Evaluating the Trusted Partner Campaign in Four Sub-Saharan African Countries

This document summarises the findings of an evaluation of the Trusted Partner Campaign, a regional project implemented by Population Services International (PSI). The campaign was developed to alter misconceptions that resulted in a heightened sense of trust between partners leading to subsequent lower consistent condom use. The evaluation, which was conducted in Lesotho, Mozambique, Uganda, and Zambia, found that the campaign has not had any impact on elevating the level of caution individuals use before having sex with regular partners. Exposure to this campaign was only associated with consistent condom use in Mozambique, and the impact was only felt among those with the highest levels of exposure to the campaign.

According to this report, qualitative research conducted in urban areas of sub-Saharan Africa has found that youth may initially use condoms with a partner, but they stop condom use once they believe they can trust their partner. Young couples feel less at risk for HIV because they trust that their partner has goodwill towards them and would not intentionally infect them. The document states that sexual trust in a partner can be conceptualised as having three parts: generalised interpersonal trust in the partner (expectations that a partner will behave benignly), caution with the partner (behavioural preference to protect oneself or feelings of vulnerability in a relationship in terms of one's sexual health), and perceived assurance (discussions of HIV testing and behavioural expectations) from the partner. Studies have found that greater feelings of caution in a relationship are associated with higher rates of consistent condom use.

The Trusted Partner Campaign was developed to alter the misconceptions that resulted in a heightened sense of trust and subsequent lower consistent condom use. The campaign involved 4 television spots that focused on profiling young people whose characteristics were appealing to the opposite sex and deemed responsible and trustworthy. The end of each spot, however, showed that looks and perceptions can be deceiving, as it is revealed that each "trustworthy" person is HIV-positive. Similar storylines were used in 5 radio spots and supporting print materials for posters and billboards. All spots were run in various languages in Lesotho, Mozambique, Uganda, and Zambia.

Evaluation Methodology
Cross-sectional surveys of youth aged 15-24 were conducted before and after the Trusted Partner Campaign aired in Lesotho, Mozambique, Uganda and Zambia. The surveys measured three aspects of trust (caution, interpersonal trust, and received assurances) as well as condom use for youth who reported having a regular partner. Factors affecting caution were examined using multiple regression analysis. The impact of the campaign on caution was evaluated using analysis of variance (ANOVA) to examine whether those who were exposed to the campaign showed significant increases in caution after controlling for other factors. Finally, logistic regression was used to investigate the determinants of condom use and whether exposure to the campaign contributed to increased condom use, after controlling for other factors.

Main Findings

Lesotho - Prior to the campaign, consistent condom use was 56.4%, and this increased to 66.9% at follow-up. One third of respondents had seen at least one advertisement, but those who were exposed to the campaign tended to have higher education and socio-economic levels compared to those who were not exposed. On levels of trust, people who were more cautious in relationships tended to be less likely to consistently use condoms. Those with positive condom attitudes were more likely to use condoms consistently.

Mozambique - Reported consistent condom use increased from 38.7% to 48.2%. Unlike Lesotho, individuals in Mozambique who had higher levels of caution had an increased likelihood of consistently using condoms with their regular partner. Individuals with high levels of exposure to the campaign were more likely to report consistent condom use compared to individuals in the baseline, although there was no difference in consistent condom use between those in the baseline and those in the follow-up with low exposure.

Uganda - There was no statistical change in reported consistent condom use between baseline and follow-up, although there was an increase in the level of interpersonal trust. The logistic regression revealed that individuals with higher levels of caution were more than twice as likely to report consistent condom use compared to those with lower levels. The addition of campaign exposure to the model did not substantially affect the relationship between sociodemographic variables and consistent condom use. Furthermore, individuals with no exposure at follow-up and those with some exposure did not report significantly different levels of condom use than those interviewed prior to the campaign.

Zambia - There was no statistical change in reported consistent condom use between baseline and follow-up. Those with higher levels of caution in the follow-up were more likely to use condoms, and males were less likely to report consistent condom use compared to females. When indicators of exposure to the campaign were included, there was no difference in reported consistent condom use between those not exposed to the campaign and those exposed.

In conclusion, the analysis has shown mixed results for the campaign in the four countries studied. There was a significant trend in condom use in Lesotho and Mozambique associated with the Trusted Partner campaign, however, exposure to the Trusted Partner campaign was not associated with higher levels of caution. The document sites two possible reasons for this. First, it is possible that caution is a risk perception that is negotiated at the couple level rather than at the individual level. Second, the data indicate that more intense exposure may have been necessary to bring about behavioural changes. Various factors that cannot be controlled with the existing data may confound the results. These include inconsistent timing of the post-campaign survey in relation to the campaign period, inconsistent measures of campaign exposure in the 4 survey questionnaires, and the possible multiplicity of HIV/AIDS campaigns and programmes during the same time period. Nevertheless, the results show a significant positive trend in condom use in Mozambique. The results also indicate that more intense exposure may be necessary to achieve impact. Regional campaigns appear to be a cost-effective way to reach large numbers of youth and further research with more consistent measures of exposure should be pursued.

This resource brief is based on Working Paper 68, "Trust, Caution and Condom Use with Regular Partners: An Evaluation of the Trusted Partner Campaign in Four Countries". To obtain a copy of the full document, please send a request to the contact details below.