Ophthalmology Registrar at Stellenbosch University and Founder of Vula Mobile, Dr William Mapham, talks about the development of the award winning mHealth app and how it’s helping to transform rural healthcare in Southern Africa. He also dispels some of the myths around the “Uberisation of healthcare” and discusses the fundamentals to solving healthcare challenges.
Tell us the story behind Vula.
As a junior doctor stationed at a rural hospital in the Transkei I experienced first-hand what it was like to have no support and access to specialist opinions. I recognised that the problem wasn’t going to be “fixed” and we needed systemic change. In 2005/2006 I moved into public health and started working with Soul City using media for health payer change. It was around the same time that mobile phones were starting to gather momentum as a form of mass media, which led to me going to the US to do a fellowship looking at mobile applications for healthcare at Columbia University. I ended up working for a start up in Washington, but my passion was still rural healthcare in South Africa. So I quit my job, moved back to SA and worked for the South African National AIDS Council (SANAC) on policy work before deciding to go back to my roots of clinical medicine.
That decision resulted in me volunteering for 10 months at an eye clinic in Swaziland. It was there that I saw patients coming in far too late with symptoms too advanced to be treated effectively with the resources at our disposal. And although there were health workers in the community who could screen people, they didn’t really know what cases should be referred. It was where I could see clearly how a mobile phone could be used to improve patient eye care in a rural public health setting.
And it started with just Ophthalmology?
The eye is obviously very important to me and it was a great way to start because it’s such a good visual specialty. Most health workers only have two weeks of eye training at medical school so there’s a huge skills gap between them and specialists. And with general medicine, case transfers are more complicated because a whole bunch of other data is required, like ECGs, etc. So Ophthalmology was where we began but we’ve since added a number of specialties on the app to broaden its application in the real world setting. Vula now includes Ophthalmology; Orthopaedics; Dermatology; Burns; HIV; Family Medicine; Internal Medicine; Neurosurgery; ENT; Cardiology; and Oncology. And in the near future we’ll be adding Surgery; Obs and Gynae; and Paediatrics.
So talk us through a Vula user experience.
Imagine you’re a newly qualified junior doctor and you’ve just been sent out to the boarder of Lesotho. You’re the only doctor there and patients are queuing up at the door. And while you’ve learnt a lot at medical school you’ll still come across a case that you just don’t know what it is. You can look up the case in your books but you’ll only get so far. So inevitably you would make a phone call to someone or send them a picture on WhatsApp asking for advice, but that method is informal and undocumented.
With Vula, that same doctor is able to follow the referral workflow we’ve developed for each of the specialties. For example, with Ophthalmology there’s a vision test and a specific questionnaire about the patient that must be completed before sending it to the on-call specialist to evaluate. We’ve specifically built an on-call system so it doesn’t just go to a random doctor to answer but instead it goes to the doctor on-call whose job it is to answer these kinds of questions. The average response time from a specialist is about 15 minutes, so instead of being put on hold or wasting time finding a second opinion, the healthcare worker is free to continue seeing other patients.
From the specialists’ perspective, Vula also takes their experience into the design. Instead of getting a phone call from a rural clinic asking for advice, they now digitally receive a package of relevant information that includes pictures, a vision test result and the patient’s history structured in a way that enables them to reply quickly.
What do you know from the data Vula has generated since its inception?
We’ve noted that about 25% of all cases, across all specialties, are actually managed at the primary level in the rural setting. This is important because it shows that Vula has helped to minimise unnecessary referrals and, more than that, if patients do get referred they are given a specific date to visit the hospital to ensure that they are seen to properly. Using that data, we were able to start a conversation with the School of Public Health to initiate an economic study to understand how much money is saved by reducing referrals by 25%.
This data is also valuable because we now have a better idea of what’s happening and we can track how the rural health workers are learning case by case. To give one example, there was a child whose eye was accidently burnt by boiling water and we taught the doctor how to manage the case on Vula. Although he referred the patient anyway in the end, at least he ensured the right thing had been done at the coal face. A while later he saw a similar case, although this time the patient was a baby, and the doctor knew what to do. So Vula is also being used as a teaching tool, which is something we didn’t predict. It’s been fascinating to watch the data come in and see how people are learning from it and how junior doctors are using it to manage more complicated cases on their own with support.
Let’s touch on the “Uberisation of healthcare.” New coverage recently will probably go a long way in deterring people from using the term but there was a time when it was readily used to illustrate the automation of processes in healthcare. Do you have any comment on that?
There’s actually a brilliant article called ‘Why there is no Uber for healthcare’ which I found absolutely fascinating. In short, you might use Uber say 100 times a year. But you’ll only see a doctor four times a year. Catching an Uber will cost you R50 a time, whereas seeing a doctor will cost you R500. Your commodity in Uber is your taxi drivers, who are skilled drivers but they haven’t gone through 10 years of training, or at the very least a minimum of six years of training as is the case when you see a doctor. As a result Uber can go viral far more easily than any disruptive technology in the health sector.
Healthcare is a very complicated sector and as a result, innovation moves very slowly. If you’re going to release a new drug or surgical tool it’ll take years to get approval. Like with Vula, despite having tons of support within hospitals it still took us two years to get any recognition from the Department of Health. And that is right because tools and innovation, like medicines and devices, need to be rigorously tested and proven before exposing the majority of patients to it. In healthcare, it’s a priority to protect people and technologies have to be designed with that objective in mind.
Vula has been the recipient of numerous awards and accolades over the last couple of years. Give us the winning formulae for designing and launching mHealth solutions based on your experience.
The initial version of Vula was basically built on a power point template. I was very fortunate to get R50,000 in funding from the Shuttleworth Foundation. And while the funding was nice what it really gave me was some credibility and the confidence to phone around although most developers laughed at my budget. I then contacted Gary Marsden, who used to run the UCT Centre in ICT for Development, and asked if he knew anyone who could help me. He was extremely helpful and gave me a list of key people, one being Debré Barrett who, at the time, was running a company called Flow Interactive, which was South Africa’s first ever user experience company. So not only was she a talented business person, she had specific expertise on how to make complicated things simple. Her advice was clear: if you’re going to design this app you need to look at who’s going to use it; why would they use it; how would it make their life easier; and if it was going to make their life easier what would it look like. She gave me a lot of guidance which helped me to know exactly what I wanted.
I used a system called Productivity on Paper, which has since been bought by a company called Marvel, which basically allowed me to create a non-functioning app which I could then send to developers and ask for quotes. And although I couldn’t afford them at least I had an idea of how much money we needed. And then I got lucky again because Debré phoned me and said that one of their designers had available time to work on Vula for a couple of months. This was amazing because they donated around R200,000 worth of design time which produced a real Android demo which went on to win the SAB Innovation Award in 2013, which was worth R1 million. We’ve kind of bankrolled our prize money since then; we won a big prize in Morocco and then another big competition last year in London which has really kept us going.
At the same time when we won the award in Morocco, Debré sold her company to Deloitte Digital and initially helped us on a part time basis which turned into full time. I’ve been very lucky to have her on the team and if you look at Vula’s growth it’s very obvious that it grew exponentially once she came on-board. When Sara Hilliard Garrett, Strategist and ex-Advocate, came on board in early 2016 Vula stared to grow even faster. The three of us with vastly different skills enables a cauldron of debate which produces robust solutions.
So to answer your question, I think the formulae for Vula can be distilled down to funding, people and more than that, experts in their field and a problem to solve. Vula was designed to solve a problem; it wasn’t built as a business where we looked at the market and how much money we could generate. Obviously we’ve matured since our start-up days in a number of ways and currently, we’re in the process of setting up a Board with a view to operate as a professional enterprise but even still, our value doesn’t just come down to profits and losses. Vula has grown way beyond what I dreamed it could be. We’re carving out our value by facilitating collaboration between public and private healthcare, we’re actively reducing unnecessary referrals which has an impact on the cost of delivering care and in February of this year, we set a new record for the number of patients helped, which was over 1,000.
Lastly, give us your real-world take on collaboration between developers and clinicians to create mHealth apps.
I was very fortunate to have a foot in both camps. And even more fortunate to have the support of people like Debré and a few other entrepreneurs who were willing to share their advice and expertise with me. So I think the collaboration stretches further than the developer and the clinician.
Vula is expensive to build and maintain and, certainly initially, we just didn’t know that because we didn’t have any experience in developing or building software. While you do get the odd exception of a clinician who has taught themselves how to program, it’s still not their core function. What clinicians are really good at is thinking about what would help them. But building software is really, really hard – something we completely underestimate as doctors. On the other hand, from a developers’ point of view it’s sometimes easy to think you’ve got the perfect system but in reality it’s not practical. Working in the public health sector for most of my life I understand how precious time is in a clinic and how simple things have to be. The minute tools get complicated or onerous it’s just not going to happen because at the end of the day, we want to help the patient – that is why we’re there. Tools have to support that purpose unequivocally.
mHealth development needs a combined approach, with numerous collaborators and contributors who are all aligned to solving a problem for an individual, a community, a region and eventually, an entire health system.