Unisanté | Data repository
Unisanté | Unisanté

Created on : 05.03.2026 Last modified : 05.03.2026

Swiss Job Exposure Matrix for active Tobacco smoking
A quantitative tool providing smoking probability for occupational groups

Switzerland
DSTE

Reference ID

10.16909-dataset-64

DOI

Producer(s)

Irina Guseva Canu

Collections

Department of occupational and environmental health (DSTE)

Metadata

Identification

Survey ID number

10.16909-dataset-64
Title

Swiss Job Exposure Matrix for active Tobacco smoking
Subtitle

A quantitative tool providing smoking probability for occupational groups
Abbreviation or Acronym

SJEM-T
Translated Title

Matrice Suisse Emploi-Exposition pour le Tabagisme Actif
Country
Name Country code
Switzerland CHE
Series Information

The Swiss Job-Exposure Matrix for Active Smoking (SJEM-T) is a standalone methodological tool derived from the Swiss Health Survey series. While the SJEM-T itself is not part of a repeated survey series, it is based on data from four waves of the ESS (2007, 2012, 2017, 2022), a nationally representative health survey conducted every five years by the Swiss Federal Statistical Office since 1992. The SJEM-T may be updated in the future as new ESS waves become available, potentially establishing a versioned series of occupational smoking exposure matrices. Current version: v1.0.0 (2025), covering the period 2007-2022.
Abstract

Active smoking remains a major confounding factor in occupational epidemiology studies. When individual smoking data are unavailable or incomplete in registry-based studies or retrospective analyses, indirect adjustment using job-exposure matrices (JEMs) provides an approach to control for smoking confounding based on occupational and demographic characteristics. The Swiss Job-Exposure Matrix for Active Smoking (SJEM-T) is a validated quantitative tool providing smoking probability estimates for specific occupational groups. The SJEM-T was developed using Swiss Health Survey data from four waves (2007, 2012, 2017, 2022), comprising approximately 60,000 workers. Smoking probabilities were estimated using logistic regression with current smoking status as the dependent variable, stratified by occupation (ISCO-88), sex, age group, and year. The matrix provides estimates for 12,160 unique strata. Dual validation (internal and criterion in independent cohorts) was performed.
Kind of Data

Aggregate data [agg]
Unit of Analysis

Occupational strata defined by the unique combination of occupation (ISCO-88 codes at 2-, 3-, and 4-digit levels), sex (male/female), age group (under 30, 30-39, 40-49, ≥50 years), and year

Version

Version Description

Version 1.0.0: Initial validated public release (2007-2022). First public version of SJEM-T providing smoking probability estimates for occupations (ISCO-88) stratified by sex, age, and period, covering 5,737 strata at 4 digits level. Based on ~60,000 workers from Swiss Health Survey (2007, 2012, 2017, 2022). Generated using logistic regression with temporal interactions. Dual validated (internal and external in independent cohorts). Includes probabilities, standard errors, sample sizes, and quality indicators. Suitable for research use requiring indirect smoking adjustment. Future updates planned as new survey waves become available.
Version Date

2025-12-31

Scope

Notes

The SJEM-T covers:
OCCUPATION: occupations at three granularity levels (2-digit: 29 groups; 3-digit: 121 groups; 4-digit: 380 groups), occupation-specific smoking prevalence, temporal trends by occupation, high-exposure occupational groups.
DEMOGRAPHICS: Sex (male/female), four age groups (under 30, 30-39, 40-49, ≥50 years), age-sex-specific smoking patterns within occupations.
TEMPORAL DIMENSION: Four periods (2007, 2012, 2017, 2022), 15-year temporal trends, period-specific estimates.
SMOKING MEASURES: Current smoking probability estimates (0-1), standard errors, quality indicators, estimation methods (modeled/imputed).
VALIDATION: Internal validation metrics (consistency, fidelity, temporal stability), external validation in independent cohorts, sensitivity analyses by sample size.

Coverage

Geographic Coverage

National coverage
Universe

Workers aged 15 years and older residing in Switzerland

Producers and sponsors

Primary investigators
Name Affiliation
Irina Guseva Canu Department of Occupational and Environmental Health (DSTE), Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland
Producers
Name Affiliation Role
Jésugnon Ezéchiel DJOHI Department of Occupational and Environmental Health (DSTE), Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland Researcher, co-author
Pedro Marques-Vidal Department of medicine, internal medicine, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland Researcher, co-author
Funding Agency/Sponsor
Name Role
State Secretariat for Education Financing
INTERCAMBIO Financing
Other Identifications/Acknowledgments
Name Affiliation
Carol Clair Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland
Aurélie Lasserre University of Lausanne, Lausanne, Switzerland
Murielle Bochud Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland
Natalie von Goetz
Céline Racine Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland
Thomas Charreau Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland

Sampling

Sampling Procedure

The SJEM-T is not a primary survey but a secondary analytical product derived from the Swiss Health Survey (ESS).

Data collection

Mode of data collection
  • The SJEM-T was developed using Swiss Health Survey data from four waves (2007, 2012, 2017, 2022). Raw data are available on request : https://www.bfs.admin.ch/bfs/fr/home/statistiques/sante/enquetes/sgb.html

Data processing

Data Editing

The SJEM‑T was built by pooling four survey waves (2007, 2012, 2017, 2022), resulting in a dataset of roughly 60,000 workers aged 15 and older with valid ISCO‑88 occupation codes. At each ISCO‑88 level, occupational strata with at least ten workers in the pooled data were modeled directly, while those with fewer than ten workers were assigned imputed smoking probabilities derived from higher‑level ISCO aggregates. Smoking probabilities were estimated using logistic regression models.
Reliability was assessed through a comprehensive validation strategy that combined internal validation using the Swiss Health Survey with criterion validation in two independent Swiss cohorts: CoLaus (n = 3,776) and SHeS‑pilot (n = 615).
All analyses and modeling were conducted using the open‑source software R, version 4.4.2.

Access policy

Location of Data Collection

Unisanté Data repository
URL for Location of Data Collection

https://doi.org/10.16909/dataset/64

Data Access

Access authority
Name Affiliation URL Email
Irina Guseva Canu Department of Occupational and Environmental Health (DSTE), Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland irina.guseva-canu@unisante.ch
Documentation and data unit (UDD) Unisanté, University Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland https://www.unisante.ch/fr/formation-recherche/bibliotheque udd.data@unisante.ch
Access conditions

Public use files, accessible to all. Only for anonymous data
Citation requirements

Guseva Canu, I., Djoki, J.E., Marques-Vidal, P. Swiss Job Exposure Matrix for active Tobacco smoking. Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland. Version 1.0 of the licensed dataset (03/2026), provided by the Unisanté Research Data Repository. DOI:10.16909/DATASET/64
Availability Status

This dataset is made available for research purpose

Disclaimer and copyrights

Disclaimer

The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
Copyright

(c) 2026, Unisanté, University Center for Primary Care and Public Health, Lausanne, Switzerland

Contacts

Contacts
Name Affiliation Email
Irina Guseva Canu Unisanté, Center for Primary Care and Public Health & University of Lausanne, Lausanne, Switzerland Irina.guseva-canu@unisante.ch

Metadata production

DDI Document ID

10.16909-dataset-64
Back to Catalog