Enhancing the Canadian Workplace Exposure Database
The Canadian Workplace Exposure Database (CWED) is a national exposure database, created as a part of CAREX Canada, that compiles and manages measurement data on exposure to known, probable and possible carcinogens, as well as non-carcinogens, in Canada. CWED currently holds over 480,000 exposure observations from 1953 to 2012, with 95% of the data from 1981-1998, collected from government agencies, researchers, and other sources from eight federal, provincial, and territorial jurisdictions. Data from CWED has been used for both cancer prevention and research. Currently, we are working to help sustain and grow the occupational exposure data holdings. We are doing this by securing the existing CWED by reviewing and updating the governance and data stewardship, making the data more widely available and known, and improving data management.
Co-Principal Investigators: Dr. Hugh Davies, Dr. Cheryl Peters, Dr. Jérôme Lavoué
Funder: Current enhancements to CWED have been funded by WorkSafeBC. CWED has been funded in the past by multiple organizations including WorkSafeBC, the Workers’ Compensation Board of Manitoba, the Workers’ Compensation Board of New Brunswick, the Workers’ Compensation Board of Nova Scotia, Workplace Health, Safety and Compensation Commission of Newfoundland & Labrador, and WorkSafe Saskatchewan.
Automating occupation and industry coding
The Canadian Partnership for Tomorrow’s Health (CanPATH, previously known as the Canadian Partnership for Tomorrow Project, CPTP) is Canada’s largest cohort representing 330,000 Canadians from 9 provinces. CanPATH is an exceptional source of data for health researchers as it comprises six regional cohorts that have been linked to allow for national analyses of health outcomes. However, occupational data has been under utilized because of lack of resources to clean and code the occupation and industry data, which has typically been done manually and is time-intensive, and because of inconsistent occupation and industry classifications used across cohorts. The purpose of this study is to enhance and test a previously developed Natural Language Processing approach that has been used to code workers’ occupations and industries (called the Automated Semantic Occupation Coding (ASOC) algorithm) on two regional cohorts (the Atlantic PATH and Alberta’s Tomorrow Project datasets). In doing so, we will harmonize the job title data for approximately 111,000 questionnaires and will ensure that job titles are appropriately coded to the Canadian NOC 2016 system. The refined algorithm can also be applied to the remaining cohorts, harmonizing job title data across CanPATH.
Principal Investigators: Dr. Ellen Sweeney
Co-Investigators: Dr. Anil Adisesh, Dr. Chris Baker, Dr. Cheryl Peters, Mr. Jason Hicks, Dr. Jennifer Vena, Dr. Grace Shen-Tu
Collaborators:: Mr. Yunson Cui, Ms. Debora Addey
Funder:: 2019-2020 CPTP Strategic Development Project