Skip to main content

Table 4 Hazard ratios for incident invasive breast cancer in women 50–69 years old, according to reduced Gail model (5-year risk) quintiles and area under the curve (AUC) for each model using continuous Gail scores

From: Prospective validation of the NCI Breast Cancer Risk Assessment Tool (Gail Model) on 40,000 Australian women

Gail Subset (model)

Frequency in rank (%)

Risk score quintile group (5-year risk)

Test for trend

(Q1 to Q5)a

AUC (95% CI)

Q1

Q2

Q3

Q4

Q5

D9

D10

Age (M1)

100

Referent

1.25 (0.95–1.64), p = 0.112

1.37 (1.03–1.83), p = 0.029

1.38 (1.06–1.81), p = 0.018

1.96 (1.51–2.56), p < 0.001

2.06 (1.52–2.79), p < 0.001

1.85 (1.33–2.58), p < 0.001

χ2 = 23, p < 0.0001

0.56 (0.53–0.58)

+ first live birth age (M2)

62

Referent

0.94 (0.70–1.26), p = 0.686

1.18 (0.90–1.54), p = 0.229

1.38 (1.05–1.80), p = 0.02

1.64 (1.27–2.12), p < 0.001

1.50 (1.09–2.06), p = 0.012

1.77 (1.32–2.38), p < 0.001

χ2 = 21, p < 0.0001

0.56 (0.54–0.58)

+ age at menarche (M3)

62

Referent

1.12 (0.83–1.51), p = 0.458

1.50 (1.13–1.98), p = 0.004

1.49 (1.13–1.97), p = 0.005

1.89 (1.44–2.47), p < 0.001

1.76 (1.28–2.43), p = 0.001

2.01 (1.48–2.73), p < 0.001

χ2 = 26, p < 0.0001

0.56 (0.54–0.59)

+ number of first-degree relatives (M4)

84

Referent

1.22 (0.90–1.67), p = 0.200

1.43 (1.06–1.93), p = 0.019

2.05 (1.55–2.71), p < 0.001

2.15 (1.63–2.84), p < 0.001

1.82 (1.30–2.54), p < 0.001

2.49 (1.82–3.39), p < 0.001

χ2 = 45, p < 0.0001

0.58 (0.56–0.60)

+ number of biopsies (M5)

79

Referent

1.06 (0.77–1.46), p = 0.730

1.69 (1.27–2.27), p < 0.001

1.91 (1.43–2.54), p < 0.001

2.28 (1.73–3.01), p < 0.001

2.22 (1.61–3.06), p < 0.001

2.34 (1.71–3.21), p < 0.001

χ2 = 52, p < 0.0001

0.59 (0.56–0.61)

+ had biopsy (M6)

91

Referent

1.07 (0.78–1.48), p = 0.672

1.70 (1.27–2.27), p < 0.001

1.90 (1.43–2.52), p < 0.001

2.29 (1.74–3.02), p < 0.001

2.24 (1.63–3.08), p < 0.001

2.33 (1.70–3.20), p < 0.001

χ2 = 52, p < 0.0001

0.59 (0.56–0.61)

+ ethnicity (M7)

98

Referent

1.07 (0.78–1.48), p = 0.672

1.70 (1.27–2.27), p < 0.001

1.90 (1.43–2.52), p < 0.001

2.29 (1.74–3.02), p < 0.001

2.24 (1.63–3.08), p < 0.001

2.33 (1.70–3.20), p < 0.001

χ2 = 52, p < 0.0001

0.59 (0.56–0.61)

+ had hyperplasia (full model) (M8)

100

Referent

1.07 (0.78–1.48), p = 0.679

1.70 (1.27–2.27), p < 0.001

1.89 (1.42–2.52), p < 0.001

2.28 (1.73–3.02), p < 0.001

2.21 (1.61–3.05), p < 0.001

2.35 (1.72–3.23), p < 0.001

χ2 = 52, p < 0.0001

0.59 (0.56–0.61)

  1. Values are shown as hazard ratio (95% CI), p value
  2. CI confidence interval, D decile, Q quintile
  3. a Tests for trend for each model across hazard functions for risk score groups Q1 to Q5