In the last 20 years, the rise of digitization in medicine has led to an abundance of information we are unprepared to utilize. Data miners are working hard to process various studies for discovery of results. More importantly, on the individual patient level, collected information may or may not be helpful given the context. A friend of mine mentioned a few months ago that he couldn’t understand why (despite being young and thus far healthy) every year he wasn’t offered every possible test to make sure something wasn’t lurking ready to attack when he tipped over a certain age. That was the real way he thought medicine should be practiced.
In a recent NPR story, the status of breast cancer in 2013 is reviewed. From what we learned in 2012, our best test for diagnosing breast cancer, the mammogram, is great at what it does: it finds tumors. However, because it is an imaging test, we can’t be sure which findings require action. There are many studies on this topic and the verdict is still hotly debated, but we should always reconsider if our current screening guidelines are doing right by our patients — knowing when to act on information is hard when you have too much of it. We live in a time where the ethics of finding anything in our collection of patient data implies action. Physicians who ignore information or miss it in the first place are too liable by our legal systems to be held negligent if something goes awry.
As a global society, the era of data collection is certainly at its height. Now the rise of the era of information extraction and knowledge discovery are most important. Without this transition from bench-to-bedside and machine-to-guidelines our science is too academic, intellectual and useless. How will we take what we’re learning about the difficulty of breast cancer diagnosis and apply it to the future of delivering quality data in all domains that can help us make the right clinical decisions at the right time?
Photo Credit: Kiki gets a mammogram by kristiewells