Katherine de Tolly
Helen Alexander
Publication Date
March 1, 2009


This 13-page document explores initial work that has been done in the Cellphones4HIV project, asking: "Can content delivery via cellphones impact on HIV-related behaviour?" Key insights to emerge from this research include the need to test assumptions before implementing projects and a series of other observations about language of choice, technical ability, and project cost - communication-related elements of which are summarised below. Challenges around measuring impact in behaviour change communication (BCC) are briefly discussed, and some of Cell-Life's upcoming initiatives are outlined.

According to one figure quoted here, there are approximately 36 million active cellphone users, and around 80% of all youth and adults have a cellphone. Capitalising on this trend, Cell-Life, a non-governmental organisation (NGO) based in Cape Town, South Africa, is exploring the range of ways in which cellphones can be used to improve the lives of people infected and affected by HIV - supporting treatment, disseminating information, providing anonymous counselling, gathering data, and linking patients to services. Cell-Life's Cellphones4HIV project is exploring a range of cellphone services to assess their viability for content delivery. These include:

  • Short messaging service (SMS) antiretroviral (ARV) reminders are provided daily to members of "adherence clubs" run by the Treatment Action Coalition (TAC) and the Department of Health at Site B clinic in Khayelitsha, Cape Town. On World AIDS Day (December 1) 2008, approximately 120 people began receiving twice-daily SMSs in English or Xhosa, at the time they should take their ARVs. The SMSs contain both a reminder to take the medication, and ARV- or HIV-related information on topics such as side effects, nutrition, and tuberculosis (TB).

    Concerns have been raised that in a society where people share cellphones, HIV-positive people may be unwilling to use services that can potentially result in others being aware of their status. And yet, "So far, none of the participants in this pilot have reported concerns around unwanted disclosure of their status as a result of the messaging..."

    Language issues have posed challenges; SMSs are limited to 160 characters, and Xhosa requires on average 20% more characters than English. In addition, there are no standard SMS abbreviations in Xhosa, and there are often no accepted terms to describe medical conditions, sexual practices, and other issues relevant to HIV. To get around the character limitation, the translated SMSs use some English words and acronyms (e.g. "ARVs") and abbreviations (largely: the removal of vowels in certain words).

    At time of writing, the pilot had only been running for 3 months. However, the adherence club coordinator did report positive feedback from people receiving the SMSs, and noted that members of the adherence clubs who were not invited to participate in the pilot have expressed displeasure. She thus requested that all 800 members be sent the SMSs; so far, only 5 people have chosen to opt out.

    Organisers note that, in addition to assessing the content and timing of the SMSs for their information utility, another issue to explore is whether receiving the SMSs makes recipients feel like they belong to something (like a group or particular community). This might represent a worthwhile benefit in itself.

    Organisers stress that the scaling up of SMS reminders will only be feasible if cellular networks provide severely discounted (or free) SMSs for such a programme, or if the SMSs can be geared toward certain groups (e.g., those who are isolated and/or are identified as experiencing difficulties with adherence).

  • Unstructured Supplementary Service Data (USSD) is a capability of all Global System for Mobile (GSM) phones generally associated with real-time or instant-messaging-type phone services. "Although few people know the acronym, many South Africans already use USSD to recharge their prepaid airtime - the technology is thus widely available and should be at least vaguely familiarly to users. Cell-Life has partnered with Soul City....As part of the current 9th season of the show, Soul City is running the 'OneLove' campaign, that aims to educate people about the dangers of multiple concurrent partnerships (MCP)."

    Cell-Life has created 2 USSD services based on the messaging of the campaign: one provides information in a traditionally didactic style (on the dangers of MCP and alcohol, etc.), and the other is a "soap opera" that seeks to convey the same messaging in story format. Cell-Life piloted these services in early March 2008 and surveyed 24 people in 3 areas in greater Cape Town. Half of those surveyed experienced one or more fails: the service never initialised, or "crashed." These were due mostly to problems with the USSD itself (either the network or the service provider hosting the USSD). In some cases, users were unable to navigate through the service due to limitations on their handsets; half of the participants also needed instructions on how to get from one screen to the next, sometimes more than once.

    The pilot showed that, on average, in a 2-minute USSD session, people could make it through about 9 screens of content. At 150 characters per screen, this makes for a total of about 1,350 characters (equivalent to 8 SMSs). Cell-Life found that user abilities varied considerably; one user got to 2 screens, another to 14, and another to 26. The services were presented in English only. While most users reportedly found the content very easy to understand, most said that they would prefer to read it in their own language.

    Evaluators concluded that USSD was not well suited for the delivery of "narrative" content, but that, rather, it should be used for providing menus that allow users to "drill down" to content they want.

  • MXit is a Java application installed on users' phones that allows for general packet radio service (GPRS) or 3G-based instant messaging [per Wikipedia, 3G or 3rd Generation is a family of standards for mobile telecommunications defined by the International Telecommunication Union]. MXit claims over 11 million users globally. Given MXit's popularity with South African youth, Cell-Life is piloting the provision of HIV-related content on MXit. Organisers e-piloted the HIV content on MXit with a group of 7 high school students in Khayelitsha, Cape Town, to test whether the content was understandable and easy to navigate. They were all existing MXit users. From a technical perspective, MXit was very stable, unlike USSD, and only one learner reported a fail during the pilot. From a content perspective, the learners found the information easy to understand and locate in the menu structure. They said they would tell their friends to use the service, and they supported the idea of Mxit chatrooms with counsellors to discuss HIV-related issues. The participants indicated that they preferred MXit as a medium because it is cheap, fast, and anonymous. They did not want to read the information in their home language, Xhosa - advocating instead that all the information be translated into MXitlanguage (akin to SMS shortenings, e.g. "sumting", "every1") because this language is universal.

    "The main strength of MXit is chat, and this needs to be explored further for counselling purposes. However not all phones are able to support MXit and technical support may be required to assist new users install the application on their phones. From the pilot it also seems that English (and particularly its shortened form) is the predominant language of MXit, though this would need to be explored further. This contrasts with stated user language preferences in relation to the USSD services."

Within the final sections of the paper are descriptions of future studies that Cell-Life plans to undertake with funding from the United States Agency for International Development (USAID). To cite only one, an intervention is planned with adolescents to assess the use of cellphones in HIV prevention. An observational study design (prospective cohort study) is envisaged, which would involve 2 groups of participants to measure desired outcomes. The intervention group will receive one SMS per day containing HIV-prevention messaging and encouragement to stay in the study and come to meetings.

Two specific technical channels Cell-Life is exploring are:

  1. Wireless Application Protocol (WAP) net: Cell-Life is looking at piloting HIV-related chatrooms through MYMsta, LoveLife's WAP offering; and
  2. Voicemail-push: a voicemail message is "pushed" into the user's voicemail inbox, and he or she is notified of its arrival by SMS. This development could help overcome barriers related to: restrictions in terms of number of characters, constraints linked to screen size and usability factors, illiteracy (the user just needs to listen), and language (content can be delivered in the user's language of choice). Also, content can be delivered by a "trusted source", such as a popular character in a local soap opera, sports stars, or a local nursing sister known to the recipient.

World Wide Web Consortium (W3C) website, accessed January 14 2010.