Article Text
Abstract
Objectives In France, two occupational exposure databases with different goals co-exist. Colchic was built to store measurements from prevention activities since 1987, while Scola contains measurements from compliance monitoring since 2007. We aimed to compare exposure levels between these databases targeting the same population of workers.
Methods Despite their different objectives, both databases share the same structure. We selected personal measurements with a sampling duration ≤12 hours of airborne chemical substances with ≥1,000 results in each database between 2007 and 2015 and combined into a single dataset. We used Tobit models to measure associations between log-transformed concentrations and six predictors (sampling year and duration, global ventilation, number of workers, personal protective equipment (PPE) and region) for each agent. We combined agent-specific models using meta-analytic approaches. We assessed average differences between Colchic and Scola for four prediction scenarios based on sampling year and duration.
Results We selected 239,968 measurements (25% from Colchic and 75% from Scola) across fifteen substances (66% organic solvents, 27% dusts, and 7% metals). The most prevalent agents were wood dust (n=42,193), respirable dust (n=25,299) and quartz (n=23,774). Non-detects represented 45% of all measurements. PPE, sampling year and duration were the main predictors of exposure levels. For 2007, predicted exposure levels in Scola were 2.3 times lower than in Colchic, regardless of sampling duration. This difference decreased in recent years: predicted exposure levels for 2015 were 1.25 times higher in Scola than Colchic for a sampling duration of 30 minutes, and 1.18 times lower in Scola than Colchic for a duration of 240 minutes.
Conclusions Colchic and Scola both represent important sources of information on historical and current occupational exposures in France. Despite a notable early difference in average exposure levels between Colchic and Scola, our results suggest that the contrasts in exposure are getting smaller over time.