Skip to main content

Table 4 Multivariate logistic regressions: Odds ratio for associations between each clinical subtype and treatment subgroups (first line of each cell in a row) and each treatment subgroup and clinical subtypes (second line of each cell in a column, in italic)

From: Multidimensional analyses to assess the relations between treatment choices by physicians and patients’ characteristics: the example of COPD

Odds ratio Trt type 1 Trt type 2 Trt type 3 Trt type 40 Trt type 5 Trt type 6
Nebulised treatments Fixed combinations LABA and inhaled corticosteroids prescribed separately Non-invasive ventilation and oxygen therapy Fixed combinations of short acting anticholinergic and β2 agonist Flu and pneumococcal vaccines
Clinical type 1
Exposed but not severely impaired patients 0.611 3 0.713 1 0.581 1 0.198 4 0.592 2 0.575 3
0.636 2 0.800 0 0.686 1 0.255 3 0.637 2 0.500 4
Clinical type 2
Overweight smokers with comorbidities 0.7810 0.9070 0.8910 3.625 4 1.0600 0.581 2
0.877 0 0.939 0 1.004 0 4.396 4 1.221 0 0.762 0
Clinical type 3
Severe airflow obstruction 1.636 3 1.605 3 1.580 1 6.214 4 1.382 1 1.606 3
1.683 4 1.455 3 1.207 0 7.381 4 1.381 2 2.095 4
Clinical type 4
Women 1.1420 0.9520 1.2930 0.8210 1.2500 1.1050
1.097 0 0.868 0 1.274 0 1.101 0 1.254 0 1.194 0
Clinical type 5
Symptoms of chronic bronchitis 1.1160 0.9860 0.8680 1.1170 1.0950 2.210 4
1.161 0 1.083 0 0.912 0 1.243 0 1.107 0 2.221 4
Clinical type 6
Elderly patients with cardiovascular comorbidity 0.686 2 1.335 1 1.0220 1.3080 0.9540 1.441 2
0.733 1 1.323 1 0.899 0 1.200 0 1.087 0 1.351 1
  1. All models included the 6 treatment or clinical subtypes as covariates. Figures in bold are significantly different from 1 and exponent indicates the class of p values of the log-likelihood test: 0 for p > 0.05, 1 for p ≤ 0.05, 2 for p < 0.01, 3 for p < 0.001 and 4 for p < 0.0001.