133 PROMIS in VWD, IPFDs and RBDs 5 PROMIS measures versus the SF-36v2 in adults with VWD Feasibility The mean number of questions answered per CAT varied from 5.5 (range 4-12) for satisfaction with social roles and activities to 8.7 (range 5-12) for anxiety, which is higher than the number of questions needed to answer for almost all individual SF-36v2 domains. In total, the mean number of PROMIS CAT items completed was 47 (range 30-79), the number of questions completed on the SF-36v2 was 36. (Table 2 and 3) For the PROMIS measures, floor effects were observed in the CATs pain interference and depression (Table 2 and 3). No ceiling effects were observed. Ceiling effects were observed on the SF-36v2 domains of physical functioning, role limitations due to physical health problems, role limitations due to emotional problems, social functioning, and bodily pain. No floor effects were observed in the SF-36v2. Construct validity All PROMIS CATs met the criteria for excellent construct validity except for the CATs anxiety and depression (Table 4 and Supplement Table S1). Due to the high correlation between these two CATs, the correlation between same PRO different PROM (SPDP) (0.67 and 0.71) was smaller than the maximum correlation between different PRO same PROM (DPSP) (0.74 for both CATs), indicating that these CATs do not meet the third criteria of construct validity according to MTMM. For both CAT Anxiety and CAT Depression, 80% of the correlations between different PRO same PROM (DPSP) and different PRO different PROM (DPDP) were smaller than the correlation between same PRO different PROM (SPDP), indicating that both CATs have a good construct validity. Reliability All PROMIS measures and SF-36v2 domains had a good group-level reliability (reliability coefficient or GLB >0.70) (Table 2 and 3). The PROMIS CATs, the PROMIS short form anger and the SF-36v2 domains physical functioning, role limitation due to physical health problems, role limitations due to emotional problems, mental health, bodily pain, and general health had an excellent reliability on individual level (reliability coefficient or GLB ≥0.90).
RkJQdWJsaXNoZXIy MTk4NDMw