Sierra Leoneans Are Collecting Ebola Data with Their Phones

While texting might be a much bigger threat to Americans' health than Ebola, in Sierra Leone, text messages could actually help tackle the spread of the disease.

A new initiative launched by IBM in West Africa promotes a "citizen engagement system" that makes use of mobile technology—which is booming in the continent—to help both collect data and offer citizens a way to communicate their concerns.

The project is a collaboration with Sierra Leone's Open Government Initiative (OGI), an organisation that promotes openness and interaction between the government and local citizens and that has already been collecting information on the Ebola situation via questionnaires and spot checks at health facilities. The OGI website gives regular updates on the outbreak, providing numbers on new and discharged cases in the developing outbreak.

It's a refreshingly transparent attempt to help explore the issues around the spread of Ebola amid a global response that has been partly clouded by misconceptions and paranoia.

The Centers for Disease Control and Prevention reports that Sierra Leone has so far seen over 3,500 reported cases of Ebola and over 1,000 deaths in an outbreak that has been ongoing since May.

In a blog post, the OGI explains that efforts to fight Ebola are now its priority. It singles out "rumors being circulated by social media" as causing anxiety both in the country and on an international level (just check your Twitter feed whenever a new case is reported outside of Africa to get a glimpse of this), and a reason why better communication between the government and citizens is needed.

The new system simply allows people to reach out via text or voice call to report their concerns over a toll-free number provided by telecom company Airtel. It's a resource for those affected to give their feedback on the crisis and how it's being handled, which in turn gives authorities insight into the situation on the ground.

IBM explained that analytics conducted in the cloud using supercomputing can pick out connections in the messages to create "opinion-based heat-maps which correlate public sentiment to location information."

For example, the system has helped pinpoint regions that urgently require basic supplies such as soap and electricity, or faster burial times for bodies. The heat map above shows various key concerns across Sierra Leone's capital city of Freetown—"electricity" is colour-coded as a need for supplies, "ambulance" represents a need for services, and "dead" is a marker of danger.

Uyi Stewart, chief scientist at IBM's Africa research lab, told Computer Weekly that the system was already helping to bust some myths around the outbreak and why certain preventative measures might not be up to scratch.

He said that while some media reports have suggested that the country's dead aren't buried immediately because of cultural practices—which increases risk of the disease spreading—they actually found another reason for the delay. Owing to the stigma around Ebola, families were waiting for tests to be done to confirm if their loved ones had in fact died of the disease. A backlog on testing resulted in a backlog on burials, revealing a hazardous trend that could then be acted upon.

The project also incorporates radio broadcasts through a collaboration with the University of Cambridge's Africa's Voices project, which allows radio shows to ask listeners questions to be answered by SMS. In this situation, radio can be used to give public service announcements relating to the outbreak and encourage listeners to give their feedback.

Of course, communication alone won't contain the outbreak; it's not yet clear how well the new system will take off and, as a result, how comprehensive the data gathered from it will be. Then it's also down to how effectively authorities and volunteers will be able to respond.

But in a crisis that has become somewhat characterised by misinformation, data-led initiatives will at least help to give a clearer picture of the real situation.