Skip to main content
. 2013 Feb 14;108(4):873–880. doi: 10.1038/bjc.2013.42

Table 3. Backward stepwise Binary Logistic Regression (BLR) statistical analysis.

Response variable Predictor variables β P-value OR (95% CI)
PgR
p53 wild type
0.72
0.030
2.05 (1.07–3.93)
 
Low DI tumours
−0.67
0.030
0.51 (0.28–0.94)
p53 wild type
ER
1.69
<0.001
5.40 (2.62–11.13)
 
Intermediate DI tumours
−1.02
0.009
0.36 (0.17–0.78)
ER
p53 wild type
1.68
<0.001
5.37 (2.38–12.14)
 
HER2
1.20
0.013
3.34 (1.29–8.63)
HER2
ER
1.38
0.002
3.96 (1.65–9.46)
 
High DI tumours
−1.55
0.003
0.21 (0.08–0.59)
Low DI tumours
PgR
−0.76
0.012
0.47 (0.26–0.84)
Intermediate DI tumours
p53 wild type
−1.17
0.001
0.31 (0.15–0.63)
High DI tumours HER2 −1.42 0.004 0.24 (0.09–0.64)

Abbreviations: CI=confidence interval; DI=DNA index; ER=oestrogen receptor; HER2=human epidermal growth factor receptor 2; OR=odds ratio; PgR=progesterone receptor.

Variables were analysed by backwards step-wise BLR. All significant associations between the predictor and response variables are shown and non-significant associations are omitted for brevity and clarity.