Table 4

Results of the logistic regression analysis

Descriptor
Variable
Odds ratio (95% CI)
P value
Likelihood ratio test
P of the test

SLR
Intensity of leg pain
1.3 (1.2–1.4)
0.000
55.2
0.000
X-rays
Intensity of leg pain
1.2 (1.0–1.2)
0.017
76.4
0.000

>2 episodes of low back pain
1.9 (1.4–2.7)
0.000



Disability
1.1 (1.0–1.1)
0.002



Chronicity*

0.000



     Subacute
2.3 (1.5–3.5)
0.000



     Chronic
2.7 (1.2–6.1)
0.022



SLR

0.002



     > 60°
0.7 (0.4–1.0)
0.038



     < 60°
1.5 (1.0–2.2)
0.085


CT scan or MRI
Sex (male)*
1.9 (1.3–3.0)
0.003
93.4
0.000

Intensity of leg pain
1.1 (1.1–1.2)
0.001



Disability
1.1 (1.0–1.1)
0.000



Chronicity*

0.000



     Subacute
2.7 (1.6–4.5)
0.000



     Chronic
5.8 (2.5–13.6)
0.000



SLR

0.002



     > 60°
0.7 (0.4–1.3)
0.230



     < 60°
1.9 (1.1–3.3)
0.016


Prescription of medication
Chronicity*

0.022
18.4
0.001

     Subacute
1.5 (0.6–3.4)
0.370



     Chronic
0.3 (0.1–0.8)
0.013



SLR

0.003



     > 60°
0.5 (0.3–1.0)
0.059



     < 60°
2.2 (0.9–5.1)
0.073


Referral to rehabilitation or physical therapy
Intensity of low back pain
1.1 (1.0–1.2)
0.015
29.5
0.000

>3 episodes of low back pain
1.5 (1.0–2.3)
0.039



Chronicity*

0.006



     Subacute
2.1 (1.3–3.4)
0.002



     Chronic
1.9 (0.8–4.7)
0.172



SLR

0.018



     > 60°
0.6 (0.4–1.1)
0.092



     < 60°
1.4 (0.9–2.2)
0.184


Referral to surgery (orthopedic surgery or neurosurgery)
Sex (male)
2.0 (1.1–3.5)
0.023
61.6
0.000

Intensity of leg pain
1.2 (1.1–1.3)
0.002



Disability
1.1 (1.0–1.2)
0.001



Chronicity*

0.000



     Subacute
3.5 (1.8–6.7)
0.000



     Chronic
9.9 (4.0–24.4)
0.000


Sick leave*
Roland-Morris
1.2 (1.1–1.2)
0.000
58.7
0.000

Intensity of leg pain
1.1 (1.0–1.2)
0.024



Employed by others
2.0 (1.1–3.9)
0.031



*The variables chronicity and SLR are coded as "dummy variables", using the categories "acute" and "not done", respectively, as reference. For "sex", the reference is the "female" category. Since sick leave was only possible in workers, the models on sick leave only included the 367 workers in the study.

For continuous variables, the table shows the OR of each one point increase in the corresponding scale. For dichotomous variables, the table shows the OR for each category

Kovacs et al. BMC Health Services Research 2006 6:57   doi:10.1186/1472-6963-6-57