–       Uganda suffers from a severe shortage of professional healthcare workers, thus programs aimed at prevention of disease are an important complement to the limited healthcare delivery system.

–       There is an average of one doctor for every 20 000 people in Sub-Saharan Africa and a high prevalence of HIV while human resources remain a challenge to tackle disease there is a pressing need for new health approaches that are focused on prevention rather than treatment and that do not depend on traditional face-to-face interaction between medical providers and the public.

–       This paper looks the mobile phone and addressing some of these challenges but more specifically looks at delivering HIV education through incentive-based sms messages with quizzes. The goal is to increase the awareness of knowledge about significant public health concerns to promote healthy behaviors and to encourage people to know and check their health status.

–       The system was deployed by a non-governmental organization from Netherlands called Text to Change (TTC) that works in partnership with healthcare NGOs in Africa to appropriate medical content for the quizzes.



–       Looks at two groups: a group of 10 000 mobile phone users in a circumscribed geographic area northwest Uganda. A total of 5000 workers in three factories located in central and southeastern Uganda.

–       questions:

  • whether interaction over SMS presents an effective user interaction mechanism for the population of users. SMS an attractive user interface for  given the ubiquity of SMS-capable phones in Africa but the question remains whether the general public can use the modality effectively.
  • Second look at the factors that influence participation in these sorts of quizzes and what if any social factors might take the place of the often0used monetary incentives to drive participation rates.

–       The WHO recommends a ratio of doctors in Sub-Saharan Africa 1:1000 while in Uganda that number is 1:10 000 however 40% nurses and 70% of doctors are concentrated in the urban settings.

–       Mobile Phones in Africa: grew from 49 million in 2002 to 280 million in 2007 and is projected to approach 600 million by 2012. Pay as you go is typically popular for the rural poor.

–       Mhealth: numerous applications have explored the possibility to addressing health challenges using mobile phones and other mobile devices in Africa other regions of the developing world.

–       Vodaphone study with the UN found that for the 5.3 billion people living in developing countries 2.3 billion mobile phone users, but only 305 million computers and 11 million hospital beds. Mobile phones have the ability to reach further into developing countries then any other technology and health infrastructure. Mhealth falls into 5 categories: remote data collection, remote monitoring, communication and training for healthcare workers, diagnostic treatment support and education and awareness which I will be focusing on.

–       SMS based applications are important for low end “brick” mobile phones widely deployed among the African poor. SMS based mhealth applications is facilitated by the availability of application frameworks such as FrontlineSMS and RapidSMS that enable the sending and receiving of group SMS messages.


SMS Quiz Application (section 4)

–       Dutch and Uganda based organization TTC were founded in 2006 with 4 primary goals:

  • 1: to gauge the knowledge of the population about HIV/AIDS in order to inform the content of future quizzes and better target interventions.
  • 2: to educate participants about safe behaviours with regard to HIV/AIDS
  • 3: to motivate people to seek HIV testing and treatment
  • 4: to reduce HIV/AIDS related stigma and discrimination.

–       TTC application can be summarized as an inecentive-based mobile phone quiz: questions are sent to the participants’ mobile phones, participants reply and those participants with the highest accuracy and participation rates are entered into drawings for free “airtime” and other prizes.

–       TTC system sends out a questions, participants answer question, TTC system check for the answer and responds:  if the answer is correct, participants get more information on the topic. If the answer is wrong participants get the correct answer and an explanation. This is repeated until the quiz is complete.

–       Each quiz must only be 160 characters due to the format of the user interface.


Method Deplyment Sites

–       Northwest Uganda, 10 000 users during a two week period TTC and the quiz was publicized through various media outlets, newspapers, radio stations within the district.

  • Quiz: multiple choice 13 question HIV/AIDS quiz

–       Second deplyment site was 3 factories in rural Uganda (as discussed above).

  • Southeastern Uganda where workers are supplied with housing, schools and health facilities. 18 true or false questions created by different healthcare NGOs from the first deplyment site. Face to face interviews with specially trained HIV/AIDS “peer educators” who are part of each factory education outreach program. 2493 participates phone numbers were collected.

–       Quiz participates are believed to have only a primary school education. English is the primary language taught in primary school through local ones also include Ganda and Luganda and where commonly spoken in areas where they quizzes were sent out.

–       Literacy rate in Uganada: 15 and up, 66.8% can read and write English but higher for males at 76.8% then for females 57.7%.

–       Messages sent in regular intervals over 2-3 weeks and responses were accepted any point between the initial distribution of a question and the completion of the quiz. Multiple answers for the same question were accepted due to the fact that participate often shared a phone with other family or friends.

–       Feedback on each question was given whether it was right or wrong as mentioned above if it was right they were entered into draws for free airtime and other prizes.

–       Messages sent to and from the short code for TTC were free.

–       Multiple choice questions had 2 to 4 answers. Participants were expect to get a score of 35% correct if they guessed at all the answers.

–       Example of questions: “What causes AIDS? (1) a virus called HIV (2) witchcraft (3) Mosquitoes/insects. Reply with “cause” and the number of your answer. “

  • When the response is received a reply is sent back as feedback. The response is something like “Too bad! The right answer is 1: AIDS is not cause by witchcraft or insects; it is caused by the ‘Human Immunodeficiency Virus’.  

–       In The factory quizzes participants if they guessed were expected to to achive a score of 50%. An example message:

  • “HIV can be prevented by using condoms correctly and consistently (1) true (2) false. Reply with “condom” and the number of your answer. The 3 factories received the same answers.

–       Each response for both deplyment sites were entered into a database with one record per response. Data was eliminated if the message was blank, an incorrect response (less then 1% of responses) as well as GIVE or TEST which were words used to transfer money to other mobile phone users. 1.8% was eliminated from district quiz while 2.4% was elimited from factory2 quiz.

–       Accuracy and Participation was focused on for the acceptable data (refer to graph entitled “Accuracy” to see results) Accuracy improved if formatting was disregarded.  Conversational responses were also disregarded for example a response “family planning is the responsibility of both the husband and wife.” In this example the correct answer was produced but formatting was not correct. In the District quizzes conversational questions appeared 9.7% of the time while in the factory districts they were 3.7% and under for the 3 factories.

–       Participation rates varied, however the district quiz offered more incentives to participate in the quiz then the factory district did. Factory quizzes increased throughout the duration of the period and new participates were added. Particiaption rates decreased for district quizzes an the study concludes that radio and newspaper ads may not have been sufficient to get the  purpose and instructions of the quiz out.

–       Major problems: formatting issues

–       Overall conclusions: found that participants in the two deplyment know much of the content  the quizzes tested. When formatting is disregarded the accuracy levels avg over the entire quiz from 84.6% to 91.9%. TTC also increased progress of individuals attending local clinics for HIV test requests in the factory quiz towns by three fold. It was concluded that users who had never used SMS could respond correctly to the quiz questions however improvements could be made on the interface while can distinguish between error messages and correct messages in the wrong format.

–       Next issue was literacy: although those that responded were usually correct what about the other 75% of the district who did not participate. The individuals who did responded were most likely English literate but what does that mean for those who do not speak English? Are they as knowledgeable as others?