Table 2

 The strength of the predictors for chest x ray indicative for pneumoconiosis (ILO profusion category ⩾1/1)

Questionnaire modelQuestionnaire + exposure modelQuestionnaire + lung function model
OR (95% CI)OR (95% CI)β*OR (95% CI)
*Regression coefficients after multiplication by the correction or shrinkage factor obtained from the bootstrapping procedure.
‡See footnotes of table 1.
The probability of having a chest x ray result that is indicative for the presence of pneumoconiosis defined as ILO profusion category ⩾1/1 can be estimated using the following formula: P (pneumoconiosis)  =  1/(1+exp(−((0.72 × age ⩾ 40 years) + (0.7 × current smoker) + (1.14 × high exposed job title) + (1.0 × work in the construction industry ⩾ 15 years) + (0.84 × “feeling unhealthy”) + (0.91 × standardised residual FEV1 ⩽ −1.0) −6.33))). Each predictor is valued as 1 when present and 0 when absent.
Predictors
Age ⩾40 years2.3 (1.0 to 5.4)2.3 (1.0 to 5.4)0.722.3 (1.0 to 5.4)
Current smoker2.5 (1.2 to 5.1)2.5 (1.2 to 5.1)0.702.4 (1.1 to 4.9)
High exposed job title‡4.1 (1.5 to 10.6)3.9 (1.3 to 11.6)1.144.0 (1.5 to 10.5)
Work in the construction industry ⩾15 years3.3 (1.2 to 9.1)3.2 (1.1 to 9.4)1.003.4 (1.2 to 9.3)
“Feeling unhealthy”2.6 (1.2 to 5.7)2.6 (1.2 to 5.7)0.842.8 (1.3 to 6.0)
Cumulative exposure index ⩾10.01.1 (0.5 to 2.4)
Standardised residual FEV1 ⩽−1.00.913.0 (1.5 to 6.3)
ROC area of the model0.79 (0.74 to 0.85)0.79 (0.74 to 0.85)0.81 (0.75 to 0.86)