Skip to main content

Table 4 Logistic regression analysis

From: Ethnicity and thrombolysis in ischemic stroke: a hospital based study in Amsterdam

  OR (95% CI) P-value
I. Univariate model   
   Ethnicity 0.34 (0.17-0.71) 0.004
   Arrival hospital after 2.5 hours of symptom onset 0.033 (0.005-0.21) < 0.001
   Age 0.99 (0.98-1.0) 0.29
   Gender 1.8 (1.1-2.9) 0.02
   Previous stroke and diabetes 0.48 (0.14-1.6) 0.24
   Blood pressure higher than 185/110 0.68 (0.35-1.3) 0.25
   NIHSS < 5 8.1 (4.2-16) < 0.001
   NIHSS > 25 4.9 (0.30-79) 0.26
   Oral anticoagulation use 0.15 (0.035-0.61) 0.008
II. Bivariate model (ethnicity is added to each model)  
   Arrival hospital after 2.5 hours of symptom onset 0.60 (0.24-1.50)* 0.28
   Age 0.30 (0.14-0.63)* 0.001
   Gender 0.36 (0.17-0.74) 0.006
   Previous stroke and diabetes 0.34 (0.17-0.71) 0.004
   Blood pressure higher than 185/110 0.33 (0.15-0.71)* 0.004
   NIHSS < 5 0.36 (0.16-0.78) 0.01
   NIHSS > 25 0.34 (0.16-0.73) 0.006
   Oral anticoagulation use 0.31 (0.15-0.64)* 0.002
III. Multivariate model 0.38 (0.13-1.16) 0.09
  1. The univariate model (I) shows the odds ratio for receiving thrombolysis for the variables separately. Next, the association with ethnicity is analysed in the bivariate model (II), using the same preselected variables. Finally, in the multivariate model (III) the variables which significantly changed the odds ratio (age; blood pressure higher than 185/110; oral anticoagulation use; arrival at the hospital after 2.5 hours of symptom onset) were entered in a model together with ethnicity.
  2. OR = odds ratio; CI = confidence intervals; NIHSS = National institute of health stroke scale; * indicates a 5% or greater change in odds ratio