In my last post, I talked about how the Healthcare industry is challenged with improving overall health while not overlooking that one person that could have been treated and healed. Big Data has the potential to ensure that Healthcare does not have to sacrifice one for the other.
Attending the UCSF Big Data Seminar recently, I was reminded of something we all know at some level – data about our overall can be expressed in many ways – family history, identified risk factors, results from any myriad of tests, image data (e.g. x-rays), environmental factors, behavior/personality traits, etc.
Yet, it is ironic that prescribed medical treatments address conditions based on 2, maybe 3, sets of those data. It prompted one speaker to ask, “Why can’t a treatment embrace all of these factors?” Being able to combine these data (provided that these areas can be collected as “datapoints”) into some type of comprehensive analysis (as opposed to focus on the change in one or two variables) can aid in deriving some true holistic treatments.
Aside from combining health-related data, another promising aspect of Big Data has been to incorporate data that may have never been previously considered, let alone examined. In Healthcare, incorporating lifestyle-type data – financial, shopping, entertainment, weather, social – can potentially provide a more holistic view of people above and beyond any medical tests. (The idea is nothing new…just remember that Wall Street began to use parking lot images to help them better estimate Walmart revenues.)
Of course, incorporating this plethora of data is not without its challenges. Privacy and legal issues abound, but imagine the types of discoveries that can be uncovered if not viewing treatment efficacy as just a “it works or it doesn’t” test. There is also the issue of how to link “health related” and non-health related” data meaningfully. Those at the seminar did agree that linking these varied datasets is a “young science”.
Another thing to consider is that regardless of how much data is used in an analysis, one still has to deal with “correlation vs. causation”. Within healthcare, that will never be sufficient; they ultimately want to find the cause of an ailment and determine the right treatment. Big Data can help to perform more holistic analyses at a faster rate, yet it cannot necessarily help to arrive at the “final”answer more quickly (i.e. the cause).
I left this seminar thinking that there are so many opportunities remaining for companies to address how Healthcare can take advantage of Big Data more effectively. At the same time, I firmly believe that the need for getting the right answer quickly is not as important as getting the right answer.
I believe that the best value proposition for Big Data in Healthcare should focus on obtaining a more complete picture of an individual’s health from both a medical and lifestyle perspective. Imagine being able to derive and test more complete treatment plans the just relying on one drug or procedure. That can go a long way to improving overall population health and decreasing costs. At the same time, the possibility of missing that one person who should have received some treatment can be decreased, since a holistic profile of a treatment candidate can be identified and used.
One can always become faster at deriving answers through practice and experience, However, getting the right answers in Healthcare is more important. Let’s have Big Data companies focus on more on the discovery, less on the time spent.