How “A Learning Healthcare System” Uses Big Data
Big data is changing the world before our eyes, transforming numerous facets of our society and economy: business, advertising, finance, Internet searching, government, criminal justice, sports, education, media, and scientific research.
One area of profound transformation is that of medicine.
EHR Adoption Led the Way for Big Data in Healthcare
An important step towards this significant transformation occurred with the 2009 American Reinvestment and Recovery Act. It established monetary incentives for doctors to adopt electronic health records. 75 percent of eligible physicians and 96 percent of eligible hospitals embraced these incentives by 2015.
Medicine, as a business, is shifting away from a health care model based on a magnitude of services towards a model based on greater efficiency and health results.
Big Data in Healthcare Enhances Clinical Trials
At the same time, there is another shift occurring within health care: a shift from evidence-based medicine in the form of clinical trials to practice-based medicine that draws its evidence from big data.
Despite the stringent scientific standards required by clinical trials, they are not without methodological weaknesses; for example, what is observed in the relatively small populations studied in trials cannot always be generalized to the broader population. In such situations, practice-based data in the form of actual patient records can be mined for valuable information, on the levels of both the individual patient and broader population. This kind of unstructured data makes up a vast 80% of medical information about patients.
The Combination of Big Data and Evidence-Based Medicine
Darren Schulte, CEO of medical search company Apixio, states big data can help to create a “learning healthcare system” in which “what actually works and what doesn’t is updated with evidence from real-world data.” Ideally and realistically, practice-based medicine using big data will be combined with evidence-based medicine using clinical trials rather than replacing it altogether.
Data Management Tools Aid in Analysis
Before data can be analyzed, it must be extracted from a great variety of sources and represented in a format that computers can understand. Through OCR, an optical character recognition technology, companies like Apixio can create computer-readable representations from clinician notes in handwriting, PDF, and other formats.
Once data is ready for analysis, it will develop individualized patient data models or be combined with population data to reveal patterns and trends on a macro scale. An individualized model refers to as a “patient object,” which Schulte explains as “a profile assembled using data derived by…mining text and coded healthcare data,” which can help establish “the basis for personalized medicine.” These models can be used to predict disease onset and recurrence, allowing doctors to take early preventative steps.
Efficiency in Healthcare Boosts Revolutionary Developments
Big data in healthcare provides our society with new methods for greater business efficiency and new capabilities for understanding illness, preventing disease, promoting health, and saving lives.