IBM’s many innovations within the field of epidemiology have led to more successful disease prevention and better overall patient care. The company’s history of working to further understand infectious diseases stretches back to the 1950s, when IBM punched cards played a supporting role in testing American medical researcher Jonas Salk’s vaccine for polio.
At the time, polio—a highly infectious, incurable disease that invades the nervous system—was one of the most feared diseases in industrialized countries, paralyzing thousands of children each year. In the largest medical experiment in history, the Salk vaccine was tested on 1.8 million children throughout the US—generating 144 million data points, all captured on IBM punched cards. Salk’s vaccine proved to be a winner, and within two years of its development, annual polio cases in the US had shrunk from 35,000 in 1953, to 5600 in 1957.
Today, through its advances in predictive modeling, public health information exchange and supercomputing, IBM continues to help governments, researchers and public health organizations around the world prepare for—and respond to—disease outbreaks.
Open source disease modeling with STEM
Developed collaboratively by IBM’s Almaden, Haifa and Watson labs, the Spatiotemporal Epidemiological Modeler (STEM) is an open source tool that has the capabilities to forecast and analyze the possible spread of infectious disease, such as bird flu, dengue fever, the H5N1 virus and others. It allows users to create spatial and temporal models of emerging infectious diseases, helping researchers and public health officials simulate the spread of disease across time and space, and allowing them to assess the likely impact of preventive measures. STEM can estimate, for example, how soon after the first case of a new influenza appears in San Francisco, California, it will peak in Mexico City, as well as data points such as the total number of potential disease outbreak cases.
STEM users can access the tool’s main components—the core representational framework, graphical user interface, simulation engine, disease model computations and various data sets—as separate plug-ins to build on existing models or create new ones. The data sets describe the geography, transportation systems and population for the world's 244 countries and dependent areas down to administrative level two—equivalent to the county level in the US— for most countries. Its compatibility with the standards-based interoperable healthcare infrastructure developed by IBM, Health Information Exchange (HIE), allows it to query clinics, hospitals, labs and other public health information sources for real-time data about the health of a population.
Shared clinical surveillance data with PHIAD
A scalable network that uses Health Information Exchange (HIE) technology and standards for the exchange of public health information, the Public Health Information Affinity Domain (PHIAD) enables clinics and labs to electronically share clinical surveillance data with public health officials for up-to-the-minute infectious disease outbreak detection. Developed by IBM researchers collaborating at the Almaden Research Center and Haifa Research Lab, PHIAD provides a web-based end-user application, and is coded to work across proprietary systems and geopolitical boundaries for truly borderless application.
Fueling collaborative disease detection in the Middle East
Development of PHIAD began in 2008 as a collaborative project with the Nuclear Threat Initiative and the Middle East Consortium on Infectious Disease Surveillance (MECIDS), a program that facilitates collaboration between Israel, Jordan and Palestine to prevent and respond to infectious disease outbreaks. Using its standards-based public health information exchange network as the foundation, IBM worked with MECIDS to customize a solution that would allow consortium members to share information, statistics and anecdotes electronically and to analyze data collaboratively and in real time. The project, which depended on a high level of partner collaboration, overcame significant political tensions and illustrates how regional disease surveillance—and equal disease prevention tools for all—are possible even in areas where there is entrenched political discord.
Curbing H1N1 in Mexico
IBM and Mexico’s Ministry of Health worked jointly to develop new models of H1N1’s spread using STEM and PHIAD when swine flu cases in Mexico City reached pandemic proportions in 2009 and 2010. Mexican health officials and IBM researchers were able to determine the “reproductive rate” of the flu—the number of secondary cases each single infected case will cause in a population with no immunity—and the likely effect of health policy decisions. For example, STEM showed that the city’s social distancing measures, which included shuttering restaurants and schools at the height of the outbreak, cut the flu’s reproductive rate by 22 percent. The project was a milestone in the Mexican government’s efforts to improve public health through the sophisticated modeling that STEM offers.