Skip to main content
. 2018 Jul 4;27(11):2815–2822. doi: 10.1007/s11136-018-1925-1

Table 2.

Internal validation of mapping algorithms

All patients Patients with HAQ less than 1 Patients with HAQ between 1 and 2 Patients with HAQ greater than 2
Models with HAQ covariate only
 RMSE (SD)
  MV probit 0.196 (0.013) 0.065 (0.003) 0.300 (0.019) 0.293 (0.039)
  Beta 0.246 (0.021) 0.094 (0.018) 0.376 (0.037) 0.343 (0.011)
Samples where RMSE was lower in the MV probit (%) 100 99.6 99.8 91.8
Models with HAQ and VAS pain covariates
 RMSE (SD)
  MV probit 0.177 (0.022) 0.088 (0.022) 0.226 (0.027) 0.277 (0.031)
  Beta 0.196 (0.016) 0.099 (0.014) 0.261 (0.028) 0.281 (0.014)
Samples where RMSE was lower in the MV probit (%) 94.2 90.8 95.2 63.8