Table 2.

Included and excluded variables used in a linear regression model to predict gains in the Early Rehabilitation Barthel Index (ERBI).

	Gains in ERBI
	β-coefficient	Significance
Included variables
ERBI on admission	−0.530	0.000
GCS on admission	0.412	0.000
Total T3	0.202	0.000
Age	−0.179	0.002
Time between injury and admission	−0.176	0.002
Excluded variables
Free T4	−0.013	0.980
TSH	0.005	0.984
Duration of acute treatment	−0.046	0.687