Skip to main content
. Author manuscript; available in PMC: 2018 Aug 1.
Published in final edited form as: Cancer Epidemiol. 2017 Jun 9;49:118–127. doi: 10.1016/j.canep.2017.05.013

Table 1.

Comparison of geospatial clusteringa by counties within United States regions, 2004–2014.

Midwest countiesb (N = 1,055) Northeast counties b (N = 217) Southern countiesb (N = 1,422) Western counties b (N = 414) Total (N = 3108) P valuec
Presented as N (%)d N (%)e
LISA
 High-high cluster 112 (25.11) 2 (0.45) 328 (73.54) 4 (0.90) 446 (14.35) <0.01
 High-low outlier 16 (34.78) 1 (2.17) 6 (13.04) 23 (50.00) 46 (1.48)
 Low-high outlier 13 (28.26) 0 (0.00) 32 (69.57) 1 (2.17) 46 (1.48)
 Low-low cluster 94 (20.75) 26 (5.74) 129 (28.48) 204 (45.03) 453 (14.58)
 Non- significant 820 (38.79) 188 (8.89) 927 (43.85) 182 (8.61) 2116 (68.08)
EB
 1st Quintile 155 (24.92) 34 (5.47) 202 (32.48) 231 (37.14) 622 (20.01) <0.01
 2nd Quintile 250 (40.26) 68 (10.95) 215 (34.62) 88 (14.17) 621 (19.98)
 3rd Quintile 246 (39.55) 69 (11.09) 269 (43.25) 38 (6.11) 622 (20.01)
 4th Quintile 223 (35.91) 40 (6.44) 325 (52.33) 33 (5.31) 621 (19.98)
 5th Quintile 181 (29.10) 6 (0.96) 411 (66.08) 24 (3.86) 622 (20.01)
Gi* (%)
 Lung cancer hotspot 42 (18.58) 70 (30.97) 65 (28.76) 49 (21.68) 226 (7.27) <0.01
 Lung cancer cold spot 337 (53.58) 0 (0.00) 148 (23.53) 144 (22.89) 629 (20.24)
 Non- significant 676 (30.04) 147 (6.53) 1209 (53.73) 221 (9.82) 2252 (72.46)
Lung cancer mortality cluster category
 Clustered 91 (25.14) 2 (0.55) 264 (72.93) 5 (1.38) 362 (11.65) <0.01
 Non- clustered 964 (35.11) 215 (7.83) 1158 (42.17) 409 (14.89) 2746 (88.35)
a

Defined as counties estimated as clustered by two of three cluster methods (local indicators of spatial autocorrelation, empirical Bayes, and Gi*)

b

US regions as determined by the US Census Bureau.

c

Significance determined using chi-square test.

d

Denotes row percentages.

e

Denotes column percentages.

LISA, local indicators of spatial association; EB, empirical Bayes.