Setting Up Cloud Transformations for Success in Healthcare

How Sentara Health Approaches Cloud-First Strategies

Sentara Health has nearly 30,000 employees and serves communities in Virginia and North Carolina. The large system recently launched a joint venture focused on cloud-first solutions to help healthcare organizations modernize their environments.

“The first bottom line is that many CIOs in healthcare do not have the skill sets in their own organizations to move the cloud,” Thomas says. “Because of what we’ve learned, we’re at the forefront of large-scale movements of systems to cloud.”

Thomas joined Sentara Health in 2018 with the goal of driving the organization to cloud-supported solutions, starting with the largest on-premises data set.

“When we started deploying our first workloads into Azure, we designed backup as a core part of it,” Thomas says. “We set all the policies up before we deployed our first solution set into Azure. That was two months, so it was a very quick evolution.”

This year, Thomas says, close to 80 percent of the organization is supported by cloud solutions. Within the next 18 months, Sentara Health will leave its last on-premises backup solutions and move it all to the cloud.

DISCOVER: Children’s hospitals find space to grow through the cloud.

Why Asante Went with a Hybrid Cloud Model

In southern Oregon, health system Asante adopted HPE’s GreenLake, a hybrid edge-to-cloud platform for its critical on-premises data services. The process of discovery, setup and handoff took about 18 months, says Asante ITS Operations Manager Tim O’Rourke.

“Our internal cloud has been a very valuable resource, spanning redundant systems across data centers in different buildings and on different sites. We have been able to deliver most of the benefits of the public cloud to our internal customers for years. Now, under GreenLake, Asante can also realize the financial benefits,” O’Rourke says.

Asante’s rural location doesn’t offer wide availability of high-capacity internet access, nor is it nearby public cloud data centers, so the health system uses cloud services only for data backup and under a Software as a Service model for some applications.

“Our first GreenLake initiative was for HPE Primera storage arrays,” O’Rourke says. Though Primera is a hardware solution, it has the performance and agility of the cloud. “The program allowed us to spec equipment at a level that we could not have capitalized in a single fiscal year.”

O’Rourke says his team first studied how the GreenLake program would work, assessing the benefits, risks, costs, support needs and lifecycle management. He discussed the implications of the shift from capitalized to operationalized server/storage infrastructure with Asante’s finance department and senior leadership.

“We had internal discussions about how on-premises cloud resources would be managed, usage reported, budgets established and project chargebacks processed under the GreenLake model,” O’Rourke says.

Deciding Cloud Deployments Based on Organizational Needs

Thomas says that Sentara Health uses a variety of solutions to address particular business needs, including Microsoft Azure for native backup, Rubrik as an instance for Unix and Dell Avamar to back up virtual machines, images and snapshot servers.

“The percentage changes because we are bringing in new solutions and retiring solutions,” Thomas says. “We are heavy users of cloud backup, and we've transitioned a lot of our on-premises to cloud backup solutions.”

It’s important that cloud support is chosen based on application and system need, Thomas adds. Considering the processes along with their risks is key, even over staff needs.

“It’s usually never an individual making recommendations, because they would be driven by the processes they create, and then we manage the risk for them,” Thomas says. “It's very easy for somebody to say, ‘I want it backed up every night,’ because they think it's free, but it’s not. We need to manage that risk, the frequency of backup and how long we maintain the backups based on the data classification and record retention policies.”

FIND OUT: Why modern data platforms are the next step in the healthcare cloud journey.

Vibrant Emotional Health began working with Hitachi in 2016 to move to public cloud and build a data warehouse and call center failover solution using Amazon Web Services (AWS). Formerly known as the Mental Health Association of New York City, Vibrant operates a 24-hour crisis center with more than 500 employees.

The organization chose to work with Hitachi due to its ongoing customer service offerings, says Vibrant Marketing and Communications Specialist Dante Worth. “One major benefit of working with Hitachi is their reliable 24/7 monitoring of our environment. Hitachi also offers an ideal cost along with cost and security reviews,” Worth says.

The data warehouse allows for Vibrant to offer continuous behavioral health services in New York City. The process of AWS implementation with Hitachi took about 45 days.

“Before getting started with Hitachi, we started by outlining requirements. We then moved into the initial architecture and design phase. Finally, we followed up by initiating the infrastructure build,” says Worth.

Supporting and Maintaining Cloud Capabilities in Healthcare

Ensuring that connectivity is resilient and has bandwidth to manage large data sets is an important step to reduce possible workflow interruptions, Thomas says.

“In the cloud, it’s really about the timing of when you back it up, because you’re not constrained by the throughput. You’re more constrained by the impact to performance, the method you use for backing up,” Thomas says. His team performs regular checkups to measure cloud health.

“We also have automated the alerting and the reporting of backups, and if a backup fails, it goes into ServiceNow as an incident and gets processed just like any other incident. Next, the team checks why the backups had issues and works through those processes to resolve them,” Thomas says. “From an operational workflow, we’ve just rolled the cloud support into our standard model. I think the big difference is that we’re not moving tapes. We’re not worried about offsiding tapes the same way.”

UP NEXT: Assess and optimize cloud security tools as part of zero-trust initiatives.