Table 2. Logistic regression estimates of known HIV status and safer injecting practices in matched and unmatched samples of PWID, Kermanshah, west of Iran 2014 .

Behavioral outcomes	Pooled sample OR (95% CI)	Match sample OR (95% CI) a
Borrowing a syringe in past month
PWID unaware of HIV status	1.00	1.00
Aware HIV+ status	1.71 (1.30, 1.86)	1.68 (1.32, 2.81)
Aware HIV- status	0.51 (1.23, 1.77)	1.54 (1.28, 2.71)
Lending a syringe in past month
PWID unaware of HIV status	1.00	1.00
Aware HIV+ status	0.25 (0.18, 0.77)	0.22 (0.10, 0.84)
Aware HIV- status	0.34 (0.18, 0.64)	0.31 (0.15, 0.60)
Sharing a paraphernalia in past month
PWID unaware of HIV status	1.00	1.00
Aware HIV+ status	0.7 (0.42, 1.75)	0.68 (0.30, 2.32)
Aware HIV- status	0.89 (0.43, 2.04)	0.78 (0.47, 3.22)
Syringe reuse in past month
PWID don’t Know HIV status	1.00	1.00
To know HIV+ status	2.31(1.19,4.51)	2.38 (1.60, 4.68)
To know HIV- status	3.44 (1.27,5.72)	3.54 (1.30, 5.42)
Consistent condom use with any type of sexual partner
PWID unaware of HIV status	1.00	1.00
Aware HIV+ status	3.34 (1.70, 5.20)	3.22 (1.50, 4.80)
Aware HIV- status	3.24 (1.90, 4.86)	3.19 (1.50, 4.60)

^aThe matched subsample was made by considering age, income, education, access to NSP facilities and city of residence.