A fundamental confusion often arises in discussions of effectiveness regarding the relative importance of methodologies and messages. Two points need to be made at the outset of this effectiveness review:

• All controlled research into the effectiveness of methods has concentrated on methodology. Implicit within this has been the assumption that there is a `core curriculum' of HIV information and skills which need to be taught. There is no discussion in the literature of the detailed content of programmes, so we have no way of comparing whether some programmes proven to be effective actually say contradictory things.
• All controlled research which demonstrates effectiveness has relied on the adaptation of the `core curriculum' of skills and information to the culture and needs of the target group. Nevertheless, similar outcomes have been seen in all controlled studies – the intervention, whatever it is, has a greater impact than doing nothing.

Further investigation of the methodology vs message question is needed in HIV prevention research before it is possible to argue either that `doing something is better than doing nothing', or that `doing nothing may sometimes be a better use of resources than doing something'. For example, investigation of the relative impact of campaigns which promote undifferentiated safer sex messages to negative and positive men versus campaigns which promote negotiated safety would be a useful way of testing this question.

Mass media campaigns

The success of mass media campaigns in increasing awareness of HIV and risk behaviours have been well reported elsewhere, and research suggests that the content of mass media messages is a crucial determinant of the success of advertising and media campaigns.

Small media

Small media refers to interventions such as leaflets, posters and advertising in small circulation or community publications.

The primary evaluative tools which have been used to assess the impact of these interventions have been investigations of numbers who have seen the leaflet or advert, its comprehensibility and its relevance to the target audience. Small media interventions may be measured for other impacts if they are part of a specific campaign with a defined set of aims and objectives. An example of this might be a leaflet or advertising campaign which seeks to introduce messages about `negotiated safety' or assumptions about HIV status. It would be possible to measure the impact compared with baseline in awareness of the ideas communicated in the leaflets or adverts, but would it be possible to correlate this awareness with changes in behaviour? Is this a reasonable expectation?

If an intervention's aims and objectives are correlated with a particular set of assumptions about how behaviour changes in response to informational cues (see Models of behaviour change above), some researchers argue that it is reasonable to assume that changes in behaviour have occurred in response to informational interventions. However, research needs to be designed in such a way as to prove that it was exposure to the leaflet or poster which was an essential component of the intervention, rather than other factors such as participation in the study which influenced behaviour.

An Australian study has shown that small media in the form of posters have little or no impact on behaviour amongst those already exposed to a message by other means. Men who had reported unprotected sex were randomised either to a cognitive behavioural intervention, to receive copies once a month of posters judged by the researchers to be good examples of gay men's safer sex promotion, or to a control group which received no intervention. Participants were asked to record their reactions to the material, and also their sexual behaviour during the follow–up period. Those randomised to the poster group reported that they found the posters attractive and that the posters put across the safer sex message effectively, yet reported little or no change in levels of unprotected sex during the follow–up period in comparison with the no–intervention control group. In contrast, the cognitive intervention group reported a much lower rate of unprotected sex (Gold).

Such findings are problematic because of the resources allocated to the development of small media interventions. They are seen as the core of AIDS education because they are low–budget and because they can slot into many other programmes of work, such as outreach or peer education.

An effective small media intervention is likely to have the following characteristics:

• Cultural sensitivity to the idioms and styles of the target audience.
• Visual impact, attractiveness.
• Tailored to the educational level of the target audience.
• The target audience has repeated exposures to the intervention.

Given these characteristics, it is amazing how many small media interventions in the HIV prevention field ignore fundamental rules of marketing and advertising, and how rarely professionals in the fields of copywriting and marketing are employed to develop such materials. This is not a consequence of budgetary restrictions, by and large, but of a failure to appreciate that health communications are no different from other forms of marketing or PR. This is not an argument for turning HIV prevention over to advertising agencies, but for allowing the finished product to be created by communications professionals rather than a committee.