Table 3.

The detrimental effect of using Revicki’s model to predict the EQ-5D-5L crosswalk from the PROMIS-29 for the United Kingdom, France, and Germany

	R2adj	ICC	Bias	nRMSE	nMAE	95% theoretical LoA	95% empirical LoA
France
Revicki	0.61	0.78	− 0.09	0.112	0.072	− 0.38; 0.20	− 0.38; 0.08
Polynomial regression	0.72	0.85	0.00	0.075	0.052	± 0.23	− 0.19; 0.17
Germany
Revicki	0.53	0.73	0.00	0.091	0.058	− 0.22; 0.22	− 0.18; 0.14
Polynomial regression	0.64	0.80	0.00	0.079	0.051	± 0.19	− 0.16; 0.13
UK
Revicki	0.68	0.82	− 0.10	0.113	0.075	− 0.39; 0.19	− 0.39; 0.07
Polynomial regression	0.74	0.86	0.00	0.076	0.053	± 0.25	− 0.20; 0.17

UK United Kingdom, adj adjusted, ICC intraclass correlation coefficient, nRMSE normalized root mean squared error, nMAE normalized mean absolute error, LoA levels of agreement