In the summer of 2013, Jose Gomez-Marquez and his colleagues at MIT and Harvard started a crowd-sourced experiment to test water quality throughout the world with cheap sensors and mobile phones. You'd snap a photo of your water sample with your feature phone through a special app the researchers had developed. The algorithms baked into it would analyze the water in real-time and upload those results to the web, paired with location data. People all over the world — in New Zealand, Panama, Nicaragua and Germany — participated.
"That gave us enough of a dry run to understand how people could be citizen sensors," says Gomez-Marquez, a medical device designer. Now Gomez-Marquez and his team are gearing up for a new experiment; this time, they'll be crowd-sourcing the presence of disease instead of dirty water. They're deploying a simple diagnostic test that detects Ebola, Dengue and Yellow fever to clinics in West Africa and Colombia. The idea is to get these '3-in-1 rapid tests,' which cost about $2, into the hands of people in villages and rural areas, so they can test themselves and then snap a photo of the results with phones.
They work like over-the-counter pregnancy tests, changing color after 15 minutes to reveal infection status, except patients don't pee on them; instead, the test strip is dipped in blood. If the patient is positive for, say, Ebola, the test strip would turn pink. If he doesn't (or the amount of virus in the blood is too low), then there's no color change. After people test themselves, at home or in a clinic, they or a health worker will snap a photo of the results and upload them to the cloud through a mobile app, all in minutes. The results will be logged and mapped using the geolocation information from the photo.
It would be like Google Flu search trends, but with real people and real results, mapping the spread of viral diseases. If these initial trials work, the team hopes to roll it out throughout South America and Africa.
"The quality of the information would increase dramatically, both for the patient and for people looking at surveillance systems," Gomez-Marquez said.
It may sound far-fetched, but the academic team is not the only one trying to use fast tests and an app to better map the spread of disease. It's the hot new pathological trend among biotech companies. LuSys Labs, a San Diego-based diagnostics company developing rapid tests for Ebola and other infectious diseases, has partnered with San Diego-based analytics startup BridgeCrest Medical to develop real-time location mapping. Colorado-based Corgenix, whose Ebola test recently got greenlighted by the FDA and the World Health Organization for emergency use, announced earlier this month that it had partnered with Fio, a data analytics company, to track patients. The quick diagnoses paired with mapping is potentially a game changer. It has the potential to do for healthcare what smartphone location tracking did for traffic reports, so health workers will be able to better see "disease traffic." And for the biotech companies, it means they'll know where their tests will sell.
As it stands now, those responding to a disease outbreak don’t have real-time data on how and where a bug is spreading. They rely on interviewing patients that test positive to tell them where they've been and who they've had contact with to build epidemiological maps, a time- and resource- intensive process that hinges on people's often faulty memories and biases. The lack of a real-time disease map helped fuel the Ebola outbreak last summer, because responders were behind the outbreaks instead of in front of them. It’s why other infectious diseases like Dengue, Yellow Fever, and Lassa often get out of hand.
If rapid-test results were uploaded to the cloud through mobile devices, analysts could generate maps of disease hot zones and even make predictions of where the disease might spread next. Businesses hope to use these data to deploy their tests to high-risk areas, and health workers on the ground would know where their help was needed most.
"As we see more outbreaks and information reported, we’ll know how to concentrate our efforts. It would benefit us in terms of manufacturing and and figuring out logistics,” said Dennis Poon, the director of business development at LuSys Labs. “In less developed countries, the combination [of rapid tests with location data] provides tremendous utility to underfunded and undermanned hospitals and clinics on the frontlines.” Governments, he said, could also use these data to monitor bioterrorism threats and public health concerns.
Insta-mapping isn't feasible with current tests. The gold standard for detecting viruses like Ebola is the polymerase chain reaction, or PCR, a test that detects viral DNA in blood; it requires lab space, expensive specialized equipment, trained technicians, and access to electricity, which might not always be readily available in disease hot zones. It can take days to get results back because samples have to be sent off to a lab. “They are not super scalable and not able to get into villages,” said Pardis Sabeti, a computational geneticist at Harvard University, who recently published a study in the journal Science about the Ebola genome. Rapid diagnostic tests, while less sensitive than PCR, are easily taken into the field. They have antibodies in their strips that bind proteins the virus being tested makes, and those antibodies are attached to nanoparticles that change color if an antibody binds a viral protein. It all happens in 15 minutes, but it can miss an infection if the virus count in the blood is too low.
It won't be perfect, of course. At first, most of these tests will be administered at health clinics, rather than in people's homes, which won’t give scientists the level of mapping granularity they’d need to predict where an outbreak might spring up next. “The challenge — and what is missing — is the ability to track the movement of people,” Delmelle added. “If we don’t know where that individual has been, we won’t be able to track down where the disease was transmitted, or to create a map that shows the areas most likely to show signs of transmission.”
If this future materializes, it’s important to keep in mind that it relies on sick people sending their diagnosis into the cloud in real time. (In a perfectly dystopian sci-fi scenario, companies would require the masses to upload their tests before they could get the results.) Then one (or several) of these companies will have access to the world’s health data. There are upsides: faster responses to outbreaks and more efficient healthcare systems, but with the benefits come massive security and privacy concerns too, especially in the developing world. In theory, the data could be used to hawk not just rapid tests but other types of medical equipment and supplies, meaning that mapping companies could become data brokers — a link between manufacturers and the markets they want to sell to.
In countries like the U.S., patients are protected by laws that prevent governments and employers from discrimination based on health information — though, even with that in place, a nurse who worked with Ebola patients, but was not at risk for the disease, found herself trapped in quarantine for days in New Jersey. In the U.S., doctors would have to ask permission before uploading our information into a database that a company would use to map diseases. Those same protections don’t always exist internationally.
“Typically we leave it up to the direction of the doctor to inform the patient. In many countries, in West Africa, for example, they probably don't ask,” LuSys Lab's Poon told me. “In the U.S. there would have to be full disclosure.”
That could create situations in which companies take advantage of already disenfranchised populations to boost their bottom line. But as Poon points out, privacy concerns are not necessarily top of mind for the populations that would benefit most from the tests LuSys Labs and others are developing. And for better or worse, that gives companies the upper hand.
Daniela Hernandez is a senior writer at Fusion. She likes science, robots, pugs, and coffee.