A publication of the Centre for Advancing Journalism, University of Melbourne

Science

Made in Melbourne: phone app for detecting pneumonia set to save thousands of lives

Health workers in Africa have started testing a low-cost mobile phone app, developed by Melbourne researchers, which could help save the lives of thousands of children by more easily detecting whether or not they have pneumonia.

Words and images by Julie Milland
 

The app, designed for non-smart or “feature” phones, helps workers better diagnose the killer illness simply by counting the number of breaths being taken by sick children.

Pneumonia, which has been the leading killer of young children for more than 30 years, causes fever and a faster breathing rate.

The app allows health workers in the field, particularly in remote areas, to accurately count breaths so they can distinguish between pneumonia and other causes of fever that less often increase breathing rates, such as malaria and colds.

The ingenious but easy-to-use app asks for the child’s age and then tells the health worker to press any numerical button on the phone every time the child breathes. The app then gives an estimate of the breathing rate after 10 breaths and again after 20 breaths, before giving a final one-minute count.

Once a faster breathing rate is detected, the workers can diagnose pneumonia and give suitable antibiotic treatment rather than misdiagnose a cold and send the child home where they may die.

https://s3.amazonaws.com/the-citizen-web-assets-us/uploads/2018/02/13231507/Julie.phoneAPP1-2.jpg

Developed by researchers led by Associate Professor Jim Black at the Nossal Institute for Global Health, the app is being tested in Uganda, South Sudan and Ethiopia.

It is available in multiple languages, including Luganda for Uganda, Dinka for South Sudan and Amharic for Ethiopia. The trial is being run by the Malaria Consortium, with money from the Bill and Melinda Gates Foundation.

Dr Black, who spent more than 10 years working as a doctor and researcher in Mozambique, noted that pneumonia kills between 1 million and 1.5 million children every year, or one child every 20 to 30 seconds. Nearly all of these deaths could be prevented and patients cured, as is the case in developed countries such as Australia.

Deaths from both dehydration (caused by diarrhoea), and pneumonia have been in the “too-hard basket” for decades, according to Dr Black, and researchers were looking at low-cost mobile phone technology as a means of assisting decision-making in health clinics.

“It gets really exciting for a technical person like me to work out how we capture the data in a way that fits perfectly in the work flow and is something that [health workers] still see as a handy and useful tool that gives them the answer on a piece of technology they already recognise, like their phone,” Dr Black said.

Deaths from pneumonia have declined little because solutions have seemed too complex, too expensive and too difficult to deliver. Mozambique, for example, only spends around $50 per person per year on health compared with 100 times that amount ($5000) in Australia.

But low-cost technologies were motivating researchers such as Dr Black to sift through once-stagnating problems because it offers the greatest potential for saving lives.

“So with the best intention in the world, they’re overprescribing and they’re giving four drugs having asked two questions. And the critical questions, which are how severe is the illness and is this a child who needs to be admitted to hospital, never get asked.” — Dr Jim Black, Nossal Institute for Global Health

Counting a person’s breathing rate sounds like it should be a straightforward task. But Dr Black said it was not so simple, with even the clinical definition of when a breath begins and ends open to debate.

During the app’s early development, he asked experienced clinicians in Melbourne to count breathing in children and found the results varied. However, accuracy was critical when designing the app for use in developing countries.

Many health and fitness apps sold in Australia are described as toys and show little evidence of effectiveness. Dr Black said developing a useful app that could actually save lives required sticking to a set of rigorous design principles and thoroughly testing the apps before they were used by health workers in the field.

His own clinical experience had given him intimate understanding of how health systems and health workers operate in developing countries.

A critical first decision was to develop the app for a feature phone rather than a smart phone. A feature phone is a cheaper class of phone owned by many people in developing countries. It has a camera and can still accept applications but usually does not have a touch screen.

https://s3.amazonaws.com/the-citizen-web-assets-us/uploads/2018/02/13231535/Julie.phoneAPP2-1.jpg

“The main thing [about feature phones] is that the battery lasts a lot longer and you can drop them,” said Dr Black. “And the battery life is the key thing because of the killer apps that everybody feels they have to have nowadays – SMS, calls, and in Africa there’s a cashless transfer system.”

But most health workers around the world have their own phone or access to a phone, and that means the latest app technology can be made available in the field on people’s existing phones.

https://s3.amazonaws.com/the-citizen-web-assets-us/uploads/2018/02/13231559/Julie.phoneAPP3-1.jpg

The apps must also meet the reality of small health clinics. Dr Black said half of the world’s health workers are deployed in small facilities at a sub-district level, where health workers are trained but are not doctors and are rarely supervised by doctors. In Africa, small clinics typically cover 30,000 people and are usually over-stretched.

“Imagine that a child has come five kilometres on foot or was carried on the mother’s back,” Dr Black said. Typically, the mother and child are forced to wait in a full clinic and are one of 100 or so consultations conducted by each health worker every day.

“The same health workers tend to also be the ones looking after [hospital patients], so it’s usually one massive clinic that starts 7 or 7.30 in the morning and they just keep going until they finish,” Dr Black added.

This reality is important for the design of the app because a parent and child will only have about two minutes with the health worker, with the mother on average being asked just two questions. While some consider “six-minute medicine” shabby in Australia, Dr Black said in developing countries it is every second that counts.

Mothers often report that their child has a cough or problems breathing, but the health worker can’t afford to spend a whole minute counting breaths because counting the breathing rate of 100 patients would take up to an extra two hours of clinic time.

“Anything that takes time in the consultation, where the health worker can’t do anything else but count the respiratory rate, is always going to have that hangover . . . they’re going to stop [measuring breaths] because they can see how many are waiting and everyone’s getting tired and hungry.”

START TO FINISH: The thinking behind the App 

Design principle

Reason

Ready to use on existing feature phones

Smart phones not widely owned

Extensive testing before deployment

Will not be used if it doesn’t work well

Tiny set-up costs

Otherwise governments can’t afford it

No ongoing costs when in use

Even 1c per text messages is too much

Minimal training requirements

Governments can’t afford it

No need for expert input from outside the point of use

Experts are not available long term

 

To overcome this time constraint, Dr Black’s group had to find the minimum time needed to count breaths while maintaining accuracy. They enlisted physiotherapy and nursing students in Melbourne to count breaths using the app and discovered that counting each breath for 20 breaths was faster and more accurate than counting for 60 seconds. These results have just been published in the International Journal of Nursing Studies.

Dr Black said that breathing rates could even be measured in the waiting room.

“Ideally, we should have an automated method of counting the respiratory rate that can be done in the waiting room by the health worker’s untrained assistant. They don’t have any medical training but you can teach them how to do it like measuring a temperature.”

Dr Black said health workers would be delighted if patients walked into the consultation with their temperatures and breathing rate already recorded.

Without tools to help diagnose whether the fever is caused by pneumonia, Dr Black said there are usually two possible outcomes.

The first is that the health worker decides to cover all possibilities by giving the child one drug for the fever, one drug for malaria, one antibiotic and vitamins for malnutrition. This costs precious money and drugs, and exposes the children to unnecessary treatment.

“If you’ve got $50 per head you can’t afford to give a child who doesn’t need an anti-malarial, an antimalarial – because you run out of drugs. And they do. Typically they run out of things half way through the month because they just haven’t got enough. These are like 10c a treatment but if you’ve got a lot of people coming through and you’re giving people unnecessary treatments then it’s really inefficient from that point of view,” he said.

“So with the best intention in the world, they’re overprescribing and they’re giving four drugs having asked two questions. And the critical questions, which are how severe is the illness and is this a child who needs to be admitted to hospital, never get asked.”

The second outcome is that the health worker incorrectly diagnoses a common cold.

“The child goes home again and dies [of pneumonia] that day or the next day. The parents think even the best health worker available to us wasn’t able to save the child. And the health worker never gets the feedback, so they never know they’ve made the wrong call. So things tend to just perpetuate themselves.”

► An edited version of this story also appeared in The Sunday Age.

About The Citizen

THE CITIZEN is a publication of the Centre for Advancing Journalism. It has several aims. Foremost, it is a teaching tool that showcases the work of the students in the University of Melbourne’s Master of Journalism and Master of International Journalism programs, giving them real-world experience in working for publication and to deadline. Find out more →

Winner — BEST PUBLICATION 2016 Ossie Awards