Changing Healthcare Outcomes with Predictive Analytics

An analytics executive at a major East Coast hospital told me something a couple years ago that left me dumbstruck at the time, and it’s stayed with me ever since.

He told me that predictive analytics was capable of forecasting who might die on any given day in the hospital. In other words, by running analytics against information collected from real-time patient data, caregivers were getting advance notice of patients most likely in need of intensive oversight and care.

This is a well-regarded hospital system known for its sophisticated use of big data and the analytics necessary to turn that data into actionable insights. The truth is though, most health care organizations aren’t able to use the massive amounts of data they’re collecting to improve patient care.

Yet the data keeps coming. The volume of data collected by healthcare organizations is increasing faster than for any other industry, and it all has to be stored in ways that it can be accessed, integrated, analyzed and of course, secured.

Beyond PHI and PII

When we talk big data in healthcare, we’re often talking about Protected Health Information, PHI, and Personally Identifiable Information, or PII. But the kind of data I referring to is part of another trend in healthcare today – evidence-based medicine.

Now, evidence-based medicine requires, of course, actual evidence.  It’s a reflection of what actually works in patient care; an active record – intelligent trial and error if you like – of what the best treatment methodologies are for that particular illness. These treatment methodologies have basis in scientific fact, and the changes we are currently witnessing are the result of the application of those facts.

Benefiting from that evidence, however, means that the data is accessible and can be integrated across the ever-present disparate silos of information. This includes integration of competing data models, formats and standards commonly found in the world of healthcare. The advent of open APIs will go a long way toward helping solve healthcare’s current interoperability problem but ignoring those issues for even a moment can jeopardize the true promise of both big data and evidenced-based medicine: enabling a fact-based, data-driven approach to medical care, improving the quality of care and overall patient health.

Predictions from the past

As Yogi Berra reportedly once said, “It’s tough to make predictions, especially about the future.” But that witticism falls completely flat when analytics are applied to big data.

Modern data-centric healthcare organizations are increasing their use of sophisticated algorithms to predict what is most likely to happen next. By pairing large data sets with analytic tools, they’ve been able to gain the type of insights promised from data lakes, large databases and streaming, sensor-based information generated via the Internet of Things. In the simplest terms, the best predictor of the near future is the recent past. This application of analytics at the hospital I mentioned above illustrates the real benefits of aggregated data.

The most significant hurdle to fully realizing the usefulness of this aggregated data lies in the ability to incorporate data from the recent past. Without universal global data standards or protocols, healthcare providers are left with the painstaking task of consolidating and analyzing data from multiple sources as well as multiple incongruent structures.

With the help of modern API tools, data analytic solutions can and will deliver the benefits of big data.

But there’s still the more prosaic problem of where and how to store all that data.

Be ready for the flood

Many organizations are finding that traditional on-premises storage devices can no longer keep up with their growing needs. The rate of data accumulation continues to rise, while varying data formats and long-term storage needs further complicate matters. In the US, most states require the retention of data for at least seven years, and HIPAA compliance requires that data be stored for similar lengths of time. That means a lot of data and a lot of storage for a very long time.

What kind of secure, cost-effective and accessible storage options can we utilize? At this point, most experts agree that cloud is the only real viable option. The cloud offers economical storage with the elasticity necessary to quickly expand as needed, without the traditional forklift upgrades required for on-site equipment. Disaster recovery is simplified as well.

[Learn more about navigating the complexity of big data transformation]

For organizations looking for a trusted partner, Rackspace offers the convenience of short spin-up cycle times and enables compliance with established industry security protocols, including HIPAA, HITRUST and SOC-2. Our experience helping organizations across all industries, including those with stringent regulatory requirements, create or transform their big data strategies makes us an ideal vendor for hybrid and cloud-based data and storage solutions.

Previous articleHow Private Cloud-as-a-Service Enables a Successful Multi-Cloud Strategy
Next articleSecurity is a Big Data Problem
Chris Christy, FACHE, serves as Director of Healthcare Solutions at Rackspace. He has been professionally engaged in Healthcare for 35 years, working in both hospitals and healthcare technology companies. He brings 16 years of healthcare provider industry experience and previously held administrative spots with nonprofit and for-profit healthcare systems. Chris served five years as Associate Administrator and Chief Operating Officer at several hospitals that were part of American Medical International, which became Tenet Healthcare Corporation. He also served nine years as Vice President for Professional Services at St. Paul Medical Center, a 600-bed facility in Dallas operated by the Daughters of Charity National Health System, now Ascension Health. In addition, he served two years as Regional Vice President for Emcare, Inc, a publicly traded Emergency Room physician group practice. Over the last 19 years, Chris has worked in the enterprise healthcare technology industry. At SAP, he was appointed Healthcare Principal and in 2011 he was named Executive Director for Accountable Care Organizations as a duel role. Subsequently, he also has had healthcare positions at Qlik Technologies and Oracle Corporation. He joined Rackspace in 2018. He is a Fellow in the American College of Healthcare Executives and has served as an Adjunct Professor for Healthcare Strategic Planning at Texas Woman’s University in Dallas. He received his Master of Science in Public Health from the University of Missouri-Columbia in Columbia, Missouri.


Please enter your comment!
Please enter your name here