The European Journal of Public Health Advance Access originally published online on June 7, 2005
The European Journal of Public Health 2005 15(3):305-312; doi:10.1093/eurpub/cki083
Miscellaneous |
A cohort study to estimate occupational mortality risks in Navarra
M.Dolores Ugarte1, Liliana Artieda2, Berta Ibáñez1, Ana F. Militino1, Mercedes Lezáun3, Marina López-Sagaseta4 and Conchi Moreno-Iribas5
1 Statistics and Operations Research Department. Public University of Navarra, Pamplona, Spain
2 Occupational Health and Research Service. Navarra Institute of Occupational Health, Pamplona, Spain
3 Research and Occupational Epidemiology Section. Navarra Institute of Occupational Health, Pamplona, Spain
4 Statistical Institute of Navarra, Pamplona, Spain
5 Non-transmissible Diseases and Vital Statistics Section, Institute of Public Health of Navarra, Pamplona, Spain
Correspondence: M. Dolores Ugarte, Statistics and Operations Research Department, Public University of Navarra, Campus de Arrosadía, 31006 Pamplona, Navarra, Spain, tel. +34 948169202, fax, +34 948169204, Email: lola{at}unavarra.es
Received June 24, 2004, accepted January 13, 2005
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Background: Few studies on occupational mortality have been conducted in Spain. The objective of this work was to analyse inequalities on global mortality and on mortality due to specific causes according to occupation in a historical cohort of males from the province of Navarra, Spain. Methods: The base population for this historical cohort comprised all employed men over age 34 from Navarra in the 1986 population register. Age-standardised point estimates and confidence intervals for occupational-specific mortality risks were computed. Results: There exist differences in mortality risks with respect to the overall risk of Navarra in certain occupational activities for several major causes of mortality. Some of the results corroborate previous findings in other works, such as the significant high risk that presents in leather, clothing workers and shoemakers when analysing kidney, bladder and other urinary malignant tumours, while others present a certain degree of novelty. Conclusion: This work contributes to filling the gap in the lack of works on occupational mortality in Spain. It also complements the information that other monitoring systems may provide on occupational health.
Keywords: historical cohort, occupational mortality, risk confidence intervals
Several studies on the monitoring of occupational safety and health have been carried out in the European Union1 in the last decade. The lack of information about the risks and damage that workers are exposed to was considered an obstacle to the defining of effective occupational health policies and plans. In order to know the health state of workers and the risks to which they are subjected, it is important to know the sources of data available that may be used. A recent study carried out by the European Agency for Safety and Health at Work summarises the type of systems used in the European Union for this purpose.2 The most widely used are worker surveys, registers of accidents, diseases and/or ill health and multi-source and policy directed systems. In Spain, the main sources for evaluating working conditions, which may be used as tools for prevention, are the National Working Conditions Survey, and the Occupational accidents and diseases statistics. All of these sources provide valuable information, but have several drawbacks1,2 that make impossible to use in isolation. To complement this information, other data sources such as the mortality registers can be used to hypothesise about which occupations have high risk of mortality for specific diseases, helping, therefore, to detect risk factors in the workplace that may affect the health of workers.3–6 When carrying out this type of register-based study, the interest relies on describing the inequalities on mortality according to occupation, because knowledge of the role of exposure to occupational factors in inducing morbid events is crucial to the prevention of disease. Unfortunately, there are several difficulties in carrying out these studies in some countries, such as Spain.1 Some works about inequalities in mortality according to the socioeconomic level in Spain have been recently presented,7,8 but there are still no studies that analyse occupational mortality. The main drawbacks are the availability of death certificates and the quality of the information on occupation in statistics of mortality.9 Hence, as occupation is frequently absent or incomplete in the mortality register, it is compulsory to link data sets,10 which makes the studies more complicated. Due to the information that these registers may provide, several efforts need to be undertaken to create data sources that allow for reliable results. In fact, the analysis of occupational mortality can be used to verify in time and space the presence of an excess of risk associated with a given occupation and also to formulate hypotheses about the risk factors, which should be confirmed by analytical studies. Hence, they can be a good starting point for posterior and more specific analysis. In this direction, we have conducted a study using a historical cohort of males from the Autonomous Community of Navarra in Spain to make evident the possibility and the need to carry out these studies, even in countries where the quality of the registers is not as good as in other countries such as Sweden or the UK. Therefore, the aim that motivated this work was two-fold. On one hand, data sets of reasonable quality needed to be constructed by using information based on both the mortality register and the 1986 population register (1986 census). On the other hand, adequate statistical techniques needed to be adapted to this context to obtain reliable results. A methodological approach to detect statistical differences among the occupation-specific risks compared to a reference population is the construction of appropriate confidence intervals for these specific relative risks. In this paper, we provide estimates and confidence intervals for the relative risks in Navarra in certain occupational activities for several major causes of mortality, and conclude that some of the occupational activities considered here present significant high risks for particular mortality causes under study.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The base population for this historical cohort comprised all men from Navarra who were over 34 years old and employed at the time of the 1986 population register. This encompassed 78 994 men followed until year-end 2001, rendering a total of 1 213 385 person-years.
Information was drawn from two data sets that were linked to constitute the basis of this register-based information system. The first data source comprised all men over 34 in the 1986 population register from where we extracted information on occupation and age. We used the 1986 population register as the start of the study because it is the oldest population register that has information about each individual with its respective personal identifier, allowing the linkage with mortality data. Further, we assumed that for an occupation to have some kind of influence on a mortality cause, it must have been held during a relatively long period. Hence, we did not include young men who were less than 34 in the 1986 population register because they presented less job stability. The occupations were originally coded using the first two digits of the National Classification of Occupations done by the Spanish Statistical Institute and approved by the Royal Decree 2240/1979. This codification consists of all specific professions, which can be grouped into nine groups of professionals. The assumption of homogeneity in the occupational risk factors among workers from the same group was considered to hold for eight out of the nine groups (see categories 1–8 in table 1). However, the code that groups all the industrial workers presents great diversity in the occupational risk factors, so we split that further into six categories for which the assumption of homogeneity seems to be hold (categories 9–14 in table 1). Regarding the quality of the occupational categories in the population register, out of the total of workers who were over 34 in 1986 (79 361), only 0.5% (367) were not classified and were, therefore, not included in the study, leading to 78 994 individuals followed up in the cohort. Table 1 details the 14 occupational categories considered, their corresponding codes according to the National Classification of Occupations and the age structure of the cohort.
|
The second data source was the Navarra mortality register comprising all male mortality cases over 34 at the time of the 1986 population register from 1986 to 2001 classified according to the ICD 10, which sum to a total of 37 774 mortality cases. The mortality causes were regrouped obtaining 20 major causes of mortality (see table 2): overall mortality (C1), overall cancer mortality (C2), specific-tumours mortality (C3–C9), and other specific main diseases and external causes (C10–C20).
|
Due to the difficulty in obtaining the occupation from the mortality register, we had to link this data set with the 1986 population register. In the merging operation, 27 290 out of the 37 774 records could be traced in the population register. Some of the non-linked records were people who immigrated after 1986. As the estimated ratio of immigration during the period was 4.6%, and assuming that mortality affects immigrated and non-immigrated people equally, then the hit completeness between the mortality register and the 1986 population register was 75.7 %, which may be regarded as satisfactory. This percentage was obtained11 multiplying by one hundred the proportion of linkage's success, i.e, the number of linked records divided by the difference between the total number of records and the number of immigrants at the mortality register.
The linkage process was carried out in two steps. In the first one, we used the DNI (identity national document) as the common field, which is unique for each individual and is composed of eight digits. However, the DNI was not available in the mortality register until 1998, and for the majority, but not for all, cases, so this criterion provided only 26.3% of the successful linkages. Therefore, it was necessary to create a personal identity code by using other fields such as ‘name’, ‘surname’, ‘date of birth’ and ‘sex’ in the second step. From this merged data set with 27 290 records, we selected those men who were employed in the 1986 population register, being a total of 7609 mortality cases. In figure 1 we provide a description of the mortality cases in the linkage process.
|
The overall person-time that each worker contributed to the study was allocated to the corresponding cells of the variables of stratification: occupation (14 categories) and age groups (35–44, 45–54, 55–64 and >64 years, respectively). Due to the lack of information about other possibilities of leaving the study during the period studied (e.g. emigration), we considered that the only cause for leaving the study before 2001 was death. Nevertheless, the annual emigration rate among Navarra's citizens during this period was very low, approximately 1.8/1000, so that the bias introduced by this fact may be neglected.
The expected number of cases in any given occupation and mortality cause was computed taking age-specific rates of this study cohort as the reference. We computed Standardized Mortality Ratios (SMRs) for all mortality causes and occupation categories, and further, we provided confidence intervals for the relative risks using the Byar approximation12 for those categories where the number of mortality cases is greater than 20, and using the exact probabilities13 when the number of mortality cases is less than or equal to 20.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
During follow-up a total of 7609 mortality cases were reported in the study cohort. Table 3 presents the SMRs for each occupational category for the following causes: overall mortality, overall cancer mortality and some specific causes of cancer mortality. Table 4 presents the results for the rest of the mortality causes studied here. We found differences in mortality risks with respect to the overall risk of Navarra in certain occupational activities for several major causes of mortality. For some of the mortality causes, the variability of the SMR is high because of the low number of cases for certain occupational categories. This is due to the small size of the Navarra population. In these categories, the confidence intervals are too wide, so that none of the risks are found to be significantly higher or lower than 1. However, we provide SMRs, confidence intervals for the relative risks, and the number of observed cases, to help to interpret the results. In the following we comment on the results that we consider more interesting relative to occupationally related mortality.
|
|
With regard to occupational categories that have high risks for some of the causes under study, we would like to highlight that some of the results found in this study seem to corroborate previous findings in other works. For example, category C12 ‘textile, leather, clothing workers and shoemakers’ presents the highest risk when analysing urinary malignant tumours (C5). These results match with those found in other Spanish studies.14 Further, in heavily industrialised countries, the attributable fraction to occupation may be as high as 25% in men15 in these specific malignant tumours.
It is also important to remark the high risk estimates obtained for category C5 ‘sales workers’ when studying leukaemia (C7), ischemic heart disease (C15) and all other diseases of the circulatory system (C16). Some studies indicate that job strain and work stress are related to elevated levels of established ischemic risk factors.16 However, more detailed work is needed to relate ‘sales workers’ to job stressors.
On the other hand, when studying respiratory diseases including asthma (C17), we found that categories C11 ‘painters, chemistry workers, rubber and plastic workers’ and C5 ‘catering workers, warehouse men, security guards and store keepers’ attain the highest estimated relative risks. In this direction, it has been shown in previous studies17 that exposure to occupational dusts, gases, vapours and fumes are associated with an increase in symptoms and pulmonary function deficit.
We also consider highly significant the results obtained when studying external causes (C20). Categories C6 ‘farmers, cattle farmers and foresters’ and C13 ‘construction workers’ attain high estimated risks, in contrast to C1 ‘professional and technical workers’, C3 ‘office workers and accountants’, C4 ‘sales workers’, C7 ‘military professionals’ and C8 ‘foremen’, for whom the estimated relative risks are very low. Recent studies18 point out that the mortality risk in Navarra is lower than in the whole of Spain in overall mortality and in the main mortality causes, except in external causes, where the risk in Navarra is higher. We consider that this excess of mortality in external causes could be partially explained by the high estimated mortality risk obtained in certain occupational categories such as ‘construction workers’ and ‘farmers, cattle farmers and foresters’. However, more studies are needed to develop this hypothesis. In this direction, it has been shown that the highest incident rate in occupational accidents in Navarra was found for the construction sector, with a rate of 195.7 per 1000 workers, and that the highest rates in severe occupational accidents were found in the agriculture sector, with a rate of 243.9, followed by the construction sector, with a rate of 194.8.19
As far as the significant low risks obtained for some categories are concerned, it seems clear that the so-called ‘white-collar workers’ have a ‘protective effect’ for some of the mortality causes studied. For example, one may note the significant low risks obtained for category C1 ‘professionals and technical workers’ when studying overall mortality, overall cancer mortality and respiratory tumours, and for category C3 ‘office workers and accountants’, when studying external causes.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
In the occupational mortality framework, the main contribution of this study has been to analyse inequalities in mortality by occupation in the province of Navarra. Hence, it helps to fill the gap in the lack of studies on occupational mortality in Spain. On one hand, we obtained a reliable historical cohort of all men from Navarra who were over 34 years old and employed at the time of the 1986 population register. This register has not only information about the occupational activity and the cause of mortality of each person but also about the study level, the relation to the professional activity, the work situation, the location of the industry or company, and other characteristics of interest, so the cohort can be used for other purposes. On the other hand, we estimated the occupation-specific relative risk for the mortality causes analysed here. We concluded that there exist differences in mortality risks with respect to the overall risk of Navarra in certain occupational activities for several major causes of mortality.
In the more general context of the study of monitoring systems in occupational health and safety, this work may bring to light the need for conducting this kind of study in small areas, focused on the effects that certain work exposures may have on mortality risk. In Spain, the development of information systems on health and population started very recently compared to other European countries, so it is important to promote this type of register-based study to complement the information that other monitoring systems may provide in the context of occupational health. The register-based studies present limitations if they are used as the only material to make decisions about the intervention policies. However, when considering these register-based studies together with other data available such as working condition surveys and occupational diseases and accidents statistics, we can provide more complete information about the state of occupational health.
The main difficulty in carrying out these register-based studies is the lack of information about occupation in most mortality registers, so the linkage of registers is compulsory. In this direction, we should encourage health administrative system professionals and occupational administrative system professionals to ensure continuous collaborations and to maximise the potential for the linkage process, because we consider the effort worthwhile.
Finally, we would like to highlight several uses that these studies based on the mortality registers may have on the knowledge of the occupational health situation and on the development of prevention policies. Firstly, the analysis on occupational mortality can be a good starting point for posterior and more specific analysis, as they may be used to detect the presence of an excess of risk associated to a given occupation, and also to formulate hypotheses about the risk factors that should be confirmed by analytical studies. Secondly, as is recommended by the European Agency for Safety and Health at Work,2 more studies on occupational health should be carried out at a regional level in large countries such as Spain, as there may exist diversity and, if all regions are studied as one, the differences may remain concealed. This could be done by using register-based data sources, because in terms of economic cost, the construction of mortality registers and population registers is compulsory at regional level, so the use of these tools is convenient as they do not add any cost for data gathering and data processing. And thirdly, the use of these information systems is particularly important in the study of cancer diseases, because the symptoms appear late when the person is no longer exposed to the risk, making it difficult to detect the relationship between occupation and specific tumours using other monitoring systems such as surveys or occupational accidents statistics.
In summary, we consider that research that can provide information on developing knowledge on occupational health and safety in order to identify risks and risk groups, trends and changes in occupation safety and health, may add valuable information to what is already known. As all the data sources on the topic of occupational health are complementary, by using all available information we may be able to build as complete an image as possible.
The authors would like to thank the Statistical Institute of Navarra for providing the data. We also thank Dr Marina Pollán (National Center for Epidemiology) and the referees for their useful comments. The work has been supported by the Navarra Institute of Occupational Health (project OTRI 2003 0704) and by the Health Department of the Government of Navarra (project res. 1878/2001).
Key points
|
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
1 García Gomez M, Gil López E. Systems of information and monitoring in occupational health. Rev Esp Salud Pública 1996;70:393–407.[Medline]
2 European Agency for Safety and Health at Work. Monitoring occupational safety and health in the European Union. Forum 2004;11:1–12.
3 Navas-Acién A, Pollán M, Gustavsson P, et al. Interactive effect of chemical substances and occupational electromagnetic field exposure on the risk of gliomas and meningiomas in Swedish men. Cancer Epidemiol Biomarkers Prev 2002;11:1678–83.
4 Steenland K, Burnett C, Lalich N, Ward E, Hurrell J. Dying for work: the magnitude of US mortality from selected causes of death associated with occupation. Am J Ind Med 2003;43:461–82.[CrossRef][Web of Science][Medline]
5 Zhou B, Li Y, Stamler J, et al. Relation of occupational change to cardiovascular risk factor levels in rural Chinese men: the people's Republic of China-United States collaborative study on cardiovascular and cardiopulmonary epidemiology. Am J Public Health 2003;93:2049–51.
6 Nurminen M, Karjalainen A. Epidemiologic estimate of the proportion of fatalities related to occupational factors in Finland. Scand J Work Environ Health 2001;27:161–213.[Web of Science][Medline]
7 Lostao L, Regidor E, Aïaich P, Domínguez V. Desigualdades sociales en mortalidad por enfermedad isquémica del corazón y enfermedad cerebrovascular en hombres en España y Francia, 1988–1990 (Social inequalities in male mortality by ischaemic disease of heart and cerebrovascular disease in Spain and France, 1988–1990). Gac Sanit 2000;14:264–7.[Medline]
8 Regidor E, Domínguez V, Calle E, Navarro P. Circunstancias socioeconómicas y mortalidad prematura por enfermedades crónicas (Socio-economic circumstances and premature mortality by chronic diseases). Med Clin 2003;120:201–6.
9 González CA, Agudo A. Occupational cancer in Spain. Environ Health Perspect 1999;107(Suppl 2):273–7.[Web of Science][Medline]
10 Ronda E, Regidor E. Fuentes de información epidemiológica de ámbito estatal en salud laboral en España (Sources of epidemiological information at the state level in occupational health in Spain.). Arch Prev Riesgos Lab 2002;5:111–9.
11 Wiklund K, Eklund G. Reliability of record linkage in the Swedish cancer-environment register. Acta Radiol Oncol 1986;25(1):11–14.[Medline]
12 Breslow NE, Day NE. Statistical Methods in Cancer Research. vol 1. Lyon, France: IARC Scientific Publications, 1994.
13 Ahlbom A. Biostatistics for Epidemiologists. Boca Raton, Florida: Lewis Publishers, 1993.
14 González CA, Riboli E. Bladder cancer among workers in the textile industry: results of a Spanish case-control study. Am J Ind Med 1988;14:673–80.[Medline]
15 Tola S. Occupational cancer of the urinary bladder. J Toxicol Environ Health 1980;6:1253–60.[Medline]
16 Siegrist J, Peter R. Job stressors and coping characteristics in work-related disease: issues of validity. Work Stress 1994;8:130–40.
17 Karjalainen A, Kurppa K, Virtanen S, Keskinen H, Nordman H. Incidence of occupational asthma by occupation and industry in Finland. Am J Ind Med 2000;37:451–8.[CrossRef][Web of Science][Medline]
18 Moreno-Iribas C, Floristán Y. Mortalidad en Navarra en el periodo 1991-2001 (Mortality in Navarra during the period 1991-2001). ANALES Sist San Navarra 2005; in press.
19 Layana E, Artieda L, Lezáun M, Beloqui A. Lesiones Profesionales en Navarra 2001 (Occupational injuries in Navarra 2001). Pamplona, Spain: Instituto Navarro de Salud Laboral, 2004.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||