Skip to main content
. Author manuscript; available in PMC: 2010 Apr 14.
Published in final edited form as: Cancer Prev Res (Phila). 2008 Sep;1(4):255–265. doi: 10.1158/1940-6207.CAPR-08-0082

Table 5.

Distribution of demographic, smoking, and clinical variables for internal and external validation data sets

Internal validation
External validation
Characteristic Cases (n = 89) Controls (n = 67) P* Cases (n = 172) Controls (n = 153) P*
% Male 55 39 0.004 13 22 0.04
Age (mean ± SD)
    Males 62.9 ± 8.9 47.6 ± 12.6 <0.0001 46.1 ± 3.0 44.1 ± 4.3 0.05
    Females 65.2 ± 8.7 50.9 ± 15.1 <0.0001 55.2 ± 9.7 51.5 ± 10.7 0.003
Smoking status, n (%)
    Never 10 (11.3) 27 (40.3) <0.0001 - -
    Former 38 (42.7) 12 (17.9) 148 (86.1) 66 (43.1)
    Current 41 (46.0) 28 (41.8) 24 (13.9) 87 (56.9) <0.0001
Asbestos, n (%) 35 (39.3) 15 (22.4) 0.025 37 (21.5) 20 (13.1) 0.05
Dusts, n (%) 20 (22.5) 16 (23.9) 0.84 11 (6.4) 11 (7.2) 0.78
COPD, n (%) 19 (21.3) 1 (1.5) <0.001 14 (8.1) 4 (2.6) 0.03
Hay fever, n (%) 7 (7.9) 10 (14.9) 0.16 50 (29.1) 50 (32.7) 0.48
Risk model validation
AUC (95% CI)
P
AUC (95% CI)
P
Overall 0.75 (0.67−0.82) 0.63 (0.57−0.69)
Gender
    Males 0.70 (0.58−0.82)
    Females 0.75 (0.64−0.85) 0.61 (0.54−0.67)
Early onset 0.64 (0.64−0.74)
Ever smokers
    African American model 0.79 (0.70−0.88) 0.66 (0.61−0.71)
    Spitz model 0.67 (0.57−0.77) 0.001 0.59 (0.52−0.65) 0.028
*

P-value from Pearson's Chi-square test (categorical variable) or two-sample t-test (continuous variable).

For the external validation, group allergies were used as a proxy for hay fever.

P-value for the comparison of two AUCs (16).