On Point: Better data analysis can help us heal our health system

Bob Lokken, The Daily Record Newswire

Data, data, data. We have data everywhere, and we are only just beginning the collection of data.

It has been estimated that from the dawn of civilization until the year 2003, humankind created about five exabytes of data, and that now we generate that much data every two days. An exabyte is a quintillion, or 10 to the 18th power. That’s a lot of data.

Much has been written recently about the emerging field of data analytics. It’s central to the way we handle information, and it’s key to operations that affect the way we work and live. One area where data analysis is changing our lives in Idaho is in the business of health care.

Data holds tremendous promise to illuminate insights, speed reaction times, enhance learning, predict outcomes, save us money and drive transformational change. But with the tremendous volumes of data we are generating in health care, and the speed and complexity of transformation, we have stretched traditional technologies to the limits of their effectiveness.

Traditional technologies, such as spreadsheets and reports, will continue to play a supporting role in decision-making. However, new and improved technologies are required to deal with our fast-paced, data-overloaded world.

This is where data analytics shine. Consider the following scenarios:

A few people in every community consume massive amounts of health care resources every year. What if we could better predict who those few people will be and proactively reach out to help them take better care of themselves?

Certain readmissions are very complex and costly. What if we could better predict who is likely to be readmitted and could focus on preventing those readmissions?

What if we could compare every procedure across St. Luke’s Health System with national trends? What sort of patterns would emerge? Would those patterns identify best practices that we could replicate across the system? Are there some patterns that could be refined to eliminate waste and gain more consistent quality outcomes?

There are six to 10 data sources that paint a picture of how we are providing patient-centric care. Epic data, patient survey data, lab data, pharmacy data, claims data and national benchmark comparisons are sources that can be used to identify successes we should celebrate, as well as areas that need improvement. From these data sources, we can derive hundreds of metrics that will help inform and shape program improvements. Data analytics can help us automate delivery of these metrics to thousands of care providers in a personalized, visual and meaningful way.

We know certain patterns of care indicate a high likelihood of pending problems. And there are missing patterns of care that concern us. What if we could pull all of the data together and scan the longitudinal records across tens of thousands of patients, in hundreds of settings, weekly or even daily, to identify and illuminate these patterns? If we could, we could certainly reduce the probability of adverse outcomes.

Science is based on observation and experimentation. Throughout history, whenever advances in technology allow us to see things not previously visible, there follows an explosion of learning. Examples include the microscope, radar and magnetic resonance imaging.

Analytic technologies provide a lens through which we can see into the digital world and begin to harness the power of the data we have only just begun collecting. This technology ranges from automating tasks that could have been manually performed if there were only enough time to do so, all the way to advanced machine learning algorithms that can scan billions of records in seconds to identify patterns and insights buried within.

This new generation of technology holds the power to mitigate the information overload and data deluge with which we struggle today. Instead, we can give people the right data at the right time and enable quick and easy access to the information, patterns and insights that really matter.


Bob Lokken is the CEO of WhiteCloud Analytics.