Global Perspectives on Digital Health

Smisha (00:00)
can we use digitization to augment community health systems, To improve the work of community health workers, access to services for rural communities,

But the truth in most settings where these very complex digital systems are being rolled out is that community health workers are not paid, they're not supervised, they're not legitimized, they are not institutionalized. And so in an innovation setup where you're just testing it out in one place, it made sense. Yeah, can we leapfrog what's happening? But these systems have started to scale without question of whether the healthcare workers are being paid.

Shubhanan Upadhyay (00:17)
Mm-hmm.

Smisha (00:35)
And that sort of continuity without question is, I think, problematic.

Shubhanan Upadhyay (00:42)
Welcome to the Global Perspectives on Digital Health podcast with me Shubs Upadhyay. This is the podcast that breaks you out of your bubble, giving you insights from people who are innovating or implementing with underserved communities around the world.

to help us take them back to our own work. I'm very excited to have our guest today, Associate Professor at Johns Hopkins University Center for Global Health, Smisha Agarwal. We are gonna be getting some great insights across Some of the challenges around monitoring and evaluation and around

the fallout of the USAID cuts that we've seen globally in the last couple of months. This is going to be a really, great if you're in research and evidence effectiveness of digital interventions, particularly for underserved communities. Let's get into it. It's going to be a good one.

Shubhanan Upadhyay (01:35)
Smisha, it's so great to have you on the Global Perspectives on Digital Health podcast. Welcome. Let's start a little bit about you and the work that you're doing at the Johns Hopkins Center for Global Health.

Smisha (01:50)
Great, thanks Shubs, I'm excited to be here. So it is not an easy name, Johns Hopkins, Center for Global Digital Health Innovation, I direct the center. It is interdisciplinary center focused on digital health that really brings together folks from different schools of Johns Hopkins, including the School of Public Health, Medicine, Nursing, Biomedical Engineering, and the School of Business. So it's really a place where

Shubhanan Upadhyay (01:55)
Mm-hmm.

Smisha (02:16)
we hope to advance digital health in a very interdisciplinary cross-disciplinary way.

work is largely focused on and I mean it has evolved. really started focusing on primary health care and on maternal and child health and chanced upon digital health. I should say that as an Indian every second person I knew was an engineer and so I'm an engineer by osmosis.

sort of leaning in that direction. But I started accidentally in digital health when I was away from India for a few years to study overseas. when I came back, I moved to a rural part of the country, so that everybody, literate or with low literacy, was using phones actually to listen to cricket updates. And as a young and enthusiastic recent business grad,

I

thought, well, let's see if we can use phones for healthcare. Right, because these are often places without access to clean drinking water, without access to electricity. And so that really became, and this is now in 2008 when M Health, the digital health was not really a field quite as such where we started a startup in our little part of Maharashtra to have community health workers monitor pregnancies and upload the data.

on a cloud where we can identify high risk pregnancies digitally and respond to them. And so that was sort of my initiation. Since then, I have worked across sectors. I have been at Hopkins for a little over a decade where we focus now on sort of broader issues of digital transformation, including the policy landscape, governance. We think a lot about effectiveness and impact.

and really sort of, you know, I think we're going to talk about this later, but really start with health first and when does, what is a good allocation of resources as sort of a driving question rather than digital first, right? And so I think within any sort of healthcare context, that's sort of...

And maybe that's like our driving principle and maybe what also sets us apart from other centers of innovation where the point of view is often we want to be innovative, but I think with our center, what we want to be is we want to drive health care impact. And if innovation gets in the way, then we want to remove the innovation and have it be innovation in service delivery or innovation in how we are thinking about it, not just tools and gadgets.

Shubhanan Upadhyay (04:46)
That resonates so much. We miss that and we breeze past the, let's get to the digital type of innovation, but innovation is so much wider. That sounds really great and really, really interesting background and great to learn about the work that you're doing at the centre as well.

Maybe before we get into evaluation of impact, I think it would be great to just talk a little bit about the context that we're in at the moment. So we're in March 2025. And of course, we can't really talk about impact and evaluation without thinking about the impact that...

the current state of the world, the current state of global health and the impact that USAID and also other countries' aid cuts have had on the sector, both from just an impact on people and service delivery side of things and also on the data aspects of that. Do you have through the work that you're doing, the research that you're doing, people you've had partnerships with, do you have a view in terms of what you've seen or heard?

on what's going on globally.

Smisha (05:50)
You know when the stop work orders came out I was in a remote part of India in the state of Jharkhand, which is one of the states with one of the highest rates of infant mortality in the country globally and we were working closely with with TDH and the ministry there to evaluate and try to understand how frontline health workers can use clinical decision support tools to accurately diagnose

and treat children presenting with fevers. And we spent a few days observing that. I'm very excited about what research can help show. This was a piece of work that was actually funded through USAID, through their DIV mechanism. And I think it was one of the most robust pieces of work because the interesting thing about DIV is that they really work collaboratively with researchers to iterate on the proposal.

Shubhanan Upadhyay (06:41)
Do you mind just eluding to

what DIV is by the way?

Smisha (06:44)
I think it stands for Development Innovation Ventures, and so I think, unlike other sort of proposals where you submit a proposal and if you get funded, you execute it. In this case, we had worked with an army of economists.

at USAID to iteratively design and redesign the evaluation approach. while it was an extremely tedious approach, it was also really helpful to come up with a proposal that's robust. it had me walk away with thinking, this is how we should work in this space.

right? Where we're acknowledging the reality, we're saying how can we still take our research approach a notch higher? How can we realistically consider the risks actually of evaluating? And so I think

What we had also done is we had looked at the evidence and they said, okay, here are sort of three concrete pieces of work that tell us that using such decision support tools can improve diagnosis and treatment of sick children. But they didn't demonstrate a significant impact because they were not large enough.

And so can we actually use that to get to a right sample? So we really iterated on the details and I think got to what was a very robust proposal that would have really changed how the field is moving. And that was then interrupted and since then it has been terminated.

And of course, it's disappointing for us as the research collaborator, but it's far more disappointing for our partners who have spent five, six years building those community-based relationships, training the healthcare workers, gaining trust. That is not something you can earn back quickly. And I'm sure you've seen a lot of other narratives about...

more treatment and service delivery based initiatives for HIV and where of course the situation very much will have an impact on human lives, will cost human lives. But I suppose...

It was not just the termination of it, but also the abruptness of it, right, which was not planned. Because I think several of us in the field would agree that some things need to be fixed with how we, the business of aid, right? A lot of things need to be fixed with the business of aid, but there is a responsible way to fix it.

Shubhanan Upadhyay (09:07)
Yeah. I guess it's that, yeah, that a recognition that things kind of need to change, but these can't change if also that abruptness has an impact on lives, infrastructure, and that more kind of intangible element, is so important, like trust.

I think this shows the different layers of impact as well, the negative impact that this has had.

Smisha (09:31)
Also that DIV was, I mean, they were doing some very responsible work, right? I mean, I can offer some critique for how USAID operated, but this particular engagement was exemplary. It's how others should actually model themselves after, Like sort of the iterative sort of development of how we approach evidence and how we are responsibly.

And I say that, I emphasize that because of the vilification of an entire group of people. But this was very extremely responsibly executed.

Shubhanan Upadhyay (09:58)
100%, yeah.

I think you've also then given us a nice segue into talking about the right approaches. So you've highlighted them through this example as well. Actually, even despite everything that's going on. This was an example of things being approached the right way.

Monitoring evaluation are kind of universally talked of as critical, but obviously really hard to get right. What's been your learnings? What can you share with us from the work that you've done

Smisha (10:30)
Yeah, we've been supporting specifically for health programs, but also specifically for digital health programs. We've been a partner to implementing agencies to support implementation and evaluation, often to not just ask the question of does it work, but also to ask questions around when does it work well, how can you make it better, what must improve to make it work, right? And so those are more implementation science questions of

of getting something to work. So it's not just a yea or nay, but it's also like a how. And I think my perspectives on it have changed. you know, one of how we approach research traditionally, a lot of our practices come from pharmaco-epi

pharmaco-epidemiology where we administer a drug and the patient takes the drug every day and what that allows for is consistency of dose, right? But in health systems research

there isn't a consistent dose. There is high levels of variability because people and processes, governments and so on are involved. And so I find that because of that evaluation becomes very hard because actually we can't measure dose very well and we're making a lot of assumptions. Often those assumptions are simply assumptions, right?

Maybe the other challenge, and this comes back to this point of health first, is this idea of plausibility. If you want to see something works, it should theoretically make sense. It should make some sense. But the idea of plausibility is not based on logic. It's based on our prior beliefs.

Shubhanan Upadhyay (12:18)
heuristics.

Smisha (12:18)
Right. so, yeah. And

so somebody who is like an optimist, a technologist might think something's plausible and somebody maybe who's a little bit more pragmatic or has seen sort of community realities might think, no, that's implausible. And so I think what lands up happening is even our starting point, our question for where we should start is often implausible. Well, I think it's implausible. For example.

If I send SMS messages to mothers about what they should eat and how they should conduct their healthcare during pregnancy, it would reduce maternal mortality. Right? And there is some plausibility there, right? We can see sort of a theory of how that might evolve. But the context, which is often not considered is

Are there services that exist? Can the mother get to them? Do they have access to food? Do they have access to the types of behaviors that you're trying to change? And then if all that is happening, is the reason for mortality knowledge? Or is it just maybe preeclampsia?

So I think there are number of ways in which we can think it, what I've learned over the years is that often our starting point, the question that we are asking and the impact we are expecting are incongruent. It doesn't make sense.

Shubhanan Upadhyay (13:45)
And,

and I guess the way, the way I'm trying to see how that kind of fits together if I'm a vendor or a, you know, a founder in a, in a digital health company, I've got my, I've got maybe a different, I'm in a different context. And in my context, it seems plausible from what my experience of life is, what I've seen to be available. And the, privileges that I have within.

society and the infrastructure that I'm within. And so it's very easy then to think about how, okay, if we just did X, Y, and Z, it could be very plausible in another context. And so I guess what I take away with this is it might be, but you really, really need to understand the context first. And you might need to do a lot more work in challenging your assumptions

Is that what you're saying? Have I missed something?

Smisha (14:34)
I think the context of course is important. That's absolutely right. But I think we also need to just change our expectations of what's feasible with technology, Like chatbots, the new sort of love of the field. The truth is that the average chatbot people connect with for 15 seconds. Is that going to change vaccine beliefs? I mean, can you imagine a 15 second chat will change a lifetime of long held beliefs?

I think that's not reasonable to expect actually increase engagement. And for developers, what I recommend is just identify impact measures that's closer to where you're intervening. And I think...

Shubhanan Upadhyay (15:19)
close to where you're intervening

and in your locus of control, guess, as well, right?

Smisha (15:22)
in your

locus of control, as close to your locus of control. I think the challenge with that is often donors need to be educated to allow for that. And because they are really looking at lives impacted.

Shubhanan Upadhyay (15:24)
Mm.

Absolutely. I think that what strikes me then is that we have this trade off because ultimately we want to within healthcare, you mentioned health first, right? We want to have an impact on clinical outcomes. But often there several steps away from what we can affect in our direct locus of control. And so the contribution lines to them are a bit weaker.

And there are multiple confounding variables that also then impact the success of those clinical outcomes. I definitely see across many, many, not just in underserved or rural context.

it's really, really hard to kind of make the leap to then the bigger clinical outcomes that everybody wants. And so maybe we want these health economic outcomes, maybe we want these clinical outcomes, but they're harder to move the needle on. it takes time to get to those. So how does that fit in kind of what you're saying?

Smisha (16:31)
Yeah, I guess, maybe this is an overgeneralization, but I think our hypothesis going in is that digitization or digitalization will achieve outcomes that are superior to a non-digitized approach. I think that hypothesis is often not the right one to start with.

I think we can establish non inferiority, right? Because I think part of it is also like the media when Chat GPT and the LLM based models came. So did the studies saying that LLM sort of was, know, patients had greater satisfaction with LLM based triage than with real doctors, but they were not.

not necessarily comparable because we're not talking about disease acuity. We're talking about sort of very early triage where patients are not presenting with anything that's quite significant. But that sort of made the headlines and then to a lay person that's like, great, like, chat GPT can take care of my healthcare, right? And that sort of started making the rounds, a future without doctors.

Shubhanan Upadhyay (17:37)
Mm-hmm. Mm-hmm.

Smisha (17:40)
I think we really, for that reason, there is a space for innovation and there is a need for pragmatism because I think with some work, Chat GPT could work in clinical settings, but we really have to define the boundaries of it. in what setting can we actually use it and so on? We have to define that.

Shubhanan Upadhyay (18:00)
I mean, the

use case and the tasks, right? And, you know, other language models are available as well. guess if this was the BBC, they'd be like, we don't want to just talk about one brand. I'm only trying to be funny So, really understanding the use case in the context.

Smisha (18:11)
I'm sorry! Yeah, absolutely!

Shubhanan Upadhyay (18:19)
really defining the task that you're trying to help with within that context, and then defining the direct kind of benefits. Of course, then you can talk about the second and third order effects of the other things that we want to get to. But with a degree of humility and pragmatism, I guess, and not jumping to the conclusions that often hype lends us to doing sometimes, right?

Smisha (18:44)
the right

outcomes are not actually better healthcare. It's often in the area of efficiency and long-term gains to the health system. those are harder to measure, so we don't look at them. There are numerous reasons

challenges, let alone the upfront costs of that heavy infrastructure. And so one thought is just thinking through what is the context within which that makes sense.

Shubhanan Upadhyay (19:07)
And that touches on what you mentioned earlier, health first, because that's, that's a type of impact. And perhaps the most important one, kind of thinking working backwards from, okay, here are the health care goals that we want to achieve. Therefore, what needs to happen to make that occur? And then

identifying which types of innovation might serve them. One of them being digital interventions, but others, like you alluded to, are service design or reshaping a process or doing some, you know, having people working in a different way. That's also innovative, And we always jump to the bells and whistles of digital. And so I think, the post I'd seen earlier today was,

an initiative that had been developed in Brazil for community health workers, supporting people who had problems with access to healthcare. I can share it in the show notes. And the learnings from that were taken by a council in London and implemented and supported community health workers to go and then do outreach to people who were struggling to access care in the community and resulted in

actual tangible clinical outcomes, vaccination uptake rates, know, other really, really beneficial impacts. And I was like, this is a great example of innovation, but no one talks about it as innovation, because there was no AI involved or digital health involved. And so you mentioned this at the beginning, and I just wanted to touch on that. Because if we think about health first, then you're thinking more broadly about what you want to do. then

Smisha (20:26)
Hahaha

Shubhanan Upadhyay (20:36)
you've understood by talking to patients, and the people who are delivering care around what they're struggling with, then you're open to which solutions then you're working with them on like, here are the ones that technology could help with. And so can you elaborate on health first have I got that right? Do you see it in a different way?

Smisha (20:56)
No, I think that's exactly right. As a center that focuses on digital health, we often have folks come to us and saying, hey, here's an innovation grant. You're the center for innovation. We want to do something innovative. Give us an idea in digital. And I think my sort of

concern often is that the starting point is I want to be innovative so I can be competitive in this ecosystem. And we need to flip that. And I think it's harder to flip that in the global ecosystem because we're not close to global communities where we are working often. And so the challenges are not as straightforward to us.

We're also two steps removed from the idea of plausibility because we don't see the reality. And so I think that often lands up being a challenge, I sort of that's really it that identifying what is the challenge and if a digital AI based intervention makes sense. think

Shubhanan Upadhyay (21:41)
Mm-hmm. Yep.

Smisha (21:58)
from an innovation lens in small settings, in smaller settings, test settings, it's okay to try innovations, right? Because we do have to have some blue sky thinking. We do have to think about ways that may seem implausible, but we're like, yes, it might be possible. Like we've certainly seen that in our lifetime. But there's a balance to be had there because often that sort of blue sky thinking then becomes the norm and we stop questioning it.

And one of those examples is can we use digitization to augment community health systems, right? To really bring community, like improve the work of community health workers, make it easier, have access to services for community members that are in rural communities, have greater continuity of care, referral, and so on.

But the truth in most settings where these very complex digital systems are being rolled out is that community health workers are not paid, they're not supervised, they're not legitimized, they are not institutionalized. And so in an innovation setup where you're just testing it out in one place, it made sense. Yeah, can we leapfrog what's happening? But these systems have started to scale without question of whether the healthcare workers are being paid.

Shubhanan Upadhyay (22:55)
Mm-hmm.

Smisha (23:13)
And that sort of continuity without question is, I think, problematic.

Shubhanan Upadhyay (23:21)
And

thinking beyond just the intervention related parts of success and leading to overall societal benefit, it's thinking about the layers around that, right? So certain kind of people related, process related or political policy related determinants of success as well, I think. And so I think you've highlighted this. think Ruchit, when I talked

to him, who is the one of the founders of Khushi Baby talked a lot about the insights they had got from working with community health workers where they had deployed in India as well. what you've said also chimes a lot with the work that the Community Health Impact Coalition advocate for very successfully. so I think,

we don't appreciate, the importance of these kind of things in determining success. All the funding goes into the tech part of it, and nothing goes to the people part of it. I mean, even with the USAID cuts, guess, you know, a lot of the data infrastructure, or the things that have been successful have relied on also people engaging with them.

And some, even with funding from USAID, are still not getting paid.

Smisha (24:30)
Yeah, I actually in the last month, I also spent some time where Khushi Baby works in Rajasthan. And of course, the work is incredible. sort of another complexity in the community health worker ecosystem is that they have too many apps. They have an app for everything. They're just re-entering the same data over and over. But where they see value is simple. They see value in knowing what they'll get paid.

Shubhanan Upadhyay (24:46)
Yeah, yeah.

Smisha (24:57)
digitally having that accountability. I've offered x services, it's performance-based financing, so I will get Y amount in my bank account this month. And because of that, that's sort of their hope to say, okay, I will use the app that allows me to have visibility into my payment.

Shubhanan Upadhyay (25:14)
Mm-hmm.

Smisha (25:16)
So I think that's maybe in terms of a very clear logic model there, right? Very clear, like here's the intervention, here's the benefit. The community health workers use it, there's accountability, and I think financial accountability is maybe one of the clearest sort of impact metrics, right? It also reduces waste, improves efficiency of system.

And governments need to come in to actually help solve for and create an environment that allows for several problems that you've identified, including salaries, because USAID and external entities might push for certain community health worker salaries. But I think the other perspective, and I say as a former health care worker from India, you can't, and this is one of the issues, you can't push for.

community health worker salaries that are higher than midwives and nurses and doctors in the country. That's sort of sort of complexity. It has to be part of the context, yeah.

Shubhanan Upadhyay (26:13)
I mean, it's part of the context, right? Yeah, yeah, it's the complexity of the context. Yeah,

yeah. So we talked about some of the kind of policy related things. When we kind of started this kind of bit of the discussion point, particular impact evaluation when we'd started, we were talking about a good approach. And I kind of want to also think about this from a researcher or a vendor who's trying to evaluate their intervention or their tool lens as well.

And thinking about, you I want to try, I'm trying to do this in the right way. I want to measure the right things. I want to engage with people and partner with them in the right way. How do I set solid endpoints that are within my locus of control?

Also on the health economic side of things, it sounded like you got somewhere with robustness there. So there's anything you can share on either of those, that would be great.

Smisha (26:59)
for developers, implementers, again, the ecosystem is in some ways led by donors, yeah, and so they have to be responsive to it, and there's often a mismatch between what one is able to prove and what is realistic. And so let's say that,

what one needs to establish to sell our vanity metrics. And I won't speak about that. But that is required, right? Like that's needed. And so that might be, let's say if you have a chat bot intervention, you get 20 million views on a meta platform.

Right. And off the 20 million, a thousand people chatted with your chat bot. Maybe 10 people chatted with your chat bot for more than 15 seconds. and so within those 10 people out of your 20 million, you are sort of able to say impact. So that's sort of the tricky thing about impact. Yeah. So I would say that I think it's really important to be very concrete about what you can realistically achieve to have sort of the

outcomes proximal to where you're intervening, to consider the context. I think thinking about university partnerships often is a great idea because especially in the innovation space, implementers don't have to find the funding for evaluations themselves.

A lot of universities would be thrilled to partner with them and look for alternate sources of financing. Or they may have university sources of financing. But finding ways to integrate with researchers, with local and global research entities is a great approach. It allows the implementers to focus on the implementation and it allows somebody who specializes in it to think about the evaluation. As in the context, yep.

Shubhanan Upadhyay (28:35)
and is in the context and is there.

And it touched on another thing. I know you didn't want to go so much into vanity metrics, but it links to what you mentioned earlier, especially for donors and what they need as well around the does it work? Yes or no? Versus also how do I continuously get better? And so I feel like there's a call to action to the donor community of like,

also understanding contexts and be feeling more real. And it's easy for me to pontificate on this and not being a donor or kind of fully getting their context. Note to self, get a donor on the show. But I think there is this, it's really hard, right? Because there is a lot of value to, okay, this funnel that you talked about, 20 million views.

Smisha (29:14)
Thank

Shubhanan Upadhyay (29:25)
1000 interacted for 15 seconds, which led to like maybe a change in something for one person. And you've got the vanity metric maybe that answered a yes or no question. But then when you look at the actual impact was very low. But again, that it's a process, right? So you could look at that as a failure, right? Or you could say, what can we learn about that? Why was this number? Why was this conversion so low? What do we need to understand about this context? And, and so

the conversation becomes less, I need to hit this number to get the next round of funding to here's what we've learned, here's what we know that will now give us the depth to get into this and have an impact in this community and in this context, much more meaningfully. What needs to happen there?

Smisha (30:09)
You know, this work or not? The scientific community, think, has done a disservice to sort of guiding policy making because we're fixated on our significance value of 0.05. And I like, is a clear yea or nay, and for some decisions that's critical.

Shubhanan Upadhyay (30:20)
Hehehehe

Smisha (30:26)
For some decisions, that's critical. think for health systems interventions, we need to take a different lens. I like the analogy to how we think about a business spreadsheet when we're doing sales projections. We have a set of assumptions, and based on assumptions, we decide a strategy. Based on the numbers, we're still being evidence-driven. We're not being binary in our consideration. We're still being evidence-driven, and we're making some assumptions to make a decision.

the truth is donors do need to make a decision, Like policy makers need to make a decision, go or no go. And if there is no evidence, they still need to make a decision.

And digitization is happening, whether or not we like it. Innovation is moving forward, whether or not it makes sense or is plausible. And folks that are policymakers, including governments, they're being asked, what do do with AI? Let's do something AI. If you're not doing AI, you're doing something wrong. You're not tapping into the new world and you must. And so there is this fear of missing out of the new world. so I think what is a good approach is

let's look at the evidence critically and then make a best informed decision. The sort of fixation on numbers and numbers alone often misses out on the right approach and question.

Because I think the right approach or question is not whether or not it's working. It's just these are, the world is evolving. Some approaches are moving forward. How do we get them to work better? And so for example, in a space of misinformation, disinformation, 15 second conversations or two hour conversations, people are accessing information online. How do we support the right types of engagement? How do we create sort of new engaging strategies?

So I think that's maybe the, those are some of the right questions, I think, to a good starting point. And by simply going yay or nay, we're actually missing out on the opportunity to just get better, improve things. Because even if we say nay, like I can show that something's not working and yet the community is gonna continue investing in it.

Shubhanan Upadhyay (32:16)
Mm-hmm.

Smisha (32:31)
right? Because there's a lot of enthusiasm for the space and so...

Shubhanan Upadhyay (32:35)
And even

the other side, right? Something might look good in terms of numbers, like that yes or no number that you wanted, i.e. could be a vanity metric. But actually, in terms of making a difference for someone on the ground, maybe it didn't do so. I guess it can work the other way as well, right? So yeah, how do we dig into that? yeah, totally. I could talk to you about this for hours. That's so interesting. So we've got some tangible takeaways

Smisha (32:52)
Yeah, absolutely.

Shubhanan Upadhyay (33:01)
both from a policy perspective and from a kind of vendor or researcher perspective as well, to try and really think about, I think what I've taken away is like, what is in my locus of control? Thinking about working, I think another takeaway, work with universities, have university partnerships locally to where you are. They're a great resource to A, understand the context, B, have the kind of methodologies that are appropriate for that setting as well.

Yeah, and you know, that's a good.

Smisha (33:26)
university professor

that's my totally unbiased recommendation.

Shubhanan Upadhyay (33:29)
Yeah, totally. Yeah.

Yeah. And then I think the third one is that when you're thinking about system change, using that yay or nay approach, you might need to think about that in a more nuanced way. Okay. Thank you. Yeah. Yeah. I mean, there's so much more, but those are the things that stood out to me. And then...

Smisha (33:42)
That's great summary.

Shubhanan Upadhyay (33:50)
I wanted to say you've also been instrumental in creating the Oxford Open Journal. Are you able to tell us a little bit about that?

Smisha (33:59)
Yeah, I would love to. I think that has been maybe one of the most delightful parts of my professional life.

There is just like aid, is a lot that's wrong about international publishing in the healthcare sector.

Shubhanan Upadhyay (34:13)
Mm-hmm.

Smisha (34:13)
I have just been really grateful to have the leadership at Oxford Open Digital Health Journal because we're asking questions that feel like they haven't been asked for decades of publishing. It doesn't mean we are answering those questions or we are even addressing the issues, but at least we're asking some of the right questions and we are taking small steps to move towards it. And so it is a journal that's focused on digital health.

we've approached it very differently. have, because it's focused on LMICs, we have a very globally representative editorial board. We've taken active steps to ensure that we also have a globally representative pool of reviewers. So it's not just US and UK based reviewers. I think the question that I found myself asking and that I had to negotiate, because Oxford Publishing House has been around for several, several decades, yeah,

that why are we forcing Spanish-speaking authors to write in English and then their materials inaccessible to people in their countries? And why are we not asking more of these questions? It's really nonsensical because, I mean, for countries, this is maybe a side comment, but for countries that have been colonized, we know that English language is basically a tool of colonization.

when people don't communicate in a language that they're familiar with, they really can't communicate how effective or creative or intelligent they are. Right. And so I think in scientific discourse, that's really critical that people are able to communicate in a language. Like I learned science in English, and even though Hindi is my mother tongue, I would not be able to communicate any science in Hindi. Right. So so it is actually really important that you're able to communicate science in the language that you that you read.

Shubhanan Upadhyay (36:00)
Mm-hmm, mm-hmm.

Smisha (36:00)
used to

it. So, and we were taking little steps to address all of those little challenges, right? We're asking, we're asking those folks that are funding supplements to make sure that at least abstracts are available in the language where the work is done. are requiring some of this. We are creating to the extent possible with my dual hat of Hopkins and the editor to

have.

training and writing workshops accessible to LMIC authors who don't have maybe as much sort of ease of access to such opportunities. so, yeah, so lots of like exciting things, exciting things going on on that side of publishing, at least starting with, asking the right question. And we're just excited because it is a very practical and applied journal that is really drawing on insights from a lot of policymakers and folks who are implementing

Shubhanan Upadhyay (36:29)
Awesome.

Smisha (36:54)
large-scale digital programs.

Shubhanan Upadhyay (36:57)
mean, I think massively needed. When we've talked before, you've talked about the little things that we can do in our ecosystem, each of us.

to improve things, right? And even more so now, given everything that's happened, because obviously the USAID cuts have happened And do you have any other insights to share or other parts of the ecosystem you think that need to make small changes?

Smisha (37:19)
Maybe if there is a silver lining to recent events, is that it has allowed a lot of people to voice their concerns, right? And I think for a lot of, and I say this to my students, we want to question the status quo.

we should be questioning the status quo because we are brought into a system of aid and of global development that has stopped questioning how things were done. And the world has progressed, but our field hasn't progressed.

And so I don't think the answers are always straightforward, right? We're in deeply entrenched policy systems that have worked in the same way. And people intend well, like that's at the core of it, right? Like the field as a whole is full of humanistic individuals who want to create value in the world.

And so one is even more nervous to like call something out because you don't want to be calling out the intention of your peers. So I think we have to find ways to have these difficult conversations where we are actually able to call out the power dynamics, call out the imbalances that we see around us and really just question the way that they, at least I think maybe the summary of all of this is like, let's just start with asking the right question, right? We don't have to have the

I think the answers are often complex, are contextual, are ever-changing, but we should at least prod on with good questions.

Shubhanan Upadhyay (38:43)
And for you, as really well articulated by the way, for you, what is the key question that we can take the time to answer that or we have the opportunity to answer?

Smisha (38:53)
don't know if there is one question.

Shubhanan Upadhyay (38:56)
Mm-hmm.

Smisha (38:57)
I mean, we've talked about a number of issues here, impact, where often I start with, is it plausible? And maybe if I was more optimistic, would it be plausible? And in the aid space, I pause often with my discomfort because

I find myself being uncomfortable in a lot of settings where, because of my experiences, yeah, growing up in India and previously where I didn't voice that discomfort, I think now I try to voice it more because I think those are good questions to ask, right? If something doesn't feel right, chances are it isn't and it's at least worth the discussion.

Shubhanan Upadhyay (39:44)
And I think, you know, if it wasn't for the abrupt nature of this, I guess the abrupt nature of this has shifted everyone into a kind of survival mode and thinking about what things almost like triaging and like what can survive, what can we salvage, et cetera. And at some point we can start having the headspace then to, have this.

type of conversation. And so how can we make sure that things are really partnerships? Addressing the power dynamics? How do we make sure that leadership is local? And yeah, I think, you know, you talked about silver lining, and hopefully that hopefully enough stuff can survive that we can ask those questions in the wake of this, right? Yeah, you put that really well.

Smisha (40:30)
Absolutely.

Shubhanan Upadhyay (40:33)
I think in just linking it to this kind of the whole, the bigger picture around the open access journal, because some of this propagates into like academia and the problems with academia, et cetera, that you've, that you're really trying to address. but I think other, other, yes or no question, right? and, and, I mean, other ways this manifests, right, is like, for example, you know, those, those people who then would benefit from, from

Smisha (40:46)
No, academia is perfect. No, no questions.

Shubhanan Upadhyay (41:01)
Oxford Open Journal would be able to publish, right? And then, I don't know, they would be able to have the opportunity to go and present their work somewhere, except they can't get a visa to go there. And so that's been playing out for years. And one of the things that I saw in the community was that, for example, the Global Digital Health Forum that happened in Nairobi for the first time. And so I was kind of like referring to like this, these are really good.

examples of small changes that when they stack up can be quite meaningful for people. people actually being at the table because of the expertise that they have within their context and making sure that that is represented on the global stage, I think is really important. So yeah.

Smisha (41:43)
Yeah,

absolutely.

Shubhanan Upadhyay (41:46)
Great, okay. I mentioned the Global Digital Health Forum. Is there another one this year? Do you know anything about whether another one is planned? Any other events? Yeah.

Smisha (41:53)
I'm not sure, yeah. It was supported by USA, but

Geneva Digital Health Day is coming up and I think I haven't been to one, but appreciate that it brings a diversity of voices and not the same old sort of old folks that have been around the block for the last two decades. So that's something to look forward to.

Shubhanan Upadhyay (42:02)
Yes.

Yeah, absolutely. And the Digital Health Day, just to, guess, like, give that a shout out is, so, Smisha you'll be attending. I'm also going to be attending as well. I'm planning a panel around something around that we talked about, which is like, how do we as a system, as an ecosystem, as an industry, think about failure differently in this kind of continuous improvement and share systemically things that have not worked and maybe someone's going through the same problem and can't

Smisha (42:39)
It's great.

Shubhanan Upadhyay (42:46)
And therefore can like, you know, that's valuable, right? And so if we think out of that, yes or no, black or white failure success thinking, we can all learn more. and then Geneva Digital Health, there's a great opportunity to meet people and talk to people, vendors, policymakers, researchers who are kind of working and implementing from a much more global perspective, particularly in LMICs are well-represented.

So if you're looking for a conference that really brings people together in this space, it's a really good one.

Smisha (43:14)
agree. There are

others also coming up. think, I don't know the exact dates, but HELINA has their annual conference. I think that's happening in Botswana this year.

Shubhanan Upadhyay (43:23)
Okay, tell us more about that

one, what's that?

Smisha (43:26)
HELINA is focused on health informatics and digital health The conference happens I think every two years but it really brings together the African diaspora and then

Shubhanan Upadhyay (43:33)
Okay.

Smisha (43:36)
Africa health tech summit will happen end of August which is in Rwanda in Kigali and that has been happening for the last few years really brings together private sector and policymakers and sort of helps push some of the conversations around public private partnerships and privatization of healthcare sector amongst much much else so there are there are a few interesting I think digital

Shubhanan Upadhyay (43:46)
Mm-hmm.

Smisha (44:03)
focused opportunities to gather outside of the digital health forum.

Shubhanan Upadhyay (44:09)
Awesome. Smisha it's been so valuable and insightful talking to you. we really got some real talk as well. so I really appreciate that instead of like just kind of esoteric sound bites. I mean, it's great to kind of get, a real understanding of some of the challenges in the field. not just kind of sugar coating things. So I think that was really,

valuable and insightful. So I'm really, grateful for your time.

Smisha (44:32)
This was really fun and look forward to many more such conversations. Great.

Shubhanan Upadhyay (44:37)
Misha.