Skip to content

Big Data in Medicine

June 1, 2011

“Big data: the next frontier for innovation, competition and productivity” from The Economist

Properly used, big data could save the American health-care system $300 billion a year and the European public sector €250 billion.

Of all the industries mentioned in this piece, health care has by far the most to gain from collecting, aggregating, and analyzing data. In fact, health care providers already collect a huge amount of data each and every day–it’s called “patient records.” Unfortunately, this data exists in unusable formats isolated within silos in the hands of people who have no idea what to do with it (namely physicians; sorry, but I’ve seen the extent of epidemiological and biostatistical education in medical schools and it’s pathetic). Forcing electronic medical record (EMR) systems upon private practices will not solve this problem because most of the EMRs are incompatible and no mechanism exists to aggregate data on a large scale, which is necessary to leverage such data for quality improvement and new scientific inquiry. If we were able to aggregate large amounts of patient data (and some health systems are doing so), then we would be able to exploit variations in care processes to determine what works and what is useless. Many European countries, due to their single-payer system, already do this and continually contribute valuable research from this population-based data. The Dartmouth Atlas of Health Care is the most notable example of such data aggregation and analysis in the US. However, this project (and many others like it) largely relies on billing data which is inaccurate and limited. We need comprehensive aggregation of patient data (demographics, diagnoses, laboratory and radiological data, and discharge information/follow-up) and advanced analytical methods (such as natural language processing) to develop a more robust understanding of patient care and potential areas for improvement.

One Comment leave one →
  1. redbird permalink
    August 22, 2011 4:18 am

    I have worked in the analyses of big data in healthcare for over 15 years. Unfortunately, because of the noise in the record keeping and the variation of the patients, the diagnoses, the sojourn differences and the general lack of follow up, the data in most places is useless for giving scientifically useful or well powered conclusions. It almost never meets the demand for “evidence-based” medicine.

    What it does do is allow better decision making. I am a big believer in Bayes and prior and post probabilities. Unfortunately many doctors became doctors because calculus 1 scared the crap out of them and they went the biology, observational path instead of engineering/science and the mathematical path.

    I do not believe the Economist number. It is not data that is costing too much money. It is the focus on rewarding procedures and identifiable and limited behaviors that incentives the wrong actions and the desire of the legal system to give black and white edges to the gray and murky world of clinical observation.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: