IRSST - Institut de recherche Robert-Sauvé en santé et en sécurité du travail

A Comparative Study of the IMIS (OSHA) and LIMS (IRSST) Exposure Databases


Good knowledge of levels of occupational exposure to chemical contaminants is crucial in any program to control, prevent and manage associated risks. Québec’s occupational health and safety research institute, the Institut de recherche Robert-Sauvé en santé et en sécurité du travail (IRSST), thus administers a Laboratory Information Management System (LIMS). This electronic database contains all analytical results on workplace exposure as measured by occupational health teams in Québec since 1985. Though the LIMS data are useful for developing exposure profiles, the scarcity of information on sampling circumstances and objectives limits their interpretation. In the USA, the Occupational Safety and Health Administration (OSHA) has maintained the Integrated Management Information System (IMIS) since 1979. This database, which contains analytical results from samples collected by OSHA inspectors since that time, includes more complete information on sampling circumstances than the LIMS. In addition, unlike the LIMS, which only reports sample concentrations, most IMIS results are calculated personal exposure measurements directly comparable to occupational exposure limits.

Given the similarity of U.S. and Québec socioeconomic activities, the general objective of this research project was to compare the LIMS data with the exposure data collected by OSHA to determine if the U.S. data can serve as a source of information on occupational exposure conditions in Québec.

Covered by the comparison were all chemical agents analyzed over a common period in the two databases. There were two main parts to the comparison. First, industries associated with exposure in the IMIS and LIMS were compared. Agent-industry pairs for which at least 10 detected values were available were identified in one database and the proportion of pairs found in the other database as well was determined. Second, exposure levels were compared. In a preliminary descriptive analysis, median levels of each chemical agent were compared, regardless of period or industry. In a second analysis, average levels reported in the IMIS and the LMIS were compared by statistical modeling taking into consideration measurement year, industry and sampling time. Results for all agents in two major chemical families, metals and solvents, were compared.

The extracts from the IMIS (352,442 records) and the LIMS (286,083 records) covered the years 1985 to 2011 and 49 common substances: 21 solvents, 15 metals, 5 gases, 4 isocyanates, 2 acids, crystalline silica and styrene. Metals data were more numerous in the IMIS (234,387 records compared to 86,054 records), whereas solvent data were more numerous in the LIMS (247,367 records compared to 71,690 records). Both databases included considerable data on lead, toluene, iron and manganese. For the 49 agents taken together, the proportion of non-detects and the proportion of values exceeding ACGIH threshold limit values (TLV®) were very similar in the two databases.

Most of the records in the extracts from the IMIS as well as the LIMS were for the manufacturing sector, with more than 70% of the measurements from the two industry groups designated as top priorities by Québec’s workers compensation board, the Commission des normes, de l'équité, de la santé et de la sécurité du travail (CNESST).[1] In general, the breakdown of the measurements by industry was similar in the two databases. Comparison of agent-industry pairs for which there were detected values was possible for 36 agents. Of the pairs identified in the IMIS, it was possible to match 61.4% with the LIMS, whereas 62.8% of the pairs identified in the LIMS could be matched with IMIS pairs. In other words, the IMIS and LIMS provide profiles of exposure within North American industry that are in general compatible but complementary.

Direct comparison of exposure levels was limited by the incomplete compatibility of the industrial classification systems used by the IMIS and LIMS and by the lack of information on sampling time in the LIMS prior to 1994. As a result, a descriptive comparison was made of 169,388 IMIS and 367,486 LIMS records, whereas for the modeling no more than 100,000 records in all were used from the two databases combined. The descriptive analyses and the modeling demonstrated that metal exposure levels, short-term as well as long-term, were lower in the IMIS than the LIMS by a factor of about two. This was true for most metals studied. For solvents, the descriptive analyses demonstrated similar short-term exposure levels in the two databases, despite substantial differences for certain agents. With respect to long-term exposure, the analyses suggest slightly higher levels in the IMIS. Substantial differences for certain solvents studied were also noted.

Despite the lack of reference data to check to what extent the IMIS and LIMS measurements are representative of occupational exposure in Québec, this study, which suggests a consistent overall portrait in terms of industries covered and exposure levels in the two databases, is reassuring in this respect. Given the scarcity of measurement data available, the results of this study provide strong support for using the IMIS and LIMS together to assess occupational exposure in Québec.

[1] See Table 5 for the list of industries in the top priority groups.




Additional Information

Category: Research Report
Research Project: 2013-0096
Online since: November 12, 2018
Format: Text