Table 5 Detection of clusters using the Scan statistic of Kulldorff and Nagarwalla (1995) – spatial and space-time analyses

From: Spatial and space–time clustering of childhood acute leukaemia in France from 1990 to 2000: a nationwide study

  

1990–2000

1990–1994

1995–2000

  

0–14 years

0–4 years

5–9 years

10–14 years

0–14 years

0–4 years

5–9 years

10–14 years

0–14 years

0–4 years

5–9 years

10–14 years

Spatial analyses

N S

25

88

3

32

512

13

9

6

37

1

1491

2

 

O S

9

13

4

5

44

4

4

2

8

2

65

2

 

E S

1.41

2.89

0.19

0.34

20.64

0.15

0.21

0.01

0.86

0.01

33.69

0.01

 

P S

0.81

0.62

0.92

0.77

0.44

0.61

0.92

0.81

0.30

0.94

0.09

0.92

Space-time analyses

N ST

3

150

1344

19

        
 

T

1992–93

1991–95

1996–99

2000

        
 

O ST

5

17

50

3

        
 

E ST

0.13

3.67

21.75

0.03

        
 

P ST

0.46

0.57

0.29

0.98

        
  1. This table gives, for each period and age group, a description of the area associated with the highest likelihood ratio (most likely cluster): the number of communes included in the most likely cluster (NS, NST), the number of observed cases (OS, OST), the number of expected cases (ES, EST) and the period associated to the excess in the space–time analyses (T).
  2. The spatial moving window was defined so that it contained up to 10% of the whole French population. In space-time analyses, up to 50% of the time period were covered.
  3. The statistical significance levels (PS, PST) were obtained with 999 Monte Carlo simulations.
Back to article page