Comparing the reliability of absenteeism from the current study with the original study  is complex because in the initial study, the average ICC,Footnote 1 ratios were used, resulting in an increase in CCIs. We believe that individual CCI measures are the appropriate CCI to address the issue of background checks, as patients in clinical practice complete iPCQ-VR once per measurement point (i.e. baseline, unloading, follow-up). In addition, the initial study measured a short-term sick leave with a 2-week recall, while we used 4 weeks. Finally, the initial study did not select a stable group of participants. Our results need to be carefully generalized, as there are constraints in our study that need to be addressed. First, one of the admission criteria for this study was that participants should be on sick leave at the beginning (part-time or full-time). However, 14% of the 1 and 8.5% sample of studies were not on sick leave at first, but full-time in the workplace. This resulted in a decrease in samples for testing, which likely had a negative impact on sick leave and presentism results. Second, we applied anchor elements to Measure 2 to detect stable and unstable (i.e. modified) samples of participants. In terms of working status and weekly working time, this resulted in better reliability results in the stable group of participants. However, the results remained the same for other iPCQ-VR points, such as short- and long-term pylon rates.
It should be noted that the Health items of the TiCP-VR in the stable sample generally showed a lower test reliability (lower Kappa values) compared to the unstable sample. Therefore, the anchor elements used in this study warrant refinement. A total of 52 participants completed the questionnaires (response rate again 71%). The reasons for the lack of response were technical problems (n-7), withdrawal authorization (n-3), telephone number (No. 2) or the unknown (n-9). The new test was submitted on average 19.6 days (SD 5.8) after the presentation of the first questionnaires.