You Can't Prove Everything, Nor Should You Try
Yesterday, Jerome Groopman and his wife Pamela Harzband, both Professors at Harvard Medical School, had an Op-Ed article in the Wall Street Journal. They argued that President Obama's claim that the healthcare system will save $80 billion by adopting Electronic Medical Records (EMR) is not supported by academic research on the value of EMRs.
I am reminded of John Brockman's website edge.org which is an intellectual visit well worth making. Brockman invites some of the most eminent scientists and thinkers to post articles on various subjects. Each year, Brockman invites about 100 of his contributors to answer a question of the year. A few years ago, Brockman asked "What do you believe to be true, even though you cannot prove it?"
Asking a scientist to answer such a question is antithetical to the scientific method. Yet most of the scientists responded about aspects of their work that they beleive will prove to be true in the future, but they have yet to prove it.
I therefore submit the following counter-argument to Dr. Groopman, whose work I greatly admire and respect:
Dr. Groopman, there have been many inventions, methods and solutions that have been adopted in human history, the benefits and consequences of which could not have been predicted beforehand. On the contrary, the history of technology is full of surprises and unexpected outcomes - good and bad. We simply cannot prove everything ahead of time, and we cannot research outcomes before they have occurred.
My next point may be anathema to scientists and physicians in this "evidence-based medicine" world - we should not necessarily try to prove outcomes before we try new approaches. Whether we like it or not, the world is a laboratory and human history is an experiment. If we try to prove every change in the world before we adopt it, we may certainly avoid some of the terrible consequences of our actions. But we would also stop progress dead in its tracks.
Electronic Medical Records fall under John Brockman's paradigm of "something we beleive even if we cannot prove it." (Yes, putting a number like $80 billion on EMR is simply political necessity, not science.) But we cannot prove ahead of time how patients will benefit from online health records as they will exist in the future. The RAND Corporation is correct in arguing that current systems simply do not exemplify what we will have in the future. EMRs today are merely islands of data. They are not the comprehensive systems that will exist in the future.
Does anyone remember what the Internet was like in the mid-90s, when all we had was email and limited text-based systems? Look at how far we have come in just over a decade. We now have live maps on our cellphones with GPS, we can read the news and watch videos anywhere and anytime. Imagine if we were first asked to PROVE that the future of the Internet would be beneficial? And yes, taxpayer dollars were used to create the Internet, the spread spectrum technology on which cellphones operate, and GPS.
There are some things in this world that we intuitively know must be done. You can't prove everything, nor should you try 100% of the time to do it.
Finally, I do wish to recommend Dr. Groopman's outstanding book, "How Doctors Think". He presents strong arguments for not being locked into specific thought patterns in diagnosis. Some day in the not too distant future, thinkers as profound as Dr. Groopman will pore over a computer screen sifting through a patient's medical records and finds nuggets of information that lead to a diagnosis that might otherwise have eluded them.
Hopefully Dr. Groopman will write a book about combining EMR diagnostics with good patient examination. No doubt there will be something in that book that reads, "Do NOT forget to make eye contact with your patient. There is nothing worse than a doctor tapping away on a keyboard, eyes glued to a screen."