From the healthcare perspective, what is big data? Members of the panel offer: a collection of data, more specifically data techniques for personalized medicine. More broadly, it enables an empirically driven healthcare system.
A disruptive opportunity that will change how we practice medicine.
Thinking by numbers, we’ll make better decisions. US healthcare data available today, due to the explosion in data collection combined with decreasing costs in storage and analysis, amounts to 150 Exabyte’s, or five times the volume of all spoken words in the history of time, points out Zack Cooper. Just wrap your head around that! He sees this data as able to inform four major healthcare challenges: how to predict risk; define better quality; optimally match patients to providers and treatments; and define costs and therefore value.
Harlan sees the collection of information from billions of patient-system interactions, or n=1 trials, that, instead of being discarded, could be used to inform better solutions, better patient health outcomes. As a consumer, tomorrow we will receive a diagnosis, be able to find out how many are like us, what was their treatment, what was their outcome, and then use that knowledge to guide our decisions.
At the point of care (where Paul would like to see its use), instead of a simplistic model of medicine based on historic inference we could align inductive and deductive reasoning to guide practice. Sarah points out that informed risk prediction, even simplistic models in the emergency department work better than unaided approaches. Zack adds that big data is used in the UK today for enabling pay for performance systems.
A fertile field, where there will be no harvest until it is cultivated.
Should we be investing in big data? No (dummy) it is a tool, not a thing in and of itself. Harlan implores us to invest in the analytics, the questions and the healthcare results we want the data to help inform. How many publications do we see in medical journals regarding novel techniques for data? According to Harlan, close to none! Implementing disruption is challenging.
Meanwhile, how do we build processes that begin with the identification of opportunity enabled by data - and end with an action that creates meaningful improvement in the care of a patient?
We have a long way to go. Realize the informatics field is foreign to doctors. Could it even fit in the medical curriculum? It would compromise already overloaded workflows. Even if we were to develop the analytics and obtain the results, would patients make good use of it? With patient compliance estimated at a mere 50%, the newly informed data will go unheeded. Innovation generally costs more, not less, at least initially. Or is this thinking due to riding the Fee-for-Service horse, and that race is still on?
The race is on: which horse are you riding?
Who will drive the use of big data to inform healthcare? For-profit ask where is the ROI? Meanwhile Zack suggests the insurers are the first to have taken advantage of these resources. They are keeping the books and next will begin to predict, then intervene.
The data is available. Healthcare practitioners, you have the perspective on the questions needing attention. Policy makers: help define a viable source of motivation. Analytically minded: stand by to moderate. It is time to put this tool to use.
* Hosted by The Yale Healthcare Ecosystem, The Yale Health & Life Sciences Club in collaboration with the Yale Healthcare Case Competition.