Creating Mobile Health Solutions for Behaviour Change: A Study of Eight Services in the mNutrition Initiative Portfolio

"Findings from this initiative validate the premise that mobile technology can accelerate progress towards Sustainable Development Goal (SDG) 2: Zero Hunger and SDG 3: Good Health and Well-Being."

The GSMA mHealth programme, under the mNutrition Initiative funded by UK aid (the United Kingdom (UK) Department for International Development, DFID), has been working with mobile network operators (MNOs) and other mobile and health sector stakeholders to support the launch and scale of mobile health (mHealth) services. The goal is to use information and communication technology (ICT) to deliver lifesaving maternal and newborn child health (MNCH) and nutrition information to women and their families across 8 sub-Saharan African markets. This report from GSMA mHealth, frog, and Altai Consulting presents key findings on the implementation of the mNutrition Initiative as well as mHealth service design and nutrition behaviour change outcomes.

Local implementing mHealth partners (and names of the mHealth services) that are examined through the case studies in the report include:

• Malawi: Airtel and Viamo (Airtel 321)
• Ghana: MTN and Mobile Content.Com Ltd. (MTN Cares 247)
• Tanzania: mHealth Tanzania Public-Private Partnership including Airtel, Vodafone, Tigo, and Zantel (Healthy Pregnancy, Healthy Baby)
• Kenya: Totohealth (Totohealth)
• Nigeria: Airtel and Viamo (Airtel 321)
• Zambia: MTN and Viamo (MTN 667)
• Uganda: Living Goods (Living Goods)
• Mozambique: Vodacom and Viamo (Vodacom 321)

GSMA and its implementing partners were motivated by the fact that an estimated 65% of the female population across the 8 mNutrition Initiative markets have access to a mobile phone, representing an addressable market of over 64.6 million women of reproductive age who can learn about MNCH and nutrition through a mobile phone.

The initiative pursued a human-centred design (HCD) and iterative product optimisation approach across all 8 markets, leveraging findings from business intelligence (BI), user experience (UX) research, and monitoring and evaluation (M&E) user feedback surveys.

For example, Airtel 321 is a mobile information service available to users on the Airtel network in Malawi. Users dial 321 and navigate through interactive voice response (IVR) menus to access voice messages containing information on various topics, including health, nutrition, gender, agriculture, and weather. Users can also access the content via USSD (Unstructured Supplementary Service Data) and SMS (Short Messaging Service). Following service launch in 2015, Viamo undertook a number of service improvement initiatives based on regular user testing of features and functionalities. Specifically, they: shortened and simplified welcome messages, adapted menu options, introduced dynamic content, and stylised content to fit with the audience context and preferences. According to GSMA, since these changes were implemented, the percentage of users who correctly recalled appropriate nutrition practices increased from 55% to 72%, while the percentage of users who reported to be implementing appropriate nutrition practices increased from 49% to 70%. "This demonstrates the power of iterative product development and the application of HCD approaches for delivering quality services that deliver value to end users and improve their nutrition behaviours."

Within the mNutrition Initiative, the GSMA mHealth programme has focused on behaviour change interventions aimed directly at consumers. Services range in degree of interactivity, from passive content push services to holistic healthcare delivery solutions empowering frontline health workers via apps. GSMA has found that, universally, the more interactivity a service offered, the more value people saw in the service.

Key findings include:

• Adopting a HCD approach to product development and optimisation led to increased user engagement: Engagement levels with mNutrition services have varied throughout the product life cycle, but most have shown continued improvement as products evolved to respond to user requirements over time.
• Overall, 69% of mHealth service users demonstrated appropriate nutrition behaviours in comparison to 56% of non-users.
• Sixty-nine percent of mHealth service users correctly recalled knowledge on all tested nutrition practices in comparison to 57% of non-users.
• In some markets, existing knowledge around certain breastfeeding practices is reasonably high, with credit to existing government breastfeeding education efforts. For example, in Uganda, 90% of non-users correctly recalled that breastfeeding should be initiated within one hour of birth. mHealth service users still demonstrated an improvement on this, with an average of 97% of mHealth service users correctly recalling this practice.
• Mobile information services have a stronger impact with poorly understood concepts; for example, in Kenya, 41% of users correctly recalled appropriate supplementation practices, in comparison to only 5% of non-users.
• Repetition of messages about key health practices reinforces the behaviour. For example, in Uganda, providing 4 messages on exclusive breastfeeding as opposed to just 2 over a 4-week period increased adherence to this practice by 8 percentage points. User research revealed that customers appreciate the reminders about certain practices and do not see it as an annoyance.
• Forty-two percent of mNutrition service users report sharing the information they learn with their family, friends, and communities. The majority of these users share information verbally (62%), while others read or listen to messages with other members of the family or community (21%).

Some lessons learned include:

• Public-private partnerships (PPPs) are key to delivering impactful and sustainable digital health solutions at scale.
• Partnership with government also adds credibility to the service, promoting trust in the service among its users. Facilitating the government's support requires demonstrating the ability of the mHealth service to drive health outcomes among the intended population as well as the cost benefits of implementing a digital solution.
• Government engagement can be complex and takes time. Establishing this relationship early on is advisable.
• Assisted registration via community health workers (CHWs) or other agents can overcome trust barriers and mitigate the potential technical challenges faced by less-tech-literate users during registration.
• When combined with on-the-ground promotion, mass media channels such as radio can play a crucial role in giving mHealth services legitimacy and awareness. For example, CHWs in Tanzania who assisted users in registering for the mHealth service report that it was easier to convince people to register for the service if they had already heard the radio advertisements. Although users recall an advertisement, they will often complete the call to action and register only when encouraged by a CHW.
• Traditional mobile-based marketing methods, such as end-of-call notifications and airtime top-up messages, have proved to be highly effective at driving service uptake, especially if delivered to customers at a moment in time when they are already interacting with their phone.
• Shortened self-registration increases registration rates; additional information can be gathered after on-boarding users.
• Even users who can read English often do not fully comprehend the meaning of a text message. Translation of messages from English to a local language in Uganda resulted in a 13-percentage-point increase in appropriately recalled nutrition knowledge among users.
• Ensuring action ability of content is essential. During content testing in Uganda, user research participants named access and cost as the main barriers for acting on messages that promote the consumption of high-quality animal proteins. To overcome this challenge, mHealth services should offer alternative, cheaper, and locally available sources - e.g. in the Ugandan context, silver fish and legumes.
• Services with a high degree of customisation, offering on-demand medical and emotional support, can become trusted partners for their customers. Dynamic content features like "recipe of the week" encourage regular usage. Across all 8 markets, users describe variants of dial-a-doc services as their ideal mHealth service.

In conclusion, GSMA notes that MNOs will play a crucial role in enabling these technologies in emerging markets. To remain relevant in the digital health space, operators should consider:

• Pursuing a holistic approach that looks at digital health as an integrated, as opposed to fragmented, portfolio of services;
• Positioning themselves at the centre of the ecosystem as ICT and digital service partners for governments, health providers, and health tech companies; and
• Building their ICT and health skills and resources by establishing strategic partnerships.