You are not logged in. LOG IN NOW >

Ugandan Program to Promote Safe Sex Makes Things Worse

BY Jessica McKenzie | Monday, July 1 2013

Even after Google told them to, some Ugandans did not want to wrap it up (robertelyov/Flickr)

A mobile health initiative meant to encourage safe sex practices in Uganda failed to effect positive change. In fact, researchers found it made the community, on average, even more promiscuous.

The study 'Mixed Method Evaluation of a Passive mHealth Sexual Information Texting Service in Uganda' was conducted by Innovations for Poverty Action researchers Julian Jamison, Dean Karlan and Pia Raffler, with funding from and

The mHealth initiative, a collaboration between Google, the Grameen Technology Center and the primary mobile service provider in Uganda, MTN, is an information service called 6001, the phone number employed by the service. Users could text questions in either English or the local language Luganda and receive responses retrieved through a key word-based algorithm.

This is an example of a passive mHealth initiative, in which targeted users must reach out and engage with the available material. An active mHealth initiative would be more like a public service announcement: information distributed widely and without prompting.

In a passive initiative, one challenge is making potential users aware of and interested in a particular service. The villages targeted for the initiative were visited up to six times by marketing teams promoting the 6001 service. These visits were supplemented by flyers and posters. The marketing campaign lasted three months. However, these efforts might not have been sufficient. Many potential users reported in interviews that once the marketing campaign ceased, they forgot all about the service.

With passive initiatives like this, one also has to assume users will ask the questions they need to ask or they will still not receive the information they need. “In our setting, individuals must first choose whether to access the service, and then what to ask, and finally what (if anything) to do with the information received.”

Compounding the problem, the algorithm in Luganda was less successful and accurate than the English language algorithm. Respondents reported getting inappropriate responses to the questions. When users shared their negative experiences with others, it reinforced the idea that 6001 was not a good service and people stopped using it.

Even where the service succeeded at awareness building, it could not empower users to act on that information. Some Ugandan women reported having problems standing up to their partners and demanding safe sex.

One woman told her interviewer:

You may tell him that lets start using condoms to protect ourselves, after getting advice from these messages, [...] because I am worried about our situation. He then asks why you are worried, how come: all along you had never gotten worried. When you tell him the source of the information, trouble then starts. For example saying that MTN does not live in my house so cannot decide for me what to or not to do.

Another reported that when she denied her husband sex, that he went and found it elsewhere, increasing the risk and promiscuity level of the community.

Another respondent questioned the usefulness of awareness when there are insufficient health services available: “Now, you have the information, and you are even told where to get further tests and treatment, but you don’t have money for treatment, or even transport to the place you have been referred to: Now have you been helped at all?”

The researchers concluded that initiatives like this could possibly benefit from increased direct engagement.

We find that easing access to this information is not sufficient to induce safer behavior, and may directly or indirectly lead to riskier choices. It may be that only in the presence of a guiding hand, such as a local health worker, can this category of information improve outcomes, but our study does not vary that feature so we can only suggest this as a potentially important mechanism . . .it would be appropriate to learn from this study that the mere introduction of an information technology, designed by development experts, but left to individuals to self-direct in terms of their usage, does not necessarily lead to the socially desired impacts set out in the original intention of a program.

This is perhaps a lesson, also, that scare tactics aren't enough for everyone. The program might also benefit from some positive condom reinforcement, too.

Personal Democracy Media is grateful to the Omidyar Network and the UN Foundation for their generous support of techPresident's WeGov section.