Skip to main content
. 2003 Oct;9(10):1287–1293. doi: 10.3201/eid0910.030182

Table 2. Baseline characteristics, estimated coefficients, and standard errors for the model.

Variable Epidemic experience
(n districts)a
 Multi-variable analysis
Ever (%) Never β SE
Absolute humidity profileb
Class 1
 15 (3)
 527
 Reference
Class 2
 9 (6)
 153
 0.59
 0.44
Class 3
 19 (8)
 228
 1.00c
 0.36
Class 4
 43 (47)
 48
 3.48c
 0.35
Class 5
 143 (61)
 93
 3.97c
 0.31
Class 6
 0 (0)
 2
 –2.44
 15.73
Class 7
 118 (47)
 132
 3.29c
 0.30
Class 8
 40 (28)
 102
 2.40c
 0.33
Class 9
 106 (48)
 117
 3.31c
 0.31
Class 10
 46 (22)
 161
 2.12c
 0.32
Class 11
 16 (29)
 39
 3.34c
 0.43
Class 12
 1 (11)
 8
 2.17d
 1.13
Class 13
 90 (48)
 97
 3.30c
 0.32
Class 14
 44 (52)
 40
 4.00c
 0.36
Class 15
 178 (74)
 64
 4.46c
 0.32
Class 16
 7 (37)
 12
 3.61c
 0.59
Class 17
 25 (54)
 21
 3.99c
 0.43
Class 18
 181 (80)
 46
 4.82c
 0.32
Class 19
 105 (56)
 84
 3.60c
 0.31
Class 20
 46 (39)
 73
 3.45c
 0.35
Land-cover type
Savanna
 646 (39)
 1006
 Reference
Dry land crop land/pasture
 44 (39)
 68
 –0.38
 0.23
Irrigated crop land/pasture
 0 (0)
 9
 –7.02
 7.14
Cropland/grassland mosaic
 198 (53)
 174
 –0.07
 0.14
Cropland/woodland mosaic
 4 (4)
 105
 –1.97c
 0.56
Grassland
 123 (67)
 61
 0.36
 0.19
Shrub land
 66 (34)
 127
 –0.52c
 0.19
Urban
 3 (43)
 4
 0.32
 0.85
Broadleaf deciduous forest
 44 (39)
 68
 0.17
 0.23
Evergreen broadleaf forest
 37 (13)
 248
 0.07
 0.24
Water bodies
 20 (38)
 33
 –0.07
 0.33
Forest wetland
 0 (0)
 24
 –2.82
 4.55
Barren/sparsely vegetated
 46 (28)
 120
 –1.07c
 0.24
Variable removede
 Log likelihood
 Change in –2 likelihood
 df
 Significance of change
Absolute humidity profile
 –2045.937
 877.963
 19
 <0.001
Land-cover type –1642.564 71.217 12 <0.001

aExcludes five districts for which environmental data were unavailable (in Sinai, Egypt, and Pemba, Tanzania).
bp value <0.01.
cp value <0.05.
dSee Figure 1 for description of profile classes.
eModel if term removed.