Researchers have studied the medical histories of the entire population of Denmark to chart how medical conditions are linked and forecast disease before it begins.
In a major advance for the field of biomedical Big Data analytics, scientists followed the medical history of some 6.2 million Danes over the course of almost 15 years. Since the dataset includes those who died in those years, that’s a sample size 600,000 people larger than the current living population of the small Scandinavian country. Using the Danish National Patient Registry, which healthcare providers are required to report to, the data scientists were given access to 65 million inpatient, outpatient and emergency room events from 1996 to 2010.
Over that long study period and with so many data points that included every demographic in the country, they were able to start seeing hidden patterns in how disease progresses from its earliest stages. They found more than 1,100 “sequential diagnostic correlations” that occurred the most frequently in the Danish population, from an early seemingly unrelated medical issue through later diagnosis of maladies like diabetes, chronic obstructive pulmonary disease, cancer, arthritis and cardiovascular disease.
See below for an example of a disease network.