Ann. Occup. Hyg., Vol. 56, No. 1, pp. 70–79, 2012� The Author 2011. Published by Oxford University Press

on behalf of the British Occupational Hygiene Societydoi:10.1093/annhyg/mer081

Development of an Exposure Measurement Databaseon Five Lung Carcinogens (ExpoSYN) forQuantitative Retrospective Occupational ExposureAssessmentSUSAN PETERS1*, ROEL VERMEULEN1,2, ANN OLSSON3,RAINER VAN GELDER4, BENJAMIN KENDZIA5, RAYMOND VINCENT6,BARBARA SAVARY6, NICK WILLIAMS7, TORILL WOLDBÆK8,JEROME LAVOUE9, DOMENICO CAVALLO10, ANDREA CATTANEO11,DARIO MIRABELLI12, NILS PLATO13, DIRK DAHMANN14,JOELLE FEVOTTE15, BEATE PESCH5, THOMAS BRUNING5,KURT STRAIF3 and HANS KROMHOUT1

1Environmental Epidemiology Division, Institute for Risk Assessment Sciences, Utrecht University,Utrecht, the Netherlands; 2Julius Center for Health Sciences and Primary Care, University MedicalCenter, Utrecht, the Netherlands; 3International Agency for Research on Cancer, Lyon, France;4Institute for Occupational Safety and Health of the German Social Accident Insurance, SanktAugustin, Germany; 5Institute for Prevention and Occupational Medicine of the German SocialAccident Insurance, Bochum, Germany; 6Institut National de Recherche et de Securite, Vandoeuvre lesNancy, France; 7Health and Safety Executive, Bootle, UK; 8National Institute of Occupational Health,Oslo, Norway; 9Research Centre of University of Montreal Hospital Centre, University of Montreal,Canada; 10Department of Chemical and Environmental Sciences, Universita degli Studi dell’Insubria,Como, Italy; 11Department of Occupational and Environmental Health, Universita degli Studi diMilano and Unit of Epidemiology, Fondazione IRCCS Ca’ Granda-Ospedale Maggiore Policlinico,Milan, Italy; 12Cancer Epidemiology Unit, CPO-Piemonte and University of Turin, Italy; 13TheInstitute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden; 14IGF-BG RCI,Bochum, Germany; 15Departement sante travail, Institut de veille sanitaire, St Maurice, France

Received 27 April 2011; in final form 30 June 2011; published online 11 October 2011

Background: SYNERGY is a large pooled analysis of case–control studies on the joint effectsof occupational carcinogens and smoking in the development of lung cancer. A quantitativejob-exposure matrix (JEM) will be developed to assign exposures to five major lung carcino-gens [asbestos, chromium, nickel, polycyclic aromatic hydrocarbons (PAH), and respirablecrystalline silica (RCS)]. We assembled an exposure database, called ExpoSYN, to enablesuch a quantitative exposure assessment.

Methods: Existing exposure databases were identified and European and Canadian researchinstitutes were approached to identify pertinent exposure measurement data. Results of individ-ual air measurements were entered anonymized according to a standardized protocol.

Results: The ExpoSYN database currently includes 356 551 measurements from 19 coun-tries. In total, 140 666 personal and 215 885 stationary data points were available. Measure-ments were distributed over the five agents as follows: RCS (42%), asbestos (20%), chromium(16%), nickel (15%), and PAH (7%). The measurement data cover the time period from 1951to present. However, only a small portion of measurements (1.4%) were performed prior to

*Author to whom correspondence should be addressed. Tel: +0031 30 253 9681; fax: +0031 30 253 9499; e-mail: s.peters@uu.nl

70

1975. The major contributing countries for personal measurements were Germany (32%), UK(22%), France (14%), and Norway and Canada (both 11%).

Conclusions: ExpoSYN is a unique occupational exposure database with measurementsfrom 18 European countries and Canada covering a time period of >50 years. This databasewill be used to develop a country-, job-, and time period-specific quantitative JEM. This JEMwill enable data-driven quantitative exposure assessment in a multinational pooled analysis ofcommunity-based lung cancer case–control studies.

Keywords: exposure database; general population; multinational; occupational epidemiology; retrospectiveexposure assessment

INTRODUCTION

Recently, a large pooled analysis of case–control studieson the joint effects of occupational carcinogens andsmoking in thedevelopmentof lungcancerwas initiated(Olssonetal.,2011).Thisprojectiscalled‘SYNERGY’.Up until now, 14 lung cancer case–control studies from13 European countries (Czech Republic, France, Ger-many, Hungary, Italy, the Netherlands, Poland, Roma-nia, Russia, Slovakia, Spain, Sweden, and the UK) andCanadahavebeenincluded.Together,thesestudiescom-prise39 518subjects:17 705casesand21 813controls.Information on job history and lifestyle factors is avail-able for all subjects. Since the occurrence of joint expo-sure to two or more occupational carcinogens in thegeneralpopulationisconsideredtobelowduetothewiderange of possible workplaces, a large pooled analysis isnecessary to provide the essential power for the analysisof possible synergistic effects.

Assessment of occupational exposures in commu-nity-based studies is complicated by the wide varietyof jobs in which the general population will havebeen employed (Kromhout and Vermeulen, 2001a).Moreover, when studying cancer, the relevant expo-sures are likely to have occurred at least a decade ormore ago due to the long latency between exposureand disease onset. Retrospective assessment of occu-pational exposures is often qualitative (yes/no) orsemi-quantitative (low/medium/high) in community-based studies. Methods used to assign those estimatesare case-by-case expert judgement, job-exposure ma-trices (JEMs), self-reported exposure, or a combina-tion of these (McGuire et al., 1998). However, toexplore exposure–response relationships, quantitativeexposure assessment is necessary (e.g. in terms of air-borne concentrations in milligrams per cubic metre).Detailed exposure monitoring is an option in cross-sectional or (prospective) cohort studies, but the retro-spective and community-based nature of case–controlstudies renders this impossible. In principle, however,quantitative exposure estimation can be achieved bymodelling of available relevant historical exposuremeasurement data.

To facilitate quantitative exposure–response model-ling within the SYNERGY project, a data-driven JEMwill be constructed using historical occupational expo-sure data from a wide variety of industries, workplaces,countries, and time periods. Such occupational expo-sure data can be found in several types of databases.Firstly, there are national exposure databases, for in-stance, the NEDB in the UK (Burns and Beaumont,1989), COLCHIC in France (Vincent and Jeandel,2001), MEGA in Germany (Gabriel et al., 2010),and EXPO in Norway (Osvoll and Woldbaek, 1999).However, the purposes of these databases are different:MEGA and COLCHIC serve social security occupa-tional accident and disease prevention programmes,while NEDB and EXPO were set up from a researchperspective. Secondly, industry-specific multinationalexposure databases exist, for instance, PAPDEM forthe paper and pulp industry (Kauppinen et al., 1997),EXASRUB for the rubber manufacturing industry(De Vocht et al., 2005), and AWE for the asphalt in-dustry (Burstyn et al., 2000). Furthermore, there aremultinational exposure-specific databases, such asthe IMA–DMP on respirable (silica) dust (Houbaet al., 2009), the WOODEX project for wood dust ex-posures (Kauppinen et al., 2006), and DERMDAT fordermal exposures (Kromhout and Vermeulen, 2001b).In addition, many (unpublished) measurement datafrom research institutes and individual research groupsexist (Burstyn et al., 2000; De Vocht et al., 2005).

The aim of this paper was to describe the structureand contents of the ExpoSYN database. Furthermore,we discuss the feasibility of creating such a multina-tional exposure database and the methodological issuesthat arose during the process.

METHODS

Five established lung carcinogens were selectedbased on their relatively high expected exposure prev-alence in the general population: asbestos, chromium,nickel, polycyclic aromatic hydrocarbons (PAH), andrespirable crystalline silica (RCS). Table 1 lists the

Development of an exposure measurement database on five lung carcinogens 71

selected agents and their subcategories for which airexposure measurements were collected. Estimatesfor the number of exposed workers in the EuropeanUnion (1990–1993) and Canada (2006) are alsoshown, as provided by the CAREX (CARcinogenEXposure) database (Kauppinen et al., 2000; CAREXCanada, 2010).

Collection of individual exposure measurement datapoints was essential since these, rather than summarystatistics, in theory, will give us the opportunity to ad-just for differences in the way the historical exposuredata were obtained, such as differences in samplingand analytical methods, measurement strategies, andsampling duration. Differences in these elementsmay result in biased estimates, so besides collectingexposure measurement data, we endeavoured to obtainancillary data to enable adequate adjustments in statis-tical modelling (Rajan et al., 1997). A format for struc-tured data entry was developed and input variableswere described in a protocol.

Table 2 presents the variables included in the Expo-SYN database. These variables represent the a prioridefined variables deemed essential for exposure mod-elling (Rajan et al., 1997). One of the key variableswas the job title, which was coded according to the‘International Classification of Occupations (ISCO),1968’ (International Labour Office, 1968; Mannetjeand Kromhout, 2003). Industry was coded by the ‘In-ternational Standard Industrial Classification (ISICRev.2)’ (Statistical Office of the United Nations,1971; Mannetje and Kromhout, 2003). Many datahad to be coded into ISCO and ISIC from text fieldsor recoded from other coding systems. Measurementstrategy was divided into a statistical or representativeapproach (where sampling is performed in order to ob-tain representative exposure data for all workers withthat job title, the preferred strategy) and a worst-case

scenario (i.e. sampling under conditions when highestexposures were anticipated). Reasons for measure-ments included survey, inspection/complaints, or com-pliance. Categorization schemes for these and othervariables were derived in an ‘expert-based’ mannerusing available information since the exact same for-mulations were not utilized in all datasets. All datain the ExpoSYN database were entered without trace-able factory or personal identifiers; both were recodedinto numeric identification codes.

Exposure data collection was performed from Sep-tember 2007 until August 2010. We focused on thecountries included in the SYNERGY project, althoughwe have also collected data from neighbouring coun-tries when available in order not to lose valuable infor-mation. Existing exposure databases were identifiedand owners of these databases were approached forcollaboration. Both personal (samples collected witha personal portable pump in the breathing zone of theworker) and stationary (samples from a fixed locationsampler) measurements were entered into the data-base. Furthermore, research institutes in Europe andCanada were approached in order to find additional oc-cupational exposure data. Due to cost and time consid-erations, our search for additional exposure data wasmore focused on personal rather than stationary meas-urements since these can be directly linked to jobs.Moreover, personal measurements more accuratelycapture a workers’ exposure, which is modified bya worker’s position in the work environment and work-ers’ behaviour that may modify exposure levels(Cherrie, 2003). No biological monitoring data wereincluded. Data entry according to the protocol was per-formed mainly (�85%) by the institutes where datawere originally stored; the rest was provided as rawdata and subsequently put in the correct format byone of the authors (S.P.). The data provided by

Table 1. Selected agents for exposure assessment in the SYNERGY project.

Selected lungcarcinogen

Specification CAREX—estimated number of exposed workers

European Uniona,139 million workers (%)

Canadab, 17 millionworkers (%)

Asbestos Chrysotile (white asbestos),amphibole, and other type/unspecified

1 200 000 (0.9) 152 000 (0.9)

Chromium Chromium VI and total chromium 800 000c (0.6) 83 000c (0.5)

Nickel Soluble nickel compounds, insolublenickel compounds, and total nickel

560 000 (0.4) 50 000 (0.3)

PAH Benzo(a)pyrene and naphthalene 980 000d (0.7) 307 000d (1.8)

RCS Respirable quartz, respirable cristobalite,respirable tridymite, and RCS

3 200 000 (2.3) 349 000 (2.1)

a1990–1993 (Kauppinen et al., 2000).b2006 (CAREX Canada, 2010).cChromium VI only.dDisregarding tobacco smoke (passive exposure at work) and diesel motor exhaust.

72 S. Peters et al.

contributors were inspected centrally before entry intothe ExpoSYN database in order to correct mistakes andto improve consistency in coding. No data were ex-cluded a priori as specific exclusion criteria for statis-tical modelling of the data differ by agent.

RESULTS

The ExpoSYN database is a non-hierarchical data-base comprising five sections: one for each selectedagent within the SYNERGY study. To date, the data-base contains a total of 356 551 measurements from19 countries. The majority (90%) of the data originatefrom national exposure databases [the main sourceswere MEGA (Germany), COLCHIC (France), NEDB(UK), and EXPO (Norway)], 1% originates from indus-try-specific databases (AWE and PAPDEM), and the re-maining 8% were collected at institutes. The number ofmeasurements differed considerably between the five

agents: RCS (n 5 148 911; 42%), asbestos (71 816;20%), chromium (57 119; 16%), nickel (52 751;15%), and PAH (25 954; 7%). The major contributingcountries to the complete database were Germany(60% of all measurements), the UK and France (both10%), Canada (7%), Norway (6%), and Italy (4%).These and other characteristics of the exposure dataare presented in Table 3. Information on measurementreason and measurement strategy was missing for 44and 15% of the data points, respectively, and differedconsiderably between agents. Missing information formeasurement reason ranged from 18% for PAH to58% for asbestos, while percentages of missing valuesfor measurement strategy ranged from 8% (RCS) to38% (asbestos). For data for which this informationwas available, surveys and inspection were the mainreasons (28 and 21%, respectively). Measurementswere reported to be mainly performed in a representa-tive way (77%).

Table 2. Variables included in the ExpoSYN database.

Type of information Variables Input

General Country Text

Centre Text

Date DD-MM-YYYY

Factory ID Text

Worker ID Text

Sample ID Text

Job characteristics Industry code ISIC rev.2

Job code ISCO 1968

Respiratory protectiveequipment (RPE)

No and yes (type of RPE optional)

Local exhaust ventilation No and yes

Indoors or outdoors Indoors, outdoors, and partially

Measurement information Personal or stationary sample Personal and stationary

Measured concentration Number

Unit of measurement Text (e.g. mg m�3)

Sampling duration Number (min)

Measured fraction Inhalable, thoracic, and respirable (for PAH:particulate and gaseous phase)

Sampler type E.g. seven-hole

Filter type E.g. glass fibre

Digestion methoda E.g. acid bomb

Analytical method E.g. X-ray diffraction

Measurement strategy Statistical approach/representative and worst case

Measurement reason Survey, inspection/complaints, and compliance

Below or above limit ofdetection (LOD)

Below LOD, above, or equal to LOD

Limit of detection formeasurement

Number

Unit of measurement LOD Text (e.g. mg m�3)

aFor chromium and nickel only.

Development of an exposure measurement database on five lung carcinogens 73

Table 3. Characteristics of the ExpoSYN database by agent.

Asbestos(n 5 71 816)

Chromium(n 5 57 119)

Nickel(n 5 52 751)

PAH(n 5 25 954)

RCS(n 5 148 911)

All agents(n 5 356 551)

Agent type (%) Chrysotile (67) Cr VI (43) Soluble Ni (1) B(a)P (65) Quartz (93)

Amphiboles (15) Total Cr (57) Insoluble Ni (,1) Naphtalene (35) Cristobalite (2)

Unspecified (18) Total Ni (99) Tridymite (,1)

Unspecified (4)

Time period (years) 1968–2009 1965–2009 1967–2009 1975–2009 1951–2009 1951–2009

Number ofmeasurements (%)

Personal 36 001 30 813 28 260 11 575 34 017 140 666 (39)

Stationary 35 815 26 306 24 491 14 379 114 894 215 885 (61)

Reason formeasurements (%)

Survey 12 262 (17) 14 446 (25) 11 535 (22) 10 495 (40) 17 743 (12) 66 481 (21)

Inspection 11 325 (16) 18 016 (32) 17 707 (34) 7725 (30) 49 419 (33) 104 192 (28)

Compliance 6750 (9) 6831 (12) 5363 (10) 2204 (8) 7710 (5) 28 858 (8)

Unknown 41 479 (58) 17 826 (31) 18 146 (34) 5530 (21) 74 039 (50) 157 020 (44)

Measurementstrategy (%)

Representative 37 893 (53) 42 815 (75) 39 960 (74) 19 392 (75) 130 408 (88) 270 468 (76)

Worst case 6665 (9) 9002 (16) 7067 (13) 5443 (10) 5080 (3) 33 257 (9)

Unknown 27 258 (38) 5302 (9) 5724 (13) 1119 (15) 13 423 (8) 52 826 (15)

Countries (%) Personalmeasurements (%)

Bulgaria — — — 36 — 36 (,1) 28 (,1)

Canada 6474 4298 5007 534 8449 24 762 (7) 14 765 (11)

Denmark 54 1108 662 206 318 2348 (,1) 2229 (2)

Finland 112 55 24 261 6 458 (,1) 306 (,1)

France 4758 9765 6583 3050 9921 34 077 (10) 19 075 (14)

Germany 29 237 28 770 28 759 13 377 115 009 215 152 (60) 44 534 (32)

Greece — — 51 — — 51 (,1) 36 (,1)

Iceland — 16 16 44 — 76 (,1) 76 (,1)

Italy 6468 1330 933 1318 2793 12 842 (4) 4920 (4)

The Netherlands 98 520 770 282 687 2357 (,1) 1797 (1)

74

S.

Peters

etal.

The exposure data cover the period from 1951 until2009, but only a small portion of data points (1.4%)were available before 1975 (see Fig. 1). The oldestdata points were of asbestos (both stationary and per-sonal) and RCS (stationary) measurements. Figure 2gives an impression of the historical coverage of theexposure data (separated by personal and stationarymeasurements) relative to the working history calen-dar years in the SYNERGY study population. Mea-surement data cover the most recent 40–50 years ofthe almost 90 working years (1922–2010) of the casesand controls.

In total, 140 666 personal data points were avail-able, mainly from Germany (32%), the UK (22%),France (14%), Norway, and Canada (both 11%), asshown in Table 3. For 90% of these data points, an IS-CO 1968 code was known. Figure 3 illustrates thecomparison between the major job groups representedin the SYNERGY study population, and the jobs forwhich measurements are available in the ExpoSYNdatabase. As expected, the majority of measurements(97%) were performed in occupations in productionor transport industries, covering nearly 600 ISCO co-des. Fifty per cent of the total working years of theSYNERGY study population involved work in these‘blue collar’ occupations.

DISCUSSION

The ExpoSYN database contains 356 551 occupa-tional exposure data points from all over Europe andCanada. Information regarding exposure measurementsfor asbestos, chromium, nickel, PAH, and RCS wascoded in a standardized format.

The database was set up to provide measurementdata for quantitative exposure assessment in a pooledmultinational occupational lung cancer case–controlstudy. Exposure assessment in such a design is com-plicated by the fact that, given the long latency timeof lung cancer, the relevant exposures will have oc-curred several decades ago. To perform retrospectiveexposure assessment in a quantitative way, historicalexposure measurements are essential. Accurately con-verting historical exposure data into currently ac-cepted exposure metrics can be a very demandingtask. Ideally, one would use exposure measurementsfrom the population under study (Dahmann et al.,2008). However, as can be seen in Fig. 2, it will takeseveral decades before exposure measurements willcover complete work histories of all subjects in retro-spective case–control studies, provided that exposuremeasurements are still being conducted.

Personal exposure monitoring became widely im-plemented in the 1970s (Cherrie, 2003). Therefore,T

able

3.

Continued

Asb

esto

s(n

57

18

16

)C

hro

miu

m(n

55

71

19

)N

ick

el(n

55

27

51

)PA

H(n

52

59

54

)R

CS

(n5

14

89

11

)A

llag

ents

(n5

35

65

51

)

No

rway

12

32

73

94

72

25

29

85

12

92

20

12

8(6

)1

59

62

(11

)

Po

lan

d4

40

45

19

54

33

12

37

4(,

1)

16

71

(1)

Ro

man

ia4

85

12

9—

24

55

08

13

67

(,1

)4

28

(,1

)

Ru

ssia

21

2—

——

—2

12

(,1

)2

03

(,1

)

Slo

vak

ia5

79

13

62

37

35

45

80

63

66

5(1

)1

65

5(1

)

Sp

ain

—1

——

—1

(,1

)0

(0)

Sw

eden

16

02

17

11

13

24

89

71

70

9(,

1)

16

41

(1)

Sw

itze

rlan

d—

188

120

9—

317

(,1

)1

31

(,1

)

UK

21

94

31

92

62

07

27

84

78

94

34

61

9(1

0)

31

13

9(2

2)

Development of an exposure measurement database on five lung carcinogens 75

exposure data prior to the 1970s are scarce; an issuethat others also faced when collecting historical ex-posure data for retrospective assessment (Kauppinenet al., 1997; Burstyn et al., 2000; Vermeulen et al.,2010). Most existing exposure databases started inthe 1980s; earlier measurement data had generallynot yet been digitized. Some exceptions are theMEGA database, for which stationary measurementswere already collected in the 1970s, and the personalasbestos exposure measurements from the BritishHealth and Safety Executive. About 30 000 data

points (8%) collected for the ExpoSYN databasewere not yet included in national or industry-specificexposure databases. Efforts were made to retrieverelevant data from hard copy records and to contactretired experts to help identify and describe old data.Although 8% is only a small portion of the completeExpoSYN database, these measurements might behighly valuable since these originated from countriesthat were otherwise not (or sparsely) covered, suchas Italy, the Netherlands, Poland, Romania, Russia,Slovakia, and Sweden and moreover comprised

Fig. 1. Distribution of exposure measurements in the ExpoSYN database over time.

Fig. 2. Distribution of exposure measurements in the ExpoSYN database over time (separated by stationary and personalsampling) compared to the working years represented in the SYNERGY population.

76 S. Peters et al.

predominantly less recent data. Searching for unpub-lished data has been recommended and previouslyproved to be a highly rewarding effort (Kauppinenet al., 1997; Burstyn et al., 2000; De Vocht et al.,2005; Vermeulen et al., 2010). Many of the researchersinvolved in obtaining the non-digitized exposure dataare, or will soon be, retired. This will likely lead toa loss of valuable information. Therefore, it is becom-ing increasingly pressing to recover useful exposuredata hidden away in archives. Industrial hygienistswho collected data and know the details of these data-sets might still be available to collaborate and providethe pertinent detailed ancillary information. Conse-quent digitization of exposure measurements will en-able and improve the use of measurement data forexposure assessment in future occupational epidemio-logical studies.

Overall, the ExpoSYN database has good coverageof the past 35–40 years, which provides opportunitiesfor estimating exposure-specific time trends. Coverageof each job by country is limited. Although exposurecircumstances may vary between- and within-jobs, in-dustries, countries, or regions, information can poten-tially be extrapolated from similar jobs withina country or between countries. Countries with fewmeasurements may especially gain from the exposureinformation provided by data from the same jobs andindustries elsewhere.

It can be questioned whether measurements inthe ExpoSYN database are really representative ofthe general working situation and if they reflect trueoccupational exposure levels considering selectionbias. Exposure measurements tend to be objective,

but identification and selection of workplaces andworkers to be measured involve substantial subjectiv-ity. Although representative sampling should be per-formed completely at random, measurements aregenerally only taken when and where exposures areexpected. In the most extreme case, a (‘white collar’)job—which normally does not involve exposure—might be monitored in a very unusual situation whenexposure occurs. That same job would not be moni-tored in the normal situation when exposure is notpresent. Collecting information about measurementreason and strategy enabled us to try to identify suchsituations. Potentially biases should be consideredwhen using data from any database. Kauppinenet al. (1997) estimated that 94% of the measurementdata in the PAPDEM database were not significantlybiased, as judged by exposure experts, whereas 5%of the samples overestimated exposure and 1% under-estimated it. For the situations where the measure-ment strategy was known in our database, 89% ofdata points were classified as being representative.

Setting up a prospective exposure database allowsfor a detailed protocol for data collection and registra-tion. Unfortunately, this is not possible when collectingmeasurement data retrospectively. In the latter case,one has to work with information often collected fortotally different goals or for different purposes. Forthe ExpoSYN database, it became apparent that itwould not be feasible to obtain all ancillary data re-quired by the database protocol for all measurements.Such missing data will complicate interpretation andcomparison of the measurement data. For example, itwas often unknown if measurements were performed

Fig. 3. Distribution of personal exposure measurement data in the ExpoSYN database over major groups of jobs (1 digit ISCO1968) compared to number of jobs represented in the SYNERGY population.

Development of an exposure measurement database on five lung carcinogens 77

in the same factory or for the same worker. Between-and within-subject or factory variability in exposureconcentrations will therefore be hard to estimate.Analyses on subsets of the database might give anidea about the variance components, under the as-sumption that these data are missing at random. Lackof information on measurement strategies or measure-ment reasons makes it impossible to ascertain the rep-resentativeness of all measurements, which could leadto biased results. However, even if this information isavailable, it is hard to standardize the concept of rep-resentativeness and it is questionable whether peopletaking measurements were always able to assess thisconsistently. Furthermore, data originated from verydifferent sources. Diversity in purpose of data collec-tion for the original databases or in sampling methodsor clustering of data in a few workplaces; all have thepotential to bias exposure measurement results (Olsenet al., 1991; Teschke et al., 2002; Agostini et al.,2010). Definitions and standardization of these as-pects of measurement strategies are clearly needed.

This is the first attempt to set up a large interna-tional exposure database covering all types of indus-tries for community-based epidemiological studies.Data collection took a large amount of time andwork. Before being entered into the ExpoSYN data-base, all data had to be transformed into a standard-ized format and therefore, (re)coding of occupationsand industries was often necessary. We aimed tocover a wide range of industries and countries, whichcomplicated coding due to the large amount of pos-sible jobs. Additionally, because of the internationalcharacter of the database, we had to deal with differ-ences in languages and coding systems. Beside somewell-known national exposure databases, not all datawere easy to trace and part of these data were notdigitized. In the future, the ExpoSYN database canbe expanded by adding other agents, although oneshould realize that the five agents selected for theSYNERGY project were relatively prevalent in oc-cupational settings (affecting 0.3–2.3% of the work-ing population, as presented in Table 1) and wereoften measured. It would be more difficult to findsufficient measurement data to study exposure to lessprevalent agents.

CONCLUSIONS

The ExpoSYN database is a unique database,which includes exposure measurement data from18 European countries and Canada. The large quan-tity of measurement results will allow for elaborationof a country-, job-, and time period-specific quantita-tive JEM for exposure assessment in a multinational

pooled analysis of community-based lung cancercase–control studies from various European countriesand Canada. Moreover, these data might be useful forfuture methodological studies in exposure assessment.

FUNDING

The SYNERGY project is funded by the GermanSocial Accident Insurance (DGUV) and is coordi-nated by the International Agency for Research onCancer, the Institute for Prevention and Occupa-tional Medicine of the DGUV, Institute of theRuhr-University Bochum (IPA) and the Institutefor Risk Assessment Sciences at Utrecht University.

Acknowledgements—The authors would like to thank the fol-lowing industrial hygienists, epidemiologists, and other ex-perts from European and Canadian research institutes fortheir contributions to this database and paper: Cheryl Peters,Martie van Tongeren, Thomas Schneider, Timo Kauppinen,Paolo Boffetta, Rodica Stanescu, Zuzana Vassanyi, Ewa Or-lowski, Slawek Brzeznicki, Jan Gromiec, Stefano Taronna,Gunilla Klingspetz, Nicole Charriere, Martin Tischer, PaulSwuste, Ton Spee, Evert Meijer, Mieke Lumens, and VirissaLenters.

REFERENCES

Agostini M, de Vocht F, van Tongeren M et al. (2010) Expo-sure to rubber process dust and fume since 1970s in theUnited Kingdom; influence of origin of measurement data.J Environ Monit; 12: 1170–8.

Burns DK, Beaumont PL. (1989) The HSE National ExposureDatabase-(NEDB). Ann Occup Hyg; 33: 1–14.

Burstyn I, Kromhout H, Cruise PJ et al. (2000) Designing aninternational industrial hygiene database of exposuresamong workers in the asphalt industry. Ann Occup Hyg;44: 57–66.

CAREX Canada. (2010) Report No: 2009/2010 Progressreport.

Cherrie JW. (2003) The beginning of the science underpinningoccupational hygiene. Ann Occup Hyg; 47: 179–85.

Dahmann D, Taeger D, Kappler M et al. (2008) Assessment ofexposure in epidemiological studies: the example of silicadust. J Expo Sci Environ Epidemiol; 18: 452–61.

De Vocht F, Straif K, Szeszenia-Dabrowska N et al. (2005)A database of exposures in the rubber manufacturingindustry: design and quality control. Ann Occup Hyg; 49:691–701.

Gabriel S, Koppisch D, Range D. (2010) The MGU—a moni-toring system for the collection and documentation of validworkplace exposure data. Gefahrst Reinhalt L; 70: 43–79.

Houba R, Vlaanderen J, Jongen R et al. (2009) Building an in-dustry-wide occupational exposure database for respirablemineral dust—experiences from the IMA dust monitoringprogramme. J Phys: Conf Ser; 151.

International Labour Office. (1968) International standardclassification of occupations. Geneva, Switzerland: Interna-tional Labour Office Publications.

Kauppinen T, Teschke K, Savela A et al. (1997) Internationaldata base of exposure measurements in the pulp, paper and

78 S. Peters et al.

paper product industries. Int Arch Occup Environ Health;70: 119–27.

Kauppinen T, Toikkanen J, Pedersen D et al. (2000) Occupa-tional exposure to carcinogens in the European Union.Occup Environ Med; 57: 10–8.

Kauppinen T, Vincent R, Liukkonen T et al. (2006) Occupa-tional exposure to inhalable wood dust in the member statesof the European Union. Ann Occup Hyg; 50: 549–61.

Kromhout H, Vermeulen R. (2001a) Application of job-exposurematrices in studies of the general population: some clues totheir performance. Eur Respir Rev; 11: 80–90.

Kromhout H, Vermeulen R. (2001b) Temporal, personal andspatial variability in dermal exposure. Ann Occup Hyg;45: 257–73.

Mannetje A, Kromhout H. (2003) The use of occupation andindustry classifications in general population studies. Int JEpidemiol; 32: 419–28.

McGuire V, Nelson LM, Koepsell TD et al. (1998) Assessmentof occupational exposures in community-based case-controlstudies. Annu Rev Public Health; 19: 35–53.

Olsen E, Laursen B, Vinzents PS. (1991) Bias and random errorsin historical data of exposure to organic solvents. Am Ind HygAssoc J; 52: 204–11.

Olsson A, Gustavsson P, Kromhout H et al. (2011) Exposure todiesel motor exhaust and lung cancer risk in a pooled analysis

from case-control studies in Europe and Canada. Am J Respir

Crit Care Med; 183: 941–8.Osvoll PO, Woldbaek T. (1999) Distribution and skewness of oc-

cupational exposure sets of measurements in the Norwegian

industry. Ann Occup Hyg; 43: 421–8.Rajan B, Alesbury R, Carton B et al. (1997) European proposal

for core information for the storage and exchange of workplace

exposure measurements on chemical agents. Appl Occup

Environ Hyg; 12: 31–9.Statistical office of the United Nations. (1971) Indexes to the

International standard industrial classification of all eco-

nomic activities. New York, NY: Publishing Service United

Nations.Teschke K, Olshan AF, Daniels JL et al. (2002) Occupational

exposure assessment in case-control studies: opportunities

for improvement. Occup Environ Med; 59: 575–93.Vermeulen R, Coble JB, Lubin JH et al. (2010) The Diesel Ex-

haust in Miners Study: IV. Estimating historical exposures to

diesel exhaust in underground non-metal mining facilities.

Ann Occup Hyg; 54: 774–88.Vincent R, Jeandel B. (2001) COLCHIC-occupational exposure

to chemical agents database: current content and development

perspectives. Appl Occup Environ Hyg; 16: 115–21.

Development of an exposure measurement database on five lung carcinogens 79