Skip to main content
. Author manuscript; available in PMC: 2017 May 1.
Published in final edited form as: AIDS Behav. 2016 May;20(5):1017–1025. doi: 10.1007/s10461-015-1186-8

Table II.

Differences by Sample Characteristics in HIV Testing and Treatment Behaviors among Key Populations in Tijuana, Mexico (N=191)

Measured Behaviors Observed Differences
Sample Method Gender MSM PWID SW SW-PWID


Previously tested for HIV χ2=4.813 ** χ2=0.979 χ2= 8.141 * χ2= 12.105 ** χ2= 2.253 χ2= 0.051
p=.028 p=.323 p=.004 p=.001 p=.133 p=.822
n=191 n=185 n=184 n=189 n=179 n=179
Community
(78/173), 45.1%
Female
(47/95), 49.5%
MSM
(26/38), 68.4%
PWID
(43/116), 37.1%
SW
(43/81), 53.1%
SW-PWID
(20/44), 45.5%
Clinic
(13/18), 72.2%
Male
(38/90), 42.2%
non-MSM
(62/146), 42.5%
non-PWID
(46/73), 63.0%
non-SW
(41/98), 41.8%
non-SW-PWID
(64/135), 47.4%
Knew HIV+ status χ2=10.497 ** χ2=5.152 ** χ2= 0.080 χ2= 1.305 χ2= 0.138 χ2= 0.022
p=.001 p=.023 p=.777 p=.253 p=.711 p=.881
n=95 n=89 n=91 n=93 n=88 n=88
Community
(16/82), 19.5%
Female
(8/49), 16.3%
MSM
(6/27), 22.2%
PWID
(14/47), 29.8%
SW
(9/45), 20.0%
SW-PWID
(5/22), 22.7%
Clinic
(8/13), 61.5%
Male
(15/40), 37.5%
non-MSM
(16/64), 25.0%
non-PWID
(9/46), 19.6%
non-SW
(10/43), 23.3%
non-SW-PWID
(14/66), 21.2%
Ever linked to HIV care χ2=3.471 χ2=0.038 χ2= 2.141 χ2= 4.212 ** χ2= 4.163 * χ2= 1.145
p=.062 p=.846 p=.143 p=.040 p=.041 p=.228
n=39 n=38 n=37 n=38 n=34 n=34
Community
(21/31), 67.7%
Female
(15/20), 75.0%
MSM
(5/5), 100.0%
PWID
(19/29), 65.5%
SW
(14/16), 87.5%
SW-PWID
(10/12), 83.3%
Clinic
(8/8), 100.0%
Male
(13/18), 72.2%
non-MSM
(22/32), 68.8%
non-PWID
(9/9), 100.0%
non-SW
(10/18), 55.6%
non-SW-PWID
(14/22), 63.6%

Based on available behavioral HIV data, participants across the pooled sample could be classified as follows: Community-derived (n=173; referent) vs. clinic-derived (n=18) sampling methods; Gender represented as female (n=95; referent) vs. male (n=90); MSM (n=38) vs. non-MSM (n=146), missing data (n=7); PWID (n=116) vs. non-PWID (n=73), missing data (n=2); sex workers regardless of gender (SW; n=81) vs. non-SW (n=98), missing data (n=12); SW-PWID (n=44) vs. non-SW-PWID (n=135), missing data (n=12). Proportions of each sub-group engaged in the target behavior are presented as: [(n1/Ntotal), %].

**

indicates a significantly smaller proportion of the referent group engaged in the behavior.

*

indicates a significantly larger proportion of the referent group engaged in the behavior.

Due to limited statistical representation, analysis of observed differences by gender excludes persons who identify as transgender (n=6 male-to-female).