Ten Lessons Learnt: Scaling and Transitioning One of the Largest Mobile Health Communication Programmes in the World to a National Government

"Implementation lessons learnt from this journey are a reality check."

Many digital health interventions struggle to achieve scale and sustainability, despite investors' increasing shift from funding pilots to scale-ups. However, in India, there are examples of digital health solutions successfully scaling across geographies to change health practices and generate demand for services. This paper offers 10 lessons learned from scaling and transitioning to the Indian government a complementary suite of mobile health (mHealth) services designed by BBC Media Action to strengthen families' reproductive, maternal, neonatal, and child health behaviours (for more, see Related Summaries, below).

The interventions were initially implmeneted in the state of Bihar. These interventions are:

• Mobile Academy, an interactive voice response (IVR)-based training course to refresh frontline health workers' (FLHWs) knowledge and improve their interpersonal communication skills;
• Kilkari, which delivers stage-based, time-sensitive, weekly audio information directly to families' phones to reinforce FLHWs' counselling; and
• Mobile Kunji, an IVR- and print-based job aid to support FLHWs' interactions with families.

The complementary suite of three mHealth services were originally designed under a single digital theory of change. They aimed to (i) improve FLHWs knowledge of preventative RMNCH behaviours and strengthen their interpersonal communications skills, (ii) equip FLHWs with a high-quality job aid to support and standardise their interactions with families, and (iii) provide new and expecting mothers and their families directly with audio information to reinforce messages communicated by FLHWs.

As of April 2019, when Mobile Academy and Kilkari were transitioned to the national government, 206,000 FLHWs had graduated, and Kilkari had reached 10 million subscribers. (Although Mobile Kunji was used by 144,000 registered FLHWs for 7 years in Bihar, was scaled to the states of Odisha and Uttar Pradesh, and, according to BBC Media Action, had a significant impact on a range of health outcomes, it was not adopted at the national level.)

Lessons learned include the following:

1. User-paid business models are challenging in low-resource settings - In 2012, BBC Media Action began road-testing a user-paid business model, which involved revenue share agreements with six mobile network operators (MNOs) in Bihar. Due to the high cost of acquiring subscribers, and the low price point that poor families could afford, this business model failed to cover all costs. But the team was able to pivot to a government-paid business model, under which the government covered all call costs, making the service free to subscribers.
2. Large complex interventions that include integrated theories of change need to be reassessed if only some of these interventions are scaled - When the national government scaled the mHealth services, it did so without Mobile Kunji, which was not scaled because it was thought that smartphone job aids would make the IVR and print-based tools obsolete. In addition, changes to the plan had to be made with Kilari: User testing in Bihar had revealed that to improve comprehension and recall, information should be limited to one, simple, doable action per Kilkari call. However, when going to scale, it was necessary to communicate several simple doable actions in each call - due, for example, to the need to address national and state government priorities.
3. Trade-offs are required between ideal solution design and affordability - In a donor-funded pilot project, it is possible to design, test, and develop an "ideal" maternal messaging programme. When the Indian government decided to scale Mobile Academy and Kilkari nationally, several important compromises to the service design were required to accommodate its budget and procurement policies and to overcome the challenges involved in trying to reach low-income illiterate women. For example, for Kilkari, trade-offs were required in three areas: number of calls; degree of language localisation; times of day when calls were made to subscribers.
4. Programme components should be reassessed before scaling - Examine technical partners' strengths and weaknesses - e.g., do they have the skills, experience, and staff capacity required for a national scale-up?
5. It can be just as challenging for a government to adopt open-source software as proprietary software - This programme negotiated one-off payments for unlimited capacity, in perpetuity software licences for Mobile Academy and Kilkari because they were cost-effective at scale - critically, ensuring that these licenses could be transferred to the Indian government at no additional cost.
6. Operational viability at scale is a prerequisite for sustainability - Mobile Academy and Kilkari were candidates for government adoption not just because these services offered solutions to known challenges but because they had demonstrated sustained demand from intended populations, and they could be set up and managed centrally with little face-to face training and were relatively straightforward and inexpensive to sustain.
7. Taking informed consent in low-resource settings is challenging - The National Health Mission in each state planned to train FLHWs to inform pregnant women and mothers of babies about Kilkari and to equip them to answer questions about the service. But because this sensitisation did not happen in some locations, families were caught off-guard when they suddenly started receiving Kilkari calls. Some immediately unsubscribed from the service or disconnected the calls, thinking they were spam. As concerns about data protection and privacy grow among the citizens of developing countries, governments and non-governmental organisations will need to identify how to take informed consent to capture personal and sensitive data in a cost-effective way, without excluding potential beneficiaries.
8. Big data offer promise, but social norms and SIM change constrain use - For example, in the Indian context, where only 14% of adult women own a smartphone, and only 56% have access to the household phone in states such as Madhya Pradesh, the mobile numbers associated with women in government databases are likely to belong to male family members.
9. Successful government engagements require significant capacity - It is critical that digital implementers and their investors are realistic about the timelines and resources required for government adoptions and scale-ups.
10. It is necessary to define governance structures and roadmaps up front - For example, agree on a shared strategy for co-management, including the establishment of a programme management unit and regular joint project review meetings.

In conclusion: "Patience and sustained investment - particularly in learning and evaluation - are required to optimise digital solutions at scale and amplify their impact....Furthermore, scale-ups and transitions are not controlled scientific experiments....Public sector procurement will cause delays, funding streams will no longer be in sync and different project components may not happen on time. Planning for these challenges and having the resilience to adapt to the unforeseeable should be regarded as a sign of mHealth maturity."