Results (responder rates) of the sensitivity analyses performed on the whole randomized population according to two different scenarios.

*Scenario 1: the probability of improvement observed in the control group was assigned to the workers not completing the baseline and/or the follow-up diary in both the IG and the control group.**Scenario 2 (worst scenario): the probability of improvement observed in the IG was assigned to the workers not completing the baseline and/or the follow-up diary in the control group, whereas the probability of improvement observed in the control group was assigned to the workers not completing the baseline and/or the follow-up diary in the IG.