What works in citizen science?

Grippenet reveals strengths and challenges

Citizen science provides a great opportunity to involve the public in research and to improve data collection or analysis through the power of the crowd. But with more and more projects appearing to demand citizens’ attention and time, is this really a worthwhile avenue for researchers to pursue? Grippenet, the French flu surveillance platform, tracks the illness through users’ self-reported symptoms. As epidemiologist Vittoria Colizza explains, involving citizens in this way yields unique and important information to complement data from traditional epidemic tracking systems.

 

 

Citizen science provides a great opportunity to involve the public in research and to improve data collection or analysis through the power of the crowd. But with more and more projects appearing to demand citizens’ attention and time, is this really a worthwhile avenue for researchers to pursue? Grippenet, the French flu surveillance platform, tracks the illness through users’ self-reported symptoms. As epidemiologist Vittoria Colizza explains, involving citizens in this way yields unique and important information to complement data from traditional epidemic tracking systems.

This article is the first of two on the connections between citizen science, social networks and epidemiology.
2 - Posting the flu - Does the web hold vital information for pandemic surveillance?

 

“This isn’t sexy—it’s the flu,” concedes Vittoria Colizza regarding Grippenet, France’s online, citizen-powered flu-tracking system. Nevertheless, the epidemiologist with the French National Institute of Health and Medical Research (Inserm) is pleased with its success: in its first year, and with only moderate publicity from the three partner institutions backing it (Inserm, the University Pierre and Marie Curie, and the Institute for Public Health Surveillance, the program has enrolled over 5,000 participants and achieved the admirable retention rate between seasons of around 70%.

In the growing fold of participatory research projects (for example, Nutrinet, a program investigating the nutrition and health of 500,000 internet users), it is difficult for any one to stand out from the crowd and attract the attention of the other crowd that will provide its data. “It’s hard to be engaging,” explains Dr. Colizza. “This takes literally seconds per week, but it demands a psychological effort” to faithfully complete a questionnaire on such a regular basis. After filling in their profile with important demographic and lifestyle information, users are prompted each week to answer a few questions about their flu symptoms, if any. The Grippenet team has taken care to minimize the trouble for respondents, to provide maps and graphs that are accessible and interesting, and even to offer some flu-related fun in the weekly newsletter: the etymology of the word flu, or how to sneeze in different languages.

Grippenet.fr logo

 

While an infectious disease surveillance system already exists in France, relying on a network of doctors to report flu and other illness among their patients, the effort of putting the Grippenet system in place is worth it to the team. (The same group runs both systems.) “There’s information that we can’t get from doctors,” Vittoria explains. “Does the person take public transport? Do they have pets at home? How many days of work did they miss due to the flu?” This information can help establish the risk factors, a big issue in flu research and epidemiology, as well as the cost to society. But the information itself comes at the cost of participants’ time, which must be donated repeatedly throughout the flu season. For this reason, cultural factors may play a role in the relative success of different countries when recruiting citizens to participate in research.

The sociology of citizen science

Grippenet is part of a European-level effort to gather flu data from the public, and some countries have had more success than others. The Netherlands and Belgium were the first to launch such a program, even before Europe came onboard. They now stand at 17,000 participants – nearly half the European total. Italy’s rate of participation, relative to its population, is significantly lower. “There are some populations that are generally more prone to respond to public needs, to make their small contribution to the public good. In other countries, this is not well understood,” Vittoria Colizza explains.

“It’s also linked to how much science is seen on TV, how much it’s diffused and taught, and to the general understanding of what science is and what researchers do,” she adds. In her native Italy, for example, she sees a constant debate about research and funding that puts scientists in a bad light. A call for citizen participation, then, may not meet with great success. “There are cultural situations where citizen science may just work better,” Colizza believes.

Citizen science & flu tracking
 

Rewarding citizen research?

The Grippenet team tested out a way to make their own citizen science initiative work better—with the help of children. Exploring different conduits of communication, with different age groups, in different settings, members of Dr. Colizza’s group set up a school program on the island of Corsica. Students of different levels completed the questionnaires with the help of their teachers and were challenged to recruit as many people as possible to Grippenet…and to beat the class next door. Participants were rewarded with small prizes: caps, pencils, USB keys.

“These were not meant really as prizes resulting from the competition, but as a little ‘plus’ for their involvement,” Vittoria explains. It’s an important point, because there is much debate surrounding the use of real incentives in motivating citizens to take part in studies. “You want useful, true information, and you don’t know how this is altered by incentives.” The American Public Health Association recently offered cash awards of up to $25,000 to its members collecting the highest number of weekly questionnaires from citizens recruited to a flu tracking study. Perhaps the fact that recruiters are rewarded, rather than participants themselves, will protect the data from any artificial effect, but, according to Dr. Colizza, the field overall is hesitant to offer incentives. Future studies might compare experimental groups, with and without rewards, to determine if this is an appropriate method to draw more citizens to participatory research.

In the meantime, Grippenet, with its 5500 active users, promises to tell even more than classic flu surveillance systems, which usually aim to describe the current state of an epidemic. The first step will be to confirm the validity of their analyses, and then to give the information back, in real time, to those who made it possible.