Like in many developed countries, in Australia, there has been a substantial increase in new HIV diagnoses among men who have sex with men (MSM) over the past decade.1 Australia has a highly concentrated HIV epidemic, where each year MSM account for more than 65% of newly diagnosed and 85% of newly acquired HIV infections.2 However, there are geographical differences in the HIV epidemiology across Australia. Victoria, with the second largest gay population in Australia, has seen the most rapidly increasing HIV epidemic.3 In parallel, there has been a rapid re-emergence of the syphilis epidemic in MSM in Victoria.4
The rise in HIV notifications in Victoria has been attributed to increases in sexual risk behavior,5 increasing rates of other sexually transmitted infections (STIs) such as infectious syphilis and gonorrhea6,7, and lack of resources to sustain health promotion and other prevention activities.8 Furthermore, although annual HIV testing rates are high in gay men, STI testing rates are lower, and despite guidelines recommending highly sexually active men to undergo HIV/STI testing every 3 to 6 months, only a small proportion do so,9 leaving a substantial window where onward transmissions can occur.
In response, there has been considerable investment in new HIV/STI prevention initiatives in Victoria in recent years, including multiple large-scale social marketing campaigns. Consumer-focused social marketing campaigns using traditional marketing techniques to promote behavior change have10 been widely adopted within the HIV/AIDS sector.11–13 Key elements used in social marketing campaigns include exchange theory (perceived costs and benefits of change), audience segmentation (messages take account of the target population's attitudes, beliefs, or behaviors), competition (behavioral options that compete with public health recommendations and services), and the “marketing mix” known as the 4 Ps: product, price, place, and promotion.10
Often there are numerous barriers to rigorous evaluation of social marketing campaigns, including the competing need to roll out campaigns quickly, especially in areas with growing epidemics, and the fact that many campaigns are delivered to an entire region or community and therefore do not allow a comparable control group.12 Alternative designs, such as cohort studies, can provide strong inferences about a campaign's success10; however, these can be costly, difficult to implement, and require lengthy follow-up. More recently, “public heath triangulation,” a process through which secondary data from multiple sources can be reviewed and used for monitoring and evaluating public health interventions, has received support.14
In February 2008, the “Drama Downunder” (DDU), a social marketing campaign, was launched in Victoria and aimed to increase health-seeking behavior and testing rates in gay men and to increase HIV/STI awareness and knowledge. We assessed the impact of the campaign by establishing an online cohort of gay men and analyzed HIV and other STI testing data from high caseload clinics before, during, and after the campaign.
In February 2008, the Victorian AIDS Council/Gay Men's Health Centre (VAC/GMHC) launched the DDU campaign that aimed to increase access to diagnosis, treatment, and care of STIs; improve HIV/STI awareness and knowledge; and, in the long term, minimize the transmission and morbidity of STIs in gay and other homosexually active men.
The campaign was informed by formative research,15 which revealed that although well informed about HIV prevention, gay men were less knowledgeable about STIs more generally. The formative research also indicated that campaign messages delivered through “traditional” gay media (e.g., gay press, posters in gay venues) were increasingly discounted as “old” information. The campaign was designed to target a largely “diffuse” gay population over a disperse geographical area (inner metropolitan Melbourne region) and subsequently used tools of mainstream advertising, as well as gay media, and included multiple advertising channels (i.e., print and radio advertisement, printed resources, outdoor advertisements, public events, and banner advertising on gay dating sites). Such an approach was novel for gay men's prevention in Australia. The campaign also used innovative methods for dissemination and engaging the target audience, including novel campaign resources (e.g., fridge magnets, drink holders, and underwear) and campaign-specific events (e.g., the “Drama Down Underwear” Show) to encourage community engagement with the campaign. The materials included images and brief messages designed to be “light-hearted” to capture the attention of the target audience (Fig. 1) with referral to the Web site (www.thedramadownunder.info) for more detailed information.
In Phase I of the campaign (February 2008–February 2009), approximately 566 public display points were established, including bus/train/tram interiors and stations and public billboards, and approximately 54,632 print resources were distributed. The development and implementation budget for the campaign was estimated at $411,755, with the majority spent on advertising ($339,145). It is also important to note that funding for DDU continued after Phase I, with similar Phase II (March 2009–May 2010) and Phase III (June 2010–June 2011) implementations.
The campaign was evaluated using the following 4 key indicators: campaign awareness, HIV/STI knowledge, health-seeking behavior, and HIV/STI testing. Online surveys and routinely collected testing data from high caseload MSM clinics were used to provide these indicators.
An online cohort of gay men, with rolling recruitment, was established to monitor and evaluate the impact of the various HIV/STI prevention initiatives funded in 2007/2008. Men were surveyed at regular intervals (3–6 monthly) beginning from September 2008, and remains ongoing. Men aged 18 years or more who self-identified as gay or homosexually active in the past 5 years and were able to provide informed consent were enrolled. The cohort used multiple recruitment methods, including recruitment at gay community venues, gay community events, participants from a recent community-based HIV prevalence study (which used a facility-based sampling method, again recruiting from gay community social venues), and snowballing (existing cohort participants could refer their gay community peers for inclusion in the cohort).
The surveys included an online participant information and consent form and asked about participant demographics, exposure to gay media, gay community attachment, a series of questions about campaign awareness and knowledge of various prevention initiatives (including the DDU campaign), sexual health knowledge, health-seeking behavior, HIV/STI testing patterns, and the extent to which they believed various prevention activities had impacted on their health-seeking behavior and HIV/STI testing patterns. Campaign awareness was assessed both unaided (if participants were able to recall correctly specific campaign messages without prompts) and aided (prompted by a campaign image). Surveys were linked by a unique code to allow individuals in the cohort to be identified and matched over time, enabling cross-sectional and/or longitudinal analyses of the data. Survey completion took approximately 15 to 20 minutes, and participants were reimbursed $30 for completing each survey round.
Men recruited into the cohort between September 2008 and April 2009 and who completed an online survey in any of the 3 survey rounds were eligible for inclusion in this analysis. Across the 3 survey time points, a total of 295 men completed 483 surveys (S1 n = 74, S2 n = 167, and S3 n = 242). Individuals' first survey was extracted and analyzed cross-sectionally (S1 n = 74, S2 n = 105, and S3 n = 116). Median knowledge scores (percentage of knowledge questions answered correctly) were calculated from 15 sexual health knowledge questions. We examined associations between campaign awareness and key characteristics using χ2 tests for categorical variables or Wilcoxon–Mann-Whitney test for median knowledge scores. Consistent with the primary aim of the campaign to increase STI testing, multivariable logistic regression was used to examine associations between campaign awareness and STI testing in the past 6 months while controlling for demographic and sexual behavior covariates. Backward elimination techniques were used.
Routinely collected HIV, syphilis, and chlamydia testing data from January 2007 to March 2010 were extracted from the Victorian Primary Care Network for Sentinel Surveillance on blood-borne viruses and STIs (VPCNSS), using data from 3 metropolitan high caseload MSM clinics that diagnosed ∼50% of all HIV cases in Victoria between 2007 and 2010. The methodology for the VPCNSS has been described in detail elsewhere.16 These 3 clinics are considered to provide “gay-friendly” services; they are responsible for delivering the bulk of primary clinical care to HIV-infected patients, and providing a substantial proportion of testing, prevention advice, and education to HIV-negative gay men locally. The clinics follow the sexually transmissible infection testing guidelines for MSM,17 which recommend testing for HIV and other STIs at least once a year for all MSM, regardless of whether they have symptoms, and 3 to 6 monthly testing for MSM reporting specific high-risk behaviors.
Time series analysis using linear regression was used to assess trends in monthly HIV, syphilis, and chlamydia testing in HIV-negative men across January 2007 to March 2010. Poisson regression was used to determine differences in the rate of change in monthly HIV, syphilis, and chlamydia tests (incidence rate ratio = IRR) before the campaign (January 2007–January 2008), during initial campaign period (Phase I, February 2008–February 2009), and during the continued campaign period (Phase II, March 2009–March 2010). As previously mentioned, as Phase II (March 2009–May 2010) of the DDU campaign continued on directly on from Phase I (February 2008–February 2009), a true postcampaign period was not possible to construct. Instead, we have chosen to use a continuation of the campaign as a comparison group rather than a true postcampaign period.
Ethics approval was obtained from the Victorian Department of Health Human Research Ethics Committee and the Monash University Standing Committee on Ethics in Research Involving Humans.
A convenience sample of 295 gay men (18–66 years of age) completed online surveys between September 2008 and April 2009. Participants were predominantly recruited through gay community venues (56%), gay events (21%), and snowballing (22%). The majority of respondents identified as gay or homosexually active (96%), were born in Australia (79%), had completed tertiary studies (72%), and resided in metropolitan Melbourne (97%) (Table 1). Over half of the men reported reading the gay press (57%), visiting gay venues (67%), and using online gay dating sites on a regular basis (at least monthly) (55%) (Table 1).
The majority of the sample (86%) were aware of the DDU campaign; 43% spontaneously recalled the DDU campaign when asked to list any 5 sexual health campaigns/campaign messages that they were aware of (unaided awareness), and an additional 45% recognized the campaign (aided awareness) when shown an campaign image. Among those aware of the campaign, campaign images were most commonly seen in the print media (79%), on billboards (51%), and outdoor advertisements (i.e., bus shelters, train stations) (58%). When prompted with a campaign image (with the campaign message blocked out), almost half of the men (49%) were able to correctly recall at least 1 campaign message, and the majority found the campaign visually appealing (63%). In univariable analyses, campaign awareness was significantly associated with reading the gay press on a regular basis (at least monthly), having had an STI test in the past 6 months, and reporting metropolitan postcode (Fig. 2).
In a multivariable analysis, men aware of the campaign were significantly more likely to read the gay press on a regular basis (prevalence ratio [PR] = 1.2; 95% confidence interval [CI] = 1.1–1.3) and report >5 sexual partners in the past 6 months (PR = 1.1; 95% CI = 1.1–1.2).
Median sexual health knowledge score was not significantly different between participants aware (median score = 10) or unaware (median score = 9) of the campaign (P = 0.14). However, men aware of the campaign were more likely to correctly answer questions relating to increasing rates of syphilis over the past 6 years (56% vs. 43%, P < 0.01), treatment of gonorrhea, chlamydia and syphilis (83% vs. 71%, P = 0.10), and preexposure prophylaxis (56% vs. 43%, P < 0.14) (Fig. 2).
Among men aware of the campaign, 16% reported that the campaign had directly prompted them to discuss HIV/STI testing and transmission with their general practitioner (GP) and 25% with their peers; 15% reported that the campaign had directly prompted them to visit a GP clinic and 17% to visit a sexual health clinic. More than one-fifth of respondents reported that the campaign had directly prompted them to search for sexual health information (22%) or visit the DDU campaign Web site (22%) (Table 1).
Correlates of STI Testing in the Past 6 Months.
Of those aware of the campaign, 62% reported an STI test in the past 6 months compared with 44% of men unaware of the campaign (Fig. 2). In the multivariable logistic regression, awareness of the campaign (aided) was independently associated with having had any STI test within the past 6 months (PR = 1.6; 95% CI = 1.0–2.4) (Table 2).
Over the whole analysis period (between January 2007 and March 2010), 18,328 HIV, 18,855 syphilis, and 18,749 chlamydia tests were conducted among HIV-negative MSM attending the 3 high caseload VCPSS clinics (Table 3). Across the 36-month analysis period, there was a significant increase in the average monthly number of tests for HIV (P < 0.01), syphilis (P < 0.01), and chlamydia (P = 0.01) (Fig. 3). In the continued campaign period compared with the precampaign period, there was, on average, an increase of 110 HIV, 122 syphilis, and 119 chlamydia monthly tests among HIV-negative MSM attending the 3 high caseload clinics (Table 3). There was an increase in the total number of tests across the precampaign, initial-campaign, and continued-campaign period (Table 3).
Although STI testing rates increased before the campaign, Poisson regression demonstrated accelerated rates of increase in testing during the initial and the continued campaign periods (Table 3). Compared with the precampaign period, significant increases in testing rate for HIV (17%; P < 0.01), syphilis (19%; P < 0.01), and chlamydia (15%; P < 0.01) were observed during the initial campaign period, and for HIV (27%; P < 0.01), syphilis (29%; P < 0.01), and chlamydia (28%; P < 0.01) during the continued campaign period (Table 3).
This evaluation has demonstrated that DDU campaign awareness was high, and 2 data sources showed some evidence of increased testing throughout the campaign periods, although causality cannot be assumed. Awareness of the DDU campaign was significantly associated with increases in STI testing in past 6 months and increases in knowledge of sexual health issues. Furthermore, ecological clinic data showed accelerated rates of HIV and other STI testing at metropolitan Melbourne high caseload MSM clinics during the initial-campaign and continued-campaign periods compared with the precampaign period. Although the combination of multiple data sources is a strength of this evaluation, the lack of longitudinal survey data beginning before campaign commencement and a reliance on routinely collected testing data with no control group means that the observed changes in testing patterns of HIV, syphilis, and chlamydia among gay men cannot be directly attributed to the campaign.
High levels of campaign awareness among the target population, both unaided (43%) and aided (86%), up to 14 months following the launch of campaign suggest the campaign was effectively promoted. Levels of campaign awareness reported in this evaluation are higher than previously reported in Australia for other sexual health promotion campaigns,18–20 yet similar to successful HIV/STI prevention campaigns targeting gay men internationally.11,21 Such high levels of campaign reach may be attributed to the use of multiple channels including mainstream media and traditional gay media, emphasizing the important relationships between the target audience and the channel selection.22 High exposure frequency and engagement with the campaign material are an important strategy for a successful HIV prevention campaign12,22 and were part of the guiding principles underpinning the DDU campaign design. In this evaluation, a high proportion of participants reported repeated exposure to campaign material, were able to recall specific campaign messages, and found the campaign visually appealing.
Survey data suggested that the campaign increased knowledge, prompted participants to seek sexual health information, and encouraged HIV/STI testing. The multivariable analysis showed campaign awareness was associated with recent STI testing, and clinic data provided further ecological-level evidence, showing steady increases in testing patterns among HIV-negative MSM during the initial and continued campaign period. Annual behavioral surveillance data across the same period (2007–2010) reported more comprehensive STI testing among HIV-negative men, with significant increases in all types of STI testing.23 Despite the already high rates of annual HIV/STI testing previously reported in Australia,24 the increase in STI/HIV testing reported in this study represents a significant change to gay men's testing patterns, with the potential to reduce rates of HIV/STI transmissions. Evaluations of 2 other social marketing campaigns in the United States (”Dogs are Talking” and “Healthy Penis”) have demonstrated similar associated increases in syphilis testing among gay men.11,25
It is important to note that increasing trends in HIV, syphilis, and chlamydia testing among MSM reported in clinic data were evident in the period before the campaign implementation, although testing rates accelerated during and following the campaign period. Other factors may have been influencing gay men's HIV/STI testing patterns before and during the campaign periods. The increase before the campaign potentially reflects enhanced patient and clinician awareness about STI testing in response to an increasing syphilis epidemic since 2002, and the release and promotion of testing guidelines specifically for MSM in 2005.26 Alternatively, around the same time, there were 2 other social marketing campaigns being implemented, the “Protection” campaign (January–September 2008), which was designed to address issues related to risk factors in settings of unprotected anal intercourse with casual partners, and the “GoTest” Syphilis Campaign (October 2007–June 2008), which was designed to raise awareness of the syphilis outbreak in Melbourne, targeting a core group of highly sexually active men. Unlike DDU, which used mainstream advertising (including advertisements at train stations, bus stops, on trams, and on large billboards in busy public locations), these campaigns targeted specific subgroups of the gay community and used “traditional” gay media channels for advertising (including gay magazines, gay radio, adverts in gay venues, and peer education), thus limiting their potential campaign reach. Both these campaigns may have had impact on HIV/STI testing patterns of the target population, thus confounding our associations between the DDU campaign awareness and increasing HIV/STI testing rates.
One of the long-term aims of DDU was to reduce STI transmission. Enhanced Victorian surveillance data that includes MSM status are available for HIV, infectious syphilis, and gonorrhea notifications. These data show the annual number of infectious syphilis notifications in MSM declined from 315 cases in 2008, 320 in 2009 to 204 in 2010, whereas gonorrhea notifications among MSM have increased from 370 in 2008, 495 in 2009 to 749 in 2010.27 HIV notifications among MSM have remained relatively stable, 181 in 2008, 193 in 2009, and 178 in 2010.27 Explanations for divergent trends in syphilis, gonorrhea, and HIV notifications include the differentiating background prevalence or transmission risks per event for the respective diseases,28 the curable nature of STIs compared with HIV (with testing facilitating treatment), and proactive strategies to integrate syphilis testing into HIV management checks.29
The strengths of this evaluation were the combination of multiple data sources, which included cross-sectional assessments of key indicators by campaign exposure for survey data, and a before and after analysis design using time series analysis for clinic data.
There are also some limitations. The timing of the evaluation (funded in April 2008) meant that it was not possible to collect true baseline survey data from the cohort before campaign implementation, which limited our ability to assess causality between the campaign and the measured outcomes. The short evaluation period (9 months), the short intervals between surveys (4–5 months), and our sample size in early survey waves limited our ability to conduct a true longitudinal analysis of campaign awareness and its impact on knowledge, health-seeking behaviors, and STI testing. Our venue-based and snowball sampling approach also means that observations are not truly independent and may have introduced biases affecting campaign recognition. Our sample also may not be representative of all gay men; however, sample demographics are largely comparable with a national Australian gay men's study.30 In addition, as with any surveys (online, face-to-face), there may have been some reporting bias. The evaluation also included analysis of routinely collected testing data from high caseload clinics, which was not tailored for the purpose of campaign evaluation and, as with any ecological data, limits the attribution of causality. Consultation data were also not available from the clinics, and thus clinician versus patient motivated testing could not be determined. Finally, in relation to the clinic data, during the campaign period, one of the sentinel clinics was unable to take on new patients for a short time period due to capacity constraints; however, increases in patient presentations were observed in other clinics. It is also important to note that funding for DDU continued after Phase I, with similar Phase II (March 2009–May 2010) and Phase III (June 2010–June 2011) implementations.
Our evaluation suggests DDU social marketing campaign engaged the target audience, increased their awareness and knowledge of HIV/STIs, and led to increased health-seeking behaviors including seeking testing for STIs. The DDU campaign represented the largest campaign specifically addressing STIs among gay men yet undertaken in Victoria. The high levels of campaign recognition and recall of campaign messages along with positive feedback about the campaign indicated strong acceptance of the campaign by the gay community and highlighted the importance of formative research to develop campaigns that will appeal and resonate with the target audience.
1. Grulich AE, Kaldor JM. Trends in HIV incidence in homosexual men in developed countries. Sex Health 2008; 5:113–118.
2. National Centre in HIV Epidemiology and Clinical Research. HIV/AIDS, viral hepatitis and sexually transmissible infections in Australia Annual Surveillance Report 2010. Sydney, Australia: National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales, 2010.
3. Guy RJ, McDonald AM, Bartlett MJ, et al.. HIV diagnoses in Australia: Diverging epidemics within a low-prevalence country. Med J Aust 2007; 187:437–440.
4. El-Hayek C, Horyniak D. HIV/AIDS Update, Victorian Infectious Diseases Bulletin (VIDB), 2009.Victoria, Australia: Department of Human Services, 2009. Report No.: 12(1).
5. Zablotska IB, Prestage G, Grulich AE, et al.. Differing trends in sexual risk behaviours in three Australian states: New South Wales, Victoria and Queensland, 1998–2006. Sex Health 2008; 5:125–130.
6. Hoare A, Wilson DP, Regan DG, et al.. Using mathematical modelling to help explain the differential increase in HIV incidence in New South Wales, Victoria and Queensland: Importance of other sexually transmissible infections. Sex Health 2008; 5:169–187.
7. Middleton MG, Grulich AE, McDonald AM, et al.. Could sexually transmissible infections be contributing to the increase in HIV infections among men who have sex with men in Australia? Sex Health 2008; 5:131–140.
8. Prestage G, Jin F, Zablotska IB, et al.. Trends in HIV testing among homosexual and bisexual men in eastern Australian states. Sex Health 2008; 5:119–123.
9. Guy R, Goller JL, Spelman T, et al.. Does the frequency of HIV and STI testing among men who have sex with men in primary care adhere with Australian guidelines? Sex Transm Infect 2010; 86:371–376.
10. Grier S, Bryant CA. Social marketing in public health. Annu Rev Public Health 2005; 26:319–339.
11. Montoya JA, Kent CK, Rotblatt H, et al.. Social marketing campaign significantly associated with increases in syphilis testing among gay and bisexual men in San Francisco. Sex Transm Dis 2005; 32:395–399.
12. Noar SM, Palmgreen P, Chabot M, et al.. A 10-year systematic review of HIV/AIDS mass communication campaigns: Have we made progress? J Health Commun 2009; 14:15–42.
13. Wei C, Herrick A, Raymond HF, et al.. Social marketing interventions to increase HIV/STI testing uptake among men who have sex with men and male-to-female transgender women. Cochrane Database Syst Rev 2011; 9:CD009337.
14. Rutherford GW, McFarland W, Spindler H, et al.. Public health triangulation: Approach and application to synthesizing data to understand national and local HIV epidemics. BMC Public Health 2010; 10:447.
15. Spina A. Evaluation of the Drama Downunder Campaign Phase I. Melbourne Victorian AIDS Council/Gay Men's Health Centre, 2009.
16. Goller JL, Guy RJ, Gold J, et al.. Establishing a linked sentinel surveillance system for blood-borne viruses and sexually transmissible infections: Methods, system attributes and early findings. Sex Health 2010; 7:425–433.
17. Sexually Transmitted Infections in Gay Men Action Group (STIGMA). Sexually Transmitted Infection Testing Guidelines for men who have sex with men. Sydney Australasian Chapter of Sexual Health Medicine/Royal Australasian College of Physicians, Royal Australian College of General Practitioners, Australasian Society of Infectious Diseases and the Australasian Society for HIV Medicine, 2010.
18. Wilkins A, Mak DB. Sending out an SMS: An impact and outcome evaluation of the Western Australian Department of Health's 2005 Chlamydia campaign. Health Promot J Austr 2007; 18:113–120.
19. Gold J, Goller J, Hellard M, et al.. Impact evaluation of a youth sexually transmissible infection awareness campaign using routinely collected data sources. Sex Health 2011; 8:234–241.
20. Guy R, Goller J, Leslie D, et al.. No increase in HIV or sexually transmissible infection testing following a social marketing campaign among men who have sex with men. J Epidemiol Community Health 2009; 63:391–396.
21. Trussler T, Marchand R. Prevention revived: Evaluating the Assumptions campaign. British Columbia, Canada: AIDS Vancouver & Health Canada, 2005.
22. Myhre SL, Flora JA. HIV/AIDS communication campaigns: Progress and prospects. J Health Commun 2000; 5(suppl):S29–S45.
23. Lee E, Holt M, Mao L, et al.. Gay Community Periodic Survey 2011. Victoria, Australia: Melbourne National Centre in HIV Social Research, The University of New South Wales, 2011.
24. Zablotska IB, Imrie J, Bourne C, et al.. Improvements in sexual health testing among gay men in Sydney, Australia, 2003–2007. Int J STD AIDS 2008; 19:758–760.
25. Stephens SC, Bernstein KT, McCright JE, et al.. Dogs Are Talking: San Francisco's social marketing campaign to increase syphilis screening. Sex Transm Dis 2010; 37:173–176.
26. Medicine ACoSH, Physicians RACo. Sexually transmissible infection testing guidelines for men who have sex with men, 2005.
27. Victorian Department of Health. Infectious Diseases Epidemiology and Surveillance Victorian Infectious Disease Bulletin. Victoria, Australia. Vol 14, Issue 2. Melbourne, Australia: Victorian Department of Health, 2011: 1–40.
28. Jin F, Jansson J, Law M, et al.. Per-contact probability of HIV transmission in homosexual men in Sydney in the era of HAART. AIDS 2010; 24:907–913.
29. Allen K, Guy R, Leslie D, et al.. The rise of infectious syphilis in Victoria and the impact of enhanced clinical testing. Aust N Z J Public Health 2008; 32:38–42.
30. Holt M, Mao L, Prestage G, et al.. Gay Community Periodic Surveys National Report 2010. Sydney, Australia: National Centre in HIV Social Research, National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales, 2010.