When Marieta was just a year old she developed a high fever and became really unwell.

Her family thought she had fallen under a curse that in parts of rural Burkina Faso is believed to happen when a bird flies over a child asleep outdoors.

Her father, Tibandiba Lankoande, spent most of his money on traditional remedies but this didn't work and on the sixth day she fell into a coma.

"That's when I heard a message on the radio explaining how to recognise the signs and symptoms of malaria," he recalled.

"If I hadn't heard that radio message she wouldn't be alive today."

'Gold-standard' trial
The radio messages were part of a huge trial led by Development Media International (DMI), which sought to find out whether mass media could have an impact on people's health.

Roy Head, DMI's chief executive, said: "If you are the minister of health of a typical African country and you have $30-$40 per head of population to spend on health, you are literally deciding who lives and who dies, and so anything that you spend your money on has to be proven.

"And it has not been proven until now in a gold-standard scientific trial that mass media can change behaviours, let alone save lives."

The organisation produced 150 different one-minute radio messages focusing on the signs of malaria, diarrhoea and pneumonia - three of the biggest killers of children under-five in the country. The messages were broadcast across seven local radio stations, 10 times a day, for three years.

The results, published in the journal BMJ Global Health, show that in the first year of the trial the number of consultations at local health clinics for symptoms of malaria went up by 56% compared with areas where the messages weren't broadcast.

Consultations for pneumonia increased by 39% and diarrhoea by 73%.

And though the researchers weren't able to prove this directly, they used computer modelling to estimate that their project saved 3,000 lives - all of children aged under five.

Local and interactive
Social psychologist, Dr Jennifer Sheehy-Skeffington, of the London School of Economics and Political Science, who looks at the psychology behind behaviour change, said she was surprised at how marked the change in behaviour was.

She told the BBC: "People were increasing by a huge amount the likelihood that they were going to get these symptoms checked out."

She feels part of the success is because rather than just relying on one-way communication, researchers developed regular interactive shows that allowed people to discuss their questions and concerns.

Dr Sheehy-Skeffington added: "They also worked with local people in developing scripts, dramas and stories that are really engaging and get people's emotions going and make it much more likely that people are going to change behaviour based on listening, as opposed to just thinking that it is just information given to them from some expert."

Media-usage habits in Burkina Faso are fairly unusual.

According to DMI's research much of its rural population listen to community radio stations in their own local languages.

The national station, on the other hand, broadcasts mainly in French.

And access to TV in rural areas is uncommon, which means the trial had a fairly captive radio audience.

'Child of the radio'
The big question is whether this type of health messaging can have the same success in other parts of the globe where multiple radio and TV stations and the internet all compete for people's attention.

And it is clear that the media cannot do it alone. Good public healthcare, sanitation and nutrition are crucial to child health.

That said, while millions of children are still dying of diseases that can be prevented or cured, tuning into the right message, at the right time, might just help save a life.

As for Marieta, everyone in her village now calls her "the child of the radio".