COSMOP: Monitoring of Mortality by Occupational Activity
As part of the epidemiological surveillance of occupational risks, the Cosmop program draws on existing data sources to produce mortality indicators by occupational activity.
What is COSMOP?
COSMOP is an epidemiological surveillance program tracking cause-specific mortality by occupational activity. Its objectives are to develop and regularly disseminate cause-specific mortality indicators by occupational activity to relevant stakeholders (government agencies, occupational physicians, occupational health and safety professionals, social partners, etc.).
The compiled indicators can serve several purposes:
Guiding prevention efforts. In particular, regarding mortality among young and working-age individuals, sharing the results with stakeholders and public authorities can lead to targeted and specific campaigns.
Objectively assess situations presumed to be problematic (clusters of deaths within workplaces). Mortality rates among workers in a company can be compared to national reference data, taking into account the occupational structure and the company’s sector.
Contribute to the identification of occupations and sectors at particular risk. The observation of unexplained excess mortality in certain occupational groups should enable the generation of hypotheses and targeted studies and guide surveillance according to the sector of activity.
In the longer term, by observing trends, assess the impact of changes in the workplace—both in terms of work processes and work organization—and the impact of prevention policies. The Cosmop program relies on existing data routinely collected as part of a long-term system. It covers all age groups starting at age 16, both genders, and all work situations in France, regardless of occupation or sector of activity.
The Cosmop program draws on existing data collected routinely as part of a long-term initiative. It covers all age groups starting at age 16, both genders, and all employment situations in France, regardless of occupation or industry.
Given its work on social differences in mortality, the Occupational Health Division of Santé publique France naturally reached out to INSEE to identify available data sources on individual socio-professional trajectories, which also contain demographic information enabling the identification of a subject’s potential death and its cause.
In agreement with INSEE’s Demography Department, an initial study was conducted using data from INSEE’s Permanent Demographic Sample.
A new analysis was then carried out, in partnership with INSEE’s Department of Employment and Labor Income and its Department of Demography, using a new sample from the “DADS Panel”: this longitudinal, open-ended sample is compiled and managed by INSEE based on the annual social data reports it receives.
Individuals’ vital status was retrieved from the National Registry of Natural Persons (RNIPP). The medical causes of the recorded deaths were then obtained from the Center for Epidemiology on Medical Causes of Death (CépiDc) at Inserm.
This work highlighted the advantages and limitations of the sources used; it also made it possible to define the indicators that can be produced from these data and to formulate recommendations for the establishment of a sustainable mortality surveillance system
Analysis of Data from the Permanent Demographic Sample (PDS)
In collaboration with INSEE’s Demography Department, an initial study was conducted using data from INSEE’s Permanent Demographic Sample.
This sample, representing 1% of the population living in France, was established in 1968 by INSEE, with the initial goal of creating a large longitudinal database representative of people living in France. The primary sources of information consist of census records and vital statistics collected for each individual included in the sample. The sample was more recently expanded to a 25th of the population (starting in 2004 for civil status records and 2008 for the census). Due to its construction and monitoring methods, this sample can be considered representative of the population and includes all types of workers (employees, self-employed individuals, farmers, etc.). The main limitations of this database relate to the lack of occupational information during the intercensal periods.
Santé publique France initially analyzed the data from this sample (1968–1990 version) as a feasibility study. The primary medical cause of the recorded deaths was identified using the CépiDc database at Inserm. The results of this analysis were published in a report in 2006.
Among men, the industrial sector stood out for a tendency toward excess mortality (except in the electrical and electronic manufacturing sector). This excess mortality varied across cancerous and non-cancerous conditions and violent deaths, depending on the specific industry. In the service sector, however, men tended to have rates comparable to or lower than those of all other sectors, with the notable exception of the food retail and hospitality sectors, where excess deaths from digestive cancers, certain non-cancerous conditions, and violent deaths were observed. The agricultural sector, meanwhile, was characterized by lower-than-average mortality from cancer and ischemic heart disease, in contrast to higher-than-average mortality from other non-cancerous conditions and suicides.
Analysis of data from the DADS panel
A new analysis was conducted in partnership with INSEE’s Department of Employment and Labor Income and its Department of Demography, using a new sample drawn from the “DADS Panel.”
The DADS Panel was established by INSEE through the accumulation, since 1976, of annual social data declarations (DADS) from a sample of employees selected based on their birthdays (1/24th sample). The scope of the DADS covers most of the private and semi-public sectors and represents approximately 80% of salaried jobs.
In 2002, the scope of this panel was expanded to 1/12, representing 8% of the salaried workforce in France.
Santé publique France analyzed the Panel data from the 1976–2002 period. Individuals’ vital status was retrieved from the National Directory of Natural Persons (RNIPP), and the primary medical cause of death was obtained from the Inserm’s CépiDc database.
Two main indicators were calculated: relative risks, which allow for comparisons between subgroups, and mortality rates, which enable the compilation of mortality profiles by cause of death, based on different exposure criteria:
having worked versus never having worked in the sector;
having one’s last known professional activity in the sector;
having one’s professional activity in the sector during the year in question.
Mortality rates were calculated for the entire working population to provide baseline data for studying worker mortality.
The results were the subject of a general report in 2018.
The results obtained were generally consistent with the observations from the first sample. Over the study period, certain sectors appeared to be at higher risk of death, particularly certain industries (such as the food, textile, woodworking, mineral products, and metal industries), the construction sector, transportation support services, and the hotel and restaurant industry. In the service sectors, the relative risks of death were lower, with the notable exceptions of transportation support services, business services, and health and social work. The analysis of mortality by cause showed that excess risks of death by cause were not the same across all sectors. Thus, excess mortality from malignant tumors was found in the mineral products and metal industries, in construction, transport support services, the hotel and restaurant sector, and business services, the latter being primarily represented by temporary workers. Accidental mortality, on the other hand, was more prevalent in the construction, retail, and automotive repair sectors, as well as in land transportation, while sectors with excess mortality due to suicide included business services, public administration, and health and social services.
Thematic analyses also yielded results for suicide mortality and cardiovascular mortality.
Cohidon C, Geoffroy-Perez B, Fouquet A, Le Naour C, Goldberg M, Imbernon E. Suicide and occupational activity in France: initial analysis of available data. Saint-Maurice: Institute for Health Surveillance, April 2010. 8 p.
Diène E, Fouquet A, Geoffroy-Perez B, Julliard S. Premature mortality from cardiovascular diseases among men by socioeconomic category and economic sector. Saint-Maurice: Institute for Public Health Surveillance; 2014. 6 p.
Reports
Cosmop Program: Monitoring of Mortality by Cause According to Occupational Activity. Analysis of Mortality and Causes of Death by Economic Sector from 1976 to 2005. Saint-Maurice: Santé publique France, 2018. 49 p.
Diène E, Fouquet A, Geoffroy-Perez B, Julliard S. Premature mortality from cardiovascular diseases among men by socioeconomic status and sector of activity. Saint-Maurice: Institute for Public Health Surveillance; 2014. 6 p.
Cohidon C, Geoffroy-Perez B, Fouquet A, Le Naour C, Goldberg M, Imbernon E. Suicide and occupational activity in France: initial analysis of available data. Saint-Maurice: Institute for Public Health Surveillance, April 2010. 8 p.
Geoffroy-Perez B. Analysis of mortality and causes of death by economic sector from 1968 to 1999 based on the Permanent Demographic Sample. Saint-Maurice: Santé publique France, September 2006, 162 p.
