Table 5 Significanta predictors of survival time until death from any cause: Cox's proportional-hazards model

From: Effects of specialisation on treatment and outcomes in screen-detected breast cancers in Wales: cohort study

 

Crude HR

95% CI

Adjusted HR b

95% CI

Adjusted P -value

Explanatory variable

 Surgeon volumed

0.91

0.85-0.097

0.90

0.84–0.97

0.008

 Hospital volumed

1.00

0.94–1.06

1.04

0.98–1.12

0.213

 Age at diagnosis (per year)

1.04

1.01–1.06

1.04

1.02–1.07

0.001

 Invasive vs in situ

2.49

1.55–3.99

2.33

1.39–3.90

0.0003

 Bilateral vs unilateral

3.35

1.82–6.15

2.93

1.54–5.87

0.005

 Mastectomy vs BCT

2.09

1.55–2.82

1.79

1.32–2.44

0.0001

Invasive tumours, N=2121 (100%)

 Surgeon volumed

0.91

0.85–0.97

0.91

0.84–0.98

0.011

 Hospital volumed

1.00

0.94–1.06

1.03

0.84–0.97

0.009

 Age at diagnosis (per year)

1.04

1.10–1.06

1.03

0.96–1.11

0.38

 Bilateral vs unilateral

3.35

1.82–6.15

2.31

1.00–1.06

0.05

 Mastectomy vs BCT

2.09

1.55–2.82

1.46

1.15–4.63

0.02

 Chemotherapy vs none

6.40

4.26–9.63

2.35

1.04–2.05

0.03

 Size

1.03

1.02–1.04

1.02

1.40–3.94

0.001

 Grade 1 vs

grade 2

1.52

1.00–2.33

1.23

1.01–1.03

0.002

grade 3

5.09

3.33–7.78

3.00

0.80–1.90

<0.0001

Missing grade 3

1.34

0.76–2.37

0.94

1.91–4.72

<0.0001

 No malignant nodes vs

     

1–4

2.31

1.63–3.28

1.65

0.52–1.71

 

5–9

7.08

4.50–11.13

3.47

1.14–2.38

 

⩾10

12.48

7.08–21.99

4.26

2.06–5.86

 
    

2.19–8.29

 
  1. CI=confidence interval; HR=hazards ratio.
  2. All tumours, N=2705 (100%).
  3. Hospital volume not significant, but included because it is one of the main explanatory variables of interest.
  4. bAdjusted for all the other variables in the model except for size, grade and malignant nodes, which were only relevant to the invasive tumours model.
  5. dFor a difference in mean volume of 10 patients per year.
Back to article page