“We use open data kits, a lot of mobile data collection, and power BI on the back-end for the analysis and dashboarding”
Pete Masters is Research and Innovation Coordinator at medical NGO Medicine Sans Frontières (MSF). He joined Computer Business Review to discuss innovation and digitalisation in some of the world’s most challenging field environments.
Pete – How has Covid affected MSF?
MSF is a billion dollar organization which has a very well established innovation tradition. We’ve really tried to ramp up our digital health promotion toolset. We think that Covid is here to stay for a while, and the way that we provide care at the moment — mostly encouraging patients to come to the clinic — isn’t going to work for Covid patients. If you have Covid patients coming to a clinic, you essentially have to close down your normal services.
As a result, we’re currently looking into what home-based care looks with Covid. How do we equip communities and households to the level where they can, as much as possible and as safely as possible, care for themselves?
This could involve a whole bunch of different technology solutions from things like community or household oxygen saturation monitors so people can actually monitor their own oxygen saturation, to working out how tele-consultations work in the kind of settings we work in. We’ve also been doing lots of work around digital health promotion like social media marketing techniques, using gamification and setting up WhatsApp hotlines and call centres to make providing care in the Covid world possible.
Does much innovation come from the field?
We have an internal innovation fund called the Sapling Nursery. It’s for ideas that come from field teams who want to improve the way that they work.
For example, a medical coordinator who was working in the Central African Republic wanted to update something that we call the Cold Chain [challenges around shipping and storing vaccines in a “cold chain” in the tropical heat of many resource-limited countries is a major cause of poor immunisation coverage rates].
So, with Cold Chain, you have your drugs or your vaccines in a box, and they have to stay at the same temperature. One of the problems with this is to measure the temperature inside the box, you used to have to open the box, which obviously drops the temperature or raises the temperature. We wanted to get to a point where we had smart labelling for cold chain stuff. We funded a project with a company who does this kind of thing, to look at how you could have a label on the outside with sensors on the inside and the label would change as the inside changed…
There are various commercial innovations that have come out of MSF.
For example, ready to use therapeutic food, etc., which is a high nutrient peanut paste, has replaced flour and oil as the thing you give out in a nutrition crises.
We think those two people on the Cold Chain project are going to take that on as a project independently, as a startup in itself. So MSF will have been a nest for that project, and it’ll fly off and do its thing. Obviously we’ll still be involved, it will still be tested within and meet the requirements of MSF projects. But as a product, it would be a sustainable business model and may become a way of actually generating cash. MSF obviously isn’t in a place to be the owner of intellectual property or a merchant of a product, so that’s probably going to be a spin off.
Do any other partnerships jump out as interesting?
We have relationships with lots of different academic institutions. The London School of Hygiene, Tropical Medicine, we work with the London School of Economics occasionally; Imperial College on some early stage innovation stuff.
We also have partnerships with some of the big tech companies.
There’s a project ongoing with Google at the moment and we’ve worked with various kinds of product design agencies.
What are you working on with Google?
So we’re working with Google on a couple of different things.
We had a failed project during the West Africa Ebola outbreak, which was an electronic medical record system and device. With Ebola, there’s a high risk zone and a low risk zone; you can’t move between them without donning or doffing your PPE and being essentially sanitised. You can’t take anything in or out that doesn’t get dumped in chlorine, so patient records were a real problem.
Essentially, what we had to do was take the consultation with a patient, go to the fence and shout the patient’s data across a gap to someone else, who then writes it down. No privacy and a huge margin for error. MSF and Google worked on an electronic version of a tablet that could actually be dumped in chlorine, that would transmit data directly between the outside and inside. However, by the time they were ready to pilot, the outbreak was over. We took too long, basically.
We reinvigorated that project in the latest Ebola outbreak in DRC without Google, building on the work that they did, and it was much more successful. But still, the developments for it came slightly too late to be proven.
I understand you use weather data quite a lot?
We’re submitting a proposal at the moment to do with climate epidemiology. We know that climate has an effect on the epidemiology of outbreaks such as malaria or dengue. However, the climate and weather data that you can access is quite crude, so although you might know the general details, the data for Congo, for the project area we work in, could be completely different. The site could be up a mountain or in a valley or by lake, which would make small changes in the climate.
We want to look at what the installation of local climate stations or local weather stations could add when compared with the kind of weather accessible data at the moment, to look at whether there’s any relationship between that local climate data and the epidemiology of diseases. If there is, this means we could predict certain aspects of diseases, like seasonality maybe, or what kind of severity an outbreak might have, which would help field teams to prepare better, to allocate resources better and to help more people through that outbreak.
So that’s an exciting one because that’s got implications to aid people effected by diseases like malaria and dengue, but maybe also for malnutrition.
What sort of tech do you think you’d be creating for that?
We don’t know yet whether it will be off-the-shelf weather stations. There are a few available, even some weather station NGOs who are trying to build out more local useful granular weather data. Or we might need to develop something.
It’s likely we’ll try and find the best thing that exists. When MSF wants to do something, we normally try to find someone else who can do it first before investing, because we’re not product designers and we like to spend as much of our money as possible on our medical operations. So if we can find a solution that either fits or can be adapted, we try and do that first. If not, then we’re always open to getting our hands dirty
How much is “digital transformation” an issue for you?
Last year there were some good examples. A hot topic in the MSF is quality of care, which is, essentially, an ethical question. We’re working in places where the available resources can be extremely scarce, sometimes in places where trained staff are scarce. So how do we retain a standard of care that’s high enough that you’re not actually doing harm to people who are coming to seek care in a facility?
This is a whole specialisation in itself, but one of the ways that’s done is through audits. You can audit all sorts of things. You can audit patient notes to see what kind of care patients were given when they were escalated. You can audit prescriptions, you can audit infection prevention, control measures. Hand washing and cleaning, for example, stopping infections spreading from patient to patient, creating outbreaks in the hospital. This helps to improve standards in these hospitals.
Last year we did a big project looking at how to take those paper audits and digitise them. There are several advantages to this. At one level in a facility, we don’t have to do a whole bunch of digitisation of your paper audit.
And we can automate most of the analysis so we can just send monthly reports on how well hand washing is being done or how a facility is looking year on year. The team needs to do the analysis, to make further improvements or to correct things that go wrong. And then, on a global level, you can look at the MSF portfolio and see where we’re struggling in whichever area.
That’s one example of a digital transformation project that uses a combination of open source and proprietary software.
We use open data kits, a lot of mobile data collection, and power BI on the backend for the analysis and dashboarding.
Do you use the IoT much?
We’ve done some interesting stuff with more biomedical devices, and we have used IoT technology to monitor compressed oxygen, to work out whether that bottle has gotten too low. The sensors can monitor how the compressor that fills the bottles is performing, it can automate a warning that is sent to the biomedical team so that the machine can be fixed.
We’ve also piloted this with some water system stuff. So looking at flow rate and turbidity, there are hundreds of different data points you can gather to make sure water is clean and safe and accessible. That’s definitely being piloted in certain places.
Are field tech skills widespread in MSF?
There’s 45,000 staff, especially on the logistics and biomed side, they come from all sorts of walks of life. You meet some really weird and wonderfully skilled people. I’d say it’s one of the unique selling points of the organisation.
At the same time, we don’t have a good way of identifying where those people are and grabbing them when we need them.
So it’s also a frustration because you find out a year after you needed someone who absolutely knows what they’re talking about with solar power, that there was a guy in Congo last year who absolutely knows what he’s talking about with solar energy.
Are you big Open Data users?
From a personal point of view, I think one of the biggest successes since I’ve been working in MSF in digital innovation is around geography and geodata, something called the Missing Maps Project. It has changed the way we treat patients.
I’m not pretending that the methodology is saving people’s lives, but the way that we can incorporate geodata in an intelligent way into our epidemiology, into our surveillance, into our operations, into our supply, into all these things has radically changed since we launched the Missing Maps project.
It’s moved MSF to accept and value open source data and the techniques within it. I think there’s massive scope to develop that. There are huge areas of the world where MSF works that aren’t on the map, where there is no data.
I’m a big advocate of open source, whether it’s software, hardware, data, and I think MSF could benefit a lot more from that, especially considering the number of skilled people that it has within the organization who are keen on it. But what we haven’t really created is an ecosystem in which they can play or experiment.
Obviously we know that for the people developing stuff themselves, the hackers, it can be dangerous. When you skip privacy or security you can end up doing some harm. I feel like if MSF put the resource into creating an ecosystem in which the constraints and the limitations were quite clear, where they could safely experiment.
I think my final question would be what tech innovation has had the most success?
I recently asked a bunch of colleagues what their top three favourite innovations were. The mapping stuff came up quite strongly. But our “kits” concept came up strongly too.
Take a cholera outbreak somewhere in South Sudan. To care for those affected, you need to go and buy ‘this’, and then we need some of ‘that’, and don’t forget ‘those’… A bunch of smart logistics people decided that we should kit boxes in a warehouse that contain everything you need to care for a cholera patient.
Then when your data shows a spike in cases, you can take those kits, put them on a plane and four hours later they’re there and then you can just start responding, instead of having to buy all these pieces piecemeal. From the word go; a forklift picks it up, takes it, puts it on the plane, all the flight scheduling is done. It’s a quite complex technological chain to make all that work. But once it was set up and working, it was absolutely ground breaking, the kits project is completely streamlined now.