Is there anything big data can’t do? Its advocates say it can fight fake news, win elections, and solve crimes. It could also change the face of medical research.
McKinsey & Company estimates that big data analytics could reduce US healthcare costs by $300 to $450 billion each year. However, a recent report by the McKinsey Global Institute suggests the US healthcare industry has captured just 10-20% of that potential untapped value. It might be a game-changer, but the journey has only just begun…
Today, using big data mined from electronic medical records (EMRs), medical databases, wearables, clinical trials and insurance claims, researchers can not only find targeted malignancies but also make trials faster, more powerful and less expensive.
Streamlining research to fight disease faster
In biomedical research, time is of the essence. Before big data, the quickest a researcher could hope to receive results was within a few days. With the help of big data, researchers can now access huge swaths of medical records and test results to make smarter, swifter decisions.
Using big data, scientists at London’s Institute of Cancer Research have reduced the time to conduct complex analysis of breast cancer cells from decades to months. The study, published in Genome Research, discovered that changes in the shape of breast cancer cells result in changes to genetic activity. The researchers used large data sets to map these links and reveal how cell changes are connected to the clinical outcomes of patients.
This means that physicians can predict how aggressive a patient’s cancer is based on the appearance of her cells, and choose treatment accordingly.
Boosting the value of clinical trials
Since the first official clinical trial in 1747, patients have been assigned to groups randomly. Today, using big data mined from electronic medical records (EMRs), medical databases, wearables, clinical trials and insurance claims, researchers can not only find targeted malignancies but also make trials faster, more powerful and less expensive.
For example, in Spain the Harmony project – part of the IMI Big Data for Better Outcomes (BD4BO) program – is using big data to create a European map of hematologic malignancies. Such tumors rank fifth in terms of frequency and third in terms of death rates.
Big data can also be used to determine risk. The Occupational Safety and Health Administration estimates that 500,000 workers are exposed to potentially toxic laser and electrosurgery smoke each year. By transforming extensive unstructured data sets into interoperable formats, researchers can predict which procedures are most likely to produce smoke, the impact of exposure, and the steps needed to protect against it.
And while the benefits of big data are often only associated with advancements in treating humans, technology extends these benefits to animals used in clinical research, too. Evaluating big data allows researchers to predict possible medical outcomes for an animal and adjust tests accordingly. By sharing huge datasets, biomedical processes are faster, causing less stress for the animals. Thanks to large open source data analysis, we could also see virtual animal models replacing their living counterparts.
Big data applications in the Internet of Things have also resulted in other alternatives, like wearable devices that monitor the effect of a drug when an animal is in its natural environment.
Big data might lead to breakthroughs in research, but the medical industry is playing catch-up. If the industry is ready to capture big data’s full potential, it faces some major barriers: a disjointed infrastructure; data that is often siloed or inaccessible; inadequate storage; and concerns about ‘de-identification’ of data and patient confidentiality.
But once research teams and bioinformaticians can spend less time structuring and organizing data and more time focusing on the results and insights, big data has the power to reshape biomedical research. And with 80-90% of value up for grabs, according to McKinsey, who knows where it could lead?