Table 4

Influence of different assumptions on the observed Spearman correlation coefficient between FINJEM and Montreal exposure estimates

Agent	Main^*	Montreal level 1-2-3^†	Montreal level 1-3-9^‡	Montreal level 1-10-100^§	FINJEM occupation^¶	Montreal medium or high^**	Montreal unweighted level^††	Prevalence×level^‡‡	Montreal median^§§
Asbestos	8	0.60	0.30^¶¶	0.53	0.60	0.80	–	0.77	0.16
Iron compounds	23	0.52	0.63	0.57	0.43	0.65	–	0.47	0.82
Lead compounds	12	0.47	0.44	0.50	0.39	0.05	–	0.31	0.62
Chromium compounds	9	0.87	0.82	0.87	0.68	0.57	–	0.40	0.74
Welding fume	17	0.65	0.62	0.61	0.64	0.90	0.87	0.53	0.74
Aromatic hydrocarbon solvents	5	0.45	0.22	0.45	0.45	0.21	–	0.67	0.22
Toluene	6	−0.24	−0.38	−0.24	−0.35	−0.21	–	−0.31	−0.15
Formaldehyde	9	0.60	0.60	0.60	0.60	0.76	–	0.88	–
Carbon monoxide	23	−0.01	−0.04	0.01	−0.08	−0.23	–	0.00	0.01
Polycyclic aromatic hydrocarbons	20	0.03	0.16	0.09	−0.04	−0.44	0.68	0.02	0.21
Benzo(a)pyrene	5	−0.35	−0.35	−0.35	−0.35	−0.18	–	−0.54	−0.35
Diesel engine exhaust	15	0.65	0.40	0.60	0.70	0.24	0.64	0.07	0.10
Gasoline engine exhaust	21	0.31	0.24	0.34	0.27	0.72	0.53	0.26	0.31
Flour dust	5	0.89	0.89	0.89	0.89	–	–	0.92	0.89

↵* Spearman correlation coefficient between Montreal and FINJEM estimates using a combination of ISCO'68 code and period as the unit of comparison, the 1-5-25 scaling scheme to average the low-medium-high exposure levels in Montreal data, only Montreal data corresponding to likelihood rating ‘probable’ or ‘definite’, only Montreal data corresponding to exposed for at least low exposure during 2 h per week and calculating a weekly average index from concentration and frequency of exposure for the Montreal estimate.
↵† The Montreal quantitative index was constructed using a linear (1-2-3) scaling scheme for the low, medium, high categories.
↵‡ The Montreal quantitative index was constructed using an exponential (1-3-9) scaling scheme for the low, medium, high categories.
↵§ The Montreal quantitative index was constructed using an exponential (1-10-100) scaling scheme for the low, medium, high categories.
↵¶ The comparison cell was based on the FINJEM job codes instead of on the ISCO'68 codes.
↵** Only Montreal jobs corresponding to exposed for at least 2 h per week at medium or higher were included in the comparison.
↵†† Montreal exposure level not weighted by frequency of exposure.
↵‡‡ Comparison of FINJEM P×L to the product of prevalence and exposure level for Montreal instead of FINJEM L to Montreal exposure level.
↵§§ Each cell-specific estimate is calculated using the median of individual jobs level instead of the arithmetic mean.
↵¶¶ Values in bold indicate important differences with the agreement seen in the main comparison effort.