 called Social Entrepreneur Fellow. And Freya has actually been a consultant to many ministries of health about integration of integrated health systems. She's a principal investigator of multiple domestic and international projects by NIH and other funders and over 20 years of experience developing and evaluating innovative health care strategies, point of care tests, and health IT applications. Her research targets are reductions in health disparities, patient-centered care, improvement of health outcomes, and affordability of care. And her current focus is really on the design and evaluation of a new IT-enabled health system, which equips patients and community health workers with really decision tools and algorithms at point of care so that when a patient and a community health care worker have a problem, they have what she's calling the health box, which gives you a sort of an approach and algorithm. And I thought it would be really interesting for her to discuss this and share this with us. So thank you Freya for coming today. Thank you. Well, welcome to Hot Topics and Global Health. And it's a very hot topic. Nice to see you all today. I'm happy to be here, and I would love to share some of the work that I'm doing now on my current project in Bangladesh, and also a little bit of the research that I've done on using interactive computer health tools and the research that led up to the development of this project. As we look at health care systems today, both globally and in the US, it's very apparent that there's potential for great improvement in health care quality and in patient experience, and also in reducing costs and addressing health disparities by integrating new technologies, home testing technologies, mobile health technologies, internet-based systems. And that's the focus of my research, how to really bring those technologies to bear to redesign health care. The project that I'm going to talk about mostly today is Health Box Bangladesh. We're looking at how you can take a village doctor. In Bangladesh, 95% of people, when they have a problem, will go to their local village doctor. And local village doctors are sometimes euphemistically called quacks. It's normal sort of terminology. And there are people that have decided to become health care providers. Maybe some of them have had some training through the Ministry of Health, but they've taken it on as a career and don't always have good training and background, but have really great connections with people in the community and have trust. The group that we're working with, ICDDRB, has been developing collaborations with the village doctors and have engaged in different training programs, which worked a bit, but there wasn't really a system in place to ensure that high quality care was delivered. And so now we're looking at really decision support tools to help the village doctors provide the best care that they can to their patients. And what's really clear in global health in general is that for the world's leading causes of death and disability, there are cost-effective prevention interventions that could easily save lives. There's evidence on many, many different interventions, but we don't have a great way to get those interventions to the people that need it. Rapid tests, diarrhea treatments, you name it, malaria bed nets for pregnancy. I mean, there's things that work that are inexpensive, but we don't have a great way of disseminating them. And so Health Box is a solution to that. The solution to bring the right care to the right people at the right time is a technology that automates the full patient experience, the initial experience between the community health worker or the village doctor and the patient. And what this does is it solves the problem of costly training because it's very difficult to get a network of health care workers trained adequately and to keep them trained. But because everything in the tool is designed to automate the education and automate the processes that they go through in testing and in referral, each time they use it with a client, they're being trained. Currently, community health workers don't have access to point of care tests. This solution enables point of care testing, both by providing step-by-step instructions to guide someone through the point of care test and the counseling that goes along with it. And we've also tried to build an immediate quality control device to have a webcam validate the results that the health worker types in. Currently, health workers, there isn't great continuity of care. Health workers may find a problem out in the community and want to refer someone and try to make the referral. But getting the information to the right referral can be difficult. The design of this will be that at the end of a visit, you'll have a health care summary that will be transmitted to the referral doctor, whether it be the telemedicine provider or the local clinic. And then currently, whoa, that's interesting. And also a great benefit of this is that it's very difficult to evaluate what's being delivered. But with this program, all of the evaluation and the continuous quality improvement efforts are documented within the care that's delivered. So what is Health Box? It's a mobile tool and program that takes a health worker and provides education, full health assessment, rapid tests, recommendations for health products, recommendations for medications, and recommendations for needed referrals. And when we first designed this, we did a lot of fieldwork in southern India. We did focus groups. We started at internet kiosks. And we did focus groups with people in the villages to find out first what their health care needs were and what their priorities were. And then we did iterative work with these communities to find out what designs they would like in the system that we developed. And we came up with a really long list of functionality that we wanted to build in to the system. The features, it's designed so that it can be used by the health worker or also by the patient. Everything is audio. So even patients with low literacy will be able to use it by themselves, some maybe with a little support from the health worker. But with a minimal level of literacy, they can navigate it by themselves. Touchscreens and the audio make it more easy to use for low literacy populations. Solving the problem of following people longitudinally is significant. And the most acceptable method was to use thumbprints for longitudinal care and to identify them in the field. It guides all the services, and as I said, automates the evaluation and integrates with electronic health records. We also realize that it's very important to be designed so that it can be used either in a connected or disconnected environment, because a lot of times you don't have power, electricity in some of the areas where you're providing services. So the tool can be used out in the field. And then when you get into a place where you have connectivity, automatically the data uploads to the central database and the updates to the program's download. This is the design. People come into the tool and they pick a healthcare provider to be their guide. And that guide takes them through the entire experience. People seem to like that feature as part of it. And initially we thought that they wouldn't like it because it would be too synthetic, but you find people really relating to the avatar on the screen when they're going through it themselves. And what we found was that people wanted, some of the people that were lower literacy and just kind of wanted a full evaluation, wanted to go through a complete health checkup. Others wanted to just go and get a rapid test. Some wanted to go in and just take care of a specific health problem. Others wanted medication refills or health products. We also created quite a few counseling modules. So if people or a health worker wants to focus on HIV prevention counseling or diarrhea prevention counseling, they would go right into the learn behaviors to stay healthy. And then importantly, they can print out their health summary so that they can take it with them to follow their plan and to provide to a health referral provider if it's not integrated. Because it's designed for low literacy populations, you'll find that they'll answer an assessment and then for example, if they have high blood pressure, they'll get feedback and it will tell them about the high blood pressure and they'll be listening to it. Or I think, let's see, I have one video link. So in creating the videos, creating this tool has been really quite a learning experience. All of the media that you need to be able to deliver the types and varieties of counseling necessary is quite a lot of work to produce and also quite expensive if you do it in the traditional way by hiring a media production company. And so what we did in this case was we train the staff in India. We have an office of nine in Bangalore and we train them to do the media development and they also are very close to the communities that need the resources and have a good cultural understanding of what needs to be built in to the tools. And we did some research looking at taking a very high production value video. We used a video for stigma and we delivered it to different focus groups. And then we took that same content and instead of creating a high production, costly video, we created a low cost video where we used illustrations told the same story and recorded audio. And if you do it that way and you have illustrated movies, you can then swap out the languages because in India there's so many different languages. And so that ended up from our initial research qualitatively, the acceptability of the high production value and the low production value videos were equivalent and the initial impact on education looked to be the same. And so that's the method that we used and I have one video that I will try to play for you. Oh, I just opened like three of them, let's see. I think, let's see, I'm not sure the speakers are turned on. I don't know if you can hear that. Oh, there's no, not a plugin. Well, I think there's subtitles so I'll go ahead and play it anyway. This is a diarrhea prevention video. So very simple, the India team kind of selected the look and feel of it. So I think you get the idea. In general, the stories are designed to illustrate the prevention messages within the context of a story and then to help guide patients or villagers what they can do differently to prevent the infection in this case or to prevent hypertension or other conditions. And at the end of watching the movie, they then are led through a planning process. So they decide what behavior they're gonna work on to stay healthy and kind of the pieces of motivational interviewing where you look at what the barriers are, you look at what the supports are, you commit to when you're gonna do it. And so this goes on also to demonstrate the skills that you need and also we'll talk about oral rehydration salts in the event that you do get diarrhea. So I'll stop that there in the interest of time. So really pretty low budget, very simple stories, but well tested in the field. People liked them, understood the concepts and ultimately the test will be, does it actually impact rates of diarrhea? We also have several testing modules. There's a great push to get new point of gear tests developed. Currently we've got an HIV test, anemia, where I wanna be, anemia, urine dip for multiple conditions. We, blood pressure of course. There's a malaria test that we've looked at that is still a little bit complex and a little bit too expensive for community health workers and so probably we'll wait till the next generation of that test comes down. But the idea is that as new point of care tests come down, we will integrate them into the tool and develop training modules so that the health workers can easily and reliably perform the test. The HIV test, one of the big problems with that, you may be aware that HIV testing is now going, completing clinical trials for approval as a home test. So I've done research with this test in low literacy populations to see how well they can perform it on their own. One of the biggest problems is interpretation of test results and one of the reasons why point of care tests don't get out into communities more often is because the politics around the central laboratory and their concern about quality control and that people won't be able to interpret the results or perform the test accurately. So in this module, we not only guide people through the test and provide them a simple kind of visual way. They don't fully have to understand how to interpret the results but they need to match up the pictures. Does yours look like this, this, this or this? And then in addition to that, what we've got a prototype of and I really want someone at Stanford in the bio design group maybe to polish it and give us a better version but we have a prototype of a device. So after the health worker or the patient clicks on the result, then they're instructed to put it into the device and the device has a webcam and an algorithm so it interprets the results and then it matches the results to the patient. So if they're both negative, you can say yes, your results are negative and this is what that means to go on to the counseling. And if it's positive, the same thing. But if it doesn't match, then you have a flag that there's something wrong and you give them information to indicate you can't rely on these results, you need to check with your health worker, you need to repeat the results. So it's a way to do real-time quality control in the field and I think that that's gonna be one of the key things that allows dissemination of point of care tests which is I think such an important part of improving global health. Finally, at the end, they have a comprehensive health summary. They can go and see where they are in their health checkup, their tests, this was when it was recommended, this was the last result. So they don't have to do everything in one sitting, they can come back and go in and pick up where they left off. So that's a general overview of the Health Box program. Where we are right now, we transferred the project that we were doing in India to Bangladesh earlier this year and we are almost done with all of the translation. We have like a couple more modules to go and we are planning to be in the field May, June to do our initial pilot testing of the tool with the village doctors. We're gonna be looking initially just at validating the accuracy of the tool, all the point of care test modules and looking at usability, acceptability, feasibility. We've done iterative prototyping along the way and we did it mostly with paper and PowerPoint prototyping to speed up the process. Now it's all been programmed. We also have a wonderful authoring tool and this is really exciting because it's so hard to, in the previous iteration that I'll talk to you about in a minute, you design this tool but if you wanna change it, it takes a lot of money and an act of God to get the programmers to go in and do the changes. And so with this, we actually have an authoring component. So now we have staff in Bangladesh and in India that are going in that are loading up audio, video, checking the text with the programming guides and we can really be much more facile in the way we customize the tool to the communities and to responses that we get. And then subsequent to the pilot, ICDDRB working with DFID is planning to expand an evaluation. So my goal is to ultimately do the final perfections of the tool through this pilot and then do a community randomized trial where we look at the impact on disease state and have a really robust evaluation. I wanna show you a little bit of data. Actually, does anybody have any questions on HealthFox before I move on to some of the data that we have on interactive computer tools in the US? I'll wait till the end for discussion. Okay. So I wanted to show you some data so that you can see both where this tool and idea came from but the impact that we've seen and the potential power of interactive computer health tools in the US. We first started with the CARE tool which was a tool that initially I conceived of when a study project respect was published and even if you're familiar with HIV literature we'll know that it's the study that showed that with 20 minutes of brief interactive counseling you could reduce incident STD infections by 30% and they compared it to counseling that a doctor provided or health provider that was more didactic and that did not have an impact but this interactive, it's kind of like motivational interviewing but basically patient-centered counseling, you know, looking at what their personal risks are, what their motivations are, what the barriers are and having them come up with a specific plan led to a 30% decrease in incidents. And when that came out I thought, oh my God, you know, we can do this in primary care, we can do this anywhere, you know? But actually that turned out not to be true. It was very difficult to implement 20 minutes of counseling, you know, in primary care and very difficult to train medical students and providers how to do it even though there's been efforts to do so. And even STD clinics where you would think it would be absolutely, you know, no brainer for them to incorporate it was difficult to incorporate because there were so many other priorities. So at that time I thought, well, you know, we'll just do it through a computer. They'll get the counseling, they'll get the testing and they can do the other stuff in the clinical visit and, you know, talk in more depth with the provider about that. So we got funding from the CDC to build this tool and we've had several iterations of development and funding. Initially we designed it just for risk reduction education and STD clinics, then we built on a rapid HIV testing component to go with it and then we built on risk reduction and medication adherence tool for positives and the next iteration is working on substance abuse as well. And we did successful demonstrations in the US looking at acceptability and effectiveness and I'll show you a little bit of this data. So the first place that we looked at was, could we use it in the urgent care clinic? Because, you know, as you know, every ER and urgent care clinic, every clinic is supposed to test anybody who's sexually active for HIV if they haven't been tested, you know. And even though that's a recommendation and even though there's plenty of data showing higher prevalence in urgent care and ER settings, it's still often not implemented. The project where I did this was at Harborview Clinic in Seattle and there they actually had a policy at the time of referring people out for HIV testing because they didn't have the counselors and they didn't, you know, it wasn't part of their spectrum of services at the time. And so we implemented the care tool with rapid HIV testing and we did a randomized trial and we just looked at what people were getting in standard care and whether or not people got HIV counseling and testing with the care tool. And what we found in standard care is just 1% of the patients and then the sample size was 239 total in this. So fairly small, that did the care tool actually and about equivalent in the other arm. But so they weren't even really diagnosing risks for HIV and referring. And in our arm, 96% of the people that enrolled in it went through the full process, got their HIV test result, got a risk reduction plan before they left. Very acceptable. Interestingly, they didn't really share it with the health providers. We gave them a printout to share if they wanted to. They didn't really bring that into the visit but they valued the service and left knowing their status. Care, HIV testing. No, no, no, no, no. It was really just looking at what's going on in standard care, what's happening because we didn't know if they were offering testing, if they were doing it on site. We knew it wasn't a policy to do it routinely but we didn't know if they would pick up risks among people that had risks and offer it. So this side in standard care at that time, it just wasn't happening. Just wasn't happening. And so this really just proved the acceptability and feasibility of doing computer assisted rapid testing. And also if you looked at the cost of delivering this as compared to a staff based model where you had an individual doing the counseling and testing because in this case, even people who'd never had computer experience before, people that were low literacy did it themselves. It was not provided by the staff. It was about half as much to provide this kind of counseling and testing. After that, I had also, I was working with a community organization called People of Color Pocan. And we were focusing on developing a mobile health program to deliver HIV counseling and testing to people in the community that might not seek out testing in other places. And in this program, I just finished a research project with the CDC, another one where we kind of identified in a randomized trial that oral fluid rapid testing was more cost effective and incentives was cost effective. So we designed the program based on those features that we'd identified worked. And they were great at finding people in the community. Sorry about that. But they really didn't have a good system to get data into a database to show what they were doing. And when the program transitioned from a grant based program from the CDC into a city funded program, they only got paid when they were able to show what they'd done. And it was a huge problem. And so after we evaluated this tool, I just gave it to them, trained them up. It took two weeks and let them go with it. And what they ended up doing was after two weeks totally loved the tool and it allowed them to focus all their time and energy in the fields and not to have to come back and enter data into a data system. They actually ended up doubling the number of people that they could reach because they spent more time out in the field. And for the first time, they were able to actually provide the reports because we automated the reports and get paid on time. So that was really a huge benefit for them. And again, because they were able to reach so many more people with the amount of time that they had, it was a reduced cost. And finally, and what time is it? Let me make sure I'm okay. All right. Finally, in the HIV clinic, we wanted to see if we could modify the tool to work with people with HIV and focusing both on sexual risk behaviors as well as medication adherence. And that study was a randomized trial where we randomized people from the Madison Clinic, which is really kind of a state of the art, best possible care clinic. So they were getting group based care, multidisciplinary. And we added the care tool for the folks that were randomized to this really lovely model of care. And we found that in the control group over the period of the study, we tested people every three months for nine months. The transmission risk went up by 2%, and in the intervention group, it went down by 22% based on behavioral, reported behaviors. Medication adherence in the intervention group went up by 19%. And importantly, the viral load in the intervention population went down, whereas in the control population, it went up. So that was good evidence that, you know, this method of interactive computer counseling could have an impact at very low cost in addition to standard, you know, high quality care, that it was offering something that the patients, you know, couldn't completely get from the providers in the current system. So that's some of the data that we have. And then the last thing that I wanted to share with you is just, you know, a model for sustainability. And this is a really difficult thing because, you know, anybody that looks at this, you know, will say, well, that's really great, you know, as far as the Bangladesh program to be able to provide all these services, but how is it gonna be paid for, and is it really feasible? And I think, you know, that's an open question that we hope to answer. But the model that we're looking at is a health worker entrepreneur model. So we plan to combine microcredit services with the health box training program. And so health workers that are interested, and by the way, in all of the community groups that we went to, there was always at least, you know, a couple people who were really excited about becoming health worker entrepreneurs. It was a model that they were very interested in trying out. So in this case, the health worker would receive a loan and training and the health box toolkit. And then they would offer free general health assessments to the community and education, which is, you know, in the qualitative work that we did, the level of health literacy was very, very low and the level of interest in learning about, you know, variety of these health topics was very, very high. So really great demand. And what we believe is if the health worker entrepreneurs deliver the assessments for free and the education for free, they'll be having, you know, an important public health impact and they'll also be building a rapport with the community and trust with the community. As I mentioned, we use fingerprint authentication to be able to follow people longitudinally. In addition to the assessment and counseling, they would then provide based on recommendations that are geared and individualized based on responses, you know, options for purchasing, testing, health products and algorithmic treatments. And it would have to be at a low subsidized cost. And what we found was in general, you know, 40 rupees or about a dollar was a very acceptable cost. Although interestingly, in some of the work that we did with the focus groups, it was clear that when people had a problem, a health problem, they would often pay quite a bit more for private care. Even if they were at a fairly low socioeconomic level, it's a common practice. So I think there's probably some room to move there but about, you know, a dollar per test, per medicine, per health product is what could be expected. And then I also think, you know, I've done a variety of iterations of cost models for this and in some scenarios where a health worker has, you know, perhaps four different people that they're working with at one time on the health box tool. And if each of those generate, you know, one to two dollars in income, then there's a point at which it can break even. But I think in most likelihood and certainly initially, we would probably have a private, you know, payer partner, public, private public partnership that would include, you know, funds from global donors, for example, that could fund the delivery of targeted services. So if you had a group that really wanted to eradicate malaria and wanted distribution of bed nets, you know, that could be offered for free to anyone and some of those funds would help subsidize the time of the health workers. And as well as partnerships with Ministry of Health where the Ministry of Health can provide the medications, you know, and some of the test costs and the health worker network, you know, could deliver the services. I think it would be a win-win and that's what we're going to be exploring further. So after the services are delivered, the health worker returns to the kiosk to get their supplies, to upload the data, to download any updates to the tool and to verify the services that were provided for payment. And it's interesting to think about the colleagues that I work with in India Institute of Technology who developed the rapid test interpretation device that I talked to you about. They also are very involved in mobile banking where people can go and use, you know, thumbprints to use. So you can imagine a system where you have a kiosk that's kind of a hub and a health worker could go in and upload their data and they could get paid, you know, and go to the cash machine and actually get money and have everything, you know, very well managed and, you know, with the potential to evaluate at any moment. And then ideally, with the income that the health workers make, part of that would go to repaying the loan so that you could then build the network. One of the other important points of this, and this is true for the US, you know, as well. I mean, to really be able to capitalize on the efficiency of this kind of system, the information flow is one of the key components. And if we can get this right, both in the US and globally, where we have, you know, mobile health and internet-based programs that are delivered to the patient, you know, in the case of low literacy populations or high-risk populations with health workers, and if that data can flow back to the village doctors and, you know, telemedicine consults, primary care clinics and hospitals, then, you know, not only will continuity of care be better, but the efficiency of care will be much better. What you see now in the telemedicine efforts out there, at least that I've observed, is they haven't, the models haven't really been perfected in that the scheduling of the patient and the provider time is very difficult, and then the provider, when they come to the visit, has to start from the very beginning and spend a lot of time doing a full, remote workup. But you can imagine, if they had a printout that summarized, you know, all of the critical information about the patient, what the recommendations were, what's already been done, you know, they could get back pretty quickly with appropriate recommendations, and so I think that's the model that we're working for. As I mentioned, we've completed four rounds of prototype testing. The state keeps on changing, but we're looking forward to being in the field in May, and then to moving on to doing randomized trials to document the effectiveness and impact. And, you know, what we believe is with this kind of integrated system that really provides, you know, targeted health education and early detection and referral for all of these conditions, we could make a large impact on global health. If we look at the WHO building blocks for improving healthcare systems, you know, this program has been designed to meet those building blocks, which is improving health service delivery, creating an effective workforce, providing the appropriate information, the appropriate health products, creating a system with financing that is sustainable, and having good stewardship. I'd also like to just acknowledge all of the many people that have been working on developing this tool, and we'll open it up to questions. I'm wondering if, oh, go ahead. Oh, no, go ahead, I'll ask, let me see. So, I'm wondering a little bit, how many of these boxes do you deploy? And then, tell me a little bit about the sustainability in the harsh weather of Bangladesh during monsoons, and it looks pretty, tell us a little bit about that. Is there a lot of these things that go out there? Right, right. It's all about sustainability. Yeah, you know, and I think when we first started designing this, you know, we thought about, are we gonna do a cheap cell phone intervention? Are we gonna do a tablet intervention? Ultimately, because of the feedback that we got in the field, particularly people that are a little bit older can't deal with, you know, seeing the cell phones, and they appreciated the larger user interface. And because the prices of these netbooks and tablets, you know, are coming down and really meeting the prices of smartphones, we decided to go with this strategy, although it's been designed so that we could put it on a smartphone. You know, it couldn't be on a low-cost phone, although some of the content could. So then the question is, you know, rugged design of tablets. And there are a couple that have been designed, Intel has one that they've designed for health workers that are, you know, supposed to be more rugged. And I think it, you know, there will be many, many issues to work out in, you know, the sustainability, you know, I mean, you do have to eventually hook up to electricity, you know, you do have to be able to have a way to keep the, you know, even a rugged laptop dry during monsoon season. You know, I mean, there'll be issues to work out. But I think, you know, as we've seen an explosion of mobile health interventions around the globe, clearly it's, you know, it's possible, it's feasible. And there've been many, many interventions that are kind of isolated on, you know, specific aspects of healthcare. And what I have really hoped to do is come up with something that was, you know, more holistic that addressed a full spectrum of health needs of, you know, an individual or family to, again, make the delivery of healthcare more efficient. You know, because if you have teams and teams of different outreach workers, some are working on vitamin A, some are working on, you know, immunization, some are working on malaria with this, you know, integrated systems, it just, it doesn't make sense. So I think, you know, this is at the beginning of, you know, it's a long development process and there'll be technology issues to work out. But I think that those problems are relatively simple, you know, to solve compared to really getting the nuances of the health delivery, you know, worked out. So, mm-hmm. Are you guys aware of your certain populations in the US? Yeah, absolutely. I mean, that's one of the things that I'm really, really excited about doing now. And especially, you know, if you think about, you know, the data on health disparities in this country and some of the work, I didn't show you this data, but, you know, well, HIV, for example, you know, we all know that blacks and Latinos have higher, you know, morbidity, mortality from HIV and also have, you know, poorer access to healthcare. And when we delivered this tool, the care tool, the earlier version of this to those populations, we showed that we were reaching, you know, higher, a different population than those that came into the clinic. You know, they had higher prevalence of HIV, they were lower literacy, higher substance use rates, even when we compared the blacks and Latinos that went to the STD clinic to those that were reached by the mobile health tool, we found, you know, higher rates and higher risks in that population. And so I think that if we take this system and we create both internet-based programs for those that are high literacy, as well as community health worker guided programs for those that are highest risk and lower literacy will be able to, you know, hit the triple or the quadruple aim if you add health disparities onto it. So it's always a matter of- What's if we have a health insurance? If we have- Yeah, no kidding. We should, it's one o'clock. One last question from a quickie. Just listening to the health debates is just so depressing. Yeah. It is, it's really scary. But you know, these change, I mean, no matter what happens, these kinds of systems and innovations, because they do lower healthcare costs and they do improve acceptability and quality, they'll happen. Actually, I want you guys to look out for in April 11th, maybe April 19th, somewhere in there, we're gonna be launching the I Know SFA campaign where women between 18 and 30 will be able to go online, get a home STD testing kit and if they're positive, get treatment at a local pharmacy and if they're asymptomatic. And we're doing that in Contra Costa, Alameda, San Francisco and San Mateo counties. So in fact, I'm gonna send you guys the Facebook because you gotta like it and then it will get out to more people. The preceding program is copyrighted by the Board of Trustees of the Leland Stanford Junior University. Please visit us at med.stanford.edu.