Validity test of the IPD-Work consortium approach for creating comparable job strain groups between Job Content Questionnaire and Demand-Control Questionnaire

Objectives: This study aims to test the validity of the IPD-Work Consortium approach for creating comparable job strain groups between the Job Content Questionnaire (JCQ) and the Demand-Control Questionnaire (DCQ). Material and Methods: A random population sample (N = 682) of all middle-aged Malmö males and females was given a questionnaire with the 14-item JCQ and 11-item DCQ for the job control and job demands. The JCQ job control and job demands scores were calculated in 3 different ways: using the 14-item JCQ standard scale formulas (method 1); dropping 3 job control items and using the 11-item JCQ standard scale formulas with additional scale weights (method 2); and the approach of the IPD Group (method 3), dropping 3 job control items, but using the simple 11-item summation-based scale formulas. The high job strain was defined as a combination of high demands and low control. Results: Between the 2 questionnaires, false negatives for the high job strain were much greater than false positives (37–49% vs. 7–13%). When the method 3 was applied, the sensitivity of the JCQ for the high job strain against the DCQ was lowest (0.51 vs. 0.60–0.63 when the methods 1 and 2 were applied), although the specificity was highest (0.93 vs. 0.87–0.89 when the methods 1 and 2 were applied). The prevalence of the high job strain with the JCQ (the method 3 was applied) was considerably lower (4–7%) than with the JCQ (the methods 1 and 2 were applied) and the DCQ. The number of congruent cases for the high job strain between the 2 questionnaires was smallest when the method 3 was applied. Conclusions: The IPD-Work Consortium approach showed 2 major weaknesses to be used for epidemiological studies on the high job strain and health outcomes as compared to the standard JCQ methods: the greater misclassification of the high job strain and lower prevalence of the high job strain.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.