of the worker; which are easily recognized; which do not have ambiguous denomination (as has. for example, the word 'solvent', when it is not related to a specific operation). But also in these cases direct questions may be misleading. For example, if a machine-tool operator says he used mineral oils and his questionnaire is reliable (i.e. complete, without doubtful or contradictory information), it is quite easy to decide that he was exposed. Much more difficult is to decide that a machine tool operator was not ex- posed, if he denies the use of mineral oils. In fact, the reliable exclusion of the presence of mineral oils in the working environment requires a very detailed description of ii. Such pro . standardize of interviev does not de - naire, but ability." For thest available, e judgemen . based on ar sible the pr document at - A desc posure form o - A list c posure basis o the qut list can This kind experts on t be consider and also he1 of discrepar ACKNOWI This work Research Cc by the Eurc Evaluation Epidemiolo( I Ferrario F,C. F. Des, which I Hogstec Epideni k r r i n o F. FL Assessir lured '2 (eds). P ment I -Formaldehyde in the ler and installer in 'erations in a 4fied mineral oils nachine tools with n the work place itor: use of aqueous , and machine rticular aqueous ecause of the ad- rn assessed for h a higher prob- 5. assessment are nits different exubjects with the ncertainties that :. Sometimes the 7t for a straightor example, for 1 industries. Exdion is possible, re very common ly used with the isily recognized; iination (as has, it is not related 'lese cases direct example, if a nineral oils and mplete. without ), it is quite easy more difficult is tor was not exoils. In fact, the ineral oils in the jetailed descrip- ASSESSMENTOF EXPOSURE TO FORMALDEHYDE. SOLVENTS AND MINERAL OILS S133 Such problems may be at least partially solved with standardized questionnaires and with better strategies of interview. In any case the adequacy of information does not depend only on the good quality of questionnaire, but also on subject recall and on interviewer ability. Is For these reasons, even when very detailed SQ are available. experts often have to express a subjective judgemen . So it is very important that the studies based on an expert assessment make as explicit as possible the procedure used in exposure evaluation.' The documentation should include: - A description of the general criteria used in ex- posure assessment, which can also be done in the form of a job exposure matrix. - A list of significant examples showing how the ex- posure judgements have been modified on the basis of the particular information contained in the questionnaire. Tables 1 to 6 show how such a list can be compiled. This kind of procedure allows a comparison between experts on the minimum level of exposure that should be considered to define a subject definitely exposed and also helps to improve understanding of the sources of discrepancy between different studies. ACKNOWLEDGEMENTS This work was supported by the Italian National Research Council under contract No. 88.00524.44 and by the European Concerted Action on Retrospective Evaluation of Occupational Exposure in Cancer Epidemiology. REFERENCES Ferrario F. Continenza D. Pisani P. Magnani C, Merletti F, Berrino F. Description of a job-exposure matrix for sixteen agents which are or may be related to respiratory cancer. In: Hogstedt C. Reuterwall C (eds). Progress in Occupational ' Epidemiologv. Amsterdam: Elsevier. 1988. pp. 379-82. Berrino F, Ferrario F, Pisani P, Macaluso M and Baldasxroni A. Assessing occupational exposure.: from job histories to SITUCturd 'ad hoc' questionnaires. In: Hemon D. Goldberg M (eds). Proceedingsof fhe Workshop: Methodology o/Asenment of Occupational Exposures in the Context of EpidemioloRical Detection of Cancer Risk.% Pari\. IS- I9 February 1988. Luxembourg: Office of the Official Publica- ' tions of the European Communitie, 1989; pp. 172-87. Srewart P. Herrick R. Blair A et a/. Highlights of the 1990 Leesburg. Virginia. International Workshop on Retrospmive Exposure Assessment for Occupational Epidemlology Studies. Scad J Work Environ Health 1991; 17: 281-85. International Agency for Research on Cancer (IARC). Monopraphs on the Evaluation of the Carcinopenrc Risk of Chrmirals IO ' Humans. Vol. 1-42. Lvon: IARC. t972-1987. International Agency for Research on Cancer (IARC). .Woonographs on the Evaluation of fhe Carcrttopenrc Risk of Chemicals 10 Humans. Vol. 1-42 (suppl. 7). O\erall evaluation of carcinogenicity. an updating of IARC Monograph$. b o n : IARC. 1987. International Labor Office. Encvclopaediaof Occuparrunal Healfh and Safely Vols I and 2. Geneva: International Labor Office. ' -1983. Candura F. Elemenfidi TecnolopiaIndustriale a Us0 hi Culforrdi Medicina del Lavoro. Pavia: Aurora. 1974. Bona M. lsnardi F A. Straneo S L. Manuale di TecnofopiaTerrrle. Rome: ESAC. 1981. Cutler J. UnderstandinR Airrraft Structures. CK, Granada Publishers, 1981; pp. 84-1 19. lo Villavecchia V. Eingenmann G. Nuovo Di:ionario di Merceoloaia e ChimicaApplieata. Milan: Hocpli. 1973. I' Menichini E. L 'Lrposizione Professionale Apli 01; Lubrflcanfi: Criteri per I'lndividuazione il Controllo e la Prevenzione dei Rischi. Rapport0 ISTlSAN 86-21. Rome: Istituto Supcriore di Sanita. 1986. Menichini E. Risulfato di un 'Indagine Conoscitiva Na:ionale Sull'uso Degli oli Lubrifcanti neRli Ambienti di Lmtortt. Rapporto ISTISAN 86-36. Rome: lstituto Superiore di Sanita 1986. l 3 &rrino F, Continenza D. Ferrario F. Baldasscroni A, Borroni A, Bodini L. Esposirione professionale a cancerogeni respintori: I'industria siderurgica. Epidemiologia e prevenzione 1985; 29 3542. l4 Bond G G. Bodnn K M, Sobel W. Shellenbergn R J. Flora G H. Validation of work histories obtained from interviews. Am J Epidemiol 1988; 121): 343-51. Ahrens W, Jockel K H, Brochard P etal. RetrospectivePcsessrnenI of asbestos exposure-I. Case-control analysis in a study of lung cancer: efficiency of job-specific questionnaire and job exposure matrices. Int J Epidemiol 1993; U(Supp1. 2): s83-sgs. Pastorino U, Bcrrino F, Gervasio .A. Paenti V. Riboli E. Crosignani P. Proportion of lung cancers due to occupational exposure. Int J Cancer 1984; 3 3 31-37, " lnternational Labor Office. Internotional Standord Clas&&cafion '* of Occupations. Geneva: ILO. 1%8. Eaumgarten M. Siemiatycki J, Gibbs G W. Validity of work histories obtained by interview for epidemiologic purposes. Am J Epidemiol 1983; 111: 583-91. .