Open access Original research

Changing trends of HIV, syphilis, HCV

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
infections and behavioural factors
among female sex workers in
Chongqing, China: findings from six
serial surveillance surveys
Ling Hu,1 Guohui Wu,2 Rongrong Lu,2 Hua Zhu,1 Hongfang Qiu,1 Dan Jing,1
Mengliang Ye  ‍ ‍1

To cite: Hu L, Wu G, Lu R, et al. ABSTRACT

Changing trends of HIV, syphilis, Strengths and limitations of this study
Objectives  To explore the prevalence and changing
HCV infections and behavioural trends of HIV, syphilis, hepatitis C virus (HCV) infections
factors among female sex ►► The study involved 6 consecutive years of sentinel
and risk behaviours among female sex workers (FSWs)
workers in Chongqing, China: surveillance among female sex workers (FSWs) in
and to provide reference and theoretical basis for
findings from six serial Chongqing with a large sample size.
surveillance surveys. BMJ Open formulating targeted interventions.
►► This study provides valuable information about the
2020;10:e036654. doi:10.1136/ Design  Six consecutive cross-­sectional surveys.
prevalence and trends of sexually transmitted infec-
bmjopen-2019-036654 Setting  Chongqing, China.
tions among FSWs in Chongqing, and helps to for-
Participants  FSWs were included if they (1) were aged
►► Prepublication history for mulate and improve relevant intervention strategies.
≥16 years, (2) provided commercial sex for money or
this paper is available online. ►► However, causality could not be determined due to
goods during the previous month and (3) were willing to
To view these files, please visit the cross-­sectional design.
the journal online (http://​dx.​doi.​ participate in the survey and could provide verbal informed
►► Recall bias and social desirability bias may be pres-
org/​10.​1136/​bmjopen-​2019-​ consent. This study included 16 791 of 16 810 participants
ent due to the use of self-­reported measures of con-
036654). recruited between 2013 and 2018.
dom use, drug use and other behaviour.
Primary and secondary outcome measures  The
LH and GW contributed equally. prevalence of HIV/syphilis/HCV infection.
Results  The HIV and HCV prevalence among FSWs in
Received 24 December 2019 populations and regions. Compared with adult
Chongqing was stable during the study period, but the
Revised 11 September 2020
prevalence of syphilis had an increasing trend, particularly women, female sex workers (FSWs) bear a
Accepted 15 September 2020
among low-­tier and middle-­tier FSWs. Improvements in higher burden of HIV, with an estimated HIV
HIV-­related knowledge, condom use, injecting drug use prevalence of 10.4% (95% CI 9.5% to 11.5%).1
and participation in HIV-­related services were observed. They are also considered one of the most
However, no change was found in the prevalence of vulnerable groups to the acquisition of other
drug use. HIV infection was correlated with no condom sexually transmitted infections (STIs), such as
use in the last commercial sex (adjusted OR (aOR) 3.48,
syphilis.2
95% CI 1.90 to 6.37) and syphilis infection (aOR 4.88,
© Author(s) (or their After the implementation of reform and
95% CI 1.95 to 12.18). Syphilis infection was correlated
employer(s)) 2020. Re-­use
with inconsistent condom use (aOR 1.30, 95% CI 1.02 to opening up in the late 1970s, the sex industry
permitted under CC BY-­NC. No has re-­emerged in China due to economic
commercial re-­use. See rights 1.65), HIV infection (aOR 5.88, 95% CI 2.40 to 14.41), HCV
and permissions. Published by infection (aOR 7.68, 95% CI 4.37 to 13.49) and sexually growth, population migration and changes
BMJ. transmitted infection (STI) diagnosis in the past year (aOR in sexual attitudes.3 4 A large number of rural
1
Department of Epidemiology 3.81, 95% CI 2.40 to 6.03). HCV infection was associated migrants moved to cities for enhanced job
and Health Statistics, School of with injecting drug use (aOR 8.91, 95% CI 4.45 to 17.86) opportunities. However, those rural women
Public Health and Management, and syphilis infection (aOR 7.88, 95% CI 4.49 to 13.83). found that the labour force was saturated or the
Chongqing Medical University, Conclusions  Comprehensive interventions targeting wages were not high enough; then, sex work
Chongqing, China FSWs, particularly low-­tier and middle-­tier FSWs, should
2
Department of AIDS/STD
with quick financial return became an option.5
be increasingly implemented to prevent and control HIV,
Control and Prevention, The estimated number of FSWs was 4 million6;
syphilis and other STIs.
Chongqing Center for Disease and approximately 6.9% of men aged 18–49
Control and Prevention, years have had commercial sex during their life-
Chongqing, China INTRODUCTION time.7 The incidence of STIs, such as syphilis,
Correspondence to The HIV/AIDS pandemic, one of the major has dramatically increased along with the pros-
Dr Mengliang Ye; public health problems all over the world, perity of the sex trade.8 9 Heterosexual sex has
​yemengliang@​cqmu.e​ du.​cn is highly heterogeneous among different replaced intravenous drug use and commercial

Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654 1

Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
blood/plasma collection as the primary transmission mode Chongqing, located in Southwest China, is one of the
of HIV in China.10 Among the newly reported HIV infec- four Chinese municipalities directly controlled by the
tions in January–October 2019, heterosexual transmission central government. It is geographically close to Yunnan
accounted for 73.7%, much higher than other transmission and Guangxi and it has large population contacts with
routes.11 A review also reported that FSWs in China had a Yunnan and Guangxi. It is the political and economic
median HIV prevalence of 0.6% (0% to 10.3%) and a rate centre of Western China and a city with obvious urban–
of positive for at least one STI of 41.5% (13% to 90.6%).12 rural dual structure and the prosperous sex industry. A
As a developing country with a population of 1.4 billion, study using the network scale-­up method estimated that
China faces more challenges in responding to the AIDS FSWs accounted for 0.4% of women aged 15–49 years
and STIs epidemic. The prevalence varies from province in Chongqing, and the clients of FSWs represented 2%
to province and Southwest provinces have a higher prev- of men aged 15–49 years.25 The HIV prevalence among
alence of STI.13 14 For example, although HIV prevalence FSWs in Chongqing in 2012 was estimated at 0.25%, lower
among FSWs was lower than 0.05% in most provinces than that among other high-­risk groups, such as men who
(such as Heilongjiang and Hebei) in 2012, it exceeded have sex with men (17.03%) and injecting drug users
1% in the same year in Yunnan and Guangxi.13 Further- (7.35%).26 However, FSWs can transmit STIs to numerous
more, sex work is still illegal in China. Thus, it takes clients and then to the general population, which pose a
place in informal venues, including night clubs, hair great challenge to public health. Heterosexual transmis-
salons, foot massage parlours and hotels, which is always sion has become the main route of HIV transmission in
combined with other works to cover up its illegality. This Chongqing since 2007.27 Of the 6352 newly diagnosed
hinders the implementation of related interventions and HIV/AIDS cases reported in Chongqing in 2015, 75.08%
reduces their effects. Different sex work settings are asso- were infected through heterosexual transmission.27
ciated with the risk of HIV/STI infection. The average Therefore, understanding the prevalence and trends of
price of each sex transaction is often used by Chinese STIs among FSWs in Chongqing is highly important.
researchers to categorise FSWs into high, middle and low On the basis of the surveillance data in Chongqing, this
tiers. The size of different-­tier FSWs may be related to the study was aimed to identify the prevalence and changing
local economic level. In Jiangmen City, more than 60% of trends of HIV, syphilis and HCV infections among FSWs
FSWs was high-­tier FSW, while the proportion of low-­tier over time. This study also explored the trends in sexual
FSWs was less than 10%.15 However, in Jianshui County, behaviours of FSWs and discussed the risk factors for preva-
the proportion of low-­tier FSWs was close to 30%.16 The lent STIs over a 6-­year period in order to provide a reference
proportion of low-­ tier FSWs may vary from region to for formulating interventions to control AIDS and STIs.
region, but previous studies have shown that they have a
higher risk of HIV infection than other FSWs.13 17
The Chinese government has provided voluntary HIV METHODS
counselling and testing, STIs services, condom promotion Study design
and peer education among FSWs through the coopera- Sentinel surveillance was conducted to collect HIV, syph-
tion of Centers for Disease Control (CDC), community ilis and HCV prevalence and high-­risk behaviour informa-
health services and medical institutions.18 The ‘Four tion through cross-­sectional surveys. The data analysed in
Frees and One Care’ policy helps treat patients with this study were sentinel surveillance data among FSWs in
AIDS who have financial difficulties by providing free Chongqing from 2013 to 2018.
antiviral drugs. In 1995, China established the national
HIV sentinel surveillance system to actively monitor HIV Participant selection and data collection
prevalence among high-­risk populations (eg, FSWs and The detailed procedures of data collection followed
injecting drug users) and guide the development of the Operational Manual for the Implementation
HIV prevention and control strategies.19–21 After years of Program of National AIDS Sentinel Surveillance estab-
development, the surveillance system has expanded its lished by the National Center for AIDS Prevention and
scope from 42 sentinel sites in 1995–1888 in 2010, and Control.28 As shown in figure 1, serial cross-­ sectional
combined biological and behavioural surveillance strat- surveys were conducted from April to June each year at
egies, which included serotesting for HIV, syphilis and seven sentinel surveillance sites in Chongqing (Yuzhong
hepatitis C virus (HCV).22 Surveillance data could help District, Jiulongpo District, Wanzhou District, Dianjiang,
understand the epidemic status and trend of HIV and Hechuan, Qijiang (Wansheng) and Youyang). The sample
STIs among FSWs, which is necessary for developing HIV size of each site was 400 participants. The entire sampling
or STIs prevention programmes.23 According to China’s process went through the following three steps. First, a
surveillance data, the overall prevalence of HIV/syph- distribution map of known commercial workplaces in the
ilis/HCV among FSWs showed a downward trend from monitoring area was drawn to construct a sampling frame
2010 to 2015.24 However, there may be diverse trends and these locations were categorised into low-­tier, middle-­
in different geographic regions of China, and there has tier and high-­tier venues on the basis of the average price
been less work on understanding trends of STIs among of each sex transaction. Then, these venues in each city
FSWs in underdeveloped southwest areas. were randomly selected through proportional sampling.

2 Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654

 Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
commercial sex, condom use with clients in the past
month, drug use during their lifetime, STI diagnosis in
the last year and participation in HIV-­related services.
In this study, consistent condom use (CCU) in the past
month was defined as always using a condom during
commercial intercourse.

Laboratory testing
In all surveys, the venous blood samples were tested for
HIV, syphilis and HCV by trained laboratory technicians.
Initial screening for HIV, syphilis and HCV antibodies
was conducted using ELISA method (ELISA-1). When
the result was negative, no further re-­examination was
carried out and infection status was recorded as negative.
When the result was positive, HIV or HCV infection was
confirmed by another ELISA method (ELISA-2) while
syphilis infection was confirmed using a non-­ specific
detection method called toluidine red unheated serum
test. The results were determined as positive only when
both tests were positive.
Figure 1  The location of seven sentinel surveillance sites in
Chongqing. Quality assurance
All investigators were strictly trained before the survey
to ensure that they familiarised the questionnaire struc-
The proportions of low-­tier and middle-­tier FSWs among ture and mastered the unified investigation standards
all participants were at least 10% and 40%, respectively. and requirements. After the investigation, the investiga-
The FSWs were recruited if they met the inclusion criteria tors carefully reviewed the questionnaires and promptly
as follows: participants (1) were aged ≥16 years, (2) corrected the missing items, wrong items and logical
provided commercial sex for money or goods during the errors. Experts from the Chongqing Center for Disease
previous month and (3) were willing to participate in the Control and Prevention went to the investigation site for
survey and could provide verbal informed consent. guidance and inspection to ensure quality. All laboratory
All participants completed an anonymous, standard tests were conducted at designated and certified laborato-
interviewer-­administered and face-­to-­face question- ries in local CDC or hospitals.
naire. After each interview, 3–5 mL of venous blood was
collected from each participant for HIV, syphilis and HCV Data analysis
antibody detection. The blood sample was linked to the The sociodemographic characteristics of participants
questionnaire by a unique identification code assigned to were presented using descriptive statistics by survey year.
the participant. In addition, the trends of HIV, syphilis, HCV infections
and behavioural factors over time were assessed using
Measures Cochran-­Armitage trend test. Multivariable logistic regres-
Questionnaire sion was conducted to identify related risk-­factors of HIV,
A questionnaire was used to collect the sociodemo- syphilis and HCV infection by using stepwise elimination.
graphic characteristics, HIV knowledge and behaviour of The ORs and 95% CIs of each significant risk factor were
the FSWs. also determined, unadjusted and adjusted for sociodemo-
1. Sociodemographic characteristics, including year of graphic factors. All statistical analyses were carried out
birth, marital status, household registration, ethnicity using Statistical Analysis Software, V.9.2 (SAS Institute).
and education level were collected.
2. The HIV-­ related knowledge of participants was as- Patient and public involvement
sessed using eight questions with ‘yes’, ‘no’ or ‘don’t Patients and/or the public were not involved in the
know’ as answers. The questions have been updated design, or conduct, or reporting, or dissemination plans
since 2016 and took new-­type drugs and intentional of this research.
transmission of HIV/AIDS into consideration. Only
correct responses were scored as 1 point, whereas in-
correct responses and ‘don’t know’ responses did not RESULTS
earn any point. The Cronbach’s alpha values were Sociodemographic characteristics
0.706 and 0.739 before and after the questions were This study included 16 791 of 16 810 participants recruited
changed, respectively. between 2013 and 2018. Table 1 depicted all the partic-
3. The participants’ behavioural factors were also collect- ipants’ demographic characteristics stratified by year of
ed, including questions regarding condom use in last survey. Nearly half of the participants (46.7%) worked

Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654 3

Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
Table 1  Sociodemographic characteristics of female sex workers stratified by survey year (N, %)
2013 2014 2015 2016 2017 2018 Total
Characteristics N (%) N (%) N (%) N (%) N (%) N (%) N (%)
Overall 2793 (16.6) 2796 (16.7) 2793 (16.6) 2799 (16.7) 2805 (16.7) 2805 (16.7) 16 791 (100.0)
Age group (years)
 <20 285 (10.2) 226 (8.1) 206 (7.4) 228 (8.1) 87 (3.1) 96 (3.4) 1128 (6.7)
 20–45 2404 (86.1) 2451 (87.7) 2442 (87.4) 2412 (86.2) 2521 (89.9) 2415 (86.1) 14 645 (87.2)
 >45 104 (3.7) 119 (4.3) 145 (5.2) 159 (5.7) 197 (7.0) 294 (10.5) 1018 (6.1)
Marital status
 Never been married 1131 (40.5) 1143 (40.9) 1102 (39.5) 1106 (39.5) 734 (26.2) 837 (29.8) 6053 (36.0)
 Married 1148 (41.1) 1104 (39.5) 1207 (43.2) 1152 (41.2) 1279 (45.6) 1278 (45.6) 7168 (42.7)
 Living together as if married 275 (9.8) 308 (11.0) 260 (9.3) 307 (11.0) 305 (10.9) 296 (10.6) 1751 (10.4)
 Divorced/widowed 239 (8.6) 241 (8.6) 224 (8.0) 234 (8.4) 487 (17.4) 394 (14.0) 1819 (10.8)
Household registration (hukou)
 Chongqing 2518 (90.2) 2559 (91.5) 2533 (90.7) 2547 (91.0) 2580 (92.0) 2537 (90.4) 15 274 (91.0)
 Other provinces 275 (9.8) 237 (8.5) 260 (9.3) 252 (9.0) 225 (8.0) 268 (9.6) 1517 (9.0)
Ethnicity
 Minority 253 (9.1) 266 (9.5) 308 (11.0) 295 (10.5) 280 (10.0) 272 (9.7) 1674 (10.0)
 Han 2539 (90.9) 2530 (90.5) 2485 (89.0) 2504 (89.5) 2525 (90.0) 2533 (90.3) 15 116 (90.0)
Education level
 Primary school or below 512 (18.3) 524 (18.7) 581 (20.8) 506 (18.1) 595 (21.2) 507 (18.1) 3225 (19.2)
 Junior middle school 1664 (59.6) 1570 (56.2) 1367 (48.9) 1509 (53.9) 1635 (58.3) 1446 (51.6) 9191 (54.7)
 High school or above 617 (22.1) 702 (25.1) 845 (30.3) 784 (28.0) 575 (20.5) 852 (30.4) 4375 (26.1)
Typology
 Low-­tier 973 (34.8) 955 (34.2) 1043 (37.3) 1193 (42.6) 1347 (48.0) 1161 (41.4) 6672 (39.7)
 Middle-­tier 1347 (48.2) 1504 (53.8) 1448 (51.8) 1236 (44.2) 1192 (42.5) 1108 (39.5) 7835 (46.7)
 High-­tier 473 (16.9) 337 (12.1) 302 (10.8) 370 (13.2) 266 (9.5) 536 (19.1) 2284 (13.6)
Local working time
 More than 1 year 651 (23.3) 566 (20.2) 751 (26.9) 673 (24.0) 888 (31.7) 946 (33.7) 4475 (26.7)
 6–12 months 635 (22.7) 672 (24.0) 899 (32.2) 1018 (36.4) 918 (32.7) 833 (29.7) 4975 (29.6)
 1–6 months 1097 (39.3) 1204 (43.1) 888 (31.8) 778 (27.8) 814 (29.0) 754 (26.9) 5535 (33.0)
 Less than 1 month 409 (14.6) 354 (12.7) 255 (9.1) 330 (11.8) 185 (6.6) 272 (9.7) 1805 (10.8)
Location of previous job
 Other provinces 334 (12.0) 257 (9.2) 354 (12.7) 305 (10.9) 353 (12.6) 293 (10.4) 1896 (11.3)
 Other cities in Chongqing 330 (11.8) 569 (20.4) 549 (19.7) 572 (20.4) 356 (12.7) 352 (12.5) 2728 (16.2)
 Current city 1699 (60.8) 1627 (58.2) 1629 (58.3) 1576 (56.3) 1894 (67.5) 1872 (66.7) 10 297 (61.3)
 No previous work 430 (15.4) 343 (12.3) 261 (9.3) 346 (12.4) 202 (7.2) 288 (10.3) 1870 (11.1)
HIV knowledge
 <6 388 (13.9) 458 (16.4) 231 (8.3) 449 (16.0) 348 (12.4) 305 (10.9) 2179 (13.0)
 ≥6 2405 (86.1) 2338 (83.6) 2562 (91.7) 2350 (84.0) 2457 (87.6) 2500 (89.1) 14 612 (87.0)

in middle-­tier venues, 39.7% worked in low-­tier venues Trends of HIV, syphilis and HCV prevalence
and 13.6% worked in high-­tier venues. The median age During the study period, the overall prevalence of HIV,
was 28 years and most (87.2%) of them were between 20 syphilis and HCV infection among the study population
and 45 years of age. The majority of the participants were was 0.27% (95% CI 0.19% to 0.35%), 1.73% (95% CI
in Chongqing households (91.0%) and belong to Han 1.54% to 1.93%) and 0.72% (95% CI 0.59% to 0.85%),
(90.0%). Less than one-­third of them had high school or respectively. Low-­ tier FSWs had higher prevalence of
above education and half (54.7%) had a junior middle HIV, syphilis and HCV, with 6-­year average prevalence
school education. Most participants have worked in the rates of 0.40%, 2.53% and 1.08%, respectively (p=0.019,
current location for 1–6 months (33.0%). p<0.001, p<0.001). The HIV prevalence among FSWs in

4 Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654

 Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
month. Approximately 3% reported drug use, and 2%
reported STI diagnosis in the past year.
The changing trends of FSWs’ behaviours were also
examined, and results showed no significant change
over time in drug use reported by FSWs. However, signif-
icant improvements over time were found in HIV-­related
knowledge, condom use, participation in HIV-­ related
services and self-­reported STI diagnosis in the last year
among the FSWs.

Correlates of HIV, syphilis and HCV infections

Multivariable logistic regression analysis results (table 3)
showed that, (1) HIV infection was significantly and posi-
Figure 2  Prevalence of HIV, syphilis and HCV among FSWs tively associated with no condom use in the last commer-
in Chongqing by survey year. Bars show prevalence (per cial sex (adjusted OR (aOR) 3.48, 95% CI 1.90 to 6.37)
cent); error bars show 95% CIs. FSWs, female sex workers;
and syphilis infection (aOR 4.88, 95% CI 1.95 to 12.18);
HCV, hepatitis C virus.
(2) syphilis infection was significantly and positively asso-
ciated with inconsistent condom use in the past month
Chongqing, which ranged from 0.14% to 0.43% during (aOR 1.30, 95% CI 1.02 to 1.65), STI diagnosis in the past
the study years, had no significant changes over time year (aOR 3.81, 95% CI 2.40 to 6.03), HIV infection (aOR
(p=0.129, figure 2). No significant time trend was found 5.88, 95% CI 2.40 to 14.41) and HCV infection (aOR 7.68,
among low-­tier/middle-­tier/high-­tier FSWs (p=0.498, 95% CI 4.37 to 13.49) and (3) HCV infection was signifi-
0.366 and 0.423, respectively). The HCV prevalence cantly and positively associated with injecting drug use
among FSWs in Chongqing ranged from 0.32% to 1.14% (aOR 8.91, 95% CI 4.45 to 17.86) and syphilis infection
between 2013 and 2018, with no significant trend. In each (aOR 7.88, 95% CI 4.49 to 13.83).
subgroup, no significant changing trend of HCV infection
by year was found (figure 3). Unlike HIV and HCV, a total
of 1.11%, 1.18%, 1.58%, 1.86%, 2.67% and 2.00% of the DISCUSSION
respondents were confirmed to be infected with syphilis FSWs are considered to be at high risk of HIV infection
from 2013 to 2018, showing an upward trend (p<0.001). and other STIs and contribute disproportionately to the
This trend was also found in low-­tier FSWs (from 1.13% transmission. Surveillance of STIs and risk behaviour
in 2013 to 2.50% in 2018, p<0.001) and middle-­tier FSWs among FSWs is highly important for the understanding
(from 1.19% in 2013 to 1.81% in 2018, p=0.026), but not of the potential of spreading HIV to the general popu-
in high-­tier FSWs (p=0.713). lation and the assessment and improvement of interven-
tions targeted at FSWs.23 Although analysing and knowing
Changing trends of behavioural factors the prevalence and changing trends of HIV, syphilis, HCV
Of the whole survey population, more than 80% exhib- and relevant behaviours among FSWs are critical, only few
ited HIV-­related knowledge and used condom in their studies have been reported in Chongqing, China. Thus,
last commercial sex (table 2). Nearly 90% received at in the present study, the prevalence and changing trends
least one of the following three HIV-­related services in of HIV, syphilis, HCV and behaviour of Chongqing FSWs
the last year: (1) condom promotion and distribution and were examined using sentinel surveillance data from
AIDS counselling and testing, (2) community drug main- 2013 to 2018.
tenance therapy and clean needle supply/exchange and The results showed that the HIV prevalence among
(3) peer education. In addition, about half of the partic- FSWs in Chongqing was relatively stable, with an annual
ipants used condom consistently with clients in the last average of 0.27% from 2013 to 2018, similar to 0.22%

Figure 3  Prevalence of HIV, syphilis and HCV among low-­tier FSWs (A), middle-­tier FSWs (B) and high-­tier FSWs (C) in
Chongqing by survey year. (A) Low-­tier FSWs; (B) middle-­tier FSWs; (C) high-­tier FSWs. FSWs, female sex workers; HCV,
hepatitis C virus.

Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654 5

6
Open access

Table 2  Behavioural characteristics of FSWs from 2013 to 2018 in Chongqing

2013 2014 2015 2016 2017 2018 Overall
Variables % (n) % (n) % (n) % (n) % (n) % (n) % (n) P value P for trend
HIV knowledge (≥6) Yes 86.1 (2405) 83.6 (2338) 91.7 (2562) 84.0 (2350) 87.6 (2457) 89.1 (2500) 87.0 (14 612) <0.001 <0.001
No 13.9 (388) 16.4 (458) 8.3 (231) 16.0 (449) 12.4 (348) 10.9 (305) 13.0 (2179)
Condom use in the last commercial sex Yes 79.7 (2167) 77.4 (2134) 93.8 (2604) 83.4 (2335) 87.0 (2439) 84.5 (2371) 84.3 (14 050) <0.001 <0.001
No 20.3 (552) 22.6 (625) 6.2 (171) 16.6 (464) 13.0 (366) 15.5 (434) 15.7 (2612)
CCU in the past month Yes 41.7 (1130) 51.2 (1406) 69.7 (1925) 60.1 (1682) 47.7 (1337) 51.4 (1441) 53.7 (8921) <0.001 <0.001
No 58.3 (1578) 48.8 (1341) 30.3 (835) 39.9 (1117) 52.3 (1468) 48.6 (1364) 46.3 (7703)
Drug use Yes 2.9 (81) 4.2 (116) 2.7 (74) 4.6 (129) 2.2 (63) 3.1 (88) 3.3 (551) <0.001 0.322
No 97.1 (2705) 95.8 (2631) 97.3 (2718) 95.4 (2670) 97.8 (2742) 96.9 (2717) 96.7 (16 183)
Injecting drug use Yes 1.3 (35) 1.2 (32) 0.6 (18) 1.5 (43) 0.7 (19) 0.4 (11) 0.9 (158) <0.001 0.002
No 98.8 (2754) 98.9 (2760) 99.4 (2774) 98.5 (2756) 99.3 (2786) 99.6 (2794) 99.1 (16 624)
STI diagnosis in the past year Yes 3.8 (106) 3.6 (100) 1.5 (43) 1.8 (50) 1.1 (31) 1.3 (37) 2.2 (367) <0.001 <0.001
No 96.2 (2671) 96.4 (2679) 98.5 (2747) 98.2 (2749) 98.9 (2774) 98.7 (2768) 97.8 (16 388)
HIV-­related services Yes 80.0 (2235) 86.0 (2405) 94.4 (2637) 93.4 (2615) 92.7 (2600) 88.1 (2470) 89.1 (14 962) <0.001 <0.001
  No 20.0 (558) 14.0 (391) 5.6 (156) 6.6 (184) 7.3 (205) 11.9 (335) 10.9 (1829)

CCU, consistent condom use; FSWs, female sex workers; STIs, sexually transmitted infections.

Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
 Table 3  Associations between behavioural characteristics and HIV, syphilis and HCV infection
HIV positive Syphilis positive HCV positive

Behavioural OR (95% CI) OR (95% CI) OR (95% CI)

characteristics Univariate Multivariate aOR (95% CI) Univariate Multivariate aOR (95% CI) Univariate Multivariate aOR (95% CI)
No condom use in the 3.95 3.97 3.48 1.06 0.54
last commercial sex (2.19 to 7.15)* (2.19 to 7.19)* (1.90 to 6.37)* (0.78 to 1.45) (0.29 to 1.00)
(yes vs no)
Inconsistent condom 2.86 1.33 1.30 1.30 0.64 0.63 0.72
use in the past month (1.50 to 5.45)* (1.05 to1.67)* (1.03 to 1.64)* (1.02 to 1.65)* (0.44 to 0.93)* (0.43 to 0.92)* (0.49 to 1.05)
(yes vs no)
Drug use (yes vs no) 0.32 1.27 3.28

Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654

(0.02 to 5.24) (0.71 to 2.28) (1.80 to 6.00)*
Injecting drug use 1.15 1.87 10.05 9.21 8.91
(yes vs no) (0.07 to 18.91) (0.76 to 4.59) (5.16 to 19.59)* (4.67 to 18.18)* (4.45 to 17.86)*
STI diagnosis in the 0.49 4.02 3.75 3.81 2.78
past year (0.03 to 7.98) (2.59 to 6.24)* (2.40 to 5.87)* (2.40 to 6.03)* (1.29 to 6.00)*
HIV-­related services 1.71 1.37 1.36
(yes vs no) (0.53 to 5.53) (0.90 to 2.08) (0.71 to 2.60)
HIV positive 8.89 9.17 5.88 1.51
(yes vs no) (3.74 to 21.17)* (3.84 to 21.91)* (2.40 to 14.41)* (0.09 to 25.29)
Syphilis positive 8.89 8.68 4.88 9.09 9.0 7.88
(yes vs no) (3.73 to 21.16)* (3.63 to 20.74)* (1.95 to 12.18)* (5.30 to 15.58)* (5.26 to 15.71)* (4.49 to 13.83)*
HCV positive 1.50 9.09 8.81 7.68
(yes vs no) (0.09 to 24.83) (5.30 to 15.59)* (5.09 to 15.25)* (4.37 to 13.49)*

aOR adjusted for sociodemographic characteristic (age, education level, marital status, typology and survey year).
*P<0.05.
aOR, adjusted OR; HCV, hepatitis C virus; STIs, sexually transmitted infections.
Open access

7
BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
among Chinese FSWs in 20149 but much higher than compromise with their clients on condom use to earn
China’s current estimated national HIV prevalence money or establish a closer long-­term relationship with
among the general population (0.0598%).29 The preva- clients. Thus, in addition to raising FSWs’ awareness
lence of HIV among FSWs in Chongqing is higher than regarding HIV/STI prevention and improving their skills
that in most provinces such as Beijing,30 Liaoning31 and to negotiate condom use, promoting HIV intervention
Hainan,32 but lower than that in Sichuan,33 Yunnan34 and programmes targeting clients of FSWs and others is also
Guangxi.35 HCV prevalence fluctuated between 2013 and essential.
2018, with a 6-­year average of 0.72%, but no significant No significant increasing or decreasing trend in drug
trend was found. This result was in concordance with that use was observed among FSWs. The average prevalence of
of a previous study conducted in Guangxi.35 However, a drug use was 3.3%, while the national prevalence among
previous study using national data reported that from FSWs was reported at 1.2%.13 The prevalence of injecting
2008 to 2012, the overall HIV and HCV prevalence among drug use seemed to be declining during the study
FSWs in mainland China showed a decreasing trend13; period. Injecting drug use may exacerbate unsafe sexual
the differences may be due to the differences in scope of behaviour and structural risks, leading to increased risk
study area and study period. of HIV and other STIs.44 45 In 2017, the detection rates of
The results also indicated a worrying upward trend HIV, syphilis and HCV among drug users in Chongqing
of syphilis infection (from 1.11% in 2013 to 2.00% in were 5.4%, 5.9% and 44.9%, respectively, much higher
2018), particularly among low-­tier and middle-­tier FSWs. than those among the general population.46 FSWs who
As the third most prevalent notifiable infectious disease inject drugs are exposed to sexual and blood transmis-
in categories A and B in China, syphilis has been on sion routes and at a higher risk of contracting HIV and
the rise since the 1990s.36 The reported total syphilis syphilis than other FSWs and injecting drug users.35 47
rate in China increased from 0.2/100 000 in 1993 to Similar to previous studies, the present study showed a
32.86/100 000 in 2013.36 The reported incidence of syph- strong association between injecting drug use and HCV
ilis in Chongqing also showed an upward trend, with a infection,48 and sharing of needles may be the main and
reported incidence of 40.38/100 000 in 2014.37 However, underlying route of HCV transmission. Drugs and prosti-
researchers found that the increase in reported cases tution involve relevant laws and ethics. Thus, HIV preven-
of syphilis in Chongqing in recent years may be related tion programmes and outreach for FSWs may require
to large-­scale syphilis screening by medical institutions, enhanced thinking and increased attention to the indi-
training on national standards for syphilis diagnosis and vidual’s external risk environment and risk factors.
reporting, and laboratory improvement.38 The increase In addition, as previous studies have shown, lower-­tier
in syphilis prevalence among FSWs may also be related to FSWs bear a heavier burden of HIV, syphilis and HCV than
the strengthening of syphilis detection, but it still needs higher-­tier FSWs,13 which may be attributed to the socio-­
attention. In agreement with previous studies, the FSWs demographic characteristics of lower-­tier FSWs and their
infected with syphilis have a higher risk of HIV or HCV clients. It’s found that low-­tier FSWs generally have older
infection, and HIV or HCV infection is a risk factor of age and lower education, which are often associated with
syphilis infection.35 FSWs with syphilis were nearly five a higher prevalence of HIV and syphilis.49 Low-­tier FSWs
times more likely to be infected with HIV, and seven times with low education probably had insufficient knowledge
more likely to be infected with HCV than FSWs without about those diseases, less awareness about sex health and
syphilis. This phenomenon may be due to their shared poor communication skills for condom use.50 And they
transmission routes, co-­risk factors and adverse interac- may be serving more clients who also had low education,
tion.39 Thus, efforts should be increased to strengthen had low awareness and were unwilling to use a condom,
the comprehensive and combined intervention of AIDS, particularly elderly man.17 35 Thus, future intervention
syphilis and other STIs for key groups. programmes should be specified and targeted based on
During the study period, the proportion of FSWs who the needs of FSWs under different tiers, and pay more
used condoms increased slightly, and condom use was attention to low-­tier FSWs and their clients. In addition,
associated with decreased HIV or syphilis infection risk. sex work is illegal and highly stigmatised in China; thus,
CCU is one of the most effective ways to prevent HIV FSWs face several barriers, such as fear of being arrested,
infection, and its protection rate can reach 80% based on fear of discrimination and awkwardness, to participant in
studies of persons during heterosexual sex with an HIV-­ AIDS-­related services.51 AIDS intervention services must
positive partner.40 The 100% Condom Use Programme further cooperate with non-­governmental organisations,
has been implemented in Chongqing since 2006 and has increase condom promotion and jointly increase aware-
shown a promotional effect. The number of FSWs who ness of other STIs such as syphilis.
practiced CCU increased. However, in 2018, only half of This study had several limitations. First, the data we
FSWs reported CCU in the last month. This proportion used were based on serial cross-­ sectional surveillance
was lower than that in Liaoning and Hainan.31 32 Previous surveys; therefore, trends of HIV, syphilis and HCV
studies indicated that clients’ low support and negative infection without exact reasons behind the changes
norms towards condom use may be the main reason for were possibly found. Second, the survey was anonymous.
the inconsistent condom use of FSWs.41–43 FSWs often Respondents who participated in a cross-­sectional survey

8 Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654

 Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
in a certain year may also participate in subsequent cross-­ REFERENCES
sectional surveys. However, estimating the proportion of 1 Shannon K, Crago A-­L, Baral SD, et al. The global response and
unmet actions for HIV and sex workers. Lancet 2018;392:698–710.
FSWs who repeatedly participated in this study was not 2 WHO. Report on global sexually transmitted infection surveillance,
possible. Third, the data were collected mainly through 2018, 2018. Available: https://www.​who.​int/​reproductivehealth/​
publications/​stis-​surveillance-​2018/​en/
self-­reported questionnaires, which raise concerns about 3 Gil VE, Wang MS, Anderson AF, et al. Prostitutes, prostitution
recall bias, reporting bias and social desirability bias. and STD/HIV transmission in mainland China. Soc Sci Med
1996;42:141–52.
For example, respondents may conceal the truth when 4 Hong Y, Li X. HIV/AIDS behavioral interventions in China: a literature
answering sensitive questions, such as drug use, which is review and recommendation for future research. AIDS Behav
2009;13:603–13.
illegal in China. 5 Hail-­Jares K, Chang RCF, Choi S, et al. Intimate-­partner and client-­
initiated violence among female street-­based sex workers in China:
does a support network help? PLoS One 2015;10:e0139161.
6 Wang L, Wang N, Wang L, et al. The 2007 estimates for people at
risk for and living with HIV in China: progress and challenges. J
CONCLUSIONS Acquir Immune Defic Syndr 2009;50:414–8.
7 Pan S, Parish WL, Huang Y. Clients of female sex workers: a
In conclusion, the HIV and HCV prevalence among FSWs population-­based survey of China. J Infect Dis 2011;204:S1211–7.
in Chongqing remained stable during the study period, 8 Su S, Chow EPF, Muessig KE, et al. Sustained high prevalence of
but syphilis had an upward trend among FSWs in low-­tier viral hepatitis and sexually transmissible infections among female
sex workers in China: a systematic review and meta-­analysis. BMC
and middle-­tier venues. Although factors such as knowl- Infect Dis 2016;16:2.
edge, condom use, injecting drug use and STI diagnosis 9 Tucker JD, Henderson GE, Wang TF, et al. Surplus men, sex work,
and the spread of HIV in China. AIDS 2005;19:539–47.
were improved, nearly half of FSWs used condom incon- 10 He N. Emerging changes and characteristics of the HIV epidemic in
sistently. Thus, more effective interventions for promoting China. Shanghai J Prev Med 2019;31:963–7.
11 CDC bureau. New progress in AIDS prevention and control in China
CCU among FSWs and their clients are needed to reduce in 2019, 2019. Available: http://www.​nhc.​gov.​cn/​jkj/​s3586/​201911/​
the spread of HIV/STIs, and more attention must be paid c238​8ce7​0bdd​404e​a6df​cd88​6591784d.​shtml [Accessed 30 Nov
2019].
to low-­tier and middle-­tier FSWs. 12 Poon AN, Li Z, Wang N, et al. Review of HIV and other sexually
transmitted infections among female sex workers in China. AIDS
Acknowledgements  The authors express their thanks to the individuals who Care 2011;23:5–25.
volunteered to participate in this study and Chinese Center for Disease Control and 13 Wang L, Tang W, Wang L, et al. The HIV, syphilis, and HCV
epidemics among female sex workers in China: results from a
Prevention, Chongqing, China for their contributions to the study.
serial cross-­sectional study between 2008 and 2012. Clin Infect Dis
Contributors  GW, RL and MY designed and coordinated the study. GW and RL was 2014;59:e1–9.
involved in the design and conducting of the survey. LH and MY analysed the data. 14 Zhang H, Hsieh E, Wang L, et al. HIV/AIDS among female sex
LH, HZ, DJ and HQ participated in the interpretation of results. LH wrote the first workers in China: epidemiology and recent prevention strategies.
Curr HIV/AIDS Rep 2020;17:151–60.
draft of the manuscript. LH, HQ, HZ, DJ, GW, RL and MY revised the manuscript. All
15 Liang J, Deng W, Huang S, et al. Estimation of the population size
authors have read and approved the manuscript. of female sex workers in Jiangmen with enumeration and census
Funding  This research was funded by the Chongqing Science and methods. J Diagnosis Ther Dermato-­venereology 2012;19:50–2.
Technology Commission of the People’s Republic of China in 2017 (Grant NO: 16 Sun K, Jing Z, Yang Q, et al. Population size of female sex workers
estimated by census method in the urban areas of Mengzi
cstc2017jcyjAX0185) and Chongqing Yuzhong District Science and Technology and Jianshui County, Honghe Prefecture. Chinese J AIDS STD
Bureau (20190116). 2019;25:294–6.
Map disclaimer  The depiction of boundaries on this map does not imply the 17 Chen X-­S, Liang G-­J, Wang Q-­Q, et al. HIV prevalence varies
between female sex workers from different types of venues in
expression of any opinion whatsoever on the part of BMJ (or any member of its southern China. Sex Transm Dis 2012;39:868–70.
group) concerning the legal status of any country, territory, jurisdiction or area or 18 National Health and Family Planning Commission of the People’s
of its authorities. This map is provided without any warranty of any kind, either Republic of China. 2015 China AIDS response progress report,
express or implied. 2015. Available: http://www.​unaids.​org/​sites/​default/​files/​country/​
documents/​CHN_​narrative_​report_​2015.​pdf
Competing interests  None declared. 19 Jia Y, Lu F, Sun X, et al. Sources of data for improved surveillance
Patient and public involvement  Patients and/or the public were not involved in of HIV/AIDS in China. Southeast Asian J Trop Med Public Health
the design, or conduct, or reporting, or dissemination plans of this research. 2007;38:1041–52.
20 Wang L, Wang N. HIV/AIDS epidemic and the development of
Patient consent for publication  Not required. comprehensive surveillance system in China with challenges. Chin
Med J 2010;123:3495–500.
Ethics approval  This survey was reviewed and approved by the Institutional 21 Sun X, Wang N, Li D, et al. The development of HIV/AIDS surveillance
Review Board of the National Center for AIDS Prevention and Control (NCAIDS), in China. AIDS 2007;21:S33–8.
Center for Disease Control (CDC), China, and the medical ethics committee of 22 Lin W, Chen S, Seguy N, et al. Is the HIV sentinel surveillance system
Chongqing Medical University (IRB number:2017016). adequate in China? findings from an evaluation of the National
HIV sentinel surveillance system. Western Pac Surveill Response J
Provenance and peer review  Not commissioned; externally peer reviewed. 2012;3:61–8.
23 Ghys PD, Jenkins C, Pisani E. Hiv surveillance among female sex
Data availability statement  No data are available.
workers. AIDS 2001;15:S33–40.
Open access  This is an open access article distributed in accordance with 24 Li P, Li D, Ge L, et al. Prevalence of HIV infection, syphilis and
the Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, hepatitis C in female sex workers in China, 2010-2015. Dis Surveill
2017;32:287–91.
which permits others to distribute, remix, adapt, build upon this work non-­
25 Guo W, Bao S, Lin W, et al. Estimating the size of HIV key affected
commercially, and license their derivative works on different terms, provided the populations in Chongqing, China, using the network scale-­up
original work is properly cited, appropriate credit is given, any changes made method. PLoS One 2013;8:e71796.
indicated, and the use is non-­commercial. See: http://​creativecommons.​org/​ 26 Ma D, Zhang Y, Zhou L, et al. Epidemic analysis on high-­risk HIV/
licenses/​by-​nc/​4.​0/. AIDS population in Chongqing from 2006 to 2012. J Third Mil Med
Univ 2015;37:689–92.
ORCID iD
Mengliang Ye http://​orcid.​org/​0000-​0001-9​ 394-​8682

Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654 9

 Open access

BMJ Open: first published as 10.1136/bmjopen-2019-036654 on 12 October 2020. Downloaded from http://bmjopen.bmj.com/ on June 19, 2023 by guest. Protected by copyright.
27 Yu H, Zhang W, Lu R, et al. Epidemiological characteristics and 40 Weller S, Davis K, Davis-­Beaty K. Condom effectiveness in reducing
secular trend of HIV infection and AIDS in Chongqing during 1995 heterosexual HIV transmission. Cochrane Database Syst Rev
and 2015. J Third Mil Med Univ 2017;39:2051–8. 2002:CD003255.
28 National Center for STD/AIDS Prevention and Control, Chinese 41 Ma Q, Jiang J, Pan X, et al. Consistent condom use and its correlates
Center for Disease Control and Prevention. Operational manual for among female sex workers at hair salons: a cross-­sectional study in
the implementation program of national AIDS sentinel surveillance, Zhejiang Province, China. BMC Public Health 2017;17:910.
2012. 42 Hao C, Guida J, Morisky DE, et al. Family network, workplace
29 Huang MB, Ye L, Liang BY, et al. Characterizing the HIV/AIDS network, and their influence on condom use: a qualitative
epidemic in the United States and China. Int J Environ Res Public study among older female sex workers in China. J Sex Res
Health 2016;13. 2015;52:924–35.
30 Sun Y, Li G, Lu H. Characteristics of female sex workers and their 43 Jie W, Xiaolan Z, Ciyong L, et al. A qualitative exploration of barriers
HIV and syphilis prevalence in Beijing, 2011-2016. Chinese J AIDS to condom use among female sex workers in China. PLoS One
STD 2019;25:260–3. 2012;7:e46786.
31 Sun L, Zhou D, Jiang F, et al. Ananalysis of HIV / AIDS sentinel 44 Wirtz AL, Peryshkina A, Mogilniy V, et al. Current and recent drug use
surveillance to high risk populations in Liaoning Province from 2012 intensifies sexual and structural HIV risk outcomes among female sex
to 2014. Chinese J Dis Control Prev 2015;19:1227–30. workers in the Russian Federation. Int J Drug Policy 2015;26:755–63.
32 Zeng X, Wang Z, Nie S. Survey on AIDS related knowledge, 45 Ulibarri MD, Strathdee SA, Ulloa EC, et al. Injection drug use as a
behaviors and infectious conditions among female sex workers in mediator between client-­perpetrated abuse and HIV status among
Hainan during 2011-2013. China Trop Med 2015;15:1207–9. female sex workers in two Mexico-­US border cities. AIDS Behav
33 Yang H, Su L, Ye L, et al. [Incidence rates of HIV-1 epidemics among 2011;15:179–85.
4 high risk groups in Sichuan province during 2011-2015]. Zhonghua 46 Jiang H, Lu R, Wu G. Analysis of AIDS sentinel surveillance results
Liu Xing Bing Xue Za Zhi 2017;38:1649–54. among drug users in Chongqing from 2010 to 2017. Mod Prev Med
34 Wang J, Jia M, Luo H, et al. [Analysis of the risky behaviors among 2019;46:1883–6.
HIV positive female sex workers]. Zhonghua Yu Fang Yi Xue Za Zhi 47 Hail-­Jares K, Choi S, Duo L, et al. Occupational and demographic
2015;49:978–82. factors associated with drug use among female sex workers at the
35 Chen Y, Shen Z, Morano JP, et al. Bridging the epidemic: a China-­Myanmar border. Drug Alcohol Depend 2016;161:42–9.
comprehensive analysis of prevalence and correlates of HIV, hepatitis 48 Goldenberg SM, Montaner J, Braschel M, et al. Dual sexual and
C, and syphilis, and infection among female sex workers in Guangxi drug-­related predictors of hepatitis C incidence among sex workers
Province, China. PLoS One 2015;10:e0115311. in a Canadian setting: gaps and opportunities for scale-­up of
36 Gong X, Yue X, Teng F, et al. Syphilis in China from 2000 to 2013: hepatitis C virus prevention, treatment, and care. Int J Infect Dis
epidemiological trends and characteristics. Chinese J Dermatology 2017;55:31–7.
2014;47:310–5. 49 Chen X-­S, Wang Q-­Q, Yin Y-­P, et al. Prevalence of syphilis infection
37 Jiang H, Guo X, Chen Y. Prevalence of sexually transmitted diseases in different tiers of female sex workers in China: implications for
in Chongqing municipality,2007-2014. Chinese J Public Heal surveillance and interventions. BMC Infect Dis 2012;12:84.
2016;32:1556–60. 50 Zeng H, Zhao Y, Meng S, et al. Exploring HIV prevention strategies
38 Guo X, Jiang H, Wu G, et al. Investigation and analysis on influencing among street-­based female sex workers in Chongqing, China. Int J
factors of syphilis and gonorrhoea epidemic report in Chongqing Environ Res Public Health 2015;12:855–70.
during 2010-2014. Lab Med Clin 2019;16:2144–6. 51 Zeng H, Zhang L, Zhao Y, et al. HIV prevention among street-­based
39 Shimelis T, Lemma K, Ambachew H, et al. Syphilis among people sex workers (SSWs) in Chongqing, China: interviews with SSWs,
with HIV infection in southern Ethiopia: sero-­prevalence and risk clients and healthcare providers. Health Soc Care Community
factors. BMC Infect Dis 2015;15:189. 2016;24:e173–80.

10 Hu L, et al. BMJ Open 2020;10:e036654. doi:10.1136/bmjopen-2019-036654