Skip to main content

Table 4 All inclusive, additive logistic regression models of demographic and hematological factors as predictors of DENV infection

From: Maintenance of demographic and hematological profiles in a long-lasting dengue fever outbreak: implications for management

   Dataset     
  Statistics ALL s1 s2 s3 s4
  AIC 412.1 97.5 107.1 103.4 98.1
Factors βo 30.194 54.258 30.300 31.291 15.871
Sex β1 −0.007 −1.006 0.1092 0.3120 0.7480
  OR 0.993 0.366 1.115 1.366 2.113
  CL- 0.6035 0.1079 0.3908 0.4545 0.6700
  CL+ 1.6342 1.2385 3.1827 4.1066 6.6626
  P 0.9784 0.1060 0.8383 0.5785 0.2018
Day β1 0.002 −0.050 −0.007 0.197 0.073
  OR 1.002 0.951 0.993 1.218 1.076
  CL- 0.9900 0.8872 0.8854 1.0630 1.0049
  CL+ 1.0145 1.0197 1.1132 1.3953 1.1520
  P 0.7310 0.1584 0.9012 0.0045 0.0356
Age β1 0.011 0.006 0.035 0.018 −0.028
  OR 1.011 1.006 1.036 1.018 0.973
  CL- 0.9985 0.9786 1.0055 0.9887 0.9426
  CL+ 1.0245 1.0339 1.0666 1.0488 1.0038
  P 0.0831 0.6755 0.0202 0.2274 0.0847
LEU β1 −4.518 −6.303 −2.310 −4.632 −6.838
  OR 0.011 0.002 0.099 0.010 0.001
  CL- 0.0025 0.0000 0.0049 0.0004 0.0000
  CL+ 0.0477 0.0840 20.160 0.2510 0.0430
  P <0.0001 0.0012 0.1327 0.0052 0.0003
PLA β1 −2.536 −5.556 −4.194 −4.392 0.947
  OR 0.079 0.004 0.015 0.012 2.578
  CL- 0.0150 0.0000 0.0003 0.0003 0.1099
  CL+ 0.4192 0.3552 0.7826 0.5545 60.4591
  P 0.0029 0.0160 0.0374 0.0236 0.5564
AL% β1 0.040 0.034 0.073 −0.090 0.262
  OR 1.041 1.035 1.076 0.914 1.299
  CL- 0.9828 0.9227 0.9101 0.8157 1.104
  CL+ 1.1029 1.1603 1.2724 1.0245 1.529
  P 0.1703 0.5592 0.3909 0.1227 0.0017
  1. All-inclusive, additive logistic regression models to evaluate the contribution of demographic and hematological factors to DENV incidence using different datasets (ALL = entire data; subsets s1 to s4 ordered by day of the first attendance at the Marilia Hemocenter; see text). AIC indicate model fit (comparable only between subsets s1 to s4). βo = intercept, β1 = slope, OR = odds ratio, and CL = 95 % confidence limits. Figures lower than critical P were in bold. Considering five models (one for each dataset), critical P = 0.05/5 = 0.01. Variable codes as in Table 1