A throng of media packed into an auditorium in Ann Arbor, Michigan, on April 12, 1955, anxiously awaiting the results of what was, at the time, the largest vaccine trial in history. Thomas Francis Jr., of the University of Michigan’s School of Public Health, held forth over the event, which aimed to present data from a year-long field study of a polio serum developed by his former protégé, Jonas Salk. The vaccine administration and analysis had been an enormous affair involving the cooperation of 20,000 physicians, 40,000 registered nurses, 14,000 school principals, 50,000 teachers and 200,000 volunteer tabulators.
The trial examined the effects of the vaccine on 1.8 million elementary schoolchildren in the United States, Canada and Finland. Francis had insisted on a double-blind method of statistical analysis so that neither patient nor administering physician knew whether an inoculation consisted of the vaccine or a placebo. To track the results, the university enlisted IBM to produce punched cards for each test subject, totaling some 144 million data points.
The goal was to provide granular detail for gauging the efficacy of the vaccine and to establish the most effective administration regimen. The media that had gathered that day — and, indeed, the entire world — wanted to know: How many shots might be needed? How much time should elapse between shots? Will the vaccine’s efficacy vary by condition or region? Answering such questions fell in large part on IBM. It was the company’s first effort to use data analysis as a means of disease prevention — and would become the first of many such efforts over the ensuing years and decades.
The date in April chosen for the announcement marked the 10th anniversary of the death of President Franklin D. Roosevelt, the most famous polio victim in history and founder of the National Foundation of Infantile Paralysis, better known as the March of Dimes. At the time, polio posed the most life-threatening and debilitating public health threat to children. It was highly infectious, incurable, capable of causing lifetime paralysis or even death, and had plagued the world for decades. In 1953, the US counted some 35,000 cases in a population that was less than half the size of the US in 2020. Societal anxiety around the vaccine’s efficacy was understandable — and tension in the auditorium on that spring day was palpable.
In a flourish before disseminating the printed report, more than 100 statisticians fed the final punched cards into the IBM machinery on site. “Safe, effective and potent,” Francis intoned, declaring the Salk vaccine 60% to 90% successful in preventing paralytic polio. One press account described the frenzied scene of reporters at the event: “To avoid a crush, public relations men from the university began throwing the releases into the crowd. But still hands grabbed at the boxes. In the next few seconds, pandemonium prevailed. Then there was a dash for the couple of dozen typewriters in the press room and a battery of telephones.”
Although the disease would not be eradicated instantly, it would become a faint memory to most Americans within a generation. For his part, Salk declared that if it had not been for the work of IBM equipment, researchers might well have taken years longer to establish the serum’s effectiveness as a widespread public health measure.
Although IBM had partnered with various medical institutions for decades previously, primarily through the use of punched cards to gather statistical data, its work on the polio vaccine marked the company’s entry into the field of epidemiology. Today, with polio lingering in some parts of the world, IBM researchers are continuing to harness access to massive computational power and using their expertise in mathematical modeling to support biologists at the University of California, San Francisco; Stanford University; and the University of Haifa, as they strive to engineer a new type of antiviral agent for viruses such as polio. Additionally, the company contributes to multiple projects and partnerships aimed at improving disease prevention and patient care.
One such effort led to the Spatiotemporal Epidemiological Modeler (STEM), an open-source tool that can help forecast the potential spread of infectious diseases such dengue fever, the H5N1 virus and others. Another produced the Public Health Information Affinity Domain (PHIAD), a scalable network that enables clinics and labs to electronically share clinical surveillance data with public health officials to help provide real-time detection of infectious disease outbreaks.
In parallel, IBM used STEM and PHIAD in partnership with Mexico’s Ministry of Health to develop new models on H1N1 when swine flu cases in Mexico City reached pandemic proportions in 2009–2010. In 2016, the company joined the fight against the Zika virus, donating highly localized data from The Weather Company, an IBM business, to the US Fund for UNICEF. And in early 2020, in collaboration with the White House Office of Science and Technology Policy, the US Department of Energy and many others, IBM helped to launch the COVID-19 High Performance Computing Consortium. The effort is slated to produce an unprecedented amount of computing power — 16 systems with more than 330 petaflops, 775,000 CPU cores, 34,000 GPUs and counting — to help researchers everywhere better understand COVID-19, its treatments and potential cures.
With data-driven epidemiology a norm in the 21st century, IBM’s efforts to pioneer and advance the field persist, including partnerships with governments, healthcare agencies and industry to crunch vaccine efficacy data, model the relative merits and effectiveness of lockdown strategies, and increase the overall health of the general population.
The paper on-ramp to the Information Age once held most of the world's data
Smart application of technology and a steadfast commitment to improving the lives of people
A Peace Corps for the corporate world, this program helped distinguish IBM’s leadership in philanthropy