Comparison of HIV Status Between Transgender Women and Men Who Have Sex With Men: A Meta-Analysis

Feature
Comparison of HIV Status Between

Transgender Women and Men Who Have
Sex with Men: A Meta-Analysis
Tian-Zhang Song, PhD
Ming-Xu Zhang, PhD
Han-Dan Zhang, MS
Yu Xiao, PhD
Wei Pang, PhD
Yong-Tang Zheng, PhD*
As key population groups in HIV infection, trans-

gender women (TGW) share the same oral and anal Tian-Zhang Song, PhD, is a student, Key Laboratory of An-
sexual practices with men who have sex with men imal Models and Human Disease Mechanisms of the Chi-
(MSM). However, the differences in HIV rates between nese Academy of Sciences/Key Laboratory of Bioactive
the two groups are still unclear. In our study, relevant Peptides of Yunnan Province, Kunming Institute of
publications were electronically searched in three da- Zoology, Chinese Academy of Sciences, Kunming, Yunnan,
tabases. After excluding irrelevant studies based on China, and at the University of Chinese Academy of Sci-
review of the title, abstract, and full text, a total of ences, Beijing, China. Ming-Xu Zhang, PhD, is a student,
24 studies, including 37,521 cases, were identified. Key Laboratory of Animal Models and Human Disease
There were no significant differences between TGW Mechanisms of the Chinese Academy of Sciences/Key Lab-
and MSM regarding HIV status and condom use. How- oratory of Bioactive Peptides of Yunnan Province, Kunm-
ing Institute of Zoology, Chinese Academy of Sciences,
ever, TGW had a higher proportion of high-risk behav-
Kunming, Yunnan, China. Han-Dan Zhang, MS, is a stu-
iors, including exchanges of sex, forced sex, receptive
dent, College of Pharmacy and Chemistry, Dali University,
intercourse, and sex work experience. Moreover, most Dali, China. Yu Xiao, PhD, is an Associate Professor, Key
curable sexually transmitted diseases were more prev- Laboratory of Animal Models and Human Disease Mecha-
alent in the TGW group compared to MSM. Although nisms of the Chinese Academy of Sciences/Key Laboratory
TGW showed a higher proportion of high-risk of Bioactive Peptides of Yunnan Province, Kunming Insti-
behavior and sexually transmitted disease incidence, tute of Zoology, Chinese Academy of Sciences, Kunming,
their HIV incidence was the same as that for MSM. Yunnan, China. Wei Pang, PhD, is an Associate Professor,
Key Laboratory of Animal Models and Human Disease
(Journal of the Association of Nurses in AIDS Care, Mechanisms of the Chinese Academy of Sciences/Key Lab-
-, 1-17) Copyright Ó 2018 Association of Nurses oratory of Bioactive Peptides of Yunnan Province, Kunm-
in AIDS Care ing Institute of Zoology, Chinese Academy of Sciences,
Kunming, Yunnan, China. Yong-Tang Zheng, PhD, is a Pro-
Keywords: condom, HIV, men, transgender
fessor, Key Laboratory of Animal Models and Human Dis-
AIDS caused by HIV was first recognized in 1981 ease Mechanisms of the Chinese Academy of Sciences/Key
Laboratory of Bioactive Peptides of Yunnan Province,
and has emerged from being a concern for ‘‘high- Kunming Institute of Zoology, Chinese Academy of Sci-
risk’’ groups to representing a worldwide pandemic. ences, Kunming, Yunnan, China. (*Correspondence to:
At the end of 2016, there were approximately 36.7 [email protected]).
JOURNAL OF THE ASSOCIATION OF NURSES IN AIDS CARE, Vol. -, No. -, -/- 2018, 1-17
https://doi.org/10.1016/j.jana.2018.04.001
Copyright Ó 2018 Association of Nurses in AIDS Care
2 JANAC Vol. -, No. -, -/- 2018
million people living with HIVaround the globe and an higher than, MSM. However, this conclusion lacked
estimated 1.0 million deaths during that year (World direct and compelling evidence. Although several
Health Organization [WHO], 2016a). Anal intercourse, studies comparing MSM and TGW have been reported,
intravenous drug injection, and maternal–neonatal most results have been conflicting and confusing.
transmission are the three most important routes of Bowers, Branson, Fletcher, and Reback (2012) and
HIV transmission. In addition, five key populations Subramanian and colleagues (2013) found a higher
who experience a significant HIV burden and influence HIV prevalence in MSM compared to TGW. However,
the dynamics of HIV epidemics were confirmed in the Griensven and colleagues (2006) and Ham, Northbrook,
WHO HIV guidelines (WHO, 2016b), including (a) Morales-Miranda, Guardado, and Kamb (2015) re-
men who have sex with men (MSM), (b) people who ported higher HIV prevalences in TGW groups.
inject drugs, (c) people in prisons and other closed set- Approximately the same prevalence in the two groups
tings, (d) sex workers, and (e) transgender people. was also reported by Castillo and colleagues (2015)
The umbrella term ‘‘transgender’’ describes per- and Chakrapani, Newman, Shunmugam, Logie, and
sons whose gender identity and expression differ Samuel (2017). It is still difficult to confirm the status
from the biological sex assigned at birth. It includes of the two groups in HIV prevalence. In our meta-
people who are transgender, transsexual, or otherwise analysis, a systemic searching procedure was performed
considered gender nonconforming. Therefore, people to compare differences between TGWand MSM groups
who were male at birth and who identify themselves in HIV status, high-risk behavior, STDs, condom use,
as female, trans, or bisexual were initially identified and other information.
as male-to-female transgender or transgender women
(TGW; WHO, 2016b). Historically, TGW were re- Methods
garded as a subgroup of MSM due to having the
same biological sex and engaging in anal sexual prac- Literature Search Strategy
tices. Ten percent of MSM were thought to be TGW
(Soto et al., 2007). However, in recent years, scien- A systematic review of the literature was performed
tists recognized the existence of differences between in March 2017 without restriction to regions, publica-
these groups, and epidemiological studies from 2010 tion types, or languages. Relevant publications were
started to compare the HIV status, sexual risk behav- electronically searched in three databases: PubMed,
iors, sexually transmitted diseases (STDs), social net- Embase, and the Cochrane Library. The following
works, and other basic characteristics between them MeSH terms and their combinations were searched
in order to better understand TGW. Finally, in 2014, in [Title/Abstract]: (HIV* or AIDS* or Human Immu-
WHO (2014) declared that the high vulnerability nodeficiency Virus or Acquired Immune Deficiency
and specific health needs of transgender people Syndrome) and (man or men or gay* or homo$sexual*
necessitated a distinct and independent status in the or male) and (trans$gender* or trans$sexual* or
global HIV response. From then on, transgender peo- LGBT* or trans$woman or trans$women or hijra* or
ple were identified as an independent group of HIV kathoey* or waria* or travesty*). In addition, the refer-
key populations, which resulted in more attention. ence lists of all retrieved original studies and relevant
According to the WHO guidelines (WHO, 2016b), reviews related to the regimen were identified for
the epidemic of HIV in MSM has been, on average, potentially inclusive trials, and related article functions
13 times greater than in the general population. Mean- were used to broaden the search. When multiple pub-
while, a meta-analysis published in 2013 indicated lications describing the same population were identi-
that the pooled global HIV prevalence among TGW fied, the most recent or complete report was included.
was 19.1%, with approximately 50 times the odds of
infection compared to the general population (Baral Selection Criteria
et al., 2013). Therefore, WHO (http://www.who.int/
hiv/topics/en/) declared that sexual risk in TGW may All available randomized controlled trials (RCTs)
be higher than in MSM, so the prevalence of HIV in and nonrandomized controlled trials (cohort or cross-
the TGW group in many countries was as high as, or control studies) that compared MSM to TGW were
Song et al. / Comparison of HIV Status Between Transgender Women and MSM 3
included. Editorials, letters to the editor, review arti- search. After excluding irrelevant studies based on re-
cles, case reports, and animal experimental studies view of the title, abstract, and full text, 24 studies ful-
were excluded. During this process, the primary inves- filled the predefined inclusion criteria and were
tigators independently reviewed and reported the re- included in the final analysis (Figure 1;
sults of the studies, and any disagreements were Andrinopoulos et al., 2015; Barrington, Wejnert,
resolved by an external content reviewer. Guardado, Nieto, & Bailey, 2012; Bowers et al.,
The primary outcomes were HIV status, high-risk 2012; Castillo et al., 2015; Castro et al., 2016;
behavior (including exchange of sex, forced sex, sex- Chakrapani et al., 2017; Chariyalertsak et al., 2011;
ual orientation, and sexual work), condom use Chemnasiri et al., 2010; Deutsch et al., 2015;
(including diligent condom use and regular unpro- Fernandes et al., 2015; Griensven et al., 2006; Ham
tected sexual behavior), and STD (including curable et al., 2015; Miller et al., 2011; Newman, Cameron,
STD, anal chlamydia, anal gonorrhea, herpes simplex Roungprakhon, Tepjan, & Scarpa, 2016; Poteat,
virus [HSV]-2, and syphilis). The secondary outcomes German, & Flynn, 2016; Sanchez, Finlayson,
were basic demographics (including age, education Murrill, Guilin, & Dean, 2010; Stahlman et al.,
level, living situation, marital status, working status) 2016; Subramanian et al., 2013; Tucker, Arandi,
and substance use (including drugs and alcohol). Bolanos, Paz-Bailey, & Barrington, 2014; Verre
et al., 2014; Wiewel, Torian, Merchant, Braunstein,
Quality Assessment and Statistical Analysis
& Shepard, 2016; Woodford, Chakrapani, Newman,
& Shunmugam, 2015; Yang et al., 2013; Zhang
The level of evidence was identified according to
et al., 2016). All were full-text articles except Ham
criteria by the Centre for Evidence-Based Medicine in
and colleagues (2015) and Griensven and
Oxford, UK (Phillips et al., 2009). For all outcomes,
colleagues (2006), which were identified as posters.
Review Manager 5.0 (RevMan 5.0, Cochrane Collabo-
Between-reviewer agreement was 96.2% for study
ration, Oxford, UK) was used to perform meta-
selection and 88% for quality assessment of trials.
analyses. The events in this article were all dichotomous
The characteristics of the included studies are
variables, which were compared using odds ratios (OR)
shown in Table 1. Participants in all studies self-
and weighted mean difference (p # .05 was considered
identified their sexual orientation as MSM or TGW.
statistically significant). All results were reported with
Information about HIV status, STD, condom use,
95% confidence intervals (CIs). The chi-squared test
substance use, high-risk behavior, attention to HIV,
with significance set at p , .10 (Chen, Yu, & Shao,
and basic characteristics were shown and compared
2015; Shey, Kongnyuy, Alobwede, & Wiysonge,
between TGW and MSM groups. Among the
2013) was used to assess statistical heterogeneity
included studies, there were only two RCTs, both
between studies, which was quantified using the I2
of low quality (Castillo et al., 2015; Deutsch et al.,
statistic. The fixed-effects model was used if I2 ,
2015). Nonrandom sampling methods were used in
50%, which would mean that no heterogeneity was
the remaining 22 studies (including respondent-
present. However, if I2 . 50 %, which would mean
driven sampling, purposive sampling, venue-based
that heterogeneity was detected, the random-effects
sampling, convenience sampling, and time-location
model was chosen (Higgins & Green, 2011). A sensi-
cluster sampling).
tivity analysis was performed to evaluate the robustness
of these conclusions. Finally, potential publication bias
Primary Outcomes
was assessed using funnel plots.
HIV status. Data from 15 studies (Barrington
Results et al., 2012; Bowers et al., 2012; Castillo et al.,
2015; Castro et al., 2016; Chakrapani et al., 2017;
Characteristics of Eligible Studies Chariyalertsak et al., 2011; Chemnasiri et al., 2010;
Deutsch et al., 2015; Fernandes et al., 2015;
A total of 1,973 studies were identified from the Griensven et al., 2006; Ham et al., 2015; Poteat
three databases during the initial systematic literature et al., 2016; Sanchez et al., 2010; Subramanian
4 JANAC Vol. -, No. -, -/- 2018
PubMed: n = 623 Embase: n = 1,320 Cochrane: n = 30
Studies identified through initial

Searches of electronic databases:
n = 1,973
Duplication: n = 661
Titles and abstracts screened:

N = 1,312
Excluded studies: n = 1,203
-Vitro experiments: n = 128
-A single patient: n = 77
-Irrelevant topics: n = 998
Full-text articles screened:
n = 109 Excluded studies: n = 85
-No comparison: n = 47
-Review or meeting abstracts: n = 22
-Editorials, letters, or short reviews: n = 5
Included studies: -Duplicated reports: n = 11
n = 24
Figure 1. Flow diagram of article selection strategy.
et al., 2013; Verre et al., 2014) that included these two groups, as shown in Figure 3B. TGW had
individuals (total of 12,537 in the MSM groups and a significantly higher proportion of forced sex than
3,422 in the TGW groups) were used in the meta- MSM (OR 0.46, 95% CI 0.30 to 0.71, p 5 .0005).
analysis of HIV status. There were no significant dif- Five studies (Chariyalertsak et al., 2011;
ferences between these two groups, as shown in Chemnasiri et al., 2010; Griensven et al., 2006;
Figure 2 (OR 0.99, 95% CI 0.71 to 1.39, p 5 .97). Stahlman et al., 2016; Yang et al., 2013) including
Thus, MSM and TGW had the same rates of HIV 4,116 individuals evaluated the proportion of
infection. individuals who only engaged in receptive anal sex.
The pooled data showed a significantly higher rate
Sexual risk behavior. The differences between in the TGW group, as shown in Figure 3C (OR
MSM and TGW in HIV risk behavior (including ex- 0.07, 95% CI 0.05 to 0.1, p , .00001). As shown
change of sex, forced sex, sexual position, and sexual in Figure 2D, pooling data from nine studies
work) were analyzed, and the results are shown in (Chakrapani et al., 2017; Chariyalertsak et al.,
Figure 3. Pooling the data from four articles 2011; Deutsch et al., 2015; Griensven et al., 2006;
(Andrinopoulos et al., 2015; Chariyalertsak et al., Stahlman et al., 2016; Subramanian et al., 2013;
2011; Fernandes et al., 2015; Sanchez et al., 2010) Tucker et al., 2014; Woodford et al., 2015; Zhang
that assessed a total of 1,853 individuals showed et al., 2016), the sexual work history in these two
significant differences between the MSM and TGW groups was assessed. Compared to the MSM group,
groups, as shown in Figure 3A (OR 0.22, 95% CI a significantly higher proportion of sexual work
0.07 to 0.75, p 5 .02). The individuals in the TGW history was shown in the TGW group (OR 0.29,
group had an obviously higher proportion of ex- 95% CI 0.15 to 0.54, p 5 .0001).
changes of sexual behavior than those in the MSM
group. The data from five studies (Chemnasiri Condom use. A questionnaire about the fre-
et al., 2010; Fernandes et al., 2015; Miller et al., quency of condom use during sexual behavior was
2011; Stahlman et al., 2016; Zhang et al., 2016) completed in six studies (Figure 4A; Barrington
assessed the history of being forced to have sex in et al., 2012; Chakrapani et al., 2017; Chariyalertsak
Table 1. Characteristics of Included Studies (n 5 24)
Study (First Level of Total

Author, Year) Evidence Participants Location Group Main Index
Andrinopoulos, 2015 4 670 El Salvador MSM, TGW and MSM Condom use, High risk behavior, Attention
to HIV, Basic characters, STD
Barrington, 2012 4 602 El Salvador TGW, MSM, HIV status, Condom use, Substance use,
Hetero- and Bisexual Attention to HIV, Basic characters
Bowers, 2012 4 912 USA TGW, MSM/W, MSM HIV status, Basic characters
Castillo, 2015 2b 662 Peru TGW, MSM HIV status, STD
Castro, 2016 4 793 Brazil TGW, MSM HIV status, STD, Attention to HIV, Basic
characters, Substance use
Chakrapani, 2017 4 600 India TGW, MSM HIV status, Condom use, Substance use,
High risk behavior, Basic characters
Chariyalertsak, 2011 4 551 Thailand TGW, MSM, Bisexual HIV status, STD, Condom use, Substance
use, High risk behavior, Attention to
HIV, Basic characters
Chemnasiri, 2010 4 827 Thailand TGW, MSM, MSW HIV status, STD, Substance use, High risk
behavior, Attention to HIV, Basic
characters
Deutsch, 2015 2b 2,499 Brazil, TGW, MSM HIV status, STD, Condom use, High risk
Ecuador, behavior, Basic characters
Peru, Thailand,
and USA
Fernandes, 2015 4 430 Brazil TGW, MSM HIV status, Condom use, STD, Substance
use, High risk behavior, Basic characters
Griensven, 2006 4 2,049 Thailand TGW, MSM, MSW HIV status, Condom use, Substance use,
High risk behavior
Ham, 2015 4 2,727 El Salvador, TGW, MSM HIV status, STD
Guatemala,
Honduras, and
Nicaragua
Miller, 2011 4 608 Guatemala TGW, MSM Attention to HIV, High risk behavior
Newman, 2016 4 408 Thailand TGW, YMSM Basic characters
Poteat, 2016 4 694 USA TGW, MSM HIV status, Substance use, Attention to
HIV, Basic characters
Sanchez, 2010 4 361 USA TGW, MSM HIV status, Condom use, Substance use,
High risk behavior, Basic characters
Stahlman, 2016 4 2,246 Coate d’Ivoire TGW, MSM High risk behavior, Basic characters
Subramanian, 2013 4 2,024 India TGW, MSM HIV status, STD, High risk behavior, Basic
characters
Tucker, 2014 4 29 Guatemala TGW, MSM High risk behavior, Basic characters
Verre, 2014 4 5,148 Peru TGW, gay/homosexual HIV status, STD
Wiewel, 2016 4 10,935 USA TGW, MSM Basic characters
Woodford, 2015 4 84 India TGW, MSM, DU, FSW High risk behavior, Attention to HIV, Basic
characters
Yang, 2013 4 238 Thailand TGW, MSM Substance use, High risk behavior,
Attention to HIV, Basic characters
Zhang, 2016 4 1,424 China Non-TGW, TGW Condom use, High risk behavior, Attention
to HIV, Basic characters
Note. MSM 5 men who have sex with men; TGW 5 transgender women; MSM/W 5 men who have sex with men and women;
MSW 5 men who have sex with women; YMSM 5 young men who have sex with men; DU 5 drug users; FSW 5 female sex workers;
STD 5 sexually transmitted disease; Basic characters 5 basic characteristics of the enrolled participants, including age, education level,
living situation, marital status, and working status.
Level of evidence based on Oxford Centre for Evidence-based Medicine–Levels of Evidence. Level 2b 5 Individual cohort study
(including low-quality randomized controlled trial; e.g., , 80% follow-up); Level 4: 5 Case-series (and poor-quality cohort and case-
control studies).
6 JANAC Vol. -, No. -, -/- 2018
et al., 2011; Deutsch et al., 2015; Fernandes et al., 0.42 to 0.96, p 5 .03). TGW were more likely than
2015; Griensven et al., 2006). The pooled data MSM to get an STD. Two studies (Andrinopoulos
representing the number of individuals whose et al., 2015; Ham et al., 2015) independently
answer was always showed no significant difference reported the rate of individuals with anal
between TGW and MSM groups (OR 1.24, 95% CI chlamydia, anal gonorrhea, or HSV-2 in the TGW
0.69 to 2.20, p 5 .47). The articles in Figure 4B group and MSM group. The meta-analyses of these
(Andrinopoulos et al., 2015; Chariyalertsak et al., three indexes are shown in Figures 5B-D, respec-
2011; Deutsch et al., 2015; Sanchez et al., 2010; tively. There were significant differences in anal
Zhang et al., 2016) showed the history of ever gonorrhea infection and HSV-2 infection between
having engaged in unprotected sexual behavior. MSM and TGW groups (OR 0.63, 95% CI 0.43 to
Notably, there were no significant differences 0.91, p 5 .02, and OR 0.36, 95% CI 0.26 to 0.50,
between the TGW and MSM groups (OR 0.67, 95% p , .00001). TGW were more likely than MSM to
CI 0.32 to 1.41, p 5 .29). be infected with gonorrhea and HSV-2. However,
there was no difference between these two groups
Sexually transmitted disease. Pooling the data in anal chlamydia infection (OR 0.70, 95% CI 0.39
from six studies (Castro et al., 2016; Chariyalertsak to 1.25, p 5 .23). Pooling data from three studies
et al., 2011; Chemnasiri et al., 2010; Deutsch et al., (Fernandes et al., 2015; Ham et al., 2015; Verre
2015; Ham et al., 2015; Subramanian et al., 2013) et al., 2014), 3,088 individuals were assessed
that assessed curable STD (including chlamydia, regarding the rate of syphilis infection. TGW had a
gonorrhea, syphilis) in 8,449 individuals showed higher rate of having been infected with syphilis
clear differences between the MSM and TGW compared to MSM, as shown in Figure 5E (OR
groups, as shown in Figure 5A (OR 0.64, 95% CI 0.44, 95% CI 0.35 to 0.55, p , .00001).
MSM TGW Odds Ratio Odds Ratio

Study or Subgroup Events Total Events Total Weight M-H. Random. 95% Cl M-H. Random. 95% Cl
Barrington, 2012 31 279 13 67 6.3% 0.52 [0.25, 1.06]
Bowers, 2012 132 371 54 255 8.0% 2.06 [1.42, 2.97]
Castillo, 2015 83 459 35 207 7.7% 1.08 [0.70, 1.67]
Castro, 2016 123 756 8 37 5.8% 0.70 [0.31, 1.58]
Chakrapani, 2017 30 247 28 257 7.1% 1.13 [0.65, 1.95]
Chariyalertsak, 2011 51 309 13 140 6.6% 1.93 [1.01, 3.68]
Chemnasiri, 2010 38 274 24 241 7.2% 1.46 [0.85, 2.51]
Deutsch, 2015 38 2160 3 339 4.2% 2.01 [0.62, 6.53]
Fernandes, 2015 25 278 37 152 7.1% 0.31 [0.18, 0.53]
Griensven, 2006 16 314 1 180 2.1% 9.61 [1.26, 73.09]
Ham, 2015 232 2247 67 321 8.2% 0.44 [0.32, 0.59]
Poteat, 2016 243 645 21 49 6.9% 0.81 [0.45, 1.45]
Sanchez, 2010 63 301 11 60 6.3% 1.18 [0.58, 2.40]
Subramanian, 2013 219 1620 39 403 8.0% 1.46 [1.02, 2.09]
Verre, 2014 221 2277 102 714 8.4% 0.64 [0.50, 0.83]
Total (95% Cl) 12537 3422 100.0% 0.99 [0.71, 1.39]

Total events 1545 456
Heterogeneity: Tau2 = 0.33; Chi2 = 90.49, df = 14 (P < 0.00001); I2 = 85%
Test for overall effect: Z = 0.04 (P = 0.97)
Favors TGW Favors MSM
Figure 2. Forest plot and meta-analysis of HIV status between MSM and TGW. Note. M-H 5 Mantel-Haenszel method;
MSM 5 men who have sex with men; TGW 5 transgender women; Events 5 number of participants infected with HIV in MSM
or TGW group.
A MSM TGW Odds Ratio Odds Ratio

Andrinopoulos, 2015 168 506 102 164 25.7% 0.30 [0.21, 0.44]
Chariyalertsak, 2011 65 309 27 140 25.2% 1.11 [0.68, 1.84]
Fernandes, 2015 44 278 115 152 25.2% 0.06 [0.04, 0.10]
Sanchez, 2010 17 250 21 54 24.0% 0.11 [0.05, 0.24]
Total (95% Cl) 1343 510 100.0% 0.22 [0.07, 0.75]

B MSM TGW Odds Ratio Odds Ratio

Chemnasiri, 2010 67 274 70 241 21.9% 0.79 [0.53, 1.17]
Fernandes, 2015 35 259 35 142 19.2% 0.48 [0.28, 0.81]
Miller, 2011 33 509 14 99 16.3% 0.42 [0.22, 0.82]
Stahlman, 2016 131 1793 99 452 23.9% 0.28 [0.21, 0.37]
Zhang, 2016 274 1363 21 61 18.7% 0.48 [0.28, 0.83]
Total (95% Cl) 4198 995 100.0% 0.46 [0.30, 0.71]

Heterogeneity: Tau2 = 0.18; Chi2 = 18.23, df = 4 (P = 0.001); I2 = 78%
C MSM TGW Odds Ratio Odds Ratio

Chariyalertsak, 2011 69 276 109 118 14.1% 0.03 [0.01, 0.06]
Chemnasiri, 2010 105 274 219 241 19.3% 0.06 [0.04, 0.10]
Griensven, 2006 255 821 390 474 25.0% 0.10 [0.07, 0.13]
Stahlman, 2016 84 1297 206 377 24.4% 0.06 [0.04, 0.08]
Yang, 2013 32 131 79 107 17.2% 0.11 [0.06, 0.21]
Total (95% Cl) 2799 1317 100.0% 0.07 [0.05, 0.10]

Test for overall effect: Z = 13.72 (P < 0.00001)
D MSM TGW Odds Ratio Odds Ratio

Chakrapani, 2017 13 300 79 300 12.2% 0.13 [0.07, 0.23]
Chariyalertsak, 2011 7 309 8 140 10.1% 0.38 [0.14, 1.08]
Deutsch, 2015 810 2160 217 339 13.6% 0.34 [0.27, 0.43]
Griensven, 2006 183 821 289 470 13.5% 0.18 [0.14, 0.23]
Stahlman, 2016 658 1778 216 449 13.6% 0.63 [0.51, 0.78]
Subramanian, 2013 86 1620 165 403 13.4% 0.08 [0.06, 0.11]
Tucker, 2014 8 13 7 8 4.7% 0.23 [0.02, 2.46]
Woodford, 2015 15 26 2 21 7.1% 12.95 [2.48, 67.57]
Zhang, 2016 69 1363 13 48 11.9% 0.14 [0.07, 0.28]
Total (95% Cl) 8390 2178 100.0% 0.29 [0.15, 0.54]

Figure 3. Forest plot and meta-analysis of HIV high-risk behavior between MSM and TGW: (A) exchanges of sexual behavior, (B)
forced sex behavior, (C) only receptive during sexual behavior, (D) sex work. Note. M-H 5 Mantel-Haenszel method; MSM 5 men
who have sex with men; TGW 5 transgender women; CI 5 confidence interval.
8 JANAC Vol. -, No. -, -/- 2018

Barrington, 2012 176 279 50 67 15.2% 0.58 [0.32, 1.06]
Chakrapani, 2017 133 300 152 300 17.2% 0.78 [0.56, 1.07]
Chariyalertsak, 2011 99 276 33 118 16.2% 1.44 [0.90, 2.31]
Deutsch, 2015 965 2160 49 290 17.2% 3.97 [2.89, 5.46]
Fernandes, 2015 107 277 44 148 16.5% 1.49 [0.97, 2.28]
Griensven, 2006 317 821 199 474 17.7% 0.87 [0.69, 1.09]
Total (95% Cl) 4113 1397 100.0% 1.24 [0.69, 2.20]


Andrinopoulos, 2015 198 506 63 164 20.6% 1.03 [0.72, 1.48]
Chariyalertsak, 2011 177 276 85 118 19.9% 0.69 [0.43, 1.11]
Deutsch, 2015 1195 2160 290 339 20.8% 0.21 [0.15, 0.29]
Sanchez, 2010 126 301 20 59 19.1% 1.40 [0.78, 2.52]
Zhang, 2016 398 1363 23 61 19.5% 0.68 [0.40, 1.16]
Total (95% Cl) 4606 741 100.0% 0.67 [0.32, 1.41]

Figure 4. Forest plot and meta-analysis of condom use between MSM and TGW: (A) always condom use, (B) always condomless
sexual behavior. Note. M-H 5 Mantel-Haenszel method; MSM 5 men who have sex with men; TGW 5 transgender women;
CI 5 confidence interval.
Secondary Outcomes of homelessness (Figure 6C) and showed no differ-

ence between MSM and TGW groups (OR 0.82,
Basic characteristics. The basic characteristics 95% CI 0.22 to 3.08, p 5 .77). Additionally, a com-
of individuals in MSM and TGW groups, including parison of the living situation in the groups showed
age, education level, living situation, marital status, that a higher proportion of TGW than MSM lived
and working status, are shown in Figure 6. A compar- alone (Figure 6D; OR 0.32, 95% CI 0.10 to 0.98,
ison of the number of individuals younger than p 5 .05). Data from six studies (Chakrapani et al.,
25 years of age between MSM and TGW groups 2017; Chariyalertsak et al., 2011; Subramanian
showed no significant difference, as shown in et al., 2013; Tucker et al., 2014; Woodford et al.,
Figure 6A (OR 0.72, 95% CI 0.38 to 1.35, 2015; Zhang et al., 2016) were included to compare
p 5 .31). The number of individuals who finished marital status in the groups in Figure 6E (OR 0.74,
high school or did not graduate is reported in 95% CI 0.36 to 1.52, p 5 .41). The proportion of in-
Figure 6B. There was a significant difference be- dividuals who had never married showed no signifi-
tween MSM and TGW groups (OR 0.49, 95% CI cant difference between the two groups. Finally, we
0.33 to 0.72, p 5 .0003). Thus, the TGW group had compared the number of unemployed individuals in
a lower education level than did the MSM group. the groups. There was no significant difference, as
Three studies (Andrinopoulos et al., 2015; Bowers shown in Figure 6F (OR 0.96, 95% CI 0.55 to 1.66,
et al., 2012; Wiewel et al., 2016) reported a history p 5 .88).

Castro, 2016 84 756 8 37 12.0% 0.45 [0.20, 1.02]
Chariyalertsak, 2011 44 309 4 138 9.3% 5.56 [1.96, 15.81]
Chemnasiri, 2010 181 274 170 241 19.0% 0.81 [0.56, 1.18]
Deutsch, 2015 525 2160 127 339 20.8% 0.54 [0.42, 0.68]
Ham, 2015 348 1895 86 277 20.3% 0.50 [0.38, 0.66]
Subramanian, 2013 64 1620 45 403 18.6% 0.33 [0.22, 0.49]
Total (95% Cl) 7014 1435 100.0% 0.64 [0.42, 0.96]


Andrinopoulos, 2015 88 455 42 207 49.6% 0.94 [0.62, 1.42]
Ham, 2015 121 1771 35 285 50.4% 0.52 [0.35, 0.78]
Total (95% Cl) 2226 492 100.0% 0.70 [0.39, 1.25]

Total events 209 77
C MSM TGW Odds Ratio Odds Ratio

Andrinopoulos, 2015 39 455 25 207 49.7% 0.68 [0.40, 1.16]
Ham, 2015 57 1375 20 286 50.3% 0.58 [0.34, 0.97]
Total (95% Cl) 1830 493 100.0% 0.63 [0.43, 0.91]

Total events 96 45
Heterogeneity: Chi2 = 0.20, df = 1 (P = 0.65); I2 = 0%

Andrinopoulos, 2015 296 455 167 207 39.5% 0.45 [0.30, 0.66]
Ham, 2015 861 2250 216 326 60.5% 0.32 [0.25, 0.40]
Total (95% Cl) 2705 533 100.0% 0.36 [0.26, 0.50]

E MSM TGW Odds Ratio Odds Ratio

Fernandes, 2015 73 278 76 152 24.8% 0.36 [0.24, 0.54]
Ham, 2015 133 1763 45 270 31.1% 0.41 [0.28, 0.59]
Verre, 2014 133 2242 75 691 44.1% 0.52 [0.38, 0.70]
Total (95% Cl) 4283 1113 100.0% 0.44 [0.35, 0.55]

Figure 5. Forest plot and meta-analysis of sexually transmitted disease between MSM and TGW: (A) STD1, (B) anal chlamydia,
(C) anal gonorrhea, (D) HSV-2, (E) syphilis. Note. M-H 5 Mantel-Haenszel method; MSM 5 men who have sex with men;
TGW 5 transgender women; STD 5 sexually transmitted disease; HSV 5 herpes simplex virus; CI 5 confidence interval.
10 JANAC Vol. -, No. -, -/- 2018
C

Chakrapani, 2017 44 300 141 300 42.8% 0.19 [0.13, 0.29]
Chemnasiri, 2010 132 274 139 241 43.3% 0.68 [0.48, 0.97]
Tucker, 2014 1 13 3 8 13.9% 0.14 [0.01, 1.68]
Total (95% Cl) 587 549 100.0% 0.32 [0.10, 0.98]

E MSM TGW Odds Ratio Odds Ratio

Chakrapani, 2017 180 300 254 300 23.5% 0.27 [0.18, 0.40]
Chariyalertsak, 2011 256 309 115 140 22.3% 1.05 [0.62, 1.77]
Subramanian, 2013 1208 1620 311 403 24.5% 0.87 [0.67, 1.12]
Tucker, 2014 13 13 7 8 4.0% 5.40 [0.19, 149.78]
Woodford, 2015 16 26 21 21 5.0% 0.04 [0.00, 0.67]
Zhang, 2016 1216 1363 50 61 20.7% 1.82 [0.93, 3.57]
Total (95% Cl) 3631 933 100.0% 0.74 [0.36, 1.52]

F MSM TGW Odds Ratio Odds Ratio

Andrinopoulos, 2015 111 506 30 164 14.2% 1.26 [0.80, 1.97]
Chakrapani, 2017 37 300 6 300 11.2% 6.89 [2.86, 16.59]
Chariyalertsak, 2011 142 309 81 140 14.5% 0.62 [0.41, 0.93]
Newman, 2016 46 304 21 104 13.4% 0.70 [0.40, 1.25]
Stahlman, 2016 1089 1793 251 452 15.3% 1.24 [1.01, 1.53]
Subramanian, 2013 62 1620 11 403 12.9% 1.42 [0.74, 2.72]
Tucker, 2014 3 10 3 8 5.2% 0.71 [0.10, 5.12]
Yang, 2013 59 131 89 107 13.2% 0.17 [0.09, 0.31]
Total (95% Cl) 4973 1678 100.0% 0.96 [0.55, 1.66]

Figure 6. Forest plot and meta-analysis of basic characters between MSM and TGW: (A) individuals . 25 years of age, (B)
education less than 12 years, (C) homeless history, (D) living alone, (E) individuals never married, (F) individuals unemployed.
Note. M-H 5 Mantel-Haenszel method; MSM 5 men who have sex with men; TGW 5 transgender women; CI 5 confidence
interval.
Substance use. Nine studies (Barrington et al., CI 0.37 to 0.70, p , .0001). The TGW group had a
2012; Castro et al., 2016; Chariyalertsak et al., higher proportion of drug users than did the MSM
2011; Chemnasiri et al., 2010; Fernandes et al., group. Additionally, the number of individuals in
2015; Griensven et al., 2006; Poteat et al., 2016; both groups who consumed alcohol more than once
Sanchez et al., 2010; Yang et al., 2013) reported a a week was assessed in this study. There was no sig-
history of drug use (Figure 7A.) There was an nificant difference, as shown in Figure 7B (OR 0.74,
obvious difference between groups (OR 0.51, 95% 95% CI 0.30 to 1.83, p 5 .51).
12 JANAC Vol. -, No. -, -/- 2018

Barrington, 2012 45 279 28 67 10.9% 0.27 [0.15, 0.48]
Castro, 2016 89 750 17 36 9.5% 0.15 [0.08, 0.30]
Chariyalertsak, 2011 38 309 16 140 10.4% 1.09 [0.58, 2.02]
Chemnasiri, 2010 142 274 155 241 13.9% 0.60 [0.42, 0.85]
Fernandes, 2015 113 278 97 152 13.2% 0.39 [0.26, 0.58]
Griensven, 2006 374 821 307 474 15.3% 0.46 [0.36, 0.57]
Poteat, 2016 111 645 10 49 9.1% 0.81 [0.39, 1.67]
Sanchez, 2010 83 301 19 60 10.7% 0.82 [0.45, 1.50]
Yang, 2013 10 131 9 107 7.0% 0.90 [0.35, 2.30]
Total (95% Cl) 3788 1326 100.0% 0.51 [0.37, 0.70]


Chakrapani, 2017 45 300 112 300 33.5% 0.30 [0.20, 0.44]
Chariyalertsak, 2011 134 309 54 140 33.3% 1.22 [0.81, 1.83]
Fernandes, 2015 186 278 98 152 33.2% 1.11 [0.74, 1.69]
Total (95% Cl) 887 592 100.00% 0.74 [0.30, 1.83]

Figure 7. Forest plot and meta-analysis of substances utility between MSM and TGW: (A) drug user, (B) drink more than once a
week. Note. M-H 5 Mantel-Haenszel method; MSM 5 men who have sex with men; TGW 5 transgender women;
CI 5 confidence interval.
Sensitivity Analysis and Publication Bias Discussion

The sensitivity analysis was performed to compare Our meta-analysis of 24 studies, including 37,521
the HIV status between the MSM and TGW groups individuals, compared HIV-related events between
(Figure 8). The two RCT studies (Castillo et al., 2015; TGW and MSM groups. The results showed that the
Deutsch et al., 2015) showed no significant difference TGW group had a higher proportion of high-risk be-
(OR 1.17, 95% CI 0.78 to 1.75, p 5 .34), and the haviors and infections of curable STDs. However,
remaining 13 studies also showed no significant there were no significant differences regarding HIV
difference (OR 0.99, 95% CI 0.71 to 1.36, p 5 .82). status or condom use between the two groups.
Publication bias is a type of bias that occurs in Compared to MSM, TGW may experience double
published academic studies. It occurs when the burdens, including not only men-to-men sexual
outcome of a study influences the decision whether orientation discrimination, but also a life-long confu-
to publish or otherwise distribute it. Figure 9 shows sion of self-sexual identification, resulting in loneli-
a funnel plot of the studies included in this meta- ness, self-abasement, and depression (WHO,
analysis that reported HIV status. Publication bias 2016b). In addition, as a subgroup of MSM in earlier
was analyzed using Egger’s test. Only 8 of 15 studies studies, TGW accounted for only approximately 10%
lie inside the 95% CIs, which strongly indicated an of the MSM (Soto et al., 2007), which could mean
obvious publication bias.
MSM TGW Odds Ratio Odds Ratio

9.1.1 RCT
Castillo, 2015 83 459 35 207 7.7% 1.08 [0.70, 1.67]
Deutsch, 2015 38 2160 3 339 4.2% 2.01 [0.62, 6.53]
Subtotal (95% Cl) 2691 546 11.9% 1.17 [0.78, 1.75]
Total events 121 38
9.1.2 Non-RCT
Barrington, 2012 31 279 13 67 6.3% 0.52 [0.25, 1.06]
Bowers, 2012 132 371 54 255 8.0% 2.06 [1.42, 2.97]
Castro, 2016 123 756 8 37 5.8% 0.70 [0.31, 1.58]
Chakrapani, 2017 30 247 28 257 7.1% 1.13 [0.65, 1.95]
Chariyalertsak, 2011 51 309 13 140 6.6% 1.93 [1.01, 3.68]
Chemnasiri, 2010 38 274 24 241 7.2% 1.46 [0.85, 2.51]
Fernandes, 2015 25 278 37 152 7.1% 0.31 [0.18, 0.53]
Griensven, 2006 16 314 1 180 2.1% 9.61 [1.26, 73.09]
Ham, 2015 232 2247 67 321 8.2% 0.44 [0.32, 0.59]
Poteat, 2016 243 645 21 49 6.9% 0.81 [0.45, 1.45]
Sanchez, 2010 63 301 11 60 6.3% 1.18 [0.58, 2.40]
Subramanian, 2013 219 1620 39 403 8.0% 1.46 [1.02, 2.09]
Verre, 2014 221 2277 102 714 8.4% 0.64 [0.50, 0.83]
Subtotal (95% Cl) 9918 2876 88.1% 0.96 [0.66, 1.39]
Total (95% Cl) 12537 3422 100.0% 0.99 [0.71, 1.39]

Test for subgroup differences: Chi2 = 0.50. df = 1 (P = 0.48). I2 = 0% Favors TGW Favors MSM
Figure 8. Effect of allocation concealment on HIV status. Note. M-H 5 Mantel-Haenszel method; MSM 5 men who have sex with
men; TGW 5 transgender women; CI 5 confidence interval; RCT 5 randomized controlled trial.
that it is more difficult for TGW to find other TGW, to ures might increase negative outcomes in the TGW
build a suitable social circle, and to communicate group. All of these strongly implied a poorer life
with others. In our study, the data collected from 24 for TGW. Second, the TGW group experienced
studies implied that TGW lived poorer, more more forced sex than the MSM group, leading to
dangerous, and more disorderly sexual lives both physiological and psychological injury
compared with those of MSM. First, although the (Figure 3B). In addition, the TGW group faced
same unemployment rate was shown (Figure 6F), significantly more physical violence, discrimination,
the TGW group had less education compared to the and stigma from their families, partners, and even
MSM group (Figure 6B), which implied a lower their sexual partners (Logie, Newman, Weaver,
salary even if employed. Moreover, compared to Roungkraphon, & Tepjan, 2016; Sanchez et al.,
the MSM group, more sex workers, who undoubtedly 2010; Zhang et al., 2016). Finally, the TGW group
had an increased risk of acquiring STDs, were had more disorderly sexual behaviors than those of
discovered in the TGW group. On the other hand, the MSM group. The highest proportions of
the cost of drugs, hormones, and transsexual proced- exchanges of sex (Figure 3A) and selling sex
14 JANAC Vol. -, No. -, -/- 2018
Begg's funnel plot with pseudo 95% confidence limits
logES 2
-2
0 .5 1
s.e. of: logES
Figure 9. Funnel plots illustrating meta-analysis of HIV status.
(Figure 3D) appeared in the TGW group in our meta- stigma of HIV key populations, RCTs were difficult to
analysis. In previous studies, TGW were identified to find. Therefore, most included studies used
have more sexual partners, more sexual behaviors, nonrandom sampling methods to collect data,
and more complicated social and sexual networks including convenience sampling, purposive sampling,
(Barrington et al., 2012; Konda et al., 2017; Tucker and respondent-driven sampling, which tended to in-
et al., 2014; Yang et al., 2013). In addition, more crease the risk of bias, as shown in Figure 9. However,
drug use in the TGW group (Figure 7A) not only the sensitivity analysis showed an acceptable result
increased the risk of infectious diseases from intrave- (Figure 8). Second, some important risk behaviors
nous injection, but also enhanced the proportion of were not included in this meta-analysis, including
multiple sex partners after drug use. number of partners, frequencies of anal sex, and psy-
In our meta-analysis, the TGW group showed a chological factors, because there were not enough
higher proportion of high-risk behaviors for HIV data from the included articles, and the data could
and STDs. These characteristics of TGW strongly not be combined. For example, one article compared
suggested that TGW should have a higher HIV prev- the number of partners in 3 months between the two
alence rate than MSM. However, based on the com- groups, while other articles compared the number in
bined result from 15 studies, this was not the case. 6 months or the last 1-year period. Therefore, it was
This phenomenon confused us and was difficult to inappropriate to combine these data.
explain. The reason for this contradiction needs to
be studied further.
Conclusion
Limitations In our meta-analysis, there was no significant dif-

ference in HIV status between the TGW and MSM
This meta-analysis may have the following limita- groups. However, the TGW group had a higher pro-
tions that must be taken into account. First, the rigor portion of high-risk behavior and prevalence of
of many of the studies included in this meta-analysis curable STDs. The reason for this contradiction needs
could be considered low. For the specificity and social to be studied further.
Disclosures dor, El Salvador. AIDS and Behavior, 19(1), 60-71. https://

doi.org/10.1007/s10461-014-0813-0
Baral S. D., Poteat T., Str€ omdahl S., Wirtz A. L.,
The authors report no real or perceived vested in- Guadamuz T. E., Beyrer C. (2013). Worldwide burden of
terests that relate to this article that could be HIV in transgender women: a systematic review and meta-
construed as a conflict of interest. analysis. Lancet Infectious Diseases, 13(3), 214-222.
https://doi.org/10.1016/s1473-3099(12)70315-8
Barrington C., Wejnert C., Guardado M. E., Nieto A. I.,
Key Considerations Bailey G. P. (2012). Social network characteristics and HIV
vulnerability among transgender persons in San Salvador:
identifying opportunities for HIV prevention strategies.
AIDS and Behavior, 16(1), 214-224. https://doi.org/10.
Transgender women (TGW) and men who have 1007/s10461-011-9959-1
sex with men (MSM) have similar oral and anal Bowers J. R., Branson C. M., Fletcher J. B., Reback C. J. (2012).
Predictors of HIV sexual risk behavior among men who have
sexual practices, and both of them are key pop- sex with men, men who have sex with men and women, and
ulation groups in HIV infection. transgender women. International Journal of Sexual Health,
TGW have similar rates of HIV infection as 24(4), 290-302. https://doi.org/10.1080/19317611.2012.
MSM. 715120
TGW have higher proportions of high-risk sex- Castillo R., Konda K. A., Leon S. R., Silva-Santisteban A.,
Salazar X., Klausner J. D., . Caceres C. F. (2015). HIV
ual behavior and prevalence of curable STDs
and sexually transmitted infection incidence and associated
than MSM. risk factors among high-risk MSM and male-to-female trans-
TGW are a smaller population than MSM gender women in Lima, Peru. Journal of Acquired Immune
(about 10% of MSM). TGW face greater bur- Deficiency Syndromes, 69(5), 567-575. https://doi.org/10.
dens and more serious injuries from familial 1097/qai.0000000000000667
and social sources, and it is more difficult for Castro R., Ribeiro-Alves M., Corr^ea R. G., Derrico M.,
Lemos K., Grangeiro J. R., . Grinsztejn B. (2016). The
them to find other TGW to build suitable social men who have sex with men HIV care cascade in Rio de Ja-
circles and to communicate with others. neiro, Brazil. PLoS One, 11(6), e0157309. https://doi.org/10.
There is a need to emphasize the importance of 1371/journal.pone.0157309
and need to protect sexual minorities and create Chakrapani V., Newman P. A., Shunmugam M., Logie C. H.,
equitable legal and social environments for Samuel M. (2017). Syndemics of depression, alcohol use,
and victimisation, and their association with HIV-related sex-
them.
ual risk among men who have sex with men and transgender
women in India. Global Public Health, 12(2), 250-265.
https://doi.org/10.1080/17441692.2015.1091024
Chariyalertsak S., Kosachunhanan N., Saokhieo P., Songsupa R.,
Wongthanee A., Chariyalertsak C., . Beyrer C. (2011). HIV
Acknowledgments incidence, risk factors, and motivation for biomedical inter-
vention among gay, bisexual men, and transgender persons
in Northern Thailand. PLoS One, 6(9), e24295. https://doi.
The work was supported by the National Natural org/10.1371/journal.pone.0024295
Science Foundation of China under Grant number Chemnasiri T., Netwong T., Visarutratana S., Varangrat A.,
81271892 (Principal Investigator [PI]: Wei Pang); Li A., Phanuphak P., . van Griensven F. (2010). Inconsistent
U1302224 (PI: Wei Pang). The authors declare no condom use among young men who have sex with men, male
competing financial interests. sex workers, and transgenders in Thailand. AIDS Education
and Prevention, 22(2), 100-109. https://doi.org/10.1521/
aeap.2010.22.2.100
Chen C., Yu X., Shao S. (2015). Effects of omega-3 fatty acid
References supplementation on glucose control and lipid levels in type
2 diabetes: a meta-analysis. PLoS One, 10(10), e0139565.
Andrinopoulos K., Hembling J., Guardado M. E., de Maria https://doi.org/10.1371/journal.pone.0139565
Hernandez F., Nieto A. I., Melendez G. (2015). Evidence Deutsch M. B., Glidden D. V., Sevelius J., Keatley J.,
of the negative effect of sexual minority stigma on HIV McMahan V., Guanira J., . Grant R. M. (2015). HIV pre-
testing among MSM and transgender women in San Salva- exposure prophylaxis in transgender women: a subgroup
16 JANAC Vol. -, No. -, -/- 2018
analysis of the iPrEx trial. Lancet HIV, 2(12), e512-e519. City House Ball Community: a comparison of men and
https://doi.org/10.1016/s2352-3018(15)00206-4 transgender women who have sex with men. AIDS and
Fernandes F. R., Zanini P. B., Rezende G. R., Castro L. S., Behavior, 14(2), 351-358. https://doi.org/10.1007/s10461-
Bandeira L. M., Puga M. A., . Motta-Castro A. R. 009-9610-6
(2015). Syphilis infection, sexual practices and bisexual Shey M. S., Kongnyuy E. J., Alobwede S. M., Wiysonge C. S.
behaviour among men who have sex with men and trans- (2013). Co-formulated abacavir-lamivudine-zidovudine for
gender women: a cross-sectional study. Sexually Transmitted initial treatment of HIV infection and AIDS. The Cochrane
Infections, 91(2), 142-149. https://doi.org/10.1136/sextrans- Database of Systematic Reviews(3), CD005481. https://doi.
2014-051589 org/10.1002/14651858.CD005481.pub3
Griensven F., Varangrat A., Wimonsate W., Tappero J. W., Soto R. J., Ghee A. E., Nunez C. A., Mayorga R., Tapia K. A.,
Sinthuwattanawibool C., McNicholl J. M., . Toledo C. Astete S. G., . Holmes K. K. (2007). Sentinel surveillance
(2006). HIV prevalence among populations of men who of sexually transmitted infections/HIV and risk behaviors in
have sex with men–Thailand, 2003 and 2005. Morbidity vulnerable populations in 5 Central American countries.
and Mortality Weekly Report, 55(31), 844-848. Journal of Acquired Immune Deficiency Syndromes, 46(1),
Ham D., Northbrook S. Y., Morales-Miranda S., 101-111.
Guardado M. E., Kamb M. (2015). HIV and STIs among Stahlman S., Liestman B., Ketende S., Kouanda S., Ky-
transgendered populations: Four country survey from central Zerbo O., Lougue M., . Baral S. D. (2016). Characterizing
America. In: Poster presented at the Conference on Retrovi- the HIV risks and potential pathways to HIV infection among
ruses and Opportunistic Infections, Seattle, Washington. transgender women in Cote d’Ivoire, Togo and Burkina Faso.
Higgins J., Green S. (2011). Cochrane handbook for systematic Journal of the International AIDS Society, 19(3 Suppl. 2),
reviews of interventions. Retrieved from http://handbook-5-1. 20774. https://doi.org/10.7448/ias.19.3.20774
cochrane.org/ Subramanian T., Ramakrishnan L., Aridoss S., Goswami P.,
Konda K. A., Castillo R., Leon S. R., Silva-Santisteban A., Kanguswami B., Shajan M., . Paranjape R. S. (2013).
Salazar X., Klausner J. D., . Caceres C. F. (2017). HIV sta- Increasing condom use and declining STI prevalence in
tus communication with sex partners and associated factors high-risk MSM and TGs: evaluation of a large-scale preven-
among high-risk MSM and transgender women in Lima, tion program in Tamil Nadu, India. BMC Public Health, 13,
Peru. AIDS and Behavior, 21(1), 152-162. https://doi.org/ 857. https://doi.org/10.1186/1471-2458-13-857
10.1007/s10461-016-1444-4 Tucker C., Arandi C. G., Bolanos J. H., Paz-Bailey G.,
Logie C. H., Newman P. A., Weaver J., Roungkraphon S., Barrington C. (2014). Understanding social and sexual net-
Tepjan S. (2016). HIV-related stigma and HIV prevention up- works of sexual minority men and transgender women in
take among young men who have sex with men and trans- Guatemala city to improve HIV prevention efforts. Journal
gender women in Thailand. AIDS Patient Care and STDs, of Health Care for the Poor and Underserved, 25(4), 1698-
30(2), 92-100. https://doi.org/10.1089/apc.2015.0197 1717. https://doi.org/10.1353/hpu.2014.0163
Miller W., Alvarez B., Boyce S., Alvarado A., Barrington C., Verre M. C., Peinado J., Segura E. R., Clark J., Gonzales P.,
Pazbailey G. (2011). P1–S2.23 Transgender persons in Benites C., . Lama J. R. (2014). Socialization patterns
Guatemala–over-exposed and under-protected – the findings and their associations with unprotected anal intercourse,
of an RDS behavioural survey. Sexually Transmitted Infec- HIV, and syphilis among high-risk men who have sex with
tions, 87(Suppl. 1), A132. men and transgender women in Peru. AIDS and Behavior,
Newman P. A., Cameron M. P., Roungprakhon S., Tepjan S., 18(10), 2030-2039. https://doi.org/10.1007/s10461-014-
Scarpa R. (2016). Acceptability and preferences for hypo- 0787-y
thetical rectal microbicides among a community sample of Wiewel E. W., Torian L. V., Merchant P., Braunstein S. L.,
young men who have sex with men and transgender women Shepard C. W. (2016). HIV diagnoses and care among trans-
in Thailand: a discrete choice experiment. AIDS and gender persons and comparison with men who have sex with
Behavior, 20(11), 2588-2601. https://doi.org/10.1007/ men: New York City, 2006-2011. American Journal of Public
s10461-015-1258-9 Health, 106(3), 497-502. https://doi.org/10.2105/ajph.2015.
Phillips B., Ball C., Sackett D., Badenoch D., Straus S., 302974
Haynes B., Dawes M. (2009). Oxford Centre for Evidence- Woodford M. R., Chakrapani V., Newman P. A., Shunmugam M.
based Medicine – Levels of Evidence. Retrieved from (2015). Barriers and facilitators to voluntary HIV testing up-
https://www.cebm.net/2009/06/oxford-centre-evidence- take among communities at high risk of HIV exposure in
based-medicine-levels-evidence-march-2009/ Chennai, India. Global Public Health 1-17. https://doi.org/
Poteat T., German D., Flynn C. (2016). The conflation of gender 10.1080/17441692.2015.1057757
and sex: Gaps and opportunities in HIV data among trans- World Health Organization (WHO). (2014). Consolidated guide-
gender women and MSM. Global Public Health, 11(7-8), lines on HIV prevention, diagnosis, treatment and care for
835-848. https://doi.org/10.1080/17441692.2015.1134615 key populations. Retrieved from http://apps.who.int/iris/
Sanchez T., Finlayson T., Murrill C., Guilin V., Dean L. (2010). bitstream/10665/128048/1/9789241507431_eng.pdf?
Risk behaviors and psychosocial stressors in the New York ua51&ua51
World Health Organization (WHO). (2016a). Global summary of

the AIDS Epidemic: 2016. Retrieved from http://www.who.
int/hiv/data/epi_core_2016.png?ua51
World Health Organization (WHO). (2016b). Consolidated guide-
lines on HIV prevention, diagnosis, treatment and care for key
populations: 2016 update. Retrieved from http://www.who.int/
hiv/pub/guidelines/keypopulations-2016/en/
Yang D., Chariyalertsak C., Wongthanee A., Kawichai S.,
Yotruean K., Saokhieo P., . Chariyalertsak S. (2013).
Acceptability of pre-exposure prophylaxis among men who
have sex with men and transgender women in Northern
Thailand. PLoS One, 8(10), e76650. https://doi.org/10.1371/
journal.pone.0076650
Zhang Y., Best J., Tang W., Tso L. S., Liu F., Huang S., .
Tucker J. D. (2016). Transgender sexual health in China: a
cross-sectional online survey in China. Sexually Transmitted
Infections https://doi.org/10.1136/sextrans-2015-052350
[Epub ahead of print].

Comparison of HIV Status Between Transgender Women and Men Who Have Sex With Men: A Meta-Analysis

Uploaded by

Copyright:

Available Formats

Comparison of HIV Status Between Transgender Women and Men Who Have Sex With Men: A Meta-Analysis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Comparison of HIV Status Between Transgender Women and Men Who Have Sex With Men: A Meta-Analysis

Uploaded by

Copyright:

Available Formats

Feature

Comparison of HIV Status Between

As key population groups in HIV infection, trans-

PubMed: n = 623 Embase: n = 1,320 Cochrane: n = 30

Studies identified through initial

Titles and abstracts screened:

Figure 1. Flow diagram of article selection strategy.

Table 1. Characteristics of Included Studies (n 5 24)

Study (First Level of Total

MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 12537 3422 100.0% 0.99 [0.71, 1.39]

A MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 1343 510 100.0% 0.22 [0.07, 0.75]

B MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 4198 995 100.0% 0.46 [0.30, 0.71]

C MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 2799 1317 100.0% 0.07 [0.05, 0.10]

D MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 8390 2178 100.0% 0.29 [0.15, 0.54]

A MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 4113 1397 100.0% 1.24 [0.69, 2.20]

B MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 4606 741 100.0% 0.67 [0.32, 1.41]

Secondary Outcomes of homelessness (Figure 6C) and showed no differ-

A MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 7014 1435 100.0% 0.64 [0.42, 0.96]

B MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 2226 492 100.0% 0.70 [0.39, 1.25]

C MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 1830 493 100.0% 0.63 [0.43, 0.91]

D MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 2705 533 100.0% 0.36 [0.26, 0.50]

E MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 4283 1113 100.0% 0.44 [0.35, 0.55]

D MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 587 549 100.0% 0.32 [0.10, 0.98]

E MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 3631 933 100.0% 0.74 [0.36, 1.52]

F MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 4973 1678 100.0% 0.96 [0.55, 1.66]

A MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 3788 1326 100.0% 0.51 [0.37, 0.70]

B MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 887 592 100.00% 0.74 [0.30, 1.83]

Sensitivity Analysis and Publication Bias Discussion

MSM TGW Odds Ratio Odds Ratio

Total (95% Cl) 12537 3422 100.0% 0.99 [0.71, 1.39]

Begg's funnel plot with pseudo 95% confidence limits

Figure 9. Funnel plots illustrating meta-analysis of HIV status.

Limitations In our meta-analysis, there was no significant dif-

Disclosures dor, El Salvador. AIDS and Behavior, 19(1), 60-71. https://

World Health Organization (WHO). (2016a). Global summary of

You might also like