Table 2 The logit coefficients of predictors from the three-step approach.

From: Prognostic subgroups of chronic pain patients using latent variable mixture modeling within a supervised machine learning framework

Variable

Class 1

Class 2

Class 3

Sex

 − 0.05

0.53***

0.67***

Age

 − 0.03***

 − 0.02***

 − 0.01***

Higher education level

0.15*

 − 0.27***

 − 0.18**

Full-time employment

 − 0.11

 − 0.79***

 − 0.60***

Visits to a pain clinic in the past year

0.10

0.93***

0.60***

Pain duration

0.00

0.01**

0.01**

Social support

 − 0.27***

0.00

0.20***

  1. Class 4 is the reference group. With listwise deletion, 9173 cases (76.5%) with complete data were used for the above regression estimation. Sex: 1 = man, 2 = woman. Educational level: 1 = primary education, 2 = secondary education, 3 = tertiary education. Full-time employment: 1 = yes, 0 = no. Visits to a pain clinic in the past year: 0 = 0-1 time, 1 =2–3 times, 2 = over 4 times. Pain duration: years of pain experience. Social support: a subscale from the Multidimensional Pain Inventory [29, with higher scores meaning more perceived social support for their pain problems.  *p < .05. **p < .01. ***p < .001.
Back to article page