Table 2.

Summary of included studies

Study design	Reference	Purpose (aim and objectives)	Recruitment and data collection methods	Sample	Eligibility criteria	Exclusion criteria
RCT	Tang et al. [34] China	To compare the effectiveness of a crowdsourced intervention versus a health marketing intervention to promote first time HIV testing among men who have sex with men (MSM) and transgender individuals in China	Online banner advertisement recruitment. Individuals were screened for eligibility, enrolled, and completed the survey then randomly assigned to either watch the crowdsourced video or the health marketing video. Follow-up text message 3 weeks after survey completion asking about HIV test uptake and test result	Total = 721 crowdsourced intervention = 352; health marketing intervention = 369	Born biologically male, having had anal sex with a man at least once, ≥ 16 years, never tested for HIV, provide valid mobile number	Duplicated mobile numbers were excluded
	Blas et al. [32] Lima, Peru	To study the association between video-based online interventions and proportions of HIV testing in gay-identified and non-gay identified MSM	Online banner advertisements to redirect to study website. After consent, participant randomly assigned to condition using computer algorithm. Baseline assessment, matched emails to those attending clinic	Total = 459, non-gay identified, 97 = video intervention, 90 = control (text) intervention; gay-identified, 142 = video intervention, 130 = control (text) intervention	(1) ≥ 18 years, (2) male and report having had sex with men, (3) be a resident of Lima, Peru, (4) answer the survey from Lima, Peru (5) HIV test over 12 months ago, (5) have a valid email address and, (6) do not report being HIV positive	Excluded 937 (916 did not meet criteria, 21 did not want to participate) leaving final sample of 459. Report only results from the gay and non-gay identified MSM group
	Hirshfield et al [26] United States of America	To assess the feasibility and efficacy of implementing an online intervention (videos/HIV prevention webpage) versus a no-content control	Online banner advertisements with additional email sent to US members of one of the websites. Online self-complete questionnaire at baseline and 60 days post baseline follow-up. Participants randomly assigned to conditions	Convenience sample: Total = 3,092: Control = 609 Prevention webpage = 609, Dramatic video only = 625, Documentary video only = 633, Both videos = 616	(1) identify as male; (2) ≥ 18 years; (3) live in the US.; (4) provide valid email; (5) report oral or anal sex with a current male partner (new or not), and oral, anal, or vaginal sex with at least one new partner (male or female) in the previous 60 days;( 6) ability to read/respond in English	(1) lived outside of the US; (2) identified as female, female-to-male transgender or male-to-female transgender. Duplicate cases were identified and excluded
BAstudy/Pre-Post study	Prati et al. [33] Italy	To investigate the effect of intervention on performance of HIV/AIDS protective behaviours	General population: computer assisted telephone survey, random digit dialling. Used Proportional quota sampling. Contacted again after 6 months. MSM participants—e-mail lists and Web-based communities. Self-administered anonymous online survey, again contacted again after 6 months. Migrant participants—three survey sites: workplace, migrant shelter/camp, and centre for the teaching of Italian as a second language. Self-administered anonymous paper and- pencil survey and again after 6 months	General population (n = 858), MSM (n = 109), and migrants (n = 211)	≥18 years. Took part in both pre/post surveys and sexually active in the previous 6 months	Not sexually active in the previous 6 months before each interview
	Chiasson et al. [24] United States of America	To compare HIV disclosure three months before and after viewing intervention video	Online banner advertisements; online self-complete questionnaire at baseline and 3 months follow-up	Convenience sample: Original sample of 3052, reduced to 442 in final sample following drop out/inclusion criteria	Limited to the 442 men who reported sex in both baseline and follow-up interviews	Not reported
Cross sectional study	Flowers et al. [22] Glasgow, Scotland	To understand the extent of self-reported exposure to intervention among men frequenting venues for gay MSM. To explore whether sexual health related behaviours varied by degree of exposure to the intervention	Men recruited from seven bars frequented by gay men and other MSM in Glasgow ten months post intervention launch	Convenience sample: 1313 men were approached and 822 participated, Final sample = 784 post exclusions	All men present or entering the venue were approached to complete a questionnaire	Final sample excluded men who identified themselves as HIV positive
	Pedrana et al. [29] Victoria, Australia	To assess intervention impact using four key indicators: intervention awareness, HIV/STI knowledge, health seeking behaviour and HIV/STI testing	Cross sectional data Multiple recruitment methods: convenience samples e.g. gay community venues, gay community events; participants from a recent community- based HIV prevalence study and snowballing. Completed online surveys, linked with unique code to allow matching, surveyed at regular intervals (3-6 monthly). Clinic data routinely collected data from Victorian Primary Care Network for Sentinel Surveillance	Cross sectional data Sample of 295 gay men Clinic data data from 3 clinics	Men, ≥ 18 years, self-identified as gay or homosexually active in the past 5 years. Men had to have been recruited between September 2008 and April 2009 and completed any of the 3 survey rounds	Not reported
	Wilkinson et al. [30] Victoria, Australia	To explore the effectiveness of DDU to increase HIV, syphilis, gonorrhea, and chlamydia testing among MSM	Survey data Surveyed annually between September 2008-August 2014. Recruitment sites varied over time, included gay venues and community events, gay sporting clubs, gay online dating sites, social media, and snowballing. Surveillance Data The Victorian Primary Care Network for Sentinel Surveillance (VPCNSS) gathered during specific periods	1228 MSM (survey 4: n = 389, survey 5: n = 743, survey 6: n = 343, survey 7: n = 353, survey 8: n = 328). (242 included in final sample)	Males, self-identifying as Gay/MSM, ≥ 18 years, completing 3 + surveys between December 2010 and August 2014	Evaluation cohort recruited pre December 2010, completed less than three surveys, self-reported HIV positive. Surveillance data Tests within 30 days of a previous test and those indicated for HIV post-exposure prophylaxis
	McOwan et al. [15] England, UK	To evaluate the effect of an HIV testing intervention specifically aimed at gay men in central London, UK who were South European Origin, Black Origin or aged under 25 years old	Convenience sample: MSM testing for HIV within one of three London clinics during 2000, lab records were located for those matching three target groups (South European origin, Black origin, ≤ 25 years)	three clinics in London- 1999 = 65 (target clinic), 239 (other clinics); 2000 = 292 (target clinic), 236 (other clinics)	MSM testing for HIV at one of three target clinics during a specific timeframe, specifically South European origin, Black origin, ≤ 25 years	Not reported
	Guy et al. [14] Victoria, Australia	To measure the extent of any change in the uptake of testing for HIV and STIs during and subsequent to the intervention	Three types of data: Sentinel surveillance data—five clinics referred to within intervention. Routine laboratory data—four clinics (pre, during and post intervention). Behavioural survey-subset of existing national survey, mainly administered at gay scene event. Surveys for 2004, 2005 and 2006 were compared	Convenience samples: those attending clinics (sentinel data/lab data), men completing Melbourne Gay Community Periodic Survey living in Victoria (numbers not explicitly stated)	Lab/Sentinel surveillance data—men attending clinic within set timeframes. Behavioural survey—only information from Victorian residents was included	Not reported
	Gilbert et al. [31] British Columbia, Canada	To describe the impact of targeted NAAT on identification of AHI and discuss the potential of social marketing interventions to optimise detection among MSM	Samples were included from 6 study clinics if sex recorded as male, transgendered or missing and were ≥ 18 years	Convenience sample: Testing rates from 6 clinics	sex recorded as male, transgendered or missing and were ≥ 18 years	Not reported
	Hilliam et al. [23] Scotland, UK	To evaluate the impact on awareness of HIV, attitudes towards testing, prevention and safer sex in both MSM and Health Professionals	Internet recruitment. Websites contained link to online survey. Self-reported online survey pre intervention (April–May 2010) and post intervention (October–November 2010). Post intervention recruitment added use of Grindr	Convenience sample: Pre-stage sample: 309 (MSM = 88; HP = 221) Post- stage sample: 980 (MSM = 775, HP = 205)	Not reported	Men who have sex with women only
	James [20] England, UK	To evaluate effectiveness of English intervention which promotes testing to men who have sex with men (MSM) and Africans	Limited information: Data from testing centres and community surveys	not explicitly stated	Not reported	Not reported
Retrospective cohort study or Cross sectional study	Erausquin et al. [25] Los Angeles County, USA	A pilot intervention to increase awareness of free testing services, provide incentives for getting test results, and improve access to treatment in Latino males	Community venues: outreach volunteers distributed cards target population to encourage testing. Routinely gathered data from clinic with addition of information of outreach card. Data from the intervention period (August–October 2004) compared to data from two comparison periods: May–July 2004 and August–October 2003	Convenience sample: Males testing for MSM within LAGLC’s Service, Prevention, Outreach, Treatment centre in West Hollywood-Fall 2003- n = 86, Summer 2004 n = 97, Fall 2004 n = 95	Results are limited to males who attended HIV testing within specific timeframes, ≤ age 25, reporting sexual activity with a male	Not reported
Non-comparative study	Brady et al. [18] England, UK	To pilot a national, free at the point of use home HIV sampling service	Testing rates were gathered during the intervention period	9,868 tests were requested over the pilot period and 6,230 (63.1%) were returned	Not reported	Not reported
	West et al. [15] England, UK	To review advertising strategies used and numbers of clients who requested POCT during NHTW	Grindr advertisements within 5 miles of clinics contained link to website including a video demonstrating POCT. Electronic records of those attending for POCT and activity data from software clinic	43 asymptomatic attendees	Not reported	Not reported
Interrupted time series	Hickson et al. [19] England, UK	Longitudinal survey to examine patterns of HIV testing and assess whether testing rates were associated with intervention periods	Internet recruitment. Invite to enrol sent to those completing a previous survey and users of two gay-dating websites. Self-reported baseline survey followed by 13 monthly follow ups	There were 3386 enrolments, following exclusions/drop outs final sample of 2047 participants	Male; England resident; ≥ 16 years; sexually attracted to/has sex with men; valid email address	Those with existing HIV-positive diagnosis and those with no or inconsistent HIV test results
	Solorio et al. [27] Seattle, USA	To assess intervention feasibility and identify processes that worked and those that did not	Convenience sample: recruited from various sites, including community events, the Internet, STD clinics, entertainment venues, and Latino newspapers and referral of peers to study. Survey every 3 months, starting with 3 months before intervention (baseline interview), 3 months into intervention and 2 months post-intervention. Self-reported questionnaires	pre-intervention assessment-50, mid-intervention assessment-44, follow-up post-intervention-41	(a) self-report Latino heritage; (b) speak Spanish; (c) biological male; (d) report sex with men in past 12 months; (e) 18-30; f) negative HIV serostatus (if known)	Not reported
Case study/illustrative example	Thackeray et al [28] USA	Provided illustrative example of the use of Social marketing theory in two case study interventions	Two case studies; illustrative example using social marketing theory on HIV testing intervention	two examples	Not reported	Not reported

Study design	Reference	Nature of intervention(s)	Control intervention	Outcome measures	Internal validity	External validity
RCT	Tang et al. [34] China	The 1 min video depicted 2 Chinese men embarking on a relationship and testing for HIV together. The 1 min health marketing video used a cartoon storyline to provide HIV education and promoting HIV testing	Noninferiority design without a control group	self-reported first-time HIV testing	+	+
	Blas et al. [32] Lima, Peru	Videos framed within Health Belief model and aimed to identify strategies to overcome reasons for not testing specific to target audience	Text used in control condition came from existing intervention to increase testing in Mexico	Intention to get tested, HIV testing	+	–
	Hirshfield et al [26] United States of America	Five study conditions: (1) dramatic video; (2) documentary video; (3) both videos; (4) prevention webpage; and (5) control (i.e., received no intervention content). The Morning After-drama (9 min) depicting 3 gay male friends, one of whom thinks he had unprotected sex with an HIV-positive man while intoxicated and seeks advice from friends. Talking About HIV—documentary (5 min) HIV positive men discuss their experiences, uses footage from a feature-length documentary, ‘‘Meth.’’	Control received no content	Self-reported HIV disclosure and other risk behaviours	+	+
BAstudy/Pre-Post study	Prati et al. [33] Italy	United Against AIDS (December 2012, 2 weeks; February–March 2013, 2 weeks) - television and radio public service announcements, print materials (e.g., posters, brochures), Web based advertisements, and cinema and newspaper advertisements. Emphasizing benefits and advantages of safer sex behaviour and getting an HIV test	Not applicable	Self reported exposure to the intervention, recent (in the previous 6 months) HIV risk behaviours and lifetime HIV testing	+	++
	Chiasson et al. [24] United States of America	The Morning after-Use of 9 min dramatic video to prompt critical thinking about HIV disclosure, HIV testing, alcohol use and risky behaviours	Not applicable	Self-reported HIV disclosure and other risk behaviours	–	+
Cross sectional study	Flowers et al. [22] Glasgow, Scotland	Social marketing intervention aimed at MSM promoting use of condoms and water-based lubricant during Anal intercourse; regular sexual health check-ups and HIV testing at least every 6 months. Materials included posters, electronic images and leaflets, with a intervention website. Posters and leaflets were distributed to both clinical and community (wider and gay scene) settings	Not applicable	Self-reported recency of HIV testing, recency of STI testing, Intention to HIV test and correct use of lubricant	++	+
	Pedrana et al. [29] Victoria, Australia	Drama Down under: Intervention aimed to increase access to treatment, increase awareness and knowledge; and minimize the transmission of HIV/STIs in MSM. Used print and radio advertisement, printed resources, outdoor advertisements, public events, and banner advertising on gay dating sites, ‘novel’ intervention resources (e.g., fridge magnets, drink holders, and underwear) and intervention-specific events (e.g., the “Drama Down Underwear” Show)	Not applicable	Self-reported Awareness of intervention, HIV/STI knowledge, Testing in past 6 months, Health seeking behaviours. Clinic data- testing rates	++	+
	Wilkinson et al. [30] Victoria, Australia	Drama down under aimed to improve screening rates and knowledge of HIV/STIs, and to reduce HIV/STIs transmission among MSM. Intervention was focused on ‘inner metropolitan Melbourne’ and included outdoor media, digital media (e.g., banners on dating Web sites), and print gay media, supported by a range of intervention material (e.g., postcards, pamphlets, fridge magnets, and underwear)	Not applicable	Self-reported HIV test in the previous 12 months, number of partners, sex with casual partners, reporting condomless sex with casual partner, recall of intervention and its message. Surveillance Data: HIV/STI monthly testing rates	++	+
	McOwan et al. [15] England, UK	Gimmie 5 min (12 weeks): Advertisements in free paper distributed on the gay scene in London, images were chosen to reflect target groups	Not applicable	Testing rates at target clinic, UAI since last test, testing as result of an advert	+	+
	Guy et al. [14] Victoria, Australia	‘Check-It-Out’ targeted MSM including specific groups (community/non community attached and ‘culturally and linguistically diverse’). Intervention aimed to increase HIV and STI testing, increase regular HIV and STI testing and promote general sexual health	Not applicable	Lab/sentinel data: number of tests conducted per month. Behaviour study changes in self-reported testing patterns	+	–
	Gilbert et al. [31] British Columbia, Canada	1) What Are You Waiting For - focused on raising awareness of rapid testing and NAAT (December 2009 to February 2010) 2) Hottest At The Start- focused on raising awareness of AHI and increased transmission risk in MSM in new relationships or engaging in risky sex.(June to August 2011)	Not applicable	Testing rates of those attending clinic	–	–
	Hilliam et al. [23] Scotland, UK	HIV Wake up Intervention (May 2010)- to inform MSM across Scotland about HIV and levels of transmission, the benefits of prevention and regular testing and where they can go to seek more information and advice. Resources included leaflets and posters, digital online banners and targeted web pages and other web media (e.g. emails targeted at Gaydar users). Materials displayed in ‘scene’ venues and wider community	Not applicable	Self reported knowledge and understanding around HIV testing, awareness and exposure to intervention, HIV testing and other risk behaviours	–	+
	James [20] England, UK	National HIV Testing week (four weeks) promoted through targeted print, social media and outdoor advertising. Stakeholders also provide expanded testing services	Not applicable	Clinic based testing rates	Not assessed- insufficient detail
Retrospective cohort study or Cross sectional study	Erausquin et al. [25] Los Angeles County, USA	Outreach cards provided at Latino-oriented gay club and event nights could be swapped for a movie pass at the time of testing. Information was also advertised on two Internet sites and in three gay/bisexual-oriented magazines. Again these included outreach cards that could be exchanged for movie passes at the time of testing	Not applicable	Testing rates of those attending clinic	–	+
Non-comparative study	Brady et al. [18] England, UK	HIV testing interventions and social media marketing were used to increase HIV testing rates, in particular those requesting self-tests	Not applicable	Testing rates	Not assessed- insufficient detail
	West et al. [15] England, UK	Grindr users within 5 miles, received link to website with POCT video, Poster interventions were also in place at the time	Not applicable	Clinic based testing rates and number of visits to website	Not assessed- insufficient detail
Interrupted time series	Hickson et al. [19] England, UK	1) ‘I Did It’ (December 2010-April 2011)-Terrence Higgins Trust (THT) intervention aimed to make MSM aware of ease and convenience of HIV testing. Used media advertisements, radio and website. 2) ‘Clever Dick/Smart Arse’ (November 2011-February 2012)-THT intervention promoting condom use (3)‘Count Me In’- GMFA, encouraged men to commit to an action plan which included HIV testing	Not applicable	Self reported HIV testing behaviour and self reported exposure to interventions	++	+
	Solorio et al. [27] Seattle, USA	Tu Amigo Pepe Spanish-language radio PSAs, a Web site, social media outreach, a mobile phone reminder system, print materials, posters in stores frequented by Latinos, and a free hotline	Not applicable	Self reported HIV testing rates, intention, experiential attitude, instrumental attitude, self-efficacy, and norms toward HIV testing	+	–
Case study/illustrative example	Thackeray et al [28] USA	One on mental health, second You Know Different—large-scale intervention focused on increasing HIV testing among African American youth	Not applicable	HIV testing rates	Not assessed- insufficient detail

++ All or most of the checklist criteria have been fulfilled, where they have not been fulfilled the conclusions are very unlikely to alter (high internal validity)

+ Some of the checklist criteria have been fulfilled, where they have not been fulfilled, or not adequately described, the conclusions are unlikely to alter (medium internal validity)

− Few or no checklist criteria have been fulfilled and the conclusions are likely or very likely to alter (low internal validity)