Big data is a huge topic in healthcare right now, so it's no surprise that so many sessions at AACC 2014 have been dedicated to the topic. The plenary session presented by Viktor Mayer-Schonberger, PHD especially focused his discussion on the abilities and limitations of a big data world.

As a market researcher, this discussion particularly hits close to home. Our business is deriving insight from healthcare data, and we're continuously tasked with seeking out new data sources.

Anyways, back to the whats and whys. A large portion of Monday's discussion was on what exactly we can answer with large data sets. His major theme was that we have to pivot our focus to answering the "whats"; "what are we observing" or "what is the trend".  With big data, we can answer these questions better than ever, because we no longer have to abide by "small data" thinking.

Small data thinking is the extrapolation of trends observed in a sample to the entire population something I'm intimately familiar with, given I'm paid to do it every day! With small data thinking, it's up to the researcher to ask the right questions. You can't go back and ask new questions, at least not without great cost. So what you end up with is snapshots of data, pieces of the bigger picture. I can't count the number of radiologists, cardiologists, GPs, etc. that we work with every year, every time, capturing another portion of their knowledge.

Big data is essentially getting the entire picture at once (though this is admittedly an oversimplification).  Unyielding, difficult to work with, but with so much potential. The ability to continuously go back and answer new questions, test millions of hypotheses.

The limitation is that the healthcare data captured currently typically cannot answer the "why" aspect, to give clarity to the cause of an observed trend. It's back to the basic statistics adage: correlation does not imply causation. So the challenge will be to accept that we won't always be able to answer why. But, is the why really that necessary? Professor Mayer-Schonberger gave a great example of this thinking. Walmart did a study of what is purchased most, just prior to a hurricane hitting a region. There were the obvious answers like flashlights and batteries, but surprisingly, poptarts topped the list! So Walmart was then interested in, well, why poptarts. But instead of spending money on why, someone had the foresight to say, "who cares, just put them closer to the cash register!"

I wonder if I could use that same line, next time a client asks why.

What drives the therapy selection test market?

View Now