Skip to main content

Table 3 Expected numbers (from model with maximum likelihood parameter estimates) and observed numbers of persons reporting symptoms in each of the three waves

From: Prior immunity helps to explain wave-like behaviour of pandemic influenza in 1918-9

POPULATION N SAW SA - S -W S - - -AW - A - - - W - - -
South Shields 462 0.08
0.00
0.90
1.00
1.73
1.00
12.73
12.00
2.26
3.00
24.30
24.00
33.25
31.00
386.74
390.00
Leicester 4619 1.02
2.00
16.75
14.00
27.80
22.00
271.80
287.00
37.40
37.00
613.20
609.00
314.53
323.00
3336.50
3325.00
Wigan 1075 0.02
1.00
0.69
0.00
1.99
0.00
46.64
44.00
1.74
2.00
73.84
79.00
99.85
117.00
850.23
832.00
Newcastle 4461 0.41
2.00
5.26
2.00
33.22
39.00
237.09
234.00
16.55
17.00
211.94
208.00
322.86
326.00
3633.67
3633.00
Manchester 4686 1.84
3.00
66.94
67.00
23.49
26.00
619.65
613.00
10.49
11.00
381.88
392.00
80.22
73.00
3501.50
3501.00
Blackburn 1284 0.37
1.00
5.34
7.00
11.64
6.00
104.12
97.00
5.63
10.00
81.97
72.00
84.54
83.00
990.39
1008.00
Widnes 3417 0.40
0.00
10.66
14.00
25.40
21.00
382.55
388.00
9.84
8.00
261.65
266.00
346.08
340.00
2380.43
2380.00
London police 746 0.18
0.00
5.81
4.00
1.82
0.00
41.69
46.00
3.49
5.00
115.47
108.00
16.96
24.00
560.58
559.00
Cambridge Uni 1766 3.00
8.00
44.08
33.00
33.42
17.00
341.88
365.00
25.59
38.00
375.38
368.00
135.31
130.00
807.33
807.00
Clifton College 451 14.43
13.00
19.26
9.00
73.26
71.00
51.85
69.00
38.87
61.00
51.88
38.00
86.61
85.00
114.84
105.00
Haileybury 515 8.98
10.00
27.38
31.00
45.68
22.00
96.60
117.00
19.19
25.00
58.47
48.00
87.76
106.00
170.94
156.00
Finchley School 1224 0.28
0.00
18.05
18.00
2.41
5.00
111.70
111.00
5.68
4.00
372.76
383.00
26.62
29.00
686.50
674.00
TOTAL 24706 31.01
40.00
221.12
200.00
281.88
230.00
2318.31
2383.00
176.70
221.00
2622.77
2595.00
1634.57
1667.00
17419.64
17370.00
  1. As before, SAW denotes the number of persons reporting symptoms in each of the summer, autumn and winter waves. S - - denotes the number with symptoms only in the summer wave ....etc. The expected numbers are based on the maximum likelihood parameter estimates for the 12 population model (simulation 1; results from simulation 2 are almost identical). The deviance corresponding to the MLE fit was 47182. It can be seen that the fit between observed and expected is less good for three of the "school" populations, reflecting their somewhat different behaviour (see also Table S2 in Additional File 1).