When you hear terms like data-driven or personalized medicine, what comes to mind? Prescription cocktails tailored to each patient’s height, weight, diagnoses, family history, etc? Gene therapies that actually target your specific genetics?
If your last trip to the doctor’s office was anything close to typical it’s probably hard for you to imagine your doctor querying a database or creating models against available patient data to tailor a treatment to your needs. The fact is we’re quite far from personalized, data-guided medicine, and getting there will require dramatic changes to the way we conduct both healthcare and analysis.
Why is that?
There are a few reasons, but they boil down to three central problems.
- Providers don’t have the time or training to do things manually.
- Prevailing methods can’t match prescriptive actions to a patient’s info.
- The results of statistics-based analysis are rarely actionable.
Today the role of data analysis in treating patients extends almost exclusively to the clinical studies that guide the treatment of patients more generally. Drug treatments are planned based on the strict analyses of trials, but their results are guided towards larger populations, with little room for personalization. While many healthcare systems endeavor to bring more data guided treatments to patients, that’s still a far cry from the day when a complex set of factors can be matched to extensive historical data to guide treatments as patients enter the system.
With truly personalized medicine, treatments would be adjusted to match unique traits of each patient with the most effective treatments in ways that we simply can’t do today. Meeting those needs will require access to extensive collections of patient data, including demographics, treatments, and outcomes. It will also require advanced data automation. This means data analysis itself will have to be automated to help guide healthcare professionals’ treatments. And much like many other professions, it’s not the realm of a physician to run through analysis after analysis for each patient they see.
In order to see real progress in modern personalized medicine the analysis itself will require automation to give healthcare professionals the right information to guide their decisions. The industry can’t staff analytics teams to meet this need, and even if it were possible to staff these offices and hospitals, the human efforts simply won’t keep with the flow of patients.
Healthcare is just one of many examples of how critical data automation will be in handling the enormous amounts of data we encounter in our daily lives. It can’t be through traditional querying and visualization, but must be something that can quickly reduce masses of data down to the information that help make us faster and more efficient.