Published on 17-Dec-2013
"Ultimately it’s about improving care. It's about preventing adverse effects, tailoring therapy, and developing new therapeutics." - Dr. Joshua Denny, Associate Professor of Biomedical Informatics, Vanderbilt University School of Medicine
PureData System for Analytics (powered by Netezza technology), PureSystems, Big Data, Big Data & Analytics, Big Data & Analytics: Operations/Fraud/Threats, Smarter Care, Smarter Planet
Considerable research is underway in healthcare to identify genetic and phenotypic markers associated with a particular medical condition or outcome. The work will help physicians better tailor care and potentially develop new therapies for disease prevention. But uncovering patterns hidden deep in medical records and DNA databanks can be challenging. Using a powerful big data platform from IBM, Vanderbilt University School of Medicine clinicians cut research timelines from nearly a year to a few weeks to help accelerate the pace of discovery and, ultimately, improve patient health.
Traditional IT systems were impeding Vanderbilt’s work to identify the genetic basis of diseases and drug response, and hindering new medical discoveries.
By moving its analytics to a powerful big data platform from IBM, clinicians can run complex algorithms on massive volumes of data in minutes, instead of hours.
The ability to run complex queries in seconds has profoundly improved the research process and is driving new discoveries that can improve patient outcomes.
|Preventing adverse effects with big data analytics|
|Instrumented||Enables researchers and clinicians to analyze clinical and DNA data for more than 2.2 million patients over 20 years from a single system.|
|Interconnected||Helps researchers and clinicians connect genetic and phenotypic markers to health outcomes.|
|Intelligent||By understanding the genetic basis of disease and drug response, physicians can proactively tailor care to improve patient outcomes.|
At Vanderbilt University School of Medicine, groundbreaking work is underway to identify the genetic basis of diseases and drug response. Through the use of big data analytics, researchers and clinicians are working together to learn which patients may be at risk for certain diseases and why some patients respond well to certain drugs. This work is heralding a new era in healthcare that will transform healthcare delivery and patient care—enabling physicians to better tailor care for each patient and potentially develop new therapies for disease prevention.
To achieve this incredible insight, Vanderbilt University School of Medicine created a database that includes billions of data points—medical entries and physician notes over the past 20 years from 2.2 million patients. Called the Synthetic Derivative, this database of de-identified information captured from the organization’s electronic medical record system provides Vanderbilt with a rich resource to research disease patterns and treatment protocols.
When combined with BioVU, Vanderbilt’s human DNA databank, researchers gained a treasure trove of information to help them uncover secrets of the human body.
However, according to Dr. Joshua Denny, associate professor of Biomedical Informatics at Vanderbilt and a nationally recognized expert in biomedical informatics, to uncover patterns within this data and test different assumptions, the organization needed to rethink its analytic platform.
Being able to connect genetic markers and phenotypes—specific traits or characteristics in individuals—to specific diseases and health outcomes requires the ability to survey billions of data points from numerous angles over and over.
With its existing platform, each iteration of an analysis often took weeks. The downstream effect on the discovery process was enormous.
“It's not just about the computational time,” says Dr. Denny. “It's the opportunity cost of not being able to iterate all the cycles quickly—meeting with other physicians, developing algorithms, testing and re-testing theories. With our previous environment, it took six to 12 months to complete all the informatics and iterations with clinical experts.”
For Vanderbilt, this pace was too slow.
“We have one of the largest repositories of DNA in the world linked to electronic medical records,” says Dr. Denny. “We see this as a strategic advantage to accelerate the pace of discovery. However, a limiting factor was our ability to test and re-test theories. One of the bottlenecks was data management.”
A platform for big data analytics
Dr. Denny discussed the challenge with colleagues at other healthcare facilities, conversations that led him to the IBM® PureData™ System, powered by Netezza® technology.
“I spent a lot of time speaking with colleagues and investigated a number of different systems and architectures before selecting the PureData System,” says Dr. Denny. “When we conducted our Proof of Concept, we saw firsthand how it could manage large-scale data. We have accrued close to 100 billion genotypes, which is fairly easy to manage with the PureData System. Queries that before took hours, and, in some cases days, now take less than a minute.”
The ability to run complex queries in seconds has profoundly improved the research process, and ultimately, accelerated the pace of discovery.
“To iterate through something on a seconds timeframe means that you can develop algorithms in real time,” says Dr. Denny. “I can sit down with a colleague for a given disease state, and, in a few hours, develop the algorithms. This capability has helped compress research timelines from up to a year to just weeks without too much difficulty.”
It is also vital in driving new discoveries as researchers and clinicians can test more theories—even longshots. Often this type of creative exploration, Dr. Denny explains, can lead to unexpected insight.
“If everything you want to test takes a week to test, you're much less likely to try something that you don't think will work,” says Dr. Denny. “But sometimes the ideas that have a lower likelihood of success yield very interesting outcomes. If you can get answers in seconds or minutes, you’re more likely to experiment.”
New care therapies improve patient outcomes
The new insights gained are in use at Vanderbilt University Medical Center as physicians leverage this research to guide patient care. For example, in one study, researchers found that patients with a specific genetic variance were more likely to have another heart attack when prescribed a particular antiplatelet drug.
“For certain patients with a specific variant, we found that one popular antiplatelet drug doesn't work,” says Dr. Denny. “These patients are at a much higher risk to have another heart attack or stroke, or die. We’ve tested more than 13,000 people for that variant now and have changed therapy on a number of patients as a result.”
In another study, researchers were able to pinpoint that individuals with a specific variant phenotype were more likely to develop arrhythmias than people without this phenotype. This knowledge, Dr. Denny says, will help physicians identify patients at higher risks for an arrhythmia and implement new protocols to prevent adverse outcomes.
“Ultimately, it’s about improving care,” Dr. Denny says. “It's about preventing adverse effects, tailoring therapy, and developing new therapeutics.”
These are just a few of the studies underway. The organization expects to conduct genome-wide association studies on 40 different diseases and 20 different drugs. Additionally, Vanderbilt has developed a method to study more than 1,000 phenotypes simultaneously to identify the range of human disease associated with specific genetic variation.
“Using the PureData System has revolutionized both the pace of discoveries that may lead to targeted therapies, and the number of people who can engage in the discovery process, through web-based, near-real-time access to these massive databases and query tools,” says Dr. Kevin Johnson, Cornelius Vanderbilt Chair of Biomedical Informatics at Vanderbilt.
Optimizing performance for big data analytics
Traditional IT systems were impeding work at Vanderbilt University School of Medicine to identify the genetic basis of diseases and drug response, and hindering new medical discoveries. By moving analytics to the IBM PureData System, part of the IBM family of expert integrated systems, the organization gained a high-performance platform to help researchers and clinicians obtain insight from billions of data points. The system’s integrated design and in-database analytics enable researchers and clinicians to run complex algorithms on massive volumes of data in minutes, instead of hours, helping reduce research timelines from up to a year to just one week.
The inside story: Getting there
The work underway at Vanderbilt University School of Medicine represents a new approach to translational medicine, which traditionally refers to the use of research to improve clinical care.
“This is another phase of translational medicine, where we’re going from practice to research and then feeding the new insight back into clinical practice,” says Dr. Denny. “And the faster that we can turn clinical medicine into a learning system, the faster we can improve care.”
● Helped reduce research timelines from up to a year to weeks—a 10-fold improvement
● Helped physicians identify which patients were most likely to have another heart attack when prescribed a particular antiplatelet drug
● Helped researchers identify individuals at higher risks for developing an arrhythmia
● IBM® PureData™ System for Analytics, powered by Netezza® technology
For more information
To learn more about how IBM can help you transform your business, please contact your IBM sales representative or IBM Business Partner.
To get involved in the conversation, visit: www.ibmbigdatahub.com
For more information about the Vanderbilt University School of Medicine, visit: https://medschool.vanderbilt.edu
Products and services used
IBM products and services that were used in this case study.
© Copyright IBM Corporation 2013 IBM Corporation Software Group Route 100 Somers, NY 10589 Produced in the United States of America December 2013 IBM, the IBM logo, ibm.com, PureData, and PureSystems are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the web at “Copyright and trademark information” at ibm.com/legal/copytrade.shtml Netezza is a registered trademark of IBM International Group B.V., an IBM Company. This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates. The performance data and client examples cited are presented for illustrative purposes only. Actual performance results may vary depending on specific configurations and operating conditions. It is the user’s responsibility to evaluate and verify the operation of any other products or programs with IBM products and programs. THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided. WAC12393-USEN-00