Microsoft Workloads on AWS

How is Responding to COVID-19 with AWS

For nearly two decades, Arcadia has specialized in the integration and normalization of health care data from a wide array of electronic health record (EHR) platforms, then enriching these sources with insurance claims and operational data, and using that data to enable customers to drive improvements in patient care quality, practice efficiency, and financial performance. With COVID-19 upending healthcare routines, in the span of just two months Arcadia has had to rapidly evolve their products and practices to assist healthcare organizations in stopping the spread of the virus. According to Arcadia, this wouldn’t have been possible without the scale and flexibility of AWS.

Shifting Arcadia’s focus to respond to COVID-19

According to Michael Meucci, chief growth officer for Arcadia, healthcare providers had been using Arcadia’s platform for important but less time-sensitive needs, like identifying patients that might need to schedule an office visit to receive routine care like a flu shot. But once COVID-19 hit, the stakes were raised even as time windows for responding were shortened. Arcadia thus needed to reconfigure and scale new portions of its platform to quickly ingest and process new data sources to help healthcare providers identify and communicate with patients in new, more time-critical ways.

First, they introduced new patient stratification algorithms to identify which patients had high, moderate, and low risk of developing COVID-19 complications. Arcadia pushed these to all of its customer environments within 10 days of the declaration of the state of emergency. Second, they created new communication tools for healthcare providers to use to reach patients through a variety of channels, using Amazon Simple Queue Service (Amazon SQS) to manage the messaging queue to route hundreds of thousands of messages through the Twilio API (which runs on AWS) for delivery in the course of a week. These messages might include content that guides patients through access information for telehealth and hotlines. Third, Arcadia deployed a daily survey tool, using Amazon CloudFront over Amazon Simple Storage Service (Amazon S3) for mass distribution of surveys and backing the processing using AWS Lambda functions, to enable healthcare providers to keep in touch with patients through remote patient monitoring. Such surveys might include a personalized interactive COVID symptom checker to allow health systems to securely capture information about patients’ symptoms and care needs.

Each of these actions, in turn, helps to minimize risks to vulnerable patient populations.

Beyond such reactive and proactive responses to COVID-19, Arcadia has been working with AWS and others on different coalitions to build a set of tools and resources for COVID-19 responders. The coalitions have a focus on using data (for example, the survey data mentioned above) to predict the spread of COVID-19 within communities, as well as making a de-identified, longitudinal set of data on patients suffering with COVID-19 available across research institutions to support efforts to develop a vaccine for COVID-19. This longitudinal approach is important because measurements on the same individuals taken over time can reveal causality, not just statistical correlation. In each case Arcadia depends upon AWS services like Amazon Elastic Compute Cloud (Amazon EC2) and others detailed below, to support their massive ramp-up in scale.

Arcadia survey tool

Arcadia’s Outreach and Engage products 

Emergency preparedness with AWS

Arcadia’s ability to transform itself to meet the needs of healthcare providers stems, in significant part, from its choice to build on AWS, according to Jonathan Cook, chief technology officer at Arcadia. “It was clear we needed to get out of the data center. Managing the data center was a poor use of our time and money. Additionally, we were seeking to become HITRUST certified, which requires a cloud partner that is ahead of the pack in certifications and understanding compliance requirements,” said Cook.

As Arcadia’s business has boomed, growing several-fold in the past year alone, so has their need for scalable infrastructure. Every time Arcadia onboards a new healthcare provider, they must process upwards of 10 years’ worth of historical data, Cook said. As a result, they have a “Black Friday” need to scale every time they onboard a new customer, seeing processing spikes of 10X or 100X the norm, to the tune of exabytes of data every night. Arcadia stores all of that data in Amazon Simple Storage Service (Amazon S3), which Cook hailed as “the tenth wonder of the world.”

Beyond customer onboarding, Arcadia handles other data that requires special processing that spikes hardware usage. Before AWS, said Cook, processing this data took days. Unfortunately, it also required running a particular piece of infrastructure non-stop at a cost of thousands of dollars per month. Once Arcadia moved away from a legacy SQL Server-based architecture to Apache Spark on Amazon EC2 clusters running on Distributed Cloud Operating System (DC/OS), their processing speed increased by a minimum of 50x, dropping from two days of processing to 20 minutes. The price? A few dollars.

This ability to quickly and cost-effectively scale has been particularly useful during the pandemic. As mentioned, to support healthcare providers on the frontlines against COVID-19, Arcadia had to write new patient stratification algorithms to help identify vulnerable patients, but it didn’t need to buy new servers which, according to Cook, is good because “I can get a server up and running and in production faster [on AWS] than I used to be able to get a price quote” from a server vendor.

As Arcadia has increased its development and deployment velocity during the pandemic, patients have benefited. Building with Amazon SQS, Amazon CloudFront, Amazon S3, and AWS Lambda, Arcadia’s healthcare customers have actively communicated with over 1.2 million people, with many more to come. In addition, six Arcadia customers have used Arcadia’s risk stratification models to assess COVID-19 risk for over 10 million unique patients.

Arcadia’s modernization journey continues

Eventually, the pandemic will come to a close, but Arcadia’s ongoing modernization journey will not. Arcadia is several years into its partnership with AWS, but continues to look for new areas in which to invest.

For example, though Arcadia has long used Microsoft SQL Server because many of its healthcare providers use it, Cook said SQL Server is “a legacy portion of the infrastructure that we’re slowly reducing.” Already the company runs ScyllaDB on Amazon EC2 clusters, and is evaluating Amazon Aurora. Similarly, Cook said, Arcadia “is trying to reduce our Windows footprint in order to be able to push into a Linux-based cluster, which gives us more power at less cost.” It’s a pragmatic approach. With the size of the company’s data growing exponentially, Cook said, there isn’t a Windows solution capable of handling their requirements. “With some of the processing that we’re doing, particularly the pre-processing, scaling vertically doesn’t make sense any longer. Additionally, there is no Windows box big enough to do it.”

For now, Arcadia is 100% focused on helping its healthcare provider customers, as well as global initiatives, fight the spread of COVID-19. At AWS, we’re grateful to be a small part of the incredible work that Arcadia and others are doing.

More to come

To see how your organization can, like Arcadia, save money while driving innovation and improving performance, please continue to join me as I highlight companies on their modernization journeys onto AWS. AWS can help you assess how your company can get the most out of cloud. Join the millions of AWS customers that trust us to migrate and modernize their most important applications in the cloud. To learn more on modernizing Windows Server or SQL Server, visit Windows on AWS. Contact us to start your migration journey today.

Matt Asay

Matt Asay

Matt Asay (pronounced "Ay-see") has been involved in open source and all that it enables (cloud, machine learning, data infrastructure, mobile, etc.) for nearly two decades, working for a variety of open source companies and writing regularly for InfoWorld and TechRepublic. You can follow him on Twitter (@mjasay).