After the 1995 ebola outbreak in Zaire, the WHO established Heymann’s EMC, which oversees the organization’s global surveillance networks. The network collects information from some 1,000 laboratory centers around the world, including more than 200 specializing in infectious diseases. “For arboviruses and hemorrhagic fevers,” explains Heymann, “we have 37 labs throughout the world constantly receiving specimens to look for ebola, dengue, and yellow fever. The results are given to WHO and to the country where the specimen originates. We then look to validate the results, and if it’s an outbreak situation, we look to get the necessary groups involved, to make sure the outbreak is addressed and contained.”
The WHO says it receives as many as five unconfirmed rumors a week of new infectious-disease outbreaks, by telephone, newspaper or e-mail. Each rumor is then investigated and a rumor/outbreak list is sent out electronically on a need-to-know basis to relevant personnel at the WHO, its collaborating centers and other public health authorities. These reports, however, are not intended for public consumption. The WHO will only post news of an outbreak on the EMC’s public Web page after confirmation. Because confirmation often requires sending specimens to a laboratory that may be outside the country of origin, the WHO system is notoriously slow at alerting the world at large to outbreaks.
To strengthen the system and accelerate its response time, the WHO has recently joined with Health Canada to create the Global Public Health Information Network (GPHIN), which Heymann says will “really be the system” because it will subsume ProMED within it. GPHIN will include all the reports from ProMED, but also incorporate information gathered from other sources.
GPHIN is the brainchild of Rudi Nowak, a Canadian physician and public health specialist and former director of Canadian quarantine operations. Nowak says that he realized the contemporary world had drastically changed the nature of his business and that the problem had to be addressed. “We’re not talking quarantine the way it used to be,” he says. “People boarded a ship in Europe, for instance, and it took them 11 or 12 days to come to North America, and some got sick and died, and those that were sick were put into quarantine when they arrived. Now people can get anywhere within hours, and they can get exposed to serious pathogens without even knowing it and get back home within the incubation period of the disease.” The way Nowak likes to put it is that quarantine has moved “from the seaport to the emergency room.”
GPHIN uses a search engine to scan the Web continuously for all information pertaining to infectious diseases, including specific sites such as ProMED. The search engine stores the findings under six headings: cholera, salmonella, hemorrhagic fevers, antibiotic resistance, encephalitis, and floods. The latter, says Nowak, “because if you have floods, cholera is just around the corner.”
GPHIN then further breaks the incoming information into three bins, depending on how urgently they have to be addressed: a “hot” bin, for the first report of outbreaks; a “standby” bin, for collateral information on existing situations; and a bin for rejected information, which can include worthless rumors and irrelevant information. GPHIN then extracts the relevant information from each report and places it in an “intelligence” report that can be scanned quickly. This step is currently carried out by humans, says Nowak, but only until commercial artificial intelligence technology is available.
When users enter the Web site, they’ll start with the latest intelligence reports, and if they want more detailed information they can go to the original reports from which the intelligence was gathered. “Initially they can simply come in and say, Let’s look at the cholera bin and see what’s happened around the world in the last 24 hours.’ They can then scan through reports as they come in.” GPHIN is also fully interactive and so, as with ProMED, anyone reading the page who has something important to add can do so.
Nowak and his collaborators officially launched GPHIN in early June, and the system began searching the Web in both French and English. Eventually, it will expand to all seven WHO official languages. What the system won’t do-unlike ProMED-is report to the public. The current system is designed for public health officials and no one else. “We cannot really be accused of being in the business of spreading rumors,” says Nowak. “The information sooner or later will be available to the general public, but it has to be verified. We don’t want to create unnecessary anxiety among the public.”
With a little luck, ProMED and GPHIN will both take care of what CDC arbovirus researcher Paul Reiter suggests has been the regrettable history of infectious disease fighting to date. To put it simply, he says, all the monitoring and surveillance has only served to mobilize the world’s resources to fight “ex-epidemics. Whenever we’ve been sent out to epidemics, we’ve always arrived when the epidemic was pretty much history.” He cites, for example, a 1993 yellow fever epidemic in Kenya, at which the medical cavalry arrived in time to see the last two cases. Or, going way back, a yellow fever epidemic in the Omo River Valley in Ethiopia between 1960 and 1962. “You had something like a million people susceptible to yellow fever,” says Reiter, “with 30,000 deaths”; no one outside Ethiopia, he says, “had the faintest idea until it was all over.” Now with the Internet the news should get out in time to make a difference. “We’re primed,” he says.