Considering a cloud-first strategy? Start by knowing that it doesn't mean cloud-only
Posted on: Wednesday, February 27, 2019 By: KorchekStaff
More and more hospitals are moving data and apps into the cloud and taking advantage of storage and compute resources via infrastructure as a service offerings.
Gartner predicted that the overall cloud computing market will expand by 17 percent annually during the next three years to $206 billion — and continued growth is expected in healthcare as well.
“We’re kind of nearing the tipping point with significant growth and acceleration around cloud, and with the frequency with which U.S. hospitals continue to invest in cloud,” said Janet King, senior director of market insights at HIMSS Media.
So it follows that the philosophy of being a “cloud-first” organization is also gaining momentum. But what does it really mean? And what does it not mean?
Let’s take a look at what executives from Rush University Medical Center, Emory University and Sisters of Mercy Health have to say about that.
Cloud-first vs. cloud-only
Just because an organization decides to become cloud-first doesn’t mean that every new app or software program they adopt will be in the cloud.
“Cloud first doesn’t mean cloud only,” said Ashish Sharma, an associate professor in the biomedical informatics department at Emory University. “There are certain apps that just don’t make sense to put in the cloud.”
Sharma said, for instance, that Emory is better off using edge computing for large-scale images rather than putting those in the cloud.
Large images are just one example and the same may be true at most health systems. But when it comes to other apps, hospitals will need a way to decide which apps and data are a fit for the cloud and which are better off on-premises.
“We are building an assessment model to determine what can be cloudified and which cannot be cloudified,” said Jawad Khan, director of data science and knowledge management at Rush University Medical Center. “We can use the same assessment model on existing apps as well as new ones.”
No one-size fits all assessment
Hospitals have to take into account their own unique situation to determine the business drivers of going to the cloud — and that is the hard part.
“Once you do that you will understand the benefits,” Sharma said. “And it’s easy to take your workloads into the cloud.”
Sharma said the cloud’s big benefit for Emory is processing at scale and the system uses cloud pipelines to access the data and computational activities that fuel its precision medicine oncology programs.
For Sisters of Mercy Health System, the cloud is where it has harnessed analytics services and capabilities instead of hosting those on-site.
“Descriptive analytics has led to the development of data science, AI and machine learning,” Mercy’s Vice President of Integrated Performance Solutions Curtis Dudley said.
Or what FeibusTech Principal Analyst Mike Feibus called the big data conundrum: “the move from just being able to read the data to that next level of AI.”
What comes next
“In the future there’s going to be more and more healthcare data in the cloud,” Mercy’s Dudley said. “The more we do, the more people want – internally and externally.”
That’s likely to prove true around the industry as well — particularly given the aforementioned predictions about how much hospitals will invest in cloud services.
“Going forward, one of the things that I would stress is that this entire approach only works when your local health IT and policies are cloud-aware,” Sharma said. “Cloud systems are not simply datacenters that happen to be in Oregon or Georgia.”
By Tom Sullivan
Healthcare IT News