Table 1 Synthesis of the characteristics of the included studies according to the type of intervention

Self-testing delivery system

MacGowan, 2020 [32]

USA

RCT

2665

MSM frequented social media, music and dating websites

The distribution of HIV self-tests via mail to MSM recruited via the internet

Not receive HIV self-tests via mail

Frequency of HIV testing (mean number of times tested)

Mean number of tests over 12 months:

Intervention: 5.3 / Control: 1.5

p < 0.001

Zhang, 2020[35]

China

RCT

216

Community and social media

Provision of free HIV self-testing (HIVST) kits + site-based HIV testing

Site-based HIV testing only

(i) HIV testing frequency (average number of HIV tests per participant);

ii) HIV testing coverage (average number of HIV tests for sexual partners of MSM participants)

(i) HIV testing frequency

Intervention: 3.75 / Control: 1.80

Standardized mean difference (SMD): 1.26 (95% CI 0.97–1.55), p < 0.001

Adjusted rate ratio: 2.10 (95% CI 1.75–2.53)

(ii) HIV testing coverage

Intervention: 2.65 / Control: 1.31

SMD 0.64 (95% CI 0.36–0.92), p < 0.001

Rodger, 2020 [34]

UK

RCT

10,111

Online platforms

Free delivery of free HIV self-testing kit to men recruited through online advertising

Control group did not receive free HIV self-testing kit

Proportion of HIV testing over 3 months follow-up

% tested

Intervention: 97% / Control: 43%

RR 2.27 (95% CI: 2.15–2.41)

Zhu, 2019[31]

China

RCT

100

MSM frequented community, commercial and online venues

Baseline procedures (video demonstration, questionnaire and self-test) + App (WeTest) usage to increase HIV testing through "oral fluid testing"

Only baseline procedures (video demonstration questionnaire and self-test)

Proportion of HIV testing

% tested

Intervention: 92% / Control: 68%

RR 1.99 (95% CI 1.03–3.84)

Katz, 2018[26]

USA

RCT

230

Sexual health center, social media and MSM frequented websites

Free access to HIV self-testing

Facility-based testing only

Frequency of HIV testing

Mean number of tests

Intervention: 5.3 (95% CI 4.7–6.0)

Control: 3.6 (95% CI 3.2–4.0)

p < 0.0001

Wang, 2018 [28]

China

RCT

430

MSM frequented bars, clubs, saunas and websites

Promotion of HIV testing (4 min video) + motivational interview (MI) + home-based HIVST-OIC (online real-time instructions and pre-test/post-test counseling)

Promotion of HIV testing in general (3 min video)

Frequency of HIV testing within the 6-months follow-up

% tested

Intervention: 89.8% / Control: 50.7%

RR 1.77 (95% CI 1.54–2.03) p < 0.001

Wray, 2018 [30]

USA

RCT

65

MSM frequented dating apps, social media websites, and community

eTEST system: detects when users open the HIV self-test (HST) kits remotely in real time. Intervention groups: 1) eTEST—self-test with monitoring and 2) HBST—self-test without monitoring)

Control group received reminders to perform the test at clinics

Proportion of HIV testing

% tested

eTEST: 100%

HBST: 95%

Control: 72%

p < 0.001

Jamil, 2017 [25]

Australia

RCT

362

Sexual health centers; community-based organizations; social media and MSM frequented websites

HIV self-test + facility-based testing

Facility-based testing only

Number of HIV tests over 12 months

Number of tests

Intervention: 701; tests per year: 4.0 (95% CI 3.7–4.3)

Control: 313; tests per year: 1.9 (95% CI 1.7–2.2)

RR 2.9 (95% CI 2.2–3.5)

Vera, 2018 [27]

UK

NRCT

232

MSM frequented sauna

Vending machine technology to distribute HIV self-test

Outreach testing by community workers

Number of HIV tests performed

Number of tests

Intervention: 265 (34 tests per month)

Control: 40 tests (6 tests per month)

Banerjee, 2020 [33]

UK

Retrospective cohort

1926

Sexual health centers

Home-based testing service available for online request as an alternative to attending a clinic of the “Umbrella” sexual health services

Clinic-based testing

Number of patients who requested home-based kits or who attended the clinic

Number of patients:

138 Home-based testing (7.2% of all MSM)

1788 Clinic-based testing (92.8% of all MSM)

Health services organization

Hu, 2021 [43]

China

RCT

187

Voluntary counseling and testing (VCT) clinic

Assisted partner notification (self-tests kits for partner or notification by a community health worker)

Passive partner notification (by index cases themselves)

Proportion of index cases who successfully had any sexual partner accessing testing

% tested

Intervention: 35% / Control: 17%

p = 0.004

Solomon, 2019 [39]

India

RCT

9,997

Community

Integrated care centers (ICC) focused on MSM that provide integrated testing, prevention and treatment services

Conventional care services

Proportion of HIV testing

% tested

Intervention: 33.9% (95% CI 32.8–35.2)

Control: 25.2% (95% CI 23.3- 27.1)

RR 1.77 (95% CI 1.30–2.41)

Read, 2018 [38]

Australia

RCT

422

Sexual health center

Access to HIV testing without seeing a clinician

Consultation with a clinician at every HIV test

Frequency of HIV testing over 12 months

Tests / person-years

Intervention: 453 / 205.6

Control: 432 / 204.0

IRR: 1.04 (95% CI: 0.89–1.22)

Galvan, 2006 [36]

USA

NRCT

3645

MSM frequented bars

To offer the Bundle protocol (package of screening tests)

To offer only routine HIV test

HIV testing uptake

% tested:

Intervention: 10.2% / Control: 8.9%

RR 1.59 (95% CI: 0.82—3.10)

Snow, 2013 [37]

Australia

NRCT

14,423

Clinics specialized in gay men's health

HIV testing after hiring a nurse specializing in sexual health

HIV testing in a clinic that did not have a specialized nurse

Proportion of HIV testing

% tested

Intervention: 57% / Control: 37%

p = 0.026

Maruyama, 2021 [44]

Tanzania

Cross-sectional

18,462 (number of tests)

Community

Differentiated service delivery (DSD) models providing client-centered HIV services tailored to key-populations

First year (year 1) of the intervention and the last year (year 4)

Annual number of HIV tests

Number of tests

1745 (year 1) / 5081 (year 2)

6471 (year 3) / 5165 (year 4)

Luo, 2020 [42]

China

Cross-sectional

548

HIV counseling and testing clinics

a) Direct notification of partners for information provided by the Index Case (IAPN);

b) Training and Provision of rapid self-test for the Index Case to take to the partner (HIVST);

c) Guidance for the Index Case inform the partner to seek testing (PR)

Testing and counseling of couples (Index case was oriented to return with partner) (CHCT)

HIV testing uptake

Intervention:

IAPN: 94.2% RR 3.9 (95% CI 1.6–9.3)

HIVST: 93.8% RR 3.4 (95% CI 0.4–30.0)

PR: 89.5% RR 1.8 (95% CI 0.9–3.6)

Control:

CHCT: 82.2%

Patel, 2021 [41]

Australia

Cross-sectional

43,560 (MSM attending testing services 2010–2018)

Clinical and community HIV testing services

Government HIV strategies that aimed to increase HIV testing, including testing targets, innovations in clinic procedures, expanded testing modalities, health promotion, and workforce development

–

% MSM attending testing services who had an HIV test in the last 12 months and annual testing frequency

2010 vs. 2018:

12-monthly HIV testing uptake:

83.9% vs. 95.1% (AAPC = 1.3%, 95% CI: 0.9–1.6, p < 0.001)

Annual HIV-testing frequency: 1.4 vs. 2.7

(AAPC = 7.5%,95% CI: 6.1–9.0, P < 0.001)

Fernandez-Lopez, 2020 [40]

Spain

Cross-sectional

–

Community attending a network of VCT services operated mainly by NGO

Community-based voluntary counseling and testing services (CBVCT) + point of care tests (POCT)

–

Increase of the absolute number of MSM tested after introduction of POCT

1995–2006: 167 to 380 MSM tested annually

2007–2018: 1201 to 7988 MSM tested annually

Social marketing campaign

Patel, 2016 [50]

India

RCT

244

MSM frequented dating websites; social media website

Internet-based avoidance/approach framed messages for HIV testing and consistent condom use based on the information motivation behavioral skills model

Follow-up period and one of the message frames

Proportion of HIV testing in the last 6 months

% tested

Baseline vs. follow-up:

31.5% Baseline; 43.8% Follow-up (p = 0.04)

At follow-up: 38% Approach; 50% Avoidance (p = 0.18)

Bauermeister, 2015 [48]

USA

RCT

104

LGBTQ pride celebrations and magazines; MSM frequented bars and clubs; websites

A web-based tailored content to promote HIV/STI testing (“Get Connected!”)

Did not receive personalized tailored content

HIV testing uptake after 30-day follow-up

% tested:

Intervention: 32.4%

Control: 22.2%

Chi-square 1.18 (non-significant)

Blas, 2010 [46]

Peru

RCT

459

MSM frequented commercial and advocacy websites

Online HIV-testing motivational videos

Online standard public health text

Proportion of HIV testing

% tested:

Non-gay identified group

Intervention: 11.3%; Control: 0% (p = 0.001)

Gay-identified group

Intervention: 5.6%; Control: 7.7%

RR 1.07 (95% CI 0.40–2.85)

Chiou, 2021 [55]

Taiwan

Prospective cohort

831

Public hospital website (control participants); MSM frequented dating apps, social media websites (intervention)

Social networking voluntary counseling and testing (VCT) model, recruiting MSM from social networking platforms and conducting web-based discussions on dating apps and Facebook. It included scheduling through social networking platforms and providing VCT at designated times and places

Traditional voluntary counseling and testing (VCT) model, recruiting MSM through a website and testing station at a gay village

% of MSM who completed VCT after some interaction with the website (control) or social networking platforms—web-based discussion (intervention)

Intervention

% completion of VCT after web-based discussion: 20.7% (215/1038)

Dating apps: 52.8% (158/299)

Facebook: 43.8% (21/48)

Line app: 60% (36/60)

Control

% completion of VCT after browsing the website: 8,56% (616/7200)

Hickson, 2015 [49]

UK

Prospective cohort

2,047

MSM frequented dating websites; HIV survey

HIV health promotion media interventions targeting MSM named “I Did It” and “Count Me In”

Subgroup 1 completed month 7 (incl. ‘I Did It’ measures)

Subgroup 3 also completed month 13 (incl. ‘Clever Dick’ and ‘Count Me In’ measures)

MSM not aware of campaigns

Rate of HIV testing on months 7 and 13

(number of tests/person-years)

Rate of HIV testing (number of tests/person-years):

- Recognized, or had seen and read, advertisement

Subgroup 1: 1.67 / Subgroup 3: 1.60

- Not seen advertisement

Subgroup 1: 1.19 / Subgroup 3: 1.12

Subgroup 1: RR: 1.37 (1.03 to 1.81), p = 0.03

Subgroup 3: RR: 1.48 (1.06–2.06), p = 0.02

Tang, 2018 [53]

China

Closed cohort stepped wedge cluster RCT

1,381

MSM frequented social media app

Multimedia HIV testing campaign, an online HIV testing service, and local testing promotion campaigns

Baseline period, before cities received the intervention

Proportion of HIV testing over 3 months follow-up

% tested

Study period: 62%

Intervention/post-intervention periods: 53%

The Proportion of individuals testing during the intervention periods within a city was 8.9% (95% CI 2.2–15.5) greater than during the control periods

Intention-to-treat analysis: the probability of an individual HIV testing during the intervention periods was higher than in the control periods (RR = 1.43, 95% CI 1.19–1.73)

Wang, 2019 [54]

China

Cross-sectional

15,932 (number of tested individuals)

MSM frequented social media app

Social media–based HIV testing promotion campaign (Blued), with a one-time mass massage push in 2015 encouraging users to get tested for HIV followed by monthly electronic banner promotions. In 2016, an online HIV testing appointment platform was embedded in the application

No comparison group, testing was assessed over time

Frequency of HIV testing

Number of tests per year

- Before campaign:

836 (2013) / 425 (2014)

- During and after campaign:

3.336 (2015) / 6.330 (2016) / 7.315 (2017)

Van Handel, 2014[47]

USA

Cross-sectional

NA (number of tests as unit of analysis)

National HIV survey

National HIV Testing Day (NHTD) Campaign (1 week—June 24–30, 2010)

Control week 1 (January 7–13, 2010)

Control week 2 (August 12–18, 2010)

Number of HIV testing events in a week

Number of tests:

Intervention—NHTD week: 5.919

Control week 1: 4.124 / Control week 2: 4.066

p < 0.001

Wilkinson, 2016 [51]

Australia

Quasi-experimental

48,892 (number of tests)

Sexual health centers, general practices specialized in gay men's health

"Drama Downunder campaign", a social marketing campaign using digital advertising material and printed material to promote regular sexual health screenings and increase sexual health awareness

Pre-DDU implementation period

Trends in number of HIV testing (January 2007–December 2013)

Difference on linear regression inclination: 1.7 (95% CI − 1.6 to 5.1)

Guy, 2009 [45]

Australia

Ecological

4,988 (survey); 2,717 (number of tested individuals)

Primary care clinics, sexual health centre, national MSM survey

Social marketing campaign (“Check-It-Out”): posters, cards, radio ads, publications and banners aimed at the gay public, and strategies for MSM not linked to the community (posters in public transport, educational spaces, ads on radio and newspapers). A website with HIV and STI information was developed

No comparison groups, HIV testing uptake during and subsequent to the campaign was compared

Proportion of HIV testing in the last 12 months (survey); number of HIV tests before and after the campaign (sentinel surveillance)

Survey -% tested

During the campaign (2005): 61.4%

After the campaign (2006): 61.9%

p = 0.34

Sentinel surveillance—number of HIV tests

April 2004 and August 2005: 3,435

202 HIV tests per month (average)

No statistically significant difference

Peer education

Rhodes, 2017[58]

USA

RCT

304

MSM frequented bars and clubs; community colleges and events; commercial venues; social media websites; mass media

HOLA engrupos, a small group peer education behavioral HIV prevention intervention designed to increase condom use and HIV testing among Hispanic/Latino gay, bisexual, and other men who have sex with men

General health education comparison intervention

Proportion of past 6-month HIV testing reported at the 6-month post intervention follow-up assessment

% tested

Baseline

Intervention: 32.45% / Control: 31.58%

6-month follow-up

Intervention: 80.26% / Control: 27.63%

OR 13.84 (95% CI 7.56–25.33) p = 0.001

Young, 2013[59]

USA

RCT

122

MSM frequented restaurants, clubs, schools, and universities; social media and websites

Delivery of HIV prevention and testing with peer leaders via Facebook groups over 12 weeks

Peer-delivered information on general health over Facebook

HIV test kit return rate

% returned test

Intervention: 15.8%

Control: 3.6%

Outlaw, 2010 [56]

USA

RCT

188

Community (MSM and general) venues

Motivational interviewing with field outreach

Traditional field outreach

Proportion of HIV testing immediately after

the session

% tested

Intervention: 49% /Control: 20%

p = 0.001

Rhodes, 2011[57] USA

Quasi-experimental

509

MSM frequented dating and social website

Cyber-Based Education and Referral/testing, an online peer education intervention based on social cognitive theory, empowerment education and natural helping

No comparison group, testing was assessed before and after the intervention

Proportion of HIV testing in the past 12 months

% tested Pretest: 44.5% / Posttest: 59.4% OR 1.8 (95% CI 1.4–2.5) p < 0.001

Sabin, 2019 [61] Vietnam

Cross-sectional

407

Community

The President’s Emergency Plan for AIDS Relief (PEPFAR), a community outreach program based on peer education, employing outreach workers to expand community based HIV prevention measures among key populations

MSM without recent exposure to the outreach programs

Proportion of HIV testing

% tested

Intervention: 67.7% /Control: 34.7%

p < 0.001

Adebajo, 2015 [60]

Nigeria

Cross-sectional

6265

HIV counseling and testing program

3 HIV counseling and testing (HCT) intervention strategies: S1 (standard mobile outreach service); S2 (integrated mobile outreach service); S3 (peer-based outreach service)

A comparison was made between the three intervention strategies described

Uptake of HCT between the three strategies

Strategy 1 vs. 2: OR 0.29 (95% CI 0.22–0.40)

p < 0.0001

Strategy 1 vs. 3: OR 9.21 (95% CI 5.57–15.23)

p < 0.0001