For over a decade I was fortunate to manage a portfolio of global programs funded through the Johnson & Johnson Foundation that have worked to educate and empower pregnant women and new mothers in low-resource settings by providing vital health information through mobile technology.
These programs have directly reached more than 10 million mothers and babies in 10 countries. In addition, we’ve worked with BabyCenter—a leading pregnancy and parenting website—to provide a free license to evidence-based and expert-reviewed mobile content to more than 120 additional organizations that are adapting and translating it for use across 58 countries and 39 languages.
These programs have contributed evidence to the academic literature, and in 2019 the World Health Organization included targeted client messaging in the area of maternal and child health to its first-ever digital health recommendations for health systems strengthening.
On the occasion of HIMSS Europe and the selection of the semi-finalists of the HIMSS Global Maternal Health Tech Challenge, I was invited to reflect on the five top lessons that I have learned during this time.
1. Meet your users where they are
Before you begin, take time to understand the existing patterns of mobile phone ownership and use among your target audience. As far as possible, select the channel they are already using so that they do not need to discover and download a new mobile application, particularly when they may be extremely sensitive to the cost of mobile data or have limited storage space on their phone. Understand their level of literacy and technical literacy.
You may need a multi-channel experience that includes options like SMS, interactive voice response, WhatsApp, and a call center to be able to cater to a range of preferences. To be as efficient as possible with human resources and funds, it can be useful to think about a blended model that escalates from more automated channels like SMS or WhatsApp bots to more face-to-face channels like a helpdesk or call center depending on the nature of the service that is being provided.
… but remember that they may not be on the phone
Even though mobile phone ownership is higher than ever before, we can’t forget that there are still women who do not have access. The GSMA 2020 Mobile Gender Gap report shows that 165 million fewer women than men own a mobile phone, and 300 million fewer women than men access the mobile internet. If we don’t want technology to inadvertently increase inequalities in health care access, it is important to think about strategies to reach those that remain disconnected.
2. Never underestimate the importance of your content
Not all mobile services are created equal! Technology is just the medium, not the intervention, and we can’t assume that two digital health services are equally effective any more than we can assume that two ads are going to be equally persuasive just because they both happen to be on TV or on the internet.
If a mobile health intervention aims to drive behavior change, we need to deliberately consider and incorporate behavior change techniques into the content and user interface, and to do extensive testing and iteration with users to deeply understand their circumstances. Simply providing information is not enough; instead we need to understand whether they have the opportunity and the motivation to make difference choices. Striking the right tone and making an emotional connection with mothers is also very important. For an example of how to do this well, I’d encourage anyone interested to visit the Mission Motherhood Messages library on the BabyCenter website to request access to the free SMS-formatted content that is available for non-profit use.
When measuring the effectiveness of your content, user engagement metrics are valuable, but it’s not just engagement with the user interface that we care about, but rather engagement with the actual behaviors that are being encouraged. We need to remember the difference and to think about how to measure both in our evaluations.
3. Evolve with your users
When we began this work, simple one-way outbound SMS and audio messages were really the only game in town, because smartphone ownership and use was still very low. That has been changing at a dramatic rate. There are now 2 billion WhatsApp users in the world, and ever-increasing growth in smartphone ownership and mobile internet use. This means that mobile services need to adapt from one-way push messaging to enable interactive conversations and rich content like images and video. It’s now not only technically possible, but also what users have come to expect. Since many new smartphone users will have acquired their phones in only the last six months to one year, their patterns of phone use may look quite different to those of more experienced users, and careful product design needs to take this into account.
4. Be smart with machine learning and artificial intelligence
When we begin interacting with users at scale, the volume of incoming information can be enormous and can’t possibly be managed without the right tools. MomConnect, a program of the National Department of Health in South Africa implemented by Praekelt.org, manages incoming questions and feedback from around 800,000-1,000,000 users at any time with just four trained helpdesk staff. It does this by using natural language processing and artificial intelligence to label incoming messages and decide which need to be escalated to a human operator and which can be managed through an automated messaging flow.
5. Don’t forget linkages to care
If a mother contacts a mobile health service—a chatbot, or mobile app, or something else—with a question or concern that indicates that she is at risk, it is critical that the service has thought about how she will be triaged and referred into care. Direct-to-consumer digital health services can only be safe and high-quality when they deliberately establish linkages to health workers and health services.
While we’re currently seeing good use of help desks and call centers to offer general advice and to encourage mothers to go into health facilities, we aren’t seeing that much direct integration. Ideally a mother should be able to be directly referred from a call center to a health facility, with the ticket that she opened with the call center forming part of her referral. Or if she contacts a helpdesk or call center with a query, she should be able to arrange for a community health worker to visit her in her home. It is this seamless integration between virtual and physical services that we need if we are going to use digital tools as a first point of access to health services.
It is also essential for those who are designing and implementing direct-to-consumer digital health services to think about the implications for health workers and health system managers. What new roles do they need to perform? For example, will they need to register mothers for the service, or be responsible for monitoring data submitted by mothers? Do they need training on these new roles? They are also important target audiences, and we need to think about them in design, testing and rollout to ensure that a new service doesn’t inadvertently add to their workload or negatively impact the overall health system.
Increasingly, direct-to-consumer digital health services are going to be core to the achievement of universal health coverage in the digital age. We don’t have enough health workers and brick ‘n’ mortar health facilities to provide quality coverage to everyone—particularly those in low-resource or remote areas—using our current models. But if we create hybrid models that blend virtual and physical services, with digital providing a first point of entry and a means to create continuity of care, then we can maximize the allocation of scarce in-person resources. This will also be a better experience for patients, who will have access to more support at their fingertips. Thinking about how direct-to-consumer digital health services fit into the overarching health system and complement in-person health services is therefore very important. This is an area that needs significant new research and inquiry. In addition to improvements in the health outcomes of mothers and babies, being able to show other health system benefits, like a reduction in health worker workload, or a reduction in the costs of service delivery can be an important way to achieve buy-in from stakeholders like national governments.