Development of crosswalk methodologies for a diesel engine exhaust job-exposure matrix and impacts on lung cancer risk estimates in workers in Ontario, Canada

Ziembicki S, Kirkham TL, Arrandale VH, Navaneelan T, Demers PA. Development of crosswalk methodologies for a diesel engine exhaust job-exposure matrix and impacts on lung cancer risk estimates in workers in Ontario, Canada. Occup Environ Med. 2026 Apr 17;83(1):42-49. doi: 10.1136/oemed-2025-110471. PMID: 41629154.

Objective: This study investigates the effect of three crosswalk methods of the Diesel Exhaust in Canada Job-Exposure Matrix (DEC-JEM) on lung cancer risk among Ontario workers.

Methods: Ontario workers (2 364 427 identified through workers’ compensation claims (1983-2019)) were followed for cancer diagnoses through linkage with the Ontario Cancer Registry (1964-2019). DEC-JEM is coded in North American Industry Classification System 2002 and National Occupation Classification-Statistics 2006 as an industry-occupation intersection structure, while the cohort is coded in Standard Industry Classification (SIC) 1970 and Canadian Classification and Dictionary of Occupations (CCDO). Three different methods were used to crosswalk. Cox-proportional hazards models of lung cancer were used to estimate HRs and 95% CIs, adjusted for age, birth year and sex. Results were validated against the CANadian Job-Exposure Matrix (CANJEM), coded directly into SIC1970-CCDO intersections.

Results: DEC-JEM₂ and DEC-JEM₃ had similar case counts and exposure distributions for each exposure level and threshold, notably lower than DEC-JEM₁. Increased lung cancer risk was observed using all methods. Differences in risk between methods were most noticeable at higher exposure levels, while little difference was observed among low exposed groups. DEC-JEM₃ had the highest risk. Risks mostly differed between DEC-JEM and CANJEM, particularly higher exposure levels (high, very high).

Conclusions: Crosswalk decisions can change a JEM’s exposure distribution, determine which cases get classified as exposed and influence disease risk estimates. Future crosswalks should consider unexposed matches and a weighted average exposure for one-to-multiple matches. This study’s crosswalk methods provide opportunities to save resources while improving exposure assessment.

Keywords: Epidemiology; Public Health Surveillance.