Abstract
Based on a cross-sectional survey with 1,022 female sex workers (FSWs) recruited from different types of commercial sex venues in Southwest China, we examined their Internet-using behaviors and explored the feasibility of Internet-based HIV/STI intervention in this population. About 75% of FSWs were Internet users; among them 57% were frequent users, and 40% had searched HIV/STI information online. Internet use was significantly associated with younger age, more schooling, higher income, and engagement in a social network of Internet users. Frequent use of the Internet was associated only with factors of the social environment, such as peers’ Internet use. Two thirds of Internet-using FSWs were willing to participate in an online HIV/STI prevention program. Multivariate analyses showed that willingness to participate in an online HIV/STI prevention intervention was significantly associated with higher Internet use and younger age. Our data suggest that Internet may offer a promising strategy to deliver low-cost HIV/STI prevention programs for FSWs in China.
Keywords: Internet use, Female sex workers, HIV/STI intervention, Feasibility, China
Introduction
As the number of Internet users has skyrocketed during the past decade, so has the use of Internet for health promotion and behavioral change [1]. Because of unique advantages, such as anonymity, broad accessibility, easy quality control, and low cost of delivery and scale-up, Internet-based prevention intervention has become a promising strategy for hard-to-reach and high-risk populations, including people at risk of HIV/STI [2]. During the past decade a multitude of online HIV/STI interventions or assessments have been developed [3]. Researchers have documented not only acceptability and efficiency of Internet-based HIV prevention survey research but also feasibility and initial efficacy of online HIV/STI prevention intervention [4, 5]. Although online HIV/STI prevention research has progressed rapidly in the past decade, limitations are apparent in the current literature. First, most existing studies were conducted in Western countries, particularly the USA, with very few initiatives in developing countries. Despite the explosive growth of Internet access in developing countries, studies to explore the feasibility of Internet-based HIV prevention in these countries have been limited. Second, most existing studies have focused mainly on MSM [6], which is not surprising since a majority of new HIV infections in developed countries occur in MSM. Some online interventions target adolescents in school [7, 8]. To date, virtually no data are available regarding Internet use among high-risk women, such as female sex workers (FSWs), particularly in developing countries, such as China, where heterosexual transmission has become a dominant mode of HIV transmission.
China has entered the third decade of the HIV/AIDS epidemic. Although the actual HIV sero-prevalence in China remains unknown, the current official government estimate of persons infected with HIV exceeds 700,000. Of reported cases, 60% occur in people 16–29 years of age [9]. In the last few years a notable shift has occurred in the HIV-transmission modes from intravenous drug use (IDU) to sexual transmission [9]. Sexual transmission accounted for about 7.2% of the total HIV infection in 2002, but the proportion jumped to 43.6% in 2005 [10]. Sexual transmission accounted for 60% of new infections in 2007 [9]. Indicative of the increase in heterosexual transmission is the rapid rise in new infections among women. The nationwide male-to-female ratio of people living with HIV/AIDS plunged from 9 in 1995 to 1.8 in 2007. The number of persons infected may be comparatively low for a country of 1.3 billion people; however, the sharp increase in new infections through sexual transmission and the alarming increase of infected women indicate a serious and rapidly deteriorating situation [9].
Considerable heterosexual transmission has occurred through commercial sex [11]. Thus millions of FSWs play a critical role in the country’s escalating HIV epidemic. As in other Asian countries, commercial sex in China is primarily establishment-based with a small proportion of “freelance” (i.e., self-employed). Typically, FSWs encounter their clients in entertainment establishments (e.g., karaoke bars (KTV), night clubs, dance halls, discos, bars) or personal service venues (e.g., hair washing rooms, barbershops, massage parlors, saunas, restaurants, mini-hotels). These FSWs are called “Xiaojie” (literally translated as miss in English) [12]. Currently, an estimated 10 million FSWs operate in a very complex hierarchy determined by venue characteristics [13]. FSWs in higher-level venues, such as night clubs and KTV, are strikingly different from FSWs in lower-level venues, such as streets and restaurants, in terms of age, income, and HIV-related behaviors [14]. In addition, notable variations exist within the same type of commercial sex venues [13, 15].
Existing literature on FSWs in China has revealed that this population is young (mostly in their early twenties) and highly mobile [13]. Most of them change work places every three months or less [15]. The literature also suggests that FSWs are at high risk for HIV. Specifically, FSWs’ consistent condom-use rates with clients are low, ranging from 15 to 35%. Many FSWs also have stable, noncommercial sex partners with whom they rarely use condoms. Because of such low rates of condom use, prevalence of STI among FSWs was as high as 30 to 85% in previous studies [13].
Despite the high HIV/STI risks in this population, prevention and intervention efforts among FSWs have been limited. Our recent systematic review regarding behavioral intervention of HIV/STI in China identified only seven intervention studies targeting FSWs [16]. These studies, mostly employed face-to-face approaches, produced positive but suboptimal outcomes, and faced several challenges in efficacy, feasibility, and sustainability [16]. First, the secrecy and stigma associated with commercial sex have kept many FSWs from participating in face-to-face HIV/STI prevention programs. In addition, FSWs’ mobility and migration rendered follow-up of this population so difficult that most of the existing studies targeting FSWs had to employ an open-cohort study design. Second, all interventions provided face-to-face education or counseling services to FSWs and were delivered individually or in small groups by trained interventionists. Such an intervention approach made it difficult to control the quality and fidelity of the program delivered by different interventionists. Furthermore, such interventions incurred a high cost in personnel and resources for community-based replications or scale-up. Challenges in existing intervention efforts have suggested an urgent need to develop an innovative intervention strategy that complements the face-to-face approach and can reach a large number of FSWs anonymously and effectively. Fast-growing Internet use in China provides a promising venue to meet such a need.
Not only does China have the largest population of all countries on earth, but it now also has the largest online population. The number of Internet users in China has increased 626 times since 1997, reaching 384 million in 2009 [17]. About 30% of the 1.3 billion Chinese people are now Internet users, and the proportion keeps increasing. About half of the Internet users in China are females, 70% of them live in the cities, and 70% are under 35 years old [17]. Thus urban youths and young adults are the major force among the Internet users. A population-based survey in seven Chinese cities revealed that of those younger than 30 years of age, 85% are Internet users; of those younger than 20 years of age, 90% are Internet users [18]. In recent years the Internet has been widely adopted by people with low levels of education. For example, 75% of current users do not have a college education, and 35% have no more than a junior high school education. Because of the low cost of Internet access, income level is not strongly connected with Internet access [17].
Within such a context of the rapid increase of heterosexual transmission of HIV and broad accessibility of the Internet in China, we conducted a study to explore Internet use among FSWs in China with an ultimate goal of developing an Internet-based HIV intervention targeting this at-risk population. Three research questions were asked in the current study. First, what is the prevalence of Internet use among FSWs and what are their online behavior patterns? Second, what factors contribute to their Internet use and frequency of use? Third, are most FSWs willing to participate in an online HIV/STI prevention program and what factors are associated with willingness to do so?
Methods
Study Sites
The current study was conducted in two cities, Guilin and Beihai in the Guangxi Zhuang Autonomous Region. Guangxi has witnessed an alarming increase in HIV prevalence in the past decade. A total of 48,073 HIV cases were reported by June 2009, placing Guangxi second among 31 Chinese provinces in terms of HIV seropositive cases. In the first half of 2009, new HIV cases were reported at a rate of 30 per day, placing Guangxi first in the nation in terms of new cases reported. The HIV epidemiological trend in Guangxi is similar to that in the rest of China. Traditionally, IDU was the major source of HIV infection, but sexual transmission has increased so rapidly that it has accounted for 65% of new infections from 2008 to 2009. HIV prevalence among FSWs ranged from 1 to 11% at sentinel surveillance sites in Guangxi [19]. Of 31 provincial regions in China, Guangxi was ranked 12th in terms of per capita income among urban residents [20]. Internet penetration in Guangxi was ranked 23rd, but its growth rate was ranked 9th [17].
Guilin and Beihai, two metropolitan areas in Guangxi, were selected as the study sites of the current study. Guilin is located in the northeast of Guangxi with a population of 1.34 million including an urban population of 620,000. Beihai is located on the Beibu Gulf in the southern part of Guangxi with a population of 1.36 million including an urban population of 550,000. Both cities have been top tourist spots in China, attracting 4–10 million tourists to each city each year. Because of the significant tourism, commercial sex flourishes in both cities. The local anti-epidemic stations have estimated at least 155 commercial sex venues with an estimated 2,000 FSWs working in each city. An additional 1,000 FSWs work in the suburban areas, totaling an estimated 3,000 FSWs in each metropolitan area.
Participant Recruitment and Data Collection Procedure
We used the data collection methods that were used in our previous study among FSWs in Guangxi [21]. Participants in this study were recruited from nine different kinds of commercial sex venues, representing different levels of the commercial sex hierarchy, including night clubs, saunas, KTV (karaoke), bars, hair salons, massage parlors, restaurants, mini hotels, and streets. The research team and local health workers identified entertainment establishments in Guilin and Beihai through ethnographic mapping. The owners/managers or other gatekeepers of these establishments were contacted for their permission to conduct research in their premises. Once we obtained permission from the gatekeepers, trained outreach health workers from the local anti-epidemic stations approached the women in the establishments to ask for their participation. A total of 1,022 women, including 515 in Beihai and 507 in Guilin, agreed to participate, provided written informed consent, and completed a self-administered questionnaire. Each participant received a small gift with cash value equivalent to US$4.50.
The survey was self-administered in paper–pencil format in private spaces in the establishments where participants were recruited. No one was allowed to stay with the participant during the survey except the interviewer who provided the participant with necessary assistance. For a small percentage of FSWs who were illiterate (less than 5%), the interviewer read the question to FSW, who marked her answer on the questionnaire (so that the interviewer could not see her answer). The questionnaire took about 45 min to complete. The study protocol was approved by the Institutional Review Boards at Wayne State University in the USA and Beijing Normal University in China.
Measures
Demographic information
Demographic information collected in the study included the participant’s age, ethnicity, residency (rural or urban household registration), education, marital status, length of working in the city (in months), working venue, and income. For the purpose of data analysis in the current study, we categorized ethnicity into Han and non-Han, education into no more than middle school versus more than middle school. Because of hierarchy of commercial sex establishments, the venues were categorized into four types by the mean income of FSWs at each venue: level one were those venues with mean income higher than 3,000 yuan each month (sauna); level two were those venues with mean income between 2,000 and 3,000 yuan (night club, KTV, bar); level three were those venues with mean income between 1,000 and 2,000 yuan (message parlor, hair salon); and level four were those venues with mean income less than 1,000 yuan (restaurant, mini hotel, and streets).
Internet use behaviors
Internet use behaviors included ever use of Internet (yes/no), hours spent online per week, frequency of accessing Internet (daily, several times a week, several times a month), locations of access (home, Internet café, workplace, library, school), having an email or messenger account (e.g., MSN, QQ) (yes/no), purposes of Internet use (e.g., news browse, search information, online shopping, movie, games, etc.), frequency of online chatting, meeting or having sex with cyber-friends (yes/no), engaging in sexual/pornography chat (yes/no), number of friends who use the Internet (none, some, most), and whether they ever searched HIV/STI information online (yes/no). Participants were categorized into two groups: never use or ever use. The users (i.e., those reported ever using the Internet) were further grouped into two groups based on the frequency of use: infrequent users (those who accessed the Internet less than once per week) and frequent users (those who accessed Internet at least once per week).
Willingness to participate in an Internet-based HIV/STI prevention
Willingness to participate in an Internet-based HIV/STI prevention program contained the following two questions: “Would you log into a HIV/STI prevention website if available?” “Are you willing to participate in an online HIV/STI prevention program?” The response options were “yes”, “no”, and “I don’t know”. In data analysis, we only included those who answered “yes” as willing to participate.
Data Analysis
First, participants’ demographic characteristics were examined, comparing Internet users and nonusers, using Chi-square (for categorical variables) and ANOVA (for continuous variables). Second, Internet use patterns of infrequent users and frequent users were compared using Chi-square and ANOVA. Third, to examine further the factors associated with Internet use, we performed multivariate logistic regression analyses. The dependent variables were ever use of Internet (among all participants) and frequent use of Internet (among Internet users), respectively. The independent variables included the individual characteristics (age, ethnicity, education, marital status, length of working, and income) and social environment (venue type and having Internet-using friends). To control for potential intra-class correlation (ICC) by venue due to cluster sampling, we used random effect models. Adjusted odds ratio (aOR) and 95% confidence intervals (95% CI) were used to examine the independent relationships between independent variables and dependent variables. Fourth, the association between participants’ willingness to participate in an Internet-based HIV/STI prevention (i.e., would log onto HIV/STI prevention website and willing to participate online HIV/STI prevention program) and key demographics as well as Internet use behaviors were examined using Chi-square and ANOVA. Finally, to examine further the factors associated with willingness to participate, multivariate logistic regression models were built. Similar to step 3, random-effect modeling was employed to control for venue-level ICC; aOR and 95% CI were used to examine individual and social environmental correlates of willingness. All statistical data analyses were performed using Stata 10.0.
Results
Demographics and Internet Use
The participants ranged in age from 15 to 50 years with a mean age of 24.9 (SD = 6.7). Most of them (84%) were of Han ethnicity, more than half of them (56%) were from rural areas, and 71% were never married. They had worked in the cities on average 44 months and made an average of 2,660 yuan each month. The income varied considerably between individuals and across different venues; about 27% FSWs worked in venues with mean income higher than 3,000 yuan/month, 57% in venues with mean monthly income between 2,000 to 3,000 yuan, and the rest in venues with monthly income less than 2,000 yuan.
As shown in Table 1, about 75% (772 out of 1,022) of FSWs in the current study ever used the Internet. Internet use was significantly associated with almost all key demographic characteristics. FSWs who were younger, non-Han, completed more than middle school, were never married, worked in the city longer, worked in higher-level commercial sex venues and had higher income were much more likely to use the Internet than their counterparts. For example, the mean age of Internet users was 23 years as opposed to 32 years in nonusers (F = 295.91, P < 0.0001). Among FSWs younger than 20 years, 95% used the Internet; among those aged 20 to 30, 82% used Internet, whereas among those older than 30, only 22% did so. Among FSWs who had more than middle school education, 87% used the Internet, whereas 72% of less educated FSWs did so (χ2 = 35.61, P < 0.0001). Similarly, 89% of single FSWs were Internet users, whereas only 45% of married FSWs were ever online (χ2 = 229.70, P < 0.0001). Among FSWs working in higher-income venues (i.e., night clubs, saunas, KTV, bars), more than 80% of them accessed the Internet compared to 51% of those working in lower-income venues, such as massage parlors and hair salons; and only 9% of FSWs in mini-hotels, restaurants, and streets ever used Internet (χ2 = 334.57, P < 0.0001). Internet-using FSWs earned an average 2,950 yuan/month compared to 1,720 yuan/month among nonusers (F = 26.91, P < 0.0001). Internet use was also strongly associated with perceived peer Internet use. For example, among FSWs without any Internet-using friends, only 25% were Internet users; among those with most friends using Internet, 93% were Internet users (χ2 = 356.94, P < 0.0001).
Table 1.
Demographic characteristics of FSWs by Internet use
| Total (n = 1,022) | Ever/never use | ||
|---|---|---|---|
| Never use (n = 236) | Ever use (n = 772) | ||
| Age, mean (SD) | 24.89 (6.67) | 32.01 (7.87) | 23.39 (5.18)**** |
| <20 year | 29.26% | 4.35% | 95.65% |
| 20–30 year | 55.48% | 17.99% | 82.01% |
| >30 year | 15.26% | 77.56% | 22.44% |
| Ethnicity (%) | |||
| Han | 84.54% | 24.77% | 75.23%*** |
| Non-Han | 15.46% | 13.92% | 86.08% |
| Residency | |||
| Urban | 44.37% | 20.67% | 79.33% |
| Rural | 55.63% | 25.09% | 74.91% |
| Education (%) | |||
| ≤Middle school | 63.89% | 28.64% | 71.56%**** |
| >Middle school | 36.11% | 13.28% | 86.72% |
| Marital status (%) | |||
| Never married | 71.82% | 10.76% | 89.24%**** |
| Ever married | 28.18% | 54.51% | 45.49% |
| Months of working in city, mean (SD) | 43.98 (35.81) | 53.08 (41.51) | 41.24 (33.46)**** |
| Venue levela (%) | |||
| Level 1 (>3,000 RMB) | 27.01% | 19.20% | 80.80%**** |
| Level 2 (2,000–3,000) | 57.05% | 11.32% | 88.68% |
| Level 3 (1,000–2,000) | 7.24% | 48.65% | 51.35% |
| Level 4 (<1,000) | 8.71% | 91.01% | 8.99% |
| Income, in 1,000 yuan, mean (SD) | 2.66 (2.35) | 1.72 (1.66) | 2.95 (2.45)**** |
| Have Internet-using friends (%) | |||
| None | 13.89% | 74.65% | 25.35%**** |
| Some | 34.05% | 27.30% | 72.70% |
| Most | 52.05% | 6.58% | 93.42% |
P < 0.005,
P < 0.0001
Venues are classified into four levels by the mean income of FSWS in each venue
FSWs’ Online Behavior Patterns
As shown in Table 2, among 772 Internet users, 57% (n = 445) were frequent users (accessed more than once a week). Frequent and infrequent users displayed significant differences in many aspects of online behaviors. For instance, about 37% of frequent users accessed Internet at home versus 22% of infrequent users. Among frequent users 58 and 96% had an Email account or online chatting accounts respectively as opposed to 29 and 84% in infrequent users. Frequent users were also more likely to engage in frequent online chatting, have cyber-friends, have met cyber-friends, have had sex with cyber-friends, and carry on sexual chat online compared to their counterparts. Among the Internet-using FSWs, about 40% had ever searched HIV/STI-related information online; 43% frequent users and 37% infrequent users had ever done so (χ2 = 114.94, P < 0.0001).
Table 2.
Internet-using FSWs online behaviors patterns
| Total (n = 772) | Infrequent usea (n = 327) | Frequent useb (n = 445) | |
|---|---|---|---|
| Online hours, mean (SD) | 7.21 (11.50) | 3.84 (4.44) | 13.60 (14.51)**** |
| Places of access | |||
| Internet café | 66.87% | 74.61% | 64.04%**** |
| Home | 30.02% | 22.29% | 36.85%**** |
| Other places | 13.40% | 15.17% | 8.54%**** |
| Have an email account | 42.41% | 28.97% | 58.01%**** |
| Have a QQ/MSN account | 85.34% | 85.19% | 96.83%**** |
| Purpose of Internet use | |||
| Online chat | 62.25% | 57.23% | 69.89%**** |
| News | 25.83% | 24.92% | 26.74% |
| Search info | 23.24% | 26.77% | 21.80% |
| Make friends | 13.23% | 9.54% | 16.85%*** |
| Online shopping | 10.01% | 8.92% | 11.46% |
| Email/MSN | 3.97% | 1.54% | 6.07%*** |
| Movie | 64.15% | 64.31% | 66.52%*** |
| Other | 16.07% | 8.00% | 20.22%**** |
| Online chatting | |||
| Never | 15.37% | 17.70% | 4.98%**** |
| Infrequent (<1/week) | 22.49% | 50.00% | 9.95% |
| Frequent (>1/week) | 59.15% | 32.30% | 85.07% |
| Have a cyber-friend | 70.44% | 67.78% | 83.33%**** |
| Have met cyber-friend | 26.96% | 20.25% | 35.07%**** |
| Have sex with cyber-friends | 5.18% | 3.70% | 7.03%* |
| Ever sexual/porn chat | 5.88% | 3.07% | 10.79%**** |
| Ever searched HIV/STI info online | 37.76% | 37.12% | 42.53%**** |
P < 0.05,
P < 0.005,
P < 0.0001
Infrequent use: access Internet less than once a week
Frequent use: access Internet more than once a week
Multivariate Models on Internet Use and Frequent Internet Use
As shown in Table 3, in the model to examine the factors associated with Internet use (ever use versus never use), age, education, venue level, having Internet-using friends and income were independently and significantly associated with Internet use. For instance, while controlling for all other factors, older age significantly decreased the odds of Internet use (aOR = 0.86, 95% CI = 0.81–0.90). Similarly, completing more than a middle school education significantly increased the odds of Internet use (aOR = 1.60, 95% CI = 1.01–2.53). Having some Internet-using friends increased the odds of Internet use by 3.92 (95% CI = 2.21–6.95), and having most Internet-using friends increased the odds by 11.65 (95% CI = 6.26–21.66).
Table 3.
Multivariate logistic regression of mixed effect model on FSWs’ Internet usea
| Total (n = 1,022) Ever/never aOR (95% CI) | Internet users (n = 772) Frequent/infrequent aOR (95% CI) | |
|---|---|---|
| Age | 0.86 (0.81–0.90)**** | 0.96 (0.91–1.01) |
| Ethnicity | 1.60 (0.86–2.98) | 1.17 (0.77–1.76) |
| Education (>middle school vs. ≤middle school) | 1.60 (1.01–2.53)* | 1.32 (0.96–1.81) |
| Marital status (married vs. unmarried) | 0.66 (0.39–1.10) | 0.83 (0.51–1.35) |
| Months of working in city | 1.00 (0.99–1.00) | 1.00 (0.99–1.00) |
| Venue level | ||
| Level 1 (>3,000 RMB) | Reference | Reference |
| Level 2 (2,000–3,000) | 1.72 (1.07–2.76)* | 1.92 (1.36–2.72)**** |
| Level 3 (1,000–2,000) | 0.82 (0.27–2.45) | 3.42 (1.47–7.94)*** |
| Level 4 (<1,000) | 0.82 (0.27–2.45) | - |
| Having Internet using friends | ||
| None | Reference | Reference |
| Some | 3.92 (2.21–6.95)**** | 1.52 (0.70–3.34) |
| Most | 11.65 (6.26–21.66)**** | 3.97 (1.85–8.51)**** |
| Income, mean (SD) | 1.24 (1.03–1.28)* | 0.99 (0.93–1.06) |
P < 0.05,
P < 0.0001
There are no frequent users in level 4 venue
Random effect model controlling for intra-class correlation (ICC) within each venue
In the model to examine the factors associated with frequent Internet use (frequent versus infrequent use), only venue level and having more Internet-using friends were independent and significant factors. Specifically, while controlling for all other factors, when compared to women in the venues with monthly income >3,000 yuan, women in venues with lower monthly income were more likely to use the Internet frequently (aOR = 1.92, 95% CI = 1.46–2.72 for monthly income between 2,000 and 3,000 yuan; aOR = 3.42, 95% CI: 1.47–7.94 for monthly income between 1000 and 2,000 yuan). Women with most friends using the Internet were 3.97 (95% CI = 1.85–8.51) times more likely to be frequent users compared to those without Internet-using friends.
Willingness to Participate in an Internet-Based HIV/STI Prevention Program
Table 4 depicts the association between willingness to participate in an Internet-based HIV/STI prevention program (would log onto a HIV/STI prevention website if available and willing to participate in an online HIV/STI prevention program) and key demographic characteristics as well as Internet-using behaviors. Internet users were much more likely to be willing to log on to the website (P < 0.0001) or participate in the online prevention program (P < 0.05) than nonusers. Women who have ever searched for HIV/STI info online were also more likely to accept the online program (P < 0.0001). Women aged 20–30 years, had worked in cities longer, or worked in higher-income commercial sex venues were more likely to report positive attitudes toward a online HIV/STI prevention program (P = 0–0.05). Education level and individual income were not significantly associated with willingness to participate in such a program.
Table 4.
Willingness for Internet-based HIV/STI prevention program among FSWs (n = 1,022)
| Would log on HIV/STI prevention website if available | Willing to participate online in HIV/STI prevention program | |||||
|---|---|---|---|---|---|---|
| Yes | No | Not sure | Yes | No | Not sure | |
| Internet use | ||||||
| Never | 41.56% | 22.51% | 35.93%**** | 54.94% | 9.44% | 35.62%* |
| Infrequent | 68.50% | 8.26% | 23.24% | 65.64% | 4.91% | 29.45% |
| Frequent | 63.60% | 11.01% | 25.39%% | 64.19% | 7.21% | 28.60% |
| Ever searched HIV/STI info online | 83.12% | 5.00% | 11.88%**** | 77.12% | 3.45% | 19.44%**** |
| Age | ||||||
| ≤20 | 58.45% | 11.49% | 30.07%* | 56.23% | 8.42% | 35.35%¶ |
| 21–30 | 63.85% | 11.87% | 24.28% | 64.98% | 5.96% | 29.06% |
| >30 | 49.67% | 18.54% | 31.79% | 65.79% | 7.89% | 26.32% |
| Education | ||||||
| ≤Middle school | 58.37% | 13.30% | 28.33% | 63.01% | 6.43% | 30.56% |
| >Middle school | 63.19% | 11.81% | 27.12% | 61.64% | 7.95% | 30.41% |
| Months of working in city, mean (SD) | 45.80 (35.69) | 43.05 (33.35) | 39.39* (36.56) | 46.10 (35.60) | 39.80 (35.30) | 39.60 (35.50)* |
| Venue level | ||||||
| Level 1 (>3,000 RMB) | 65.19% | 10.37% | 24.44%**** | 64.31% | 5.96% | 29.74%**** |
| Level 2 (2,000–3,000) | 60.42% | 12.33% | 27.26% | 61.46% | 6.94% | 31.60% |
| Level 3 (1,000–2,000) | 66.18% | 8.82% | 25.00% | 65.22% | 8.70% | 26.09% |
| Level 4 (<1,000) | 38.20% | 25.84% | 35.96% | 61.80% | 8.99% | 29.21% |
| Income, in 1,000 yuan, mean (SD) | 2.77 (2.37) | 2.51 (2.34) | 2.50 (2.34) | 2.75 (2.41) | 2.12 (1.63) | 2.61 (2.38) |
P < 0.1,
P < 0.05,
P < 0.0001
The multivariate logistic regression models in Table 5 show the results of further analysis of the relationship between willingness to participate in an online HIV/STI prevention program and the above key demographics and online behaviors. For both outcome variables assessing willingness, Internet use, having ever searched HIV/STI info online, and age were three independent and significant predictors. For example, compared to nonusers, Internet users were two to three times more likely to log on to an HIV/STI prevention website (aOR = 3.02 for infrequent users and aOR = 2.36 for frequent users, P < 0.0001). Women who had ever searched HIV/STI info online were much more likely to log on to the website (aOR = 4.70, 95% CI = 3.28–6.74) and willing to participate in the online program (aOR = 2.75, 95% CI = 1.98–3.82). Age was the only demographic characteristic that remained significant in the models (aOR = 1.05 and 1.06).
Table 5.
Multivariate logistic regression analysis on willingness to participate in Internet-based HIV/STI prevention program by key demographics and Internet-use patternsa (n = 1,022)
| Would log-on to HIV/STI prevention website if availableb aOR (95% CI) | Willing to participate online HIV/STI prevention programb aOR (95% CI) | |
|---|---|---|
| Internet use | ||
| Never use (reference) | Reference | Reference |
| Infrequent use | 2.98 (1.90–4.68)**** | 2.06 (1.33–3.18)*** |
| Frequent use | 2.33 (1.48–3.69)**** | 2.12 (1.36–3.30)*** |
| Ever searched HIV/STI info online | 4.71 (3.29–6.74)**** | 2.75 (1.98–3.82)**** |
| Age | 1.03 (0.99–1.08)** | 1.05 (1.01–1.09)*** |
| Education (>middle school vs. ≤middle school) | 1.02 (0.75–1.38) | 0.86 (0.64–1.16) |
| Months in city | 1.00 (1.00–1.01) | 1.00 (1.00–1.01) |
| Venue level | ||
| Level 1 (>3,000 RMB) | Reference | Reference |
| Level 2 (2,000–3,000) | 0.85 (0.53–1.37) | 1.01 (0.66–1.57) |
| Level 3 (1,000–2,000) | 1.49 (0.69–3.24) | 1.22 (0.59–2.51) |
| Level 4 (<1,000) | 0.52 (0.19–1.46) | 0.86 (0.33–2.22) |
| Income | 1.00 (0.93–1.07) | 1.03 (0.97–1.11) |
P < 0.01,
P < 0.005,
P < 0.0001
Random effect model controlling for intra-class correlation (ICC) within each venue
The dichotomous value was coded by grouping those answer “yes” into 1 and those who answer “no” and “I don’t know” into 0
Discussion
The above data revealed that most FSWs (75%) were Internet users; among them 56% were frequent users, more than 40% had searched HIV/STI information online, 65% would log on to an HIV/STI prevention website if available, and 64% were willing to participate in an online HIV/STI prevention program. Such high rates of Internet use and willingness to participate in an online HIV/STI intervention suggest that the Internet could be a promising venue to deliver the intervention for this population.
Internet use varied significantly by participants’ demographic and social factors (Table 1). Age is the most significant demographic predictor in FSWs’ Internet use. The mean age difference between the users and non-users was 9 years (23 vs. 32 years). Among FSWs younger than 30 years, 90% used the Internet compared to only 22% in those older than 30 years. Education and income were two other factors associated with Internet use (Table 3). Such findings were consistent with national representative surveys in China, which reported that among urban residents, 85% of those younger than 30 accessed the Internet and Internet users had higher level of education and income [17, 18]. Thus in terms of Internet use, FSWs were very similar to other female young adults in the city. Other than demographic characteristics, factors of social environment also play significant roles in FSWs’ Internet use. FSWs who work in entertainment venues (e.g., karaoke, bar, and night club) were much more likely to use the Internet than those working in other venues (e.g., mini-hotels, restaurants, streets). Peer influence significantly affected Internet use; specifically, having some Internet-using friends increased the chance of Internet use by four times and having mostly Internet-using friends increased the chance by 11 times. Notably, age, education, and income were only significantly associated with Internet use, but not frequent use, which was affected only by factors of social environment, including type of venue and peer Internet use.
Our study also provides important data on patterns of Internet use. For example, most FSWs accessed the Internet in Internet cafés, and frequent users spent much longer hours online. Most of them had a messenger account (e.g., MSN, QQ), and half of them had an email account. The most popular online activities were online chatting (70%) and movies (67%). Such findings were also consistent with earlier reports on Internet use in the Chinese population for which the Internet is more a communication and entertainment tool than an information tool. A majority of Chinese prefers instant communication (e.g., phone, text message, online messenger) over delayed communication (e.g., Email, voicemail) [18]. These data provide valuable information for the development of a culturally appropriate Internet-based HIV/STI prevention program for this population. Our data also reveal that although only a small proportion (7%) of FSWs were engaged in pornography or sexual chat online, more than 40% of FSWs had sex with cyber-friends. Further study is necessary to determine whether the Internet has become a “risk environment,” in which commercial sex and other sexual risks are mingled.
Internet use served as the single most important predictor in willingness to participate in an online HIV/STI prevention program. The level of willingness did not differ between frequent or infrequent users. Even among nonusers, 42–55% expressed positive attitudes toward the online prevention program, and about one third of them were not sure. Such high rates of willingness suggest that Internet could serve as a convenient resource as well as effective intervention platform for FSWs in China. Because of the severe stigma attached to commercial sex, high mobility and large size of this population, Internet-based HIV/STI prevention for FSWs is especially beneficial.
Existing Internet-based interventions underscore the importance of tailoring online programs to the target populations [4, 5, 22]. Our data suggest that the online program should be tailored to young FSWs working in higher-income commercial sex venues. Since FSWs Internet use are strongly affected their peers; the Internet may not be an appropriate venue for recruitment, instead, peer outreach or snowballing may be an effective approach to sample large numbers of Internet-using FSWs. Location of access and privacy protection should also be considered. Because of the high mobility of this population, most of them do not have personal computers. Internet cafés remain the most popular locales for Internet access. However, as the low-cost “netbook” (laptop for surfing the Internet) and “3G” (third-generation) cell phones enter the market, we expect that more and more FSWs will access the Internet at home or in private settings. Email counseling that was proved effective in Western countries might not work for Chinese FSWs because only 42% of Internet users had an email account. Instead, instant messaging or online forums may be effective approaches to reach this population [23, 24]. The program should also be designed to fit the educational background of this population; it can also incorporate different modules to fit characteristics of the subgroups.
Despite the high promise of an Internet-based intervention for this population as suggested in the data, a number of challenges exist. For example, young and more educated FSWs with higher income were more likely to use the Internet; this group may also have lower sexual risks compared to their counterparts, e.g., older FSWs walking the streets. Further study is needed to determine how to reach the group at high risk for HIV. Adoption of Internet use through peer influence may be an effective approach but deserves further assessment. As many scholars have noted, the Internet shows promise in reaching some groups in a cost-effective manner, but it does not necessarily fit all populations [2, 5]. Interventions with face-to-face interactions are still better for motivating and training individuals [25]. Internet-based programs may supplement instead of replace in-person programs.
A number of limitations exist in the current study. First, our study was conducted in Guangxi, a multiethnic region of China. Current findings may not be generalizable to other areas of China. Second, we might have oversampled the FSWs working in higher-income commercial sex venues. Our sampling scheme was based on the map of commercial sex venues maintained by the local CDC. The map was updated during the formative phase of our project. Perhaps women in the lower level of the commercial sex hierarchy were less visible (most did not have regular venues) and therefore were not included in our ethnographic mapping. However, the demographic characteristics of the current sample were consistent with other studies of FSWs in urban areas of China [26–28]. Third, in the analysis of factors associated with the willingness to participate in online HIV/STI interventions, factors such as FSWs’ HIV-related risk behaviors and perceptions were not included. These factors along with others may affect FSW’s current behaviors to search HIV/STI information online as well as their acceptance of Internet-based interventions.
Despite these limitations, this paper represents the first study on Internet use among FSWs in China. The high prevalence of Internet use and the willingness to participate in an online HIV/STI prevention program suggest the feasibility of delivering such intervention for this highly mobile and stigmatized population. With expanding access to the Internet in China and other developing countries, online delivery of HIV/STI prevention program in resource-limited settings appears promising [29].
Acknowledgments
The study was supported “in part” by NIH Research Grant R01AA018090 by the National Institute for Alcohol Abuse and Alcoholism (NIAAA). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIAAA/NIH.
Contributor Information
Yan Hong, Department of Social and Behavioral Health, School of Rural Public Health, Texas A&M Health Science Center, 1266 TAMU, College Station, TX, USA.
Xiaoming Li, Prevention Research Center, Carman and Ann Adams, Department of Pediatrics, Wayne State University, School of Medicine, Detroit, MI, USA.
Chen Zhang, Department of Social and Behavioral Health, School of Rural Public Health, Texas A&M Health Science Center, 1266 TAMU, College Station, TX, USA.
References
- 1.Bennett GG, Glasgow RE. The delivery of public health interventions via the Internet: actualizing their potential. Annu Rev Public Health. 2009;30:273–92. [DOI] [PubMed] [Google Scholar]
- 2.Swendeman D, Rotheram-Borus MJ. Innovation in sexually transmitted disease and HIV prevention: internet and mobile phone delivery vehicles for global diffusion. Curr Opin Psychiatry. 2010;23(2):139–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pequegnat W, Rosser BR, Bowen AM, et al. Conducting Internet based HIV/STD prevention survey research: considerations in design and evaluation. AIDS Behav. 2007;11:505–21. [DOI] [PubMed] [Google Scholar]
- 4.Noar SM, Black HG, Pierce LB. Efficacy of computer technology-based HIV prevention interventions: a meta-analysis. AIDS. 2009;23:107–15. [DOI] [PubMed] [Google Scholar]
- 5.Ybarra ML, Bull SS. Current trends in Internet- and cell phone-based HIV prevention and intervention programs. Curr HIV/AIDS Rep. 2007;4:201–7. [DOI] [PubMed] [Google Scholar]
- 6.Rietmeijer CA, McFarlane M. STI prevention services online: moving beyond the proof of concept. Sex Transm Dis. 2008; 35:770–1. [DOI] [PubMed] [Google Scholar]
- 7.Lou CH, Zhao Q, Gao ES, Shah IH. Can the Internet be used effectively to provide sex education to young people in China? J Adolesc Health. 2006;39(5):720–8. [DOI] [PubMed] [Google Scholar]
- 8.Halpern CT, Mitchell EM, Farhat T, Bardsley P. Effectiveness of web-based education on Kenyan and Brazilian adolescents’ knowledge about HIV/AIDS, abortion law, and emergency contraception: findings from TeenWeb. Soc Sci Med. 2008;67: 628–37. [DOI] [PubMed] [Google Scholar]
- 9.China Ministry of Health, WHO. Joint report on HIV/AIDS Epidemic Update in China 2005–2007. Beijing: China Ministry of Health; 2007. [Google Scholar]
- 10.China Ministry of Health, UNAIDS, WHO. Update on the HIV/STD Epidemic and Response in China 2005. Beijing: China Ministry of Health; 2006. [Google Scholar]
- 11.Yang H, Li X, Stanton B, Liu H, Liu H, Wang N, et al. Heterosexual transmission of HIV in China: a systematic review of behavioral studies in the past two decades. Sex Transm Dis. 2005;32(5):270–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Liao SS, He Q, Choi K, et al. Working to prevent HIV/STDs among women in the sex industry in a rural town of Hainan, China. AIDS Behav. 2006;10(suppl):S35–45. [DOI] [PubMed] [Google Scholar]
- 13.Hong Y, Li X. Behavioral studies of female sex workers in China: a literature review and recommendation for future research. AIDS Behav. 2008;12(4):623–36. [DOI] [PubMed] [Google Scholar]
- 14.Huang Y, Henderson GE, Pan S, Cohen MS. HIV/AIDS risk among brothel-based female sex workers in China: assessing the terms, content, and knowledge of sex work. Sex Transm Dis. 2004;31:695–700. [DOI] [PubMed] [Google Scholar]
- 15.Hong Y, Li X, Yang H, Fang X, Zhao R. HIV/AIDS-related risks and migratory status among female sex workers in a rural Chinese county. AIDS Care. 2009;21:212–20. [DOI] [PubMed] [Google Scholar]
- 16.Hong Y, Li X. HIV/AIDS behavioral interventions in China: a literature review and recommendation for future research. AIDS Behav. 2009;13(3):603–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.China Internet Network Information Center. The 23rd report on China Internet use and development. http://www.cnnic.cn/index/0E/00/11/index.htm. Accessed 5 April 2010.
- 18.Guo L (2007). Survey report on Internet usage and impact in seven Chinese cities, November 2007, China Academy of Social Sciences. http://www.markle.org/downloadable_assets/china_internet_survey_11.2007.pdf. Accessed 5 April 2010. [Google Scholar]
- 19.Guangxi Center for Disease Control and Prevention. Update on HIV/AIDS epidemic in Guangxi. In: Presented at workshop of NIAAA venue-based HIV and alcohol risk reduction among female sex workers in China, 19–21 July 2009, Guilin, Guangxi. [Google Scholar]
- 20.China Statistics Press. China statistics year book 2008. http://www.stats.gov.cn/tjsj/ndsj/2008/left_.htm. Accessed 6 April 2010.
- 21.Li X, Wang B, Fang X, et al. Short-term effect of a cultural adaptation of voluntary counseling and testing among female sex workers in China: a quasi-experimental trial. AIDS Educ Prev. 2006;18(5):406–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kok G, Harterink P, Vriens P, et al. The gay cruise: developing a theory- and evidenced-based Internet HIV-prevention intervention. Sex Res Social Policy. 2006;3:52–67. [Google Scholar]
- 23.Wynn LL, Foster AM, Trussell J. Can I get pregnant from oral sex? Sexual health misconceptions in e-mails to a reproductive health website. Contraception. 2009;79(2):91–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Moskowitz DA, Melton D, Owczarzak J. PowerON: the use of instant message counseling and the Internet to facilitate HIV/STD education and prevention. Patient Educ Couns. 2009;77(1):20–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Temesgen Z, Knappe-Langworthy JE, St. Marie MM, et al. Comprehensive Health Enhancement Support System (CHESS) for people with HIV infection. AIDS Behav. 2006;10:35–40. [DOI] [PubMed] [Google Scholar]
- 26.Rogers SJ, Ying L, Xin YT, Fung K, Kaufman J. Reaching and identifying the STD/HIV risk of sex workers in Beijing. AIDS Educ Prev. 2002;14:217–27. [DOI] [PubMed] [Google Scholar]
- 27.Lau JT, Tsui HY, Siah PC, Zhang KL. A study on female sex workers in southern China (Shenzhen): HIV-related knowledge, condom use and STD history. AIDS Care. 2002;14:219–33. [DOI] [PubMed] [Google Scholar]
- 28.Yang X, Xia G, Li X, Latkin C, Celentano D. Social influence and individual risk factors of HIV unsafe sex among female entertainment workers in China. AIDS Educ Prev. 2010;22:69–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Internet World Stats. World Internet users and population stats. http://www.internetworldstats.com/stats.htm. Accessed 6 April 2010.