Only a handful of technologies have the power to drive the digital transformation of robust, traditional systems, allowing for fast pacing, like those that exist in healthcare.
This trailblazing technology relies on the most advanced services to offer flexibility, automatization, and the much-needed personalization in the medical landscape. It can be successfully applied to a number of different organizations – from start-ups to enterprises, from payers to providers and product companies.
Under the hood, serverless technologies rely on physical servers, but they are managed by Cloud providers such as Amazon, Microsoft, Google, etc. and not the company itself. This means that the process of maintaining and provisioning of infrastructure remains separated from the internal processes.
The separation of concerns is one of the key reasons many agile and fast-growing health tech solutions utilize serverless technologies. When relying on serverless architecture, companies can build and run applications without having to manage redundant and scalable infrastructure and the associated security.
This is especially important today as healthcare organizations extend applications across wider geographies and more complex relationships. Consequently, scalability, redundancy, cost savings, and speed are more critical for success.
Take the example of a patient portal application. In contrast to traditional architectures, when serverless technologies are employed, the portal can be managed with multiple services that scale based on utilization and demand. Furthermore, authentication can be leveraged by using managed authentication services, such as Cognito, while for the database, AWS Aurora Serverless can be utilized. The results are lower costs of building the portal, easier maintenance, and ongoing scalability in use.
As shown in Figure 1 this application utilizes AWS EC2, however, this can be achieved also by utilizing AWS Lambda instead of running applications in containers on EC2. By utilizing AWS Lambda, security concerns are passed to the Cloud, whereas with EC2 you need to take care of the security on the instance level.
AWS Cognito provides user management and authentication for the application.
As for the database in this application, we used AWS Aurora Serverless – MySQL compatible relational database – which can automatically scale compute capacity up or down. Classical managed database instance Amazon RDS is a deployed database server with selected instance type and size. RDS does offer auto-scaling for read replicas, however, the process can take a few minutes while Aurora Serverless performs the scaling process most commonly within 30 seconds. For unpredictable and variable workloads, such as for this particular application, it is preferable to use Aurora Serverless.
In Figure 2 we can see serverless architecture used for reading and processing of data utilizing Lambda functions.
In this article, we’ll discover what serverless architecture means for your organization and learn how to ask the critical questions that will set you up for success in the future of serverless application design.
Getting Started with Serverless
When talking to healthcare companies that are migrating or deploying their solutions in the cloud, we learned that they are often unsure about how to start the sometimes bumpy road of transformational journey.
First, companies weigh the resources and capabilities needed to optimize the cloud model’s solution. The idea of dismantling and transforming the current solution in a fast-paced environment raises concerns. Usually, this process requires heavy refactoring.
Secondly, companies are worried about achieving and maintaining compliance in the cloud – especially concerning HIPAA and FDA CFR Part 11 – while not being directly under the control of the on-premise physical servers.
Third, but not least, is cost-efficiency.
Addressing each of these concerns is indeed important. The med-tech solution development asks for patience and an in-depth understanding of the field.
This doesn’t mean that every one of these questions can’t be untangled, thus bringing a more simplified way of functioning to a company. So let’s get the facts out and start addressing the concerns one by one.
1. Improving Scalability and Speed of Delivery
Serverless technologies can bring many benefits when set right. Applications built on serverless infrastructure scale automatically with the growth of demand and usage. Also, in a matter of a few two-week sprints, infrastructure and functionality can be provisioned to a cloud environment — most commonly by utilizing an infrastructure as code paradigm, which stands for provisioning and managing infrastructure by descriptive code that enables version control, testing, and graceful rollbacks. Such tools are Terraform, Azure Resource Manager, AWS CloudFormation, Pulumi, etc.
Scalable serverless solutions can be used across the digital health landscape. For example, centers for Medicare & Medicaid Services use the AWS Cloud services for hosting an infrastructure that is capable of handling hundreds of thousands of simultaneous users.
For the companies that deal with large chunks of data, the same powerful AWS features can be used to build a data warehouse to efficiently collect, safely store, and cost-effectively process and analyze a large quantity of data.
2. Ensuring Compliance
Healthcare is getting on board the digitalization train through initiatives like interoperability — a goal that serverless cloud architectures are perfectly positioned to enable.
In cases like AWS, for example, the ultimate goal is to have scalable and redundant services while maintaining HIPAA and other regulatory compliance.
The AWS HIPAA program enables covered entities—and those business associates subject to the U.S. Health Insurance Portability and Accountability Act of 1996 (HIPAA)—to use the secure AWS environment to process, maintain, and store protected health information (PHI). Based on customer feedback, AWS is trying to add more services to the HIPAA program, including serverless technologies. –source
The complete list of HIPAA eligible services can be found on: https://aws.amazon.com/compliance/hipaa-eligible-services-reference/
Utilizing these services provides HIPAA compliance out of the box, significantly reducing effort in development. Consequently, this allows companies to focus on the functionality of their product rather than spending time and effort in designing and building HIPAA compliant infrastructure.
3. Optimizing Costs
Since there is no need to perform backend configuration, developers can release and improve products easily, reducing costs and increasing the speed to go live.
Many times the infrastructure is overly complex, and the monthly cost of running and maintenance is fixed – and unacceptably high.
When our DevOps at Vicert worked on the transformation of one of our customers’ digital health solutions, they migrated them to the cloud, which resulted in significant uptime and resilience improvement and a significant reduction in maintenance cost. The client increased their monthly service uptime to 99,9% while achieving more resilient and scalable architecture of the migrated application.
By deploying in the cloud and utilizing specialized serverless services (for example, like AWS Lambda), enhanced security and faster development are both possible – resulting in savings, faster and simpler deployment, higher reliability, and service uptime.
Is Serverless Right for Your Healthcare Goals?
While serverless technologies have countless applications for an organization like yours, it’s important to identify the best use case. Answer these few critical questions to ensure maximum benefits for the use case.
Do you have unpredictable, fluctuating workloads or frequent low usage?
Healthcare workloads fluctuate for many reasons, including shifting patient volumes, expanding into new and uncharted markets (think telehealth), potential future pandemics, or even something as simple as new projects where costs need to be controlled.
If you’re working with this type of uncertainty, serverless technology could be a good fit.
Are you processing large volumes of data distributed across disparate sources?
This situation is becoming increasingly common in healthcare, thanks to the rise of telehealth. For example, remote surgery requires sharing of hi-res images and video, often across long distances. Population health initiatives require organizations to process large volumes of data, both on the patient-facing side and in back-end processes.
By utilizing AWS AppFlow, AWS DataSync, and AWS Data Exchange, you could easily create an ingestion layer that is capable of ingesting data from various sources.
For the processing of data you can use AWS Glue and AWS Step functions as described in the “AWS serverless data analytics pipeline reference architecture” article:
AWS Glue ETL builds on top of Apache Spark and provides commonly used out-of-the-box data source connectors, data structures, and ETL transformations to validate, clean, transform, and flatten data stored in many open-source formats such as CSV, JSON, Parquet, and Avro. AWS Glue ETL also provides capabilities to incrementally process partitioned data.
Step Functions is a serverless engine that you can use to build and orchestrate scheduled or event-driven data processing workflows. You use Step Functions to build complex data processing pipelines that involve orchestrating steps implemented by using multiple AWS services such as AWS Glue, AWS Lambda, Amazon Elastic Container Service (Amazon ECS) containers, and more.
As for the consumption layer you can use Athena, which is an interactive SQL service that enables you to run complex SQL queries on terabytes of data stored in the AWS S3. Supported data formats are: CSV, JSON, XML Avro, Parquet, and ORC.
Are you building modular solutions that incorporate external services?
AI and machine learning (ML) are being touted as central to the modernization of healthcare. If you’re going down this path, serverless technology is something you should consider.
This is especially true if you need to deploy digital applications that will need to be consumed by downstream systems such as EMRs — revenue cycle applications — and visualization tools.
For business intelligence capabilities you can use Amazon QuickSight, which enables you to create interactive dashboards with automated ML insights such as forecasting, anomaly detection, and narrative highlighting.
To build and train ML algorithms you can use AWS SageMaker.
Are you in a process of enabling the transformation of care?
Care transformation is an umbrella term for initiatives that improve the health of populations, support the experience of both patients and providers, and do it in a cost-effective way.
These goals have led to considerable disruption in the healthcare space — changes that will increasingly rely on new approaches to technologies like the serverless cloud to empower aspirations, including:
- New models of care delivery
- Supporting patients as active participants in making care decisions
- “Hospitals without walls” initiatives
Are you a startup or exponentially growing?
As a startup, you want to move fast and develop basic functionality as early as possible.
However, you shouldn’t lose focus on longer-term goals such as maintainability and ease of use.
Serverless architecture is a good fit for start-ups because it modernizes the process of app development and speeds up go-to-market. With the utilization of managed services, the cost of go-to-market is significantly lower while having infinite scalability.
Conclusion
Serverless technology is positioned to address some of the most pressing challenges in the industry as healthcare companies are gaining the benefits of numerous technological advancements. Single points of failure that are a risk to service continuity can be eliminated by moving away from bare metal servers.
Updating, maintenance, and disaster recovery are handled by the cloud provider, which enables the company to have more time to care about the product itself, rather than having to use overhead for maintaining the infrastructure on its own.
Serverless technologies are more than just an advancement in app development. Serverless cloud computing provides improved utilization, freedom for the developer to focus on a specific function, enhanced security, infinite scalability, resilience, and shorter time to production.
About Milos Kowatschki
Milos Kowatschki is the Engagement Manager at Vicert, a health tech software development company based in San Francisco. He is responsible for managing client relationships and collaborating with multi-functional teams to achieve clients’ business goals.
Vicert deploys nimble and specialized teams with deep experience in IT and digital health. With more than 300 health tech solutions delivered over the past 20 years, Vicert’s expertise in software development and digital health allows Vicert to provide one of the best ROIs in the industry.
Vicert is a proud sponsor of Healthcare Scene
A very detailed plus briefly described article about the benefits and the implementation of serverless technologies for the healthcare sector. It shows that the trend for serverless technologies is growing.