Table 2

Cox proportional hazards model of survival in all patients

Variable

B

SE

Wald

df

Sig.

Exp(B)


Age at diagnosis

.019

.012

2.661

1

.103

1.019


Category

-.566

.355

2.539

1

.111

.568


Tumour size

.022

.007

8.981

1

.003*

1.022


Tumour grade

.097

.095

1.027

1

.311

1.101


Histological type

-.062

.105

.353

1

.552

.939


Tumour invasion

.009

.159

.003

1

.956

1.009


Nodal status

.131

.089

2.165

1

.141

1.140


Distant metastasis

.107

.272

.154

1

.695

1.113


ER receptors status

.039

.036

1.161

1

.281

1.040


PR receptors status

.119

.051

5.351

1

.021*

1.126


HER2 status

-.090

.056

2.591

1

.107

.914


Hormonal therapy

-.264

.389

.462

1

.497

.768


Chemotherapy

-.007

.079

.008

1

.927

.993


Radiotherapy

-.003

.010

.071

1

.790

.997


Surgical treatment

.011

.209

.003

1

.957

1.011


The output of this analysis includes the unstandardized regression coefficient (B), the standard error of B and its Wald test significance value, the degrees of freedom and the significance value of the coefficient. If the significance value for the coefficient is more than 0 050, then the co-variate effect cannot be assumed to be different from 0. The predicted change in the hazard per unit increase in the co-variate is Exp (B); if the value is less than 1, then the direction of the effect is towards reducing the hazard rate. To be considered to have a significant effect on the hazard rate, the 95 per cent confidence interval for the Exp (B) should not include 1.

Category = less than 40 yrs group vs. more than 40 yrs group

(*) = significant values.

df = degree of freedom

B = unstandardized regression coefficient;

Exp (B) = predicted change in the hazard per unit increase in the co-variate

Kheirelseid et al. BMC Cancer 2011 11:383   doi:10.1186/1471-2407-11-383

Open Data