Published on 15-Mar-2012
The nature of Harvard’s pharmacoepidemiology research has been highly complex and computationally intensive since its inception. Today, however, the sheer volume of information available – and the increasing detail in which the information is recorded – has dramatically increased the analytic and computational challenges.
The lab at Brigham and Women’s Hospital was looking to find a platform for computational pharmacoepidemiologic analytics that would address rapidly emerging trends.
IBM Netezza partnered with the lab and installed an IBM Netezza data warehouse appliance. IBM Netezza data warehouse appliances are purpose-built to make advanced analytics on data simpler, faster and more accessible.
By utilizing IBM Netezza technology in their pharmacoepidemiology research studies, the Harvard Medical School division will be able to: • Increase the speed of computationally-intense analysis of claims data • Accelerate testing of new, more sophisticated algorithms • Facilitate automation of continuous drug safety and effectiveness monitoring
In this video, Sebastian Schneeweiss, MD, ScD at Harvard Medical School, practicing at Brigham & Women's Hospital, talks about how his team uses Netezza to research the effectiveness and safety of medications in routine care. They've partnered with Netezza to start a computational pharmacoepidemiology program, leveraging high performance analytics on complex computational algorithms to better understand medications and how they're performing. The Harvard Medical School® Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women’s Hospital in Boston, Massachusetts, is a globally recognized leader in drug safety and effectiveness research. Created in 1998, the division is led by Dr. Jerry Avorn, Professor of Medicine at Harvard. The division is part of Brigham and Women’s Hospital’s Department of Medicine, and performs advanced analytics on patient health claims data. Through computationally intensive ‘big data’ analytics, this research brings to light insights into how drugs compare to each other in terms of safety and effectiveness.
Sebastian Schneeweiss, MD, ScD
Associate Professor of Medicine and Epidemiology, Harvard Medical School Vice Chief, Division of Pharmacoepidemiology, Dept. of Medicine, Brigham & Women's Hospital
We’re embedded in a very clinical environment, and we do a broad base of studies that is related to researching the effectiveness and the safety of medications in routine care. We analyze large randomized trials, we study adherence patterns of medications, particularly in elderly patients using Medicare and Medicaid databases.
Despite the fact that we traditionally thought we had powerful information technology, we soon encountered limitations that made it very hard to say the least; to run our algorithms in these large databases it often took a weekend to run analyses which made it impractical for us.
The insurance databases that we work with in pharmacoepidemiology and comparative effectiveness research, they become bigger and bigger, so there are no longer 500,000 or a million lives, you’re now talking tens of millions of lives, approaching 100 million lives in this database covered.
At the same time, these databases become richer; it’s no longer naked claims, it is enriched with lab test results, with bits and pieces of clinical result data that are linked and merged to these large databases.
The analytic procedures that we are using in order to come close to causal inference with our statements, whether it’s direct causes, this and that affect, become more and more involved, more and more intensive, computationally more intensive, which actually motivated us to start together with Netezza’s help the computational pharmacoepidemiology program.
Within two days, the Netezza machine was up and running – it was delivered at the dock of our data center in Needham, it was set up and two days later our statistical analysts were already working on the machine. I don’t think the transition team worked full time over those two days really, so that was a big surprise.
We were really impressed, I think it was at least ten times improvements in speed and this is without optimizing the code at all.
During the few weeks that we’ve had the machine now, I have not heard anything that we have invested much time in maintaining the machine - which is important to us because we don’t have, as an academic institution, a standing budget for computer or IT maintenance.
And I thought, “Gosh this is the perfect marriage,” because it is a hardware and a software solution that is specialized at optimizing database applications. That is exactly what we are doing as pharmacoepidemiologists – work with large databases that we have to manipulate extensively with complicated statistical procedures, and we need these results fairly quickly. We cannot wait a weekend for an answer.
Since we’ve had the machine, we are not disappointed. We are excited about the improvements that we have observed so far and I am optimistic that once we optimize our codes to the Netezza machine we will see even more improvement in the performance of these algorithms that we are using, and that quite frankly many other people are using in our field, whether that is in academia or that is in industry, we are all relying on these large databases and we need to find ways to analyze these databases not only quickly but also the right way. And analyzing the right way is computationally intensive which is where machines like Netezza have come in.
Products and services used
IBM products and services that were used in this case study.
IBM Netezza 1000