Table 2.

Results of maximum likelihood analysis of main / internal validation study data on 924 women (n_m = 624; n_v = 300) at their 4th study visit: Estimates of primary model parameters

Assuming differential misclassificationa
Variable	$\widehat{\beta }$ (std. error)	Estimated OR (95% CI)
Race/Ethnicity (Black vs. White/Hispanic)	1.13 (0.22)	3.10 (2.01 – 4.77)
Risk Group (IDU vs. sex)	0.62 (0.24)	1.86 (1.17 – 2.95)
HIV Status (positive vs negative)	0.20 (0.25)	1.22 (0.75 – 1.98)
Age (years)	−0.07 (0.02)	0.94 (0.90 – 0.97)

Assuming non-differential misclassificationb
Variable	$\widehat{\beta }$ (std. error)	Estimated OR (95% CI)
Race/Ethnicity (Black vs. White/Hispanic)	1.16 (0.22)	3.18 (2.07 – 4.86)
Risk Group (IDU vs. sex)	0.98 (0.21)	2.67 (1.76 – 4.04)
HIV Status (positive vs negative)	−0.10 (0.22)	0.90 (0.59 – 1.38)
Age (years)	−0.08 (0.02)	0.93 (0.89 – 0.96)	< 0.001

^aSensitivity and specificity assumed to vary with the binary variables HIV risk group, HIV status, and Age (> median vs. < median), via model (14)

^bNo covariates affecting sensitivity and specificity; this assumption is not supported by the data (P<0.001)