Table 1 Surgeon volume and patient demographic characteristics, follow-up duration and survival

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

Surgeon volumea (N)

High (6) (50)

Medium (4) (11–49)

Low (15) (10)

Total (25)

Test, P -value

Age at diagnosis (years)

    

Spearman's rank

N (%)

2092

536

76

2704

correlation

 Mean (s.d.)

58.6 (5.6)

58.7 (5.7)

58.5 (4.9)

58.5 (5.6)

P=0.48

 Range

41.3–81.0

49.2–84.9

41.4–72.9

41.3–84.9

 

Townsend deprivation score

    

ANOVA

 1st quartile (most affluent)

482 (23)

167 (31)

27 (36)

676 (25)

P=0.001

 2nd quartile

553 (26)

106 (20)

11 (14)

670 (25)

 

 3rd quartile

511 (24)

142 (26)

21 (28)

674 (25)

 

 4th quartile

539 (26)

118 (22)

17 (22)

674 (25)

 

 Missingb

7 (0.3)

3 (0.6)

0

10 (0.4)

 

Follow-up (years)

    

Spearman's rank

N (%)

2086

534

75

2696

correlation

 Median (IQR)

3.0 (1.9–4.2)

3.0 (1.5–4.2)

3.1 (2.0–4.9)

3.0 (1.8–4.2)

P<0.001

 Missingb

6 (0.29)

2 (0.37)

1 (1.3)

9 (0.33)

 

Recurrence

    

Wilcoxin's rank sum

N (%)

87 (4.2)

26 (4.9)

5 (6.9)

118 (4.4)

P=0.69

All-cause mortality

    

Wilcoxin's rank sum

N (%)

146 (7.1)

46 (8.6)

11 (4.5)

203 (7.5)

P=0.76

Breast cancer mortality

    

Wilcoxin's rank sum

N (%)

86 (4.1)

28 (5.2)

6 (7.9)

120 (4.4)

P=0.9996

  1. ANOVA=analysis of variance; IQR=interquartile range.
  2. All tumours N (%): 2704 (100), one patient with missing surgeon ID.
  3. aSurgeon volume analysed as a continuous variable.
  4. bMissing values not used in analyses.