Dealing with Healthcare’s Data Explosion

By Toby Cosgrove, MD

American healthcare is in danger of being overwhelmed by data. Medical knowledge will soon double every 73 days. An estimated 800,000 papers are published yearly in 5,600 medical journals. At the same time, doctors and hospitals are now expected to collect and report a growing mass of data to the government – well over 100 measures at last count.

Absorbing new knowledge and collecting current information requires more time, personnel and money than most providers can afford. Right now, the data explosion threatens to destabilize American medicine.

Constantly changing regulations

The Affordable Care Act (ACA) was the biggest thing to hit healthcare in the past 100 years. It has generated 16,000 new regulations – a stack of paper seven feet high. And they keep changing.

Managing these data and regulations are an enormous new burden. No single doctor, practice or hospital can afford to handle all this data on their own, so we’re seeing doctors joining practices, practices joining systems, and systems joining into regional networks. The cost and benefits of data management are being distributed across large numbers of participants. Of course, better data management is only one of the many good reasons for consolidation. Patients, providers and society at large will ultimately benefit from this trend, in the same way we’ve all gained from larger networks of trade and transportation.

Watson accurately parses patient data

The massive influx of new medical information is why Cleveland Clinic has embraced IBM’s Watson technology for our medical school. Watson has learned how to perform a history and physical exam and perform lab studies better than we can do as doctors.

Watson has been “educated” by doctors and students from Cleveland Clinic’s Lerner College of Medicine, and elsewhere, to recognize natural-language medical terminology in its many variations using machine learning and neurolinguistic programming. Last year, we collaborated on a study with IBM Research and its Watson Group, to learn whether or not the super-computer could extract an accurate problem list from patients’ electronic medical records (EMRs).

Watson passed the test with flying colors – showing that it could successfully generate an open-ended medical problem list from the contents of patient EMRs with a high degree of accuracy. The techniques they used are believed to be broadly applicable to EMRs from any hospital or EMR vendor system.

In the future, computers like Watson will take on more of the management and interpretation of patient data – freeing up physician bandwidth and allowing us to focus more attention on the individual patient and his or her unique problems.

Dr. Cosgrove is CEO and President of Cleveland Clinic.