Researchers at Western Sydney University (a participant of Digital Health CRC) have developed a mobile app to help patients better manage Type 2 Diabetes. In doing so, they have taken inspiration from GPS technology and another app they designed to help Sri Lankan farmers avoid growing unwanted crops…ultimately helping to prevent farmer suicides.
In 2011, Athula Ginige – Professor of Information Technology with the School of Computing, Engineering & Mathematics at Western Sydney University (WSU) – undertook a research project in Sri Lanka with other WSU colleagues, to look at the problem of over-production of crops by local farmers.
“All of the farmers were growing the same crop at the same time, but they were only finding this out once they brought the harvest to market” he says.
“Then they couldn’t sell their crop due to oversupply – and around 4000 farmers a year were committing suicide as a result.”
“We looked at the same problem in India in 2016, because 20,000 farmers there were committing suicide, with some reports placing farmer suicide rates much higher – although it is important to emphasise that these suicides may not all have been due to the poverty cycle and farmers being unable to sell their crops.”*
In response to the problem, Professor Ginige and his colleagues developed a mobile app to help coordinate the supply of crops to market and inform farmers of potential market saturation before it occurred.
“There were all sorts of solutions being proposed by the government, but they were not getting to the root cause” he says.
“We needed to find a way of letting the farmers know what was being grown by other farmers elsewhere, and help them avoid growing over-produced crops in the first place.”
To do that, they turned to GPS – the tried and true technology used for providing directions to millions of road users around the world.
“GPS presents the same information as traditional maps – it shows the roads and all the locations they lead to – but then it takes things to the next level” Professor Ginige says.
“If you want to get to location ‘x’, for example, at every juncture the GPS will tell you to go straight, left or right.”
“It turns what would otherwise be standard topographical information into actionable information. It empowers people to get to their location by the easiest route and to avoid traffic snarls.”
Once the WSU team made actionable information available to the farmers, it was a game-changer.
“Farmers can now ask the app what is the best crop for them to grow, based on their local climate” he says.
“The app will tell them what other farmers are growing, how much particular crops will cost to produce, and if the market for that crop is saturated.”
“It can also aggregate farmer data to produce wider statistics on crop production and availability across the country, including what crops are being planted in a particular area, what crops are being purchased and consumed in each region, and what farming processes are proving to be inefficient.”
Taking actionable insight into the health space
With the app for farmers going commercial in Sri Lanka, India, Fiji, Timor and South Africa, Professor Ginige and his colleagues could see potential for a similar type of app in the fight against Type 2 diabetes – one that could provide patients with actionable insight to enable them to manage their condition more effectively.
Fast-forward two years, and that app is now being trialled with a 41-patient cohort in Western Sydney.
Utilising wearable devices and the patient’s smartphone, the app captures real-time data on the patient’s daily exercise, dietary habits, emotional responses to hunger, blood glucose levels, and other parameters.
It then uses this data to feed actionable insights back to the patient, to help them better manage their condition.
When they commence the trial, a variety of information about the patient is collected – including age, gender, weight, height, current level of physical activity, ethnicity, income, education, diabetes symptoms, food consumption time, and night snacking time.
During an initial consultation, their healthcare provider reviews these parameters, and then values and updates them based on their face-to-face meeting with the patient.
The patient then starts using the app in their day-to-day life.
The app encourages the patient to regularly input their food intake, exercise and activity, blood glucose levels and wider lifestyle factors.
In this way, it gathers data on their planned and unplanned exercise, their hunger-based emotions (such as how often they feel hungry or full), the frequency of home cooked meals vs eating out, their calorie counts, and macro and micro-nutrient consumption.
You’ve got mail
Once the app captures this critical information, it turns it into actionable insight for the user – including by sending them text messages.
On average, around 21 text messages are generated weekly for each patient, providing them with tailored information on how to reduce their portion sizes when eating, how to keep within their daily calorie limits, or suggesting they may need additional exercise.
These messages are generated by comparing the difference between the patient’s goals to their real-time lifestyle data, which is being collected by the app.
The objective is to encourage the patient to meet the goals they have set, and thus achieve positive change.
If the goals are being met on a regular basis, the app takes things one step further – encouraging the patient to modify their goals to help them further improve their lifestyle behaviour.
“By helping patients to easily act upon the data gathered, our app can make a big difference in the self-management of their condition and their health outcomes” Professor Ginige says.
“Importantly, the patient is empowered by having the option to manage their own goals, which puts them in charge of managing their own condition.”
“They are able to override the goals set by their healthcare provider. This is deliberately put in place to mirror real-world conditions, where healthcare providers encourage patients to take control and actively self-manage their Type 2 diabetes.”
Bridging the patient-clinician gap
The app also acts as a supporting digital tool for interactions between the healthcare provider and patient.
Information gathered through the app is aggregated and supplied as a summary report to both parties. The healthcare provider can then create customisable reports to obtain an holistic view of the patient’s health condition and their progress on self-care.
This enables the healthcare provider to understand what behavioural advice has been successfully followed and what advice should be modified to meet the patient’s needs.
Revised daily actions and actionable insights are then generated if required, for delivery to the patient by text messaging.
The real-time data collection of the app also makes for more effective consultations between the healthcare provider and patient.
“At the moment, a patient meets every month or two with their clinician and has to try to remember what type of food they have eaten, their average daily intake of carbohydrates, and how much exercise they have done since the last consult” Professor Ginige says.
“But it is often hard to remember this information…and it is old information by the time the patient and clinician meet, so they can’t really act on it effectively.”
“Additionally, there is often very little time in a consultation to undertake a detailed history of the patient’s diet, exercise regime and other factors.”
“Healthcare providers often struggle to provide the optimal diabetes care in a busy clinical environment with a high patient load – but by generating detailed reports in advance of the consultation, our app can assist.”
“Clinicians can also more readily track how their patient is going, and they can provide rapid feedback to the patient on where they can amend their diet or lifestyle.”
“If someone’s not travelling right, we can intervene very quickly.”
Looking to the future
While the app is being trialled with a small group at present, Professor Ginige says there is potential to scale it up.
“Our plan is to distribute the app more widely across Australia, by working with the Primary Health Networks to integrate all Type 2 diabetes patients nationally” he says.
“We could then track or identify diabetes hotspots, show where people are sourcing their food and what food-types they are eating, and determine where improved community infrastructure is needed to encourage exercise, for example.”
“We could also analyse data to assess what specific patient behaviour is occurring in what areas, and enable health services to implement tailored interventions for particular population catchments.”
“We could then assess how the intervention is working using the real-time data collected by the app, and make changes to a community-wide intervention very quickly.”
“At the moment, people need to do a survey to get this type of data. The app enables us to get data dynamically and it could really help inform population level intervention.”
“The whole idea is to keep people as long as possible outside the hospital system.”
“We believe this type of app will change not only how we respond to Type 2 diabetes, but to all chronic diseases.”
Professor Ginige says the app could prove to be very effective in other countries too.
“The next most probable place to trial the app would be in India, given it is the diabetes capital of the world.”
“But we would first need to rethink the value system of specific countries, how the information is presented in those countries to ensure the best uptake, and how the app would interact with the local health systems, local diets and other factors.”
A cautionary tale
Professor Ginige says the app provides a cautionary tale about the need to better structure patient data.
“Currently, about 80% of patient information is unstructured and, in turn, unmineable” he says.
“Data is no use if we can’t act on it. With any digital health technology, there is a need to get data into a useable format and make it actionable.”
“We need to turn straight information into actionable information, just like the GPS does.”
“Our work in this space is laying the foundation for knowledge-driven healthcare, and a shift in emphasis away from curative care to preventive care.”
“By prompting all of us to lead healthier lives – and giving us actionable information in real-time on how to achieve this – this type of technology holds real promise for a healthier future.”
The trial will be completed by November this year. Its success will then be assessed, and a decision made as to whether it should proceed to a larger trial and wider collaboration internationally.