A revolutionary warning system established by the Public Health Agency of Canada has earned global recognition. GPHIN2 – an enhanced and automated version of Health Canada’s earlier Global Public Health Intelligence Network – was recently launched at the United Nations.
GPHIN2 is a Web-based system that identifies threats to public health. It monitors and analyses more than 10,000 information sources in six languages – English, French, Arabic, Chinese, Russian and Spanish. On average, each day, information from nearly 20,000 news reports is retrieved based on specific search criteria. The data is then analysed and passed on to public health professionals around the world.
GPHIN2 is based on Global Intelligent Information Management (GIIM), an IT platform for decision makers/managers, created by Montreal–based software developer Nstein Technologies.
“We target 2,000 companies and government agencies world-wide,” said Laurent Proulx, chief technology officer at Nstein. “The whole idea is to move companies from a reactive to a proactive – and in some cases, even to a predictive – mode.”
Proulx said most information in organizations is unstructured and not consulted after three months, and unstructured information grows at a rate of 40 per cent a year. “There’s a big gap between the total information available out there…and what’s actually used in the decision-making process.”
He said the GIIM platform enables companies to collate and harness unstructured information for decision making. GIIM, Proulx said, has three functions:
? Data collection – Information is assembled from public sources such as the Web, news feeds or video. GIIM can also be connected to private information sources, such as company intranets, and can acquire specialized feeds from those who create them.
? Conversion – Data is converted and structured.
? Extraction – Advanced analytic tools are used to extract intelligence and generate multilingual real-time intelligence.
According to Proulx, unstructured information is accessed, analysed and transformed to enable several applications, including GPHIN. He said the GIIM platform can also be used to ensure compliance with regulations, or to create a customer satisfaction dashboard through a quick review of e-mails. This dashboard, he said, could be the basis of new product development or service improvements.
GPHIN2 is the latest incarnation of the GPHIN platform launched by Health Canada seven years ago. “Health Canada wanted a system that could relay information on potential outbreaks directly to international public health bodies,” said Abla Mawudeku, GPHIN manager. In the absence of such a system, she said, it would take months for information about an outbreak to filter through the various jurisdictions and levels of government and finally get to the World Health Organization (WHO). “Sometimes information was just not passed on.” With GPHIN, she said, intelligence is transmitted swiftly to stakeholders around the world. “If there’s an outbreak in Congo, relevant information can be retrieved from local sources and relayed to international public health officials in seconds. These officials, in turn, can then verify the facts with the concerned country or region.”
Mawudeku, who is an epidemiologist, said GPHIN monitors infectious diseases – not just in humans, but in animals and plants – and also tracks chemical incidents, radioactive exposures, dangerous products and natural disasters.
“Essentially we monitor everything from bugs to bombs. But always from the standpoint of how it impacts public health.” GPHIN, Mawudeku said, can track classical outbreak metrics – such as the number of infected people and resulting deaths – as well as the magnitude of the threat. During the 2003 SARS outbreak, she said, information was retrieved not just from affected countries but from regions across Asia. “That helped us determine the event’s geographic distribution.”
The system also tracks “remedial” measures adopted by countries or by public carriers, such as airlines, to protect travelers. More than 40 per cent of WHO information on potential public health risks comes from GPHIN – information which, when tested with member countries, usually proves accurate.
According to Mawudeku, GPHIN retrieved the first suggestive report on SARS in November 2002. It was an article in Chinese on how an unusual number of otherwise healthy people were visiting hospital emergency rooms with acute respiratory illness symptoms. That information was disseminated to public health authorities. A month later, she said, another article in Chinese was identified on how a large number of people in China’s Guangdong province were falling ill.
She said it was only in January 2003 that the first English article was retrieved. “Even that didn’t mention an outbreak, but was about the increase in anti-viral drug sales by a pharmaceutical company. From that we deduced something unusual was going on.” While the earlier GPHIN system worked, it required many analysts to review and make sense of information coming in.
“The system was very cumbersome,” Mawudeku said. Given the volume of reports scanned, she said, it was only possible to translate the title, create a two-line summary of contents and send it to users. “We realized we needed a more efficient system…one capable of managing massive amounts of information, and assisting in its translation and timely dissemination.” Nstein, she said, provided technology that organizes unstructured information and translates it in near real time.
