Table 1 Additional simulations on different working schedules to calculate measurement fluctuation

To gain insight into the effect of a certain working schedule, we conducted additional simulations on other types of working weeks and calculated the total CIs according to step 3 which is described in this section. In these simulations, one measurement per half-, 1-, 2- and 3-h time slot during the week was assumed. The results were compared with the total CI based on the simulations on the 40-h working week we actually used in our analyses. This showed that the CIs would be a bit lower in most cases. The differences vary between − 0.21 and + 0.10 if 50 GPs are assumed and between − 0.12 and + 0.10 if 100 GPs are assumed (Table 4 in the Appendix).