When Josh Gluck joined Pure Storage this past April, he arrived well-acquainted with the most pressing data-management issues affecting healthcare IT leaders today. He’d spent the previous two decades in hands-on leadership roles in the field, most recently as deputy CIO for Weill Cornell Medicine in New York City and, prior to that, as director of IT at New York Presbyterian Hospital. He continues to teach health policy and management at New York University. As Pure’s VP of global healthcare technology strategy, Gluck has wasted no time establishing himself as a go-to thought leader on everything having to do with, and made possible by, enterprise-level flash storage. He took our questions on flash and why it promises to remain a hot topic among healthcare IT leaders for many years to come.
Given the technological advances we’ve all seen over the past five or so years—AI being merely the most conspicuous—these are heady days for healthcare IT leaders at all levels, from project managers to CIOs. What opportunities to make a difference are ripest for the picking?
Healthcare is in an exciting time of accelerated change. Methods of care delivery and reimbursement are rapidly changing, and the opportunity to create significant impact on patients’ lives through the use of technology is real. Healthcare CIOs and other IT leaders really need to focus on efficiency and driving down cost to ensure the financial health of their organization. They need to exercise agility to support changing clinical and administrative workflows. They need to innovate to build a data platform that will unlock the value of the provider organization’s assets at the point of care. And they need to do all this while improving security and reducing risk.
In a nutshell, provider organizations are focused on improving their patients’ lives while maintaining the financial viability of their organization—and IT continues to play a central role in supporting and sustaining that focus.
What challenges are proving the most difficult to deal with?
Budgets are always getting tightened. Everyone is being asked to do more with less. Technical debt is a real thing for a lot of organizations. They’ve built individual application siloes over time, and those siloes are aging. At the same time, folks have a tendency to resist change. And in a very fast-paced environment that’s focused on patient safety, a lot of technical teams are struggling with the pace of change. They’re struggling to acquire new skills and keep up with new technologies.
The value-based reimbursement structures under Medicare’s Quality Payment Program are forcing hospitals to consider what happens to the patient after he or she is discharged to another care setting. This challenges hospital and health-system CIOs to deal with mountains of data from outside their own institution. What are you seeing there?
Without a doubt, there’s a need for more data in order to operate in this new paradigm. And the complexity doesn’t just come from the amount of data. It’s coming from internal systems, it’s coming from external care-delivery partners, it’s coming from multiple sources. It’s even coming straight from patients, with wearables and home monitors and all of that. And it’s coming in structured and unstructured form.
Apart from the volume of the data and the velocity at which it’s coming in, one of the real challenges comes from figuring out how to integrate all the partners who are responsible for the delivery of care into a single platform—one that makes it easy for end users to get their data in and out. And one that doesn’t incur a high cost for pulling the data all together. U.S. healthcare is no longer a fee-for-service world. You can’t look at a single event of care as if it’s disconnected from the patient’s other events of care. You now need to pull together systems and data from all across the continuum of care, in real time, so that you can make sure your [payer] contracts stay viable and make sure your patients are optimally cared for over time.
This is a real challenge for a lot of organizations. A lot of people think infrastructure doesn’t really matter. They think they can take what they built 10 years ago and connect it to some new web-enabled platform. And that just isn’t the case. With the geography of care expanding well beyond the hospital, we’re finding infrastructure really does matter. A data platform that enables data access across the continuum of care is a requirement for success.
How important is it to have the right data platform for provider orgs that want to maximize the potential of emerging mobile-health technologies?
If you’re going to support potentially hundreds of clinicians in the field who are suddenly scanning thousands of patients more than have ever been scanned before, you’ve got to have a data platform that can scale appropriately to support that exponential increase in demand. A dynamic, scalable platform is also going to enable healthcare organizations to improve the patient experience. For example, you may want to expose that data back to the patient once they’re back home and they want to look at their medical images. And that would be difficult to do in the traditional way, where you have siloed storage as just another bolt-on system.
It’s important to have a data platform that allows you to join multiple workloads together and that lets you expose data across different applications and keep it all within the same experience for both patients and clinicians. Trying to achieve that with traditional hard disk storage would be extremely expensive. It would be hard to live up to the performance expectations of both clinicians and patients at the same time. Building out an all-flash platform lets you provide the performance requirements. And the all-flash platform provides the security and availability that are needed to support mobile health.
What are you seeing in the IT departments of provider organizations moving into genomics and precision medicine?
CIOs of academic medical institutions need to do everything they can to empower their researchers who are developing genetic tests. That will mean building infrastructures that can support genomic data pipelines and really improve time-to science performance.
CIOs working in the healthcare provider setting today need to be aware of what’s coming down the pike and be ready to support what their organizations think is the best way to leverage precision medicine or a genomics plan. It could involve incorporating into their EMR a whole bunch of data that they hadn’t expected. It could be connecting to outside providers that will run genomics tests for them. All in all, they really need to think about the infrastructure they use in their environment and how much it matters. They need to think about how the infrastructure can enable their ability to access this data quickly or share it among their partners in the continuum of care.
How could an all-flash data platform help a provider organization maximize its return on predictive analytics and other data-intensive applications?
We’ve seen organizations start out their journey into predictive analytics by taking the data from their existing systems and loading it into Excel. Or putting it into a SQL database and running batch queries. Twenty-four hours later, they get their query results. That might have worked 10 years ago. But now what we’re finding is that organizations are hiring data scientists. And the data scientists want to rapidly prototype different predictive models and get the models into the hands of clinicians, administrators, researchers—whoever can get value out of data-rich insights.
Trying to do that on a traditional disk array makes little sense today. Those arrays were purpose-built for batch jobs. And the environment we’re talking about now is one in which everything happens in real time. An all-flash platform gives that kind of performance. And it does so in a way that’s cost-effective because of the consolidation that you can do.
Let’s drill down into, specifically, Pure Storage’s FlashBlade technology vis-à-vis the present boom in the sheer volume of healthcare data. How can this product and associated Pure Storage technologies help healthcare IT leaders “do more with less” as they seek to formulate winning data-management strategies?
We’re in a time of renaissance in healthcare. And Pure Storage built its technology from the ground up. We don’t have the technical debt that traditional spinning disk suppliers have. Some of those traditional disk vendors have gone out, purchased flash technology and tried to integrate it into their portfolio. What they’ve done is integrate it into their product portfolio, which makes it really confusing to customers. The customer has to try to figure out which array to use at which time. And they wind up with an environment that has a whole bunch of different types of arrays. The management and administration overhead is significant.
Pure Storage offers two purpose-built solutions that can handle all workflows in a healthcare environment—FlashArray and FlashBlade. We believe organizations should buy their storage capacity once and, with it, get controller upgrades as part of the purchase. In other words, we offer evergreen storage. So not only have we reinvented the technology. We’ve also reinvented the way healthcare organizations can acquire and maintain their data platform investment.
In addition, we’ve seen that folks really want to have an option for consumption-based storage. We think those organizations that are looking into the public cloud are doing so because of some of the financial drivers. So now our Evergreen Storage Service (ES2) offering allows for a consumption-based model of on-premises storage.