Written by Aida Akl at Voice of America, May 8, 2015
In some low-income countries, where there is a shortage of doctors, the odds are there are children who may never actually get to see a physician. But a unique collaboration between the University of Vermont’s medical faculty, software developers and tech professionals could provide these communities with a viable health care solution.
The fruit of that labor is the Medical Evaluation and Diagnostic System for Infants, Children and Newborns or MEDSINC – a mobile, integrated medical platform that could help community health workers determine the severity of a child’s sickness and make treatment recommendations accordingly.
Barry Finette, Professor of Pediatrics, Microbiology and Molecular Genetics at the University of Vermont and co-developer of MEDSINC, discussed the platform in an interview with TECHtonics.
Q. How did this idea come about?
FINETTE: … There are a number of reasons as a pediatrician and as a physician. One thing that is inherent in when you do this job is that you have a strong belief that access to health care is a basic human right, and… the work I’ve done in resource-poor countries in particular – I’ve witnessed the tragedy associated with children dying young simply because they’re poor, and to try to work out new methods that can reverse that as soon as possible.
Q. What is MEDSINC?
FINETTE: … It’s something that is still being developed … We already built a prototype. And this is its next version that will lead hopefully to implementation in the near future.
… The major goal of this platform is to allow people, specifically community healthcare workers who are unskilled at doing clinical assessments of children … to determine how sick a child is and automatically generate treatment recommendations for those community health care workers to implement. So the idea is to basically disrupt the healthcare system that’s in place in lower income countries.
And the reasons we’re looking to use this approach is because in these countries there is in general one physician for every 20-100,000 people. So there may be only one physician for every 2-10,000 children under five years of age. So there’s no way that any of these children will ever actually see a doctor. But there may be one community health care worker for every 10 or 15 children that they oversee. So that is basically the goal of this.
Q. So is it an app or a platform?
FINETTE: It’s a platform … It has many different functions … And it doesn’t use actually the operating system in the mobile device on which it’s loaded. So by definition, it’s more of a platform than an application.
Q. How can people with mobile phones access the platform?
FINETTE: The way that we are envisioning this to be used is on any type of mobile device … Any device that can download a website will allow them to have a fully functional platform on the MEDSINC platform.
Q. What do health workers see or get once they access the platform?
FINETTE: … We’ve employed an approach that’s used by any type of certified pediatrician.
… When a pediatrician or a family doctor is assessing a child – whether they’re sick or not – they determine four things … Is this child in respiratory distress and if so, how severe, whether they are in mild, non-mild, moderate to severe respiratory distress. They do the same thing relative to determining if the child is dehydrated, if the child is at risk for infection, and to what degree the child might be malnourished.
… That’s exactly what the MEDSINC platform does. It determines the severity of those four areas. And we do it by having the user walk through currently about 22-25 simple questions in which they just answer by pushing a button on the platform that are related to the patient’s current history of illness.
…. And we teach [community health workers] how to … determine basic vital signs, as well as some basic physical exam features. And when those are all inputted, then an integrated assessment is determined, as well as a treatment recommendation based on the findings.
Q. All of this happens in real time?
FINETTE: It’s all real time. Embedded in the platform are illustrations and videos to assist the user to learn. So it’s actually a self-learning system as well … The more the user uses it, the more efficient and the more accuracy it obtains … It’s been taking even unskilled workers maybe about 10 or 15 minutes to go through the platform for each patient.
Q. Will community health workers be able to connect with medical professionals, if needed?
FINETTE: … Once the MEDSINC platform is on the device, it is fully capable of doing any clinical assessment and generating treatments without any Internet or wireless connection … If there is a wireless connection or a cellphone connection, we are going to build a feature that will allow them to then transmit the data that they acquired to anybody who they decide that would need to … a local nurse or even better, probably a local physician working at a clinic or hospital.
Q. Will that help in compiling a local medical database?
FINETTE: Absolutely …There is … public health data that is acquired and … saved on the device until it’s sent to the Internet. And then that will be downloaded into a clinical database that we will be developing.
Q. And is this going to be confidential information?
FINETTE: Yes. The patient identification will be confidential. So there won’t be any concerns for misuse in that way. If for some reason in the future, a region or a country or a specific group would be interested in having patient identification, then that’s something that we could probably adapt in the future. But we have to follow all the laws in that particular country region as well.
Q. Is the platform available in languages other than English?
FINETTE: … The user interface has very little words to it. So actually having us adapt the user interface words to any language is not something that would be that difficult.
Q. Have you done any trials yet?
FINETTE: We’ve done trials both locally at the University of Vermont medical center as well as in Bangladesh, in Dhaka.
… Our initial results were mainly looking to test the accuracy and robustness of our platform itself in making any clinical assessment as close to a doctor as possible.
… Locally, we’ve seen a correlation of the clinical assessments between the MEDSINC users who are not physicians to board-certified physicians to be over 90 percent correlation. And in Bangladesh, with the use of community healthcare workers there, we’ve seen over an 80 percent correlation with the local doctors, as well as our doctors.
… These were health workers that had no specific training in doing clinical assessment of children. They did not at that time know how to determine the heart rate of a child …. But we taught them how to do such things quickly. And we have some tools on the platform to help them do that.
Q. Do you worry that non-skilled workers using the platform might make the wrong diagnosis?
FINETTE: … The platform is just a tool for them to use … I don’t believe that … they’d feel that it would make them into being a doctor per se, but we would hope that they would feel more confident about the tool as they use it.