Table 2.
Characteristics and findings of quantitative studies included in the systematic review
| Author, year | Country and setting | Design | Sample size | Participants’ characteristics | Awareness | Willingness to use | Factors associated with willingness to use |
|---|---|---|---|---|---|---|---|
| Ayala et al., 2013 | 145 countries- Africa, Asia, Europe and Latin America. | Online survey. | 2774 MSM.α | Age range was 12–90 years. | 69.8% of the respondents were aware.α | 80.8% were willing to use PrEP. α | PrEP stigma (β: −0.51; 95% CI= −0.55 to −0.48, p<0.001), outness (β: −0.15; 95% CI= −0.18 to −0.12, p<0.001), and knowledge about PrEP (β: −0.14; 95% CI= −0.18 to −0.10, p<0.001) were negatively correlated with acceptability of PrEP. Acceptability of PrEP was positively correlated with having experienced service provider stigma (β: 0.12; 95% CI=0.02–0.23, p=0.021). Respondents in high-income countries reported lower acceptability of PrEP than those from LMIC. |
| Ding et al.,2016 | Shanghai, China. | Survey. | 1,033 MSM. | 76% self- identified as gay and 2.5% were already using PrEP. | Not reported. | 19.1% willing to use PrEP. | Willingness to use PrEP was associated with older age (≥ 45 years (Adjusted Odds Ratio (AOR):2.18; 95% (Confidence Interval (CI)=1.13–4.23, p=0.006), immigration to Shanghai (21.5% among immigrants vs 15.9% among local residents; AOR: 1.69; 95% CI=1.16–2.45, p=0.026), two or more male sex partners in the past 6 months (AOR:1.53; 95% CI=1.07–2.17, p=0.02). Condom use at last anal sex with man were significantly less willing to use PrEP (AOR:0.68; 95% CI=0.47–0.97, p=0.034). Education, occupation, gay sexual identity, and marital status were not associated with willingness to use it. |
| Draper et al., 2016 | Yangon and Mandalay in Myanmar. | Survey. | 434 GMTβ | Not reported. | 5% aware. | 62% were willing to use PrEP among 434 HIV undiagnosed GMT. | Willingness to use PrEP was associated with reporting never/occasional use of condoms compared to always/mostly used, with casual partners (adjusted odds ratio (AOR: 2.02; 95% CI=1.00–4.10), residence Mandalay (AOR:1.79; 95% CI=1.05-3.03), perceiving as likely to become HIV positive (AOR:1.82; 95% CI=1.10–3.02), having had more than one regular partner (AOR=2.94; 95% CI=1.41–6.14) or no regular partners (AOR:2.05; 95% CI=1.10–3.67) or more than five casual partners (AOR:2.05; 95% CI=1.06–3.99) or no casual partners (AOR:2.25; 95% CI=1.23–4.11) in the past three months. Those reporting concerns about PrEP side-effects due to long-term use were less likely to be willing to use it (AOR:0.35; 95% CI=0.21–0.59). |
| Eisingerich et al.,2012 | Peru, India, South Africa. | Survey | 383 MSMγ | Mean age of MSM not reported; 39% were aged 16-24, and 6% were aged ≥41 years. | Not reported | 69% reported ‘yes, definitely’ and 25% ‘yes, probably’ across India, Peru and South Africaβ. | 42–69% reported that PrEP would give them “a lot of hope”. 3–8% reported that PrEP would be “very embarrassing” to take. Indian and Peruvian MSM preferred bimonthly injection in the buttocks while South African MSM preferred daily pill to arm injection. Of those willing to use PrEP, 32–72% were willing to use PrEP despite side effects; 39–88% were willing to use it despite having to pay, 32–85% were willing to use it even if having to use condoms, and 55–88% were willing to use it with regular HIV testing. |
| He et al., 2014 | China. | Survey. | 1323 MSMδ | Mean age=28 years. | Overall, 31.4% had heard of PrEP. | Not reported. | Factors affecting use were not reported; however, the study reported that information regarding PrEP should be promoted through media to make sure MSM in China can get the information quickly and easily. |
| Hoagland et al., 2016 | Brazil. | Cross-sectional study. | 1131 MSMε | Median age=29 years 46.8% were HIV positive. | 61.3% were aware. | 82.1% were willing to use PrEP. | Willingness to use PrEP was higher among those aware of PrEP compared to those unaware of it (85.4% vs 76.9% ;p<0.001), among those with more years of schooling (78.1% among those with <12 years vs 84.5% among those with ≥12 years of schooling; p=0.006), and those with a recent STD diagnosis in last 12 months compared to those without (68.8% vs 60.2%; p=0.02). Willingness to use PrEP was not associated with age. Compared with those aged 18–24 years, willingness to use PrEP increased marginally among 25–35year olds (81.02%, vs 81.7%; p=0.85) and among those aged ≥36years (81.02% vs 84.1%; p=0.35). Willingness to use was not associated with male gender compared to transgender (81.8% vs 89.3%; p=0.16), a negative compared to a positive HIV test result (62.7% vs 46.8%; p=0.99), or failure to perform a test (62.7% vs 65.4%; p=0.40). 75.8 % reported they would use PrEP even if they had to pay for it. |
| Jackson et al., 2012 | Guangxi, Sichuan and Chongqing, China. | Survey. | 570 MSM | Mean age=27.6 years; age range=18–62 years, and 76.8% were urban dwellers. | Not reported. | 63% had high willingness to use PrEP, while 22.8% had lower willingness to use it. | Willingness to use PrEP was associated with urban compared than rural residence, higher education attainment (2.2% among primary school, vs 10.3% among middle school, vs 38.2% high school vs 49.3% among those with undergraduate or higher education; p<0.001), lower monthly personal income (37.9% among those earning 1,000 Yuan or less vs 2.8% among those earning 5,000 Yuan or more. Occupational status and previous experience of STI were not associated with willingness to use PrEP. Stigma of PrEP was a potential barrier, while perceived benefits of PrEP was a facilitators of potential use. |
| Ko et al., 2016 | Taiwan. | Online survey. | 1151 MSM | Mean age=25.9 years, age range=18–53 years, most were from the north (48.5%), had professional qualification (61.2%) and were employed (57.0%). | Not reported | 56% were willing to use PrEP. | Of those willing to use PrEP, 70% were willing to take pills before and after sex, 61% were willing to take PrEP to prevent getting HIV, 43.7% were willing to take a pill daily, 44.4% were willing to take PrEP even if it was not 100% effective, and only 23% were willing to self-pay Taiwan $ 340 for PrEP. Willingness to use PrEP increased with tertiary compared with secondary education (30.7% vs 2.2%; p<0.05), and among those with professional qualification (54.8% vs 12.2%; p<0.05), and a past history of receiving HIV non-occupational PEP (5.9% vs 3.2%; p value <0.01). There was no difference in age or employment between participants who were willing to use PrEP and those who were not. |
| Lim et al., 2016 | Kualar Lumpur, Malaysia. | Survey (online). | 990 MSMϕ | 80.4% self-identified as homosexual and 16.6% as bisexual. Age range=16–68. Overall 19.6% were aged <25 years. In addition, 87.2% had post-secondary education and 85.2% were in part-time or full time employment. | 44% were aware of PrEP. | 39% were willing to use PrEP. | Recent STI diagnosis in the past 12 months was associated with high likelihood to use compared to those with no such diagnosis (43.3% vs 36.1%; p=0.003). Malay participants more likely to use PrEP (48.6%) compared to Chinese (32.7%), Indian (32.8%) and mixed and other races (36.7%; p<0.001). Willingness to use PrEP was not associated with age, residence in Kuala Lumpur, education, employment status, or income. A third (35.6%) were willing to pay for PrEP. However, of these the majority (88.3%) were not willing to spend over 200 RM (USD 50) on PrEP per month. Of the 603 participants who reported not willing to use PrEP, the reasons offered were side effects (18.6%), fear that PrEP won’t work (9.8%), worry about forgetting to take medication (8.3%), or what other people might think of them (5.8%), failure to afford PrEP (8.8%), or the fact that they always use a condom and therefore would not need PrEP (11.4%). |
| Oldenburg et al., 2016 | Ho Chi Minh city, Vietnam. | Survey. | 300 MSMγ | 93.7% were HIV negative, and 27% were aged 15–19 years. | Not reported. | 95.4% were willing to use PrEP daily.η | Overall, 56.7% willing to take PrEP given side effects, and 27.7% preferred a PrEP lubricant to a pill. Previous contact with Peer Health Educators was associated with higher willingness to use (AOR: 2.28; 95% CI=1.25–4.14, p<0.05). |
| Peinado et al., 2013 | Lima, Iquitos and Pucallpa, Peru. | Survey (secondary analysis). | 532 MSM and TGι | Median age=28 years; range 16–68 years. | Not reported. | 96.2% were willing to use oral PrEP while 91.7% were willing to use rectal PrEP | After adjustment for age, city, and education, only being receptive most of the time (AOR: 9.1; 95% CI=1.8–46.5, p=0.01) and exclusively receptive (AOR:7.5; 95% CI=1.6-53.2, p=0.01) during anal intercourse, compared to being versatile, were independently associated with acceptability to use oral PrEP. |
| Sineath et al., 2013 | Thailand. | Survey (online) | 404 MSM.φ | Mean age was=25 years. | 7% were aware of PrEP. | 36% were willing to use after PrEP was described. | Of those willing to use PrEP 65% indicated they would be willing to pay for it. Overall, 34% “didn’t want to have to take medication every day” and 28% “didn’t want to go see the doctor every three months”. In addition, 35% believed condoms were more effective than PrEP. |
| Wei et al., 2011 | Guangxi, China. | Survey (face to face). | 650 MSM. | Mean age=28 years | 19.7% had heard about PrEP. | 91.9% were willing to use PrEP if free and safe. | Side effects and efficacy of PrEP were reported as influencing willingness to use. |
| Wheelock et al., 2013 | Bangkok and Chiang Mai, Thailand. | Survey. | 260 MSM. | 4% and 54% were 16–18 and 19–24 years, respectively. Eligible participants were at least 16 years. 94% had post-secondary education. | Not reported. | 39.2% reported they would ‘definitely’ and 49.2% would ‘probably’ use PrEP. | Of those willing to use PrEP, 58.8% were ‘definitely’ while 35% were ‘probably’ willing to use PrEP despite having to pay 500 Baht a month for it. 2.7% reported that taking PrEP would be ‘very embarrassing’ and 5.8% reported that it would be ‘fairly embarrassing’. Daily pill was the preferred route of administration followed by a monthly injection in the arm. After learning of potential mild side effects, 24.6% were ‘definitely’ and 56.5% ‘probably’ willing to use PrEP. |
| Xia et al., 2016 | Wuhan and Shanghai, China. | Survey. | 487 MSM | Mean age=28; range 18–62; years. 31.7% were aged 18–24 and 53.5% were aged 25–34 years. 81.1% self-identified as gay, and 16.2% as bisexual. 73% were educated to college level, 61.4% were employed, 51.1% earned between 2001–5000 RMB and 7% had been diagnosed with an STI in the last year. | 19.1% aware. | 71.3% willing to use. | Willingness to use PrEP was associated with marital status: 84.4% of those married/cohabiting were willing to use PrEP versus 67.5% of unmarried/divorced or widowed (p=0.001). Bisexual (77.2%) were more likely to use it compared to gay participants (71.1%) or /other/unsure (38.5%; p=0.017). Willingness to use was associated with taking an STI test in the last 12 months compared to those that didn’t (76% vs. 63.4%; p=0.007) but was inversely associated with being diagnosed with an STI in last 12 months compared to those not diagnosed, though not significant (67.6% vs 72.6%; p=0.065). Men using the internet were more likely to report willingness to use PrEP compared to those who heard about PrEP face-to-face (75.2% vs 66.4%; p<0.05). Willingness to use was not associated with age or duration of residency in the city. |
| Xue et al. 2015 | China. | Survey (online). | 760 MSMκ | 77.2% self-identified as homosexual and 20% remainder as bisexual | 72.8% aware of, or fully understood PrEP. | 32.1% would possibly use PrEP. | 61% (305/500) would possibly take PrEP orally daily. Factors that were identified by participants as preventing willingness to use PrEP were: side effects (60.8%), low self-risk assessment (54.2%), privacy and confidentiality (41.6%), the perception that PrEP is not 100% effective (38.3%), cost (28.7%), inconvenience of taking daily medication (68.7%), and reporting that risk behaviors were not happening daily (59%). |
| Yang et al., 2012 | Chiang Mai, Thailand. | Survey. | 131 MSMλ | Mean age=23.7; range 18–49 years.μ 13% self-identified as heterosexual, 16% as bisexual and 71% as gay. | 66% aware of PrEP. | 41% willing to use PrEP.ν | Willingness to use PrEP among MSM was associated with having zero regular partners in the preceding 6 months vs. one or more partners (OR: 2.25; 95% CI=1.09–5.11, p=0.04); regularly planned sex vs. unplanned sex (OR:2.83; 95% CI=1.12–7.12, p=0.01); infrequent sex (once per month or less) vs. two or more sexual encounters per month (OR:2.36; p=0.02); a lifetime history of STIs vs. no history of STIs (OR 3.78, 95% CI=1.42–10.47, p<0.01); age 25 years or older vs. age less than 25 years (OR:2.30; 95% CI=1.10–4.79, p=0.02); and being “very confident” in the ability to take daily, oral medicines for 1 year vs. not being “very confident” (OR:2.63; 95%CI=1.12–6.24, p=0.01). In contrast, willingness to use was not associated with a lifetime history of HIV testing vs. no history of HIV testing (OR:1.95; 95% CI=0.89–4.29, p=0.07) or receptive anal sex positioning vs. insertive or versatile positioning (OR:0.47; 95% CI=0.17–1.19, p=0.08) |
| Zhang et al., 2013 | Chongqing, Guangxi, and Sichuan, China. | Survey. | 1402 MSMο | Age range=18–74 years. 18–24 years comprised 41.5% of the sample. Majority (75.1%) resided in urban areas. 70% self-identified as homosexual and 21% as bisexual. | 22% were aware of PrEP | 64% were willing to use PrEP if safe and effective. | Proportion willing to use PrEP increased to 71% if it were to be made free, and to 77% if it were free and had been used by people known to participants. However, only 30% and 37% were willing to use it if it had to be taken once daily or a weekly respectively. Willingness to use PrEP was associated with lower education up to middle school compared to those with college education and above (68.4% vs 59.5%; p=<0.001), married marital status compared to never married (69.7% vs 62.4%; p=0.035); moderate (1000-3000) monthly income (compared to lower earnings of <1000; (p=0.013) but not compared to high monthly income of >3000 (p=0.109); and STI history compared to those without STI history (71.9% vs 62.6%; p=0.027). Participants who did not or rarely found sexual partners on the internet were more likely to be willing to use PrEP compared with higher risk participants, who often or sometimes found sexual partners on the internet. Willingness to use was not associated with age or residence or sexual identity. |
| Zhou et al., 2012 | Beijing, China. | Survey. | 152 MSMπ | Age range=18–61 years. 84.9% self-identified as homosexual and 15.1% as bisexual. | 11.2% aware of PrEP. | 67.8% were willing to ‘definitely’ or ‘probably’ take PrEP if available. | Willingness to use PrEP was associated with young age <30 years versus ≥30 years (68.8% vs. 83.9%; p=0.04). Willingness to use PrEP was not associated with years of education (80% among those with <12 vs 68.1% among those with >12 year of education; p=0.09), marital status (single/divorced/ widowed versus married/cohabiting [73.1% vs 77.1%; p=0.60), local Beijing residence versus non-Beijing residence (63.3% vs 77.1%; p=0.13), lower monthly income (RMB) <2000 versus >2000 (77.1% vs 72%; p=0.47; bisexual orientation versus homosexual (73.7% vs 74.7%; p=0.89], or previous diagnosis of STD in the past 6 months versus no such diagnosis (88.9% vs 72.4%; p=0.15. Participants expressed worry about side effects (63.8%), lack of prevention efficacy in PrEP (44.1%), diet and sleep disruption by PrEP (44.7%), development of resistance from PrEP (21.7%), being treated as an AIDS patient by people (20.1%), being refused sex by male partners after using ARV drugs (14.5%) or not being able to afford ARV drugs (26.3%). |
α Total participants in this study were 3748, and were from 145 countries globally, including Asia (26%), Caribbean (2%), Eastern Europe and Central Asia (17%), Latin America 567 (15%), Middle East and North Africa (2%), Oceania (6%), sub-Saharan Africa (5%), and western and Northern Europe and North America (26%). Awareness and willingness to use data reported here relate to 2774 LMIC participants only; global awareness and willingness to use PrEP were 72% and 82%, respectively.
β Participants included gay men, other men who have sex with men and transgender participants (GMT). Among 434 of 520 were HIV undiagnosed GMT and 17% (n = 86) were HIV positive.
χ The overall sample was 1790, which included MSM, FSWs, IDUs in Peru, Ukraine, India, Kenya, Botswana, Uganda, and South Africa. However, MSM (n = 383) were sampled in Peru, India, South Africa.
δ 1407 MSM were approached, but only 1323 questionnaires completed and analyzed.
ε The overall sample was 1187 of whom 95.3% were male and 4.7% were transgender participants.
ϕ A total of 2,644 participants were screened from whom the 990 were included.
γ This was an exclusive sample of MSM who were also sex workers.
η Among the 93.7% (n = 281) HIV-negative individuals in the study.
ι Proportion of MSM vs. TG was not stated.
φ 470 MSM took part in the survey but 404 completed the survey and were included in the analysis.
κ A total of 887 MSM started to fill questionnaire, but only 760 qualified questionnaires were analyzed.
λ 326 individuals completed the screening questionnaire out of which 238 MSM and TG were eligible and completed the survey (131 MSM and 107 TG)
μ Mean age reported here is that of MSM participants only.
ν Willingness reported here is among MSM participants, and excludes transgender participants.
ο 1407 MSM were recruited, but 1402 completed the questionnaires and were analyzed in the study.
π 159 MSM were enrolled, but only 152 used for analysis as 7 were deleted for not having sex with men in the past 6 months.
FSW: female sex worker; GMT: gay, men who have sex with men and transgender; HIV: human immunodeficiency virus; MSM: men who have sex with men; IDU: injecting drug user; PEP: post-exposure prophylaxis; Taiwan $: Taiwan dollar; PrEP: pre-exposure prophylaxis; RMB: Ren Min Bi (currency of People’s Republic of China); STI: sexually transmitted infection; STD: sexually transmitted disease; TG: transgender.