Skip to main content
. 2011 May 27;6(5):e20104. doi: 10.1371/journal.pone.0020104

Table 5. Estimated effects of each other predictor when added singly to the primary model from Table 4.

Odds 95% confidence interval
Predictor Value Ratio Lower Upper P-value
Sex Male 1.21 0.94 1.57 0.14
Race/ethnicitya African American 0.79 0.62 1.01 0.055
(vs Caucasian) Hispanic 1.26 0.91 1.75 0.17
Other 1.23 0.85 1.79 0.27
HIV-infectedb Yes 1.17 0.64 2.1 0.61
HIV-infected, by treatment erab to 1995 0.68 0.17 2.7 0.58
(vs uninfected) 1996 on 2.1 0.97 4.4 0.059
HIV-infected, by treatment erab to 1995 0.37 0.00 568 0.79
(vs uninfected) 1996–1999 3.0 0.57 16.1 0.19
2000 on 1.49 0.34 6.6 0.60
Reported HCV risk factor Transfusion 0.99 0.72 1.37 0.96
(vs injection drug use) Needlestick 0.88 0.61 1.28 0.50
Other/None 1.21 0.96 1.53 0.11
Smoking Yes 1.07 0.83 1.40 0.59
Alcohol consumptionb,c Moderate 0.99 0.52 1.88 0.99
(vs none) Heavy 1.16 0.72 1.89 0.54
Injection drug useb Yes 0.92 0.69 1.23 0.58
Body Mass Index per 5 Kg/m2 1.05 0.97 1.14 0.24
Body Mass Index 25–30 1.09 0.88 1.36 0.42
(vs <25) >30 1.19 0.95 1.50 0.14
a

This is an alternative finer breakdown instead of an addition to the primary model.

b

These are time-varying covariates, with potentially differing values at each time step.

c

This model was fitted on HALS participants only, because other studies lacked complete histories.

Abbreviations: HIV: human immunodeficiency virus; HCV: hepatitis C virus; HALS: Hepatitis C and Alcohol Study [16].