Table 4.

Binary logistic regression of the most sensitive and specific combination of pain drawing criteria

Pain drawing criteria	OR/unit (95% CI)	Log likelihood	p-value
Total number of marks	1.7 (1.3, 2.3)	- 25	< 0.001
Length of the longest mark	1.1 (1.0, 1.1)		0.029
Symmetric patterns (yes/no)	6.5 (1.6, 26.9)		0.010

This tabulated group lists the three best pain drawing criteria (two continuous plus one binary) with the highest power to distinguish somatoform-functional from somatic-nociceptive pain, the criteria are included in a model as separate variables.