In this section, we're going to be talking about some early attempts at organizing the what and how of digital health. You've seen some of the reasons why we undertake the development of these taxonomies in an emergent field and now we'll hear more about some of the more successful efforts to do so. Early in the course of this fields genesis, we work closely with colleagues at the World Health Organization to try and begin to articulate what it is we were talking about beyond the apps that were being discussed widely or the technologies that were being presented at all of these digital health or mHealth festivals, if you would, or conferences. We began to clarify and disentangle the platform or technology from the project and from the broader strategy because what we were ultimately asking ourselves for an agency like the World Health Organization, who issues guidance to implementers and implementing agencies around which health strategies have the strongest evidence base and deserve investment for broader scale up and an application, which of these technologies should be invested in? As we began to think about, well, is WHO recommending platforms? Is WHO advocating for particular projects? Or is WHO looking to recommend strategies? Of course, we settled on the last of the three where it really was about being technology agnostic to some extent. But thinking about the function, the use, and the purpose of technology, as we were trying to collect evidence around that particular strategy. In this case, in this particular example, we see a platform, that is Dimagi's Commcare, a platform that's used to create applications for frontline workers. A particular project, m4 Reproductive Health in Madagascar, where this platform or technology was being used in Madagascar for reproductive health education and then the strategy, which is the hardware agnostic intervention, and that is interactive text messaging to improve information family planning method choice. As we scoured the Internet for publications and evidence about this particular strategy, we were able to look at not just Dimagi's Commcare, but other technologies and other hardware software combinations that were being used to send text messages focusing on improving family planning method choice. The technology stack is the what, the project is the how, and the strategy is broadly the why of the what and the how together. Now another example, this is again the same three layer stack, the technology, strategy, and project, starting with the technology but that is familiar to many of us so WhatsApp, a communications software that's available on any smartphone, the strategy, and that is the interactive text messaging to improve family planning method choice, and the project here being MAMA. You'll see that I've flipped the stacks a little bit to keep you on your toes, but you can see here that technology and the project, so the particular MAMA project that was implemented in four different countries was an example of a particular strategy that is, once again, interactive text messaging to improve family planning choice. We can take the lessons learned from this example and combine it with the lessons from this example to better understand whether or not this type of strategy is one that we can recommend to government and implementing agencies looking to strengthen family planning method choice. As we are generating research evidence, as we conduct randomized trials and qualitative research, what we're asking is, does this combination of software, hardware for a particular purpose to be used to deliver say, information. Does it solve the problem? In the earlier example and an inadequate amount of information was available to make a family planning decision in this particular instance in front of us now, we're asking about, does this combination of technology and software improved compliance to drug regimen as prescribed by a provider? The question that a normative agency will be asking, whether it's the FDA or the World Health Organization is, does providing a text message as a reminder to take a drug lead to increased use. It's not asking whether WhatsApp is an effective technology or whether text messaging AT&Ts text messaging is an effective strategy. But it's asking whether this channel or this approach used to improve drug adherence. Does it actually have that outcome? Now, all of this work and thinking through, how do we talk about this? How do we articulate this? Ultimately combined in the first-ever digital health guidelines that were formulated over the course of two to three years with experts from around the globe, culminating in the 2019 publication of this document that's available from the WHO, the recommendations covered a total of 10 different domains or strategies. That you can see here articulated ranging from birth and death notification to client communication and health worker decisions board via mobile devices. What you'll notice here is the complete absence of any particular name or type of hardware other than of course, mobile devices. But we're not talking about particular chipsets or particular channels of technology architecture. We're not talking about any application per se. As long as the application or the software or the combination of the different technologies and applications accomplish this strategy that you see in front of you, they were eligible to be considered for this evidence synthesis and recommendation development. Now, I know I may be beating this horse to death but for some people, this is an exceptionally hard concept to wrap one's head around. But I like to use this water analogy and I think for many this helps to crystallize that the difference of the architecture of the enabling environment, the actual technologies, and the functions that we talk about when we're talking about this use case. In this case with the water example, you've got the objective or the broad goal of providing water to households for a variety of use cases. Water in and of itself coming through the tap in your home can be used for washing, cleaning, and cooking. Just as three specific use cases. Though that water can be delivered through a number of fixture pores or channels. If you would, the tap, the shower, the kitchen sink. But in order for all of the systems to work, they depend on plumbing or software or hardware that really enables that water to get to its destination and then the water supply, that is the enabling environment. The fact that there is clean water being distilled or processed to just supply through these pipes. As we think about the different ways in which this analogy translates over to technology stack. It's the digital health strategy that platform or technology as the fixtures. The information architecture, that is the plumbing, and then the water supply being either the proven intervention so that is messaging that we know to work or we can be talking about a broader policy environment that allows for this to exist. Now, I want to step back a moment because one of the things we understand with digital health is that there are a lot of interventions that we know to work preventive and curative interventions, whether these are vaccines or trained birth attendants. Extensive research has been done to demonstrate the efficacy of these interventions, and one of the things that we struggle with is getting these interventions of known efficacy to reach the levels of high-quality coverage that can impact all of those members of the population that need access to that intervention. On the road to that high level of coverage, we run into health system constraints, so these could be failure to maintain quality, lack of demand, stigma around health care seeking, and so as we think about these potholes in that highway to high level of coverage, one of the things that we think about digital health is a way to bridge those gaps or accelerate our progress through that pothole, if you would, to reach high levels of coverage and quality, so what does this mean? This means as we're measuring impact or we're trying to assess value of these interventions, we don't necessarily have to go back to the drawing board and demonstrate that a vaccine that has been shown to be efficacious continues to be efficacious when we use a digital health intervention to improve that program. What we do have to show is that the digital health intervention can increase coverage, improve quality, improve the demand for these services that are known to be efficacious, so it's really important to understand this because we very often fall back into that mold of asking, well, does it save lives or does it reduce morbidity and mortality? When really that is a known outcome that we've measured in the past related to the vaccine or the trained healthcare delivery, but what we are doing now with the introduction of a digital health intervention is optimizing that delivery. We're overcoming some of these challenges, and so we should be focusing on measuring whether, in fact, the intervention has helped us overcome the challenge and not necessarily improve the health outcome. Which is a much more expensive and time-consuming prospect when it comes to selecting your research goals. What we see here is one of the first efforts to create a taxonomy, and this is the one that was used by the mHealth Alliance and UNICEF and WHO, and it covered a range of different facets, and so we had a focus on the reproductive, maternal, and child health continuum, we had the health area, the type of care, the constraint that was being addressed, and so what we would do is go through this matrix and describe the technology according to these various features or facets of that particular technology. What was the stage of its technologic maturity? What types of technologies were being used? Who was the target user? When it was initially used was a successful first effort to try and boil the ocean, if you would, of all of the innovation that was going on in this space, and this emerged into another approach which was a more cookbook version that's been highly used in the literature which is the 12 common applications, and so we really boiled it down to these 12 ingredients that were often blended together in various digital health strategies, and so, as you can see with both of these, there's a combination of hardware, software, strategy, target audience, and so it very rapidly ran into some challenges as we began to work through this, and one of the things we wanted to do is try and distill and clean this up as this field continued to grow and accelerate. We began to work with UNICEF to try and include taxonomies that they had developed that allowed us to look at various touchpoints within a health system, so being able to understand who are the various actors that were using this system across the continuum of care. But very rapidly we began to see how this was overwhelming for technologists and developers trying to use this taxonomy to describe what it is they were doing, but this was another effort that was out there very colorful effort.
