How to Implement DevOps in Healthcare
Greg Sienkiewicz | July 16, 2019 | 13 Min Read
Implementing DevOps can enable your healthcare organization to deliver new digital solutions and services at a faster rate and higher quality, but where do you start and how do you achieve success with implementation? We walk you through the benefits of implementing DevOps in healthcare and describe what our own journey looked like building a DevOps practice, including our strategy, lessons learned and key considerations.
Many healthcare organizations have recently come to realize the potential business impact of establishing a DevOps (Development and Operations) function as they begin to launch data-driven initiatives and adopt innovative technologies to provide better care to patients while containing costs. With healthcare companies working to adopt similar technologies that will eventually become the “norm”, they are looking to DevOps to innovate and stay competitive.
We’ve seen increased DevOps implementation in industries such as Financial Services, Industrial/Manufacturing and Retail/Consumer industries. Yet Healthcare, an industry with great cause and incentive for implementing DevOps, lags in adoption. Many healthcare companies are unsure of where to start, what’s involved, how much it will cost and how to achieve success.
To help answer some of these questions, we’ll lay out the benefits of implementing DevOps in healthcare, walk you through our own journey establishing a DevOps function at Macadamian in order to better serve our healthcare clients and discuss how you can get started implementing DevOps.
The Benefits of DevOps in Healthcare
If you’re not familiar, a DevOps unit leverages a set of cultural philosophies, practices and tools to facilitate closer collaboration between development and IT teams within an organization. This is done to improve various aspects of software product development and delivery such as time to market, testing automation, solution scalability, security, user satisfaction and competitive advantage. As mentioned earlier, DevOps can be applied to various industries, but here are a few core benefits of implementing DevOps in healthcare.
Increases Patient Engagement and Satisfaction
By practicing Continuous Integration (CI) and Continuous Delivery (CD), and leveraging a microservices architecture, solutions and/or features desired most by patients are able to be deployed in a timely fashion. Solutions like a new check-in system to facilitate faster and more convenient processing of inpatients (when they are most likely unwell and are experiencing feelings of anxiety, sadness, pain, etc). It could be implementing a new feature on a mobile patient portal that allows patients to be notified immediately when test results are in, or perhaps a solution that monitors appointment availability with specialists and automatically backfills them from a waiting list when cancellations occur.
When healthcare organizations are able to deploy more technology that either improves operational efficiency or directly addresses a patient need, and they’re able to do so at a faster rate without compromising patient security, the quality of care that patients receive increases as does patient satisfaction and engagement.
Strengthens Competitive Advantage
We all know time to market is a challenge for healthcare innovation. This is largely due to regulation and high-stakes security measures that must be followed and maintained. Additionally, value-based care models have encouraged investment in new technologies only if they can demonstrate their impact on care quality and patient outcomes. When DevOps is implemented, automated infrastructure management and testing allow the software development cycle to be shortened; bugs are addressed and updates are deployed more quickly. Leveraging a microservices architecture also enables flexibility and efficiency to deliver healthcare solutions faster. Decreased time to market is a major advantage as healthcare organizations begin to adopt similar tools and technologies in their digital initiatives and need to differentiate from competitors through some other means. Companies implementing a DevOps model can improve their service provided to patients by releasing new and improved solutions in significantly less time than before – and before the competition is able to do so.
Facilitates Regulatory Compliance & Information Security
One of the core practices of the DevOps model is implementing infrastructure as code. This involves using configuration management techniques and implementing environment and resource creation as code, which allows organizations to automate configuration changes by making them repeatable and standardized. Implementing policy as code allows teams to establish and monitor compliance (ie HIPAA compliance) at scale while also automatically identifying compliance risks.
Supports Data Science Initiatives
Data scientists utilize a variety of languages, toolkits and frameworks for the development of machine learning models. Execution of machine learning models for deep learning requires scalable compute infrastructure, operating on specialized virtual machines equipped with specific hardware such as GPUs and machine learning software dependencies. Provisioning, deploying, managing and scaling this type of compute infrastructure is a technically involved undertaking. This is where the DevOps function can provide support to data science teams by ensuring data is constantly up to date and accurate for utilization. This is especially advantageous and critical in healthcare when your data science initiatives use data that could yield an outcome that directly impacts patients. For instance, using DevOps to automate extract, transform, load (ETL) of patient data into databases so that it’s of high quality and accuracy for machine learning algorithms that analyze medical images to diagnose patients.
Our Journey Implementing DevOps
Our journey developing the DevOps expertise, in retrospect, has been a multi-year iterative business process change initiative, which was a much larger undertaking than we originally envisioned. From the onset, we started from a strong position in that we already had a user centered design (UCD) practice led by our User Experience (UX) team which functioned in lock-step with the project management and software engineering teams. This allowed us to operate in a cross-functional, collaborative fashion as an Agile development organization.
Our Motivation for Moving to a DevOps Model
Our strong, cohesive, design-driven and agile software delivery model functioned wonderfully for projects, especially those for independent software vendor (ISV) customers. In such cases, the customer’s teams operationalized and deployed the end solutions, and as partners, our contribution was the delivery of the source code package. The model of solutions delivery shifted however with the advent of “cloud native” software, in which the deployments do not happen infrequently, but rather as needed in response to business needs as a turnkey deployed deliverable.
This shift to the cloud motivated us from a business perspective to pivot towards DevOps in order to better serve our clients, including healthcare software vendors and also medical device companies looking to connect to the cloud and integrate collected patient data with provider systems. To help paint a better picture of what’s involved in establishing an organizational DevOps function, here’s what our journey looked like, including our strategy, lessons learned and key considerations.
Phase 1: Try, Fail, Learn
Our first attempt at incorporating DevOps into our delivery model failed spectacularly. The focus of the initiative was technology and tools, which resulted in philosophical disagreements and fragmented practices. Instead of improving the way that we worked and delivered projects, the teams exerted effort to prove that their preferred toolset or technology stack was the right solution. As a result, a biased view was developed and emphasis was placed on how we could make our processes fit the tools, as opposed to how we could leverage the tools to enable a more efficient delivery methodology and support our organizational mission of helping make tons of people healthier. Ultimately, the outcome was frustration and very little added value for our customers.
Meanwhile, the need for the delivery of operational, deployed solutions and cloud adoption continued to accelerate. With this realization, we regrouped and went back to our core strengths as a user centered and agile organization. We dove in and performed a retrospective that resulted in a laundry list of lessons learned. We refocused on the first principle of customer centricity and mapped value streams via a journey map of the project delivery process. Some of the key lessons learned as a result of this retrospective were:
- Developers should participate in the operational tasks
- Security and compliance requirements should be factored into the process at an early stage
- The current software architecture might not fit a DevOps enabled software development practice
- DevOps tools should be an aid, not a crutch
- Scrum masters should participate in the process as the agile project lead
Phase 2: Establish a Tiger Team
Armed and reinvigorated with newfound insights, we refocused on how to overcome the biggest hurdle: the cultural change surrounding DevOps. After many discussions, we opted for a new approach. We cultivated a “tiger team” of talent from various teams with representatives from each craft and different physical office locations, such as to align the working group with our model for the global delivery practice. They were then strategically placed on a greenfield project and challenged to define the best practices for delivering a complex software system while utilizing DevOps. Without the pressures of large scale thinking, some of our smartest people focused on designing how they would want to work and how they believed a project should be delivered. Ultimately, they designed for a single project, but the end result was a model for a DevOps practice that could then be applied to any future projects.
Phase 3: Get Team Buy-In & Scale
As the small team disbanded and as individuals moved onto new projects, they took with them this model of the DevOps methodology and the processes that they had designed and proven to be effective as part of the “tiger team”. With each new project, the new way of working was organically incorporated into the delivery practice and these team members became evangelists for this new software delivery methodology. We believe that one of the key reasons why we were able to successfully integrate DevOps culture and our redesigned processes was because they were established by the working team themselves. As opposed to receiving drastic change from upper management who aren’t always “in the field”, our system for implementing DevOps was built and vetted by colleagues who were hyper-aware of how the team works on a day-to-day basis. This undoubtedly contributed to adoption amongst the larger organization.
Kaiser Permanente’s digital technology team also found success in implementing DevOps by deploying a similar “tiger team” approach in their journey modernizing the company’s software engineering processes. Kaiser’s VP of Digital Presence Technology gathered a separate team to build and grow their own DevOps practices to align their patient care delivery model to customer journey maps. Then they enlisted a few “early adopter” squads to integrate the processes amongst other teams. In conjunction with leveraging DevOps practices to accelerate solution turnaround, staff with various skill sets from different departments also engaged in human-centered design practices to help develop Kaiser’s digital services. Through the close collaboration of IT and other teams, “a deep mutual appreciation between business and IT” was formed. This allowed Kaiser to build 10 new technology services for its clinics in 10 months and deploy 82 production releases in a year. The result? In 2018, over 6.5 million Kaiser members were registered to receive self-care online and over 60% of patient encounters occur via their website or mobile apps. Patients filled 28 million prescriptions online and viewed over 50 million lab results online without having to visit a lab or a clinician’s office.
Phase 4: Establish a DevOps CoP for Continuous Improvement
As momentum grew, a community of practice (CoP) formed via which people had a venue to contribute and enact the practice of continuous improvement to the DevOps practice at Macadamian. This forum also allowed for people to ask questions and increase their understanding of how DevOps works within the company. This CoP remains today and refinement of our DevOps function is continuous.
In terms of the cost of DevOps implementation, our most significant expense was time – the time it took to identify best practices that work for our software delivery team and the time it took to disperse knowledge and put into practice what our tiger team had established across teams. This investment in change management is significant upfront, but didn’t increase as the DevOps practices were scaled.
At the end of this nearly four-year arduous journey, DevOps is now part of the Engineering service model and operational culture at Macadamian. Our DevOps experts work as part of the Delivery teams, embedded throughout the software development life cycle (SDLC). In this role, the DevOps practitioners support the delivery model via design and maintenance of the continuous integration and continuous deployment (CICD) tooling and workflows, enabling the team to accelerate the development and deployment of healthcare solutions for our valued customers.
Considering A DevOps Partner for Your Digital Health Transformation
As you can see, establishing a DevOps team internally and successfully implementing its practices to yield a positive business impact takes a lot of time, effort, as well as trial and error. Although the “fail fast, recover faster” motto goes hand in hand with DevOps practices, healthcare is a life-impacting, high-risk industry. Many times the technology being developed/managed directly impacts people’s health and lives. With this being the case, healthcare organizations can’t afford to experience frequent errors. Moving cautiously and slowly so as to not pose major safety or data security risks to patients could mean that it may take even longer to build out a DevOps function internally.
Many organizations also lack the necessary resources to make such a drastic organizational shift. For one, to implement DevOps, organizations need to invest in certain tools such as configuration management tools and CD (Continuous Delivery) platforms. But aside from tech investments, a DevOps team requires people with a specific set of skills and expertise which either needs to be developed internally or brought in. It’s important that these people understand the nuances of the healthcare industry, be familiar with the DevOps culture and also have the ability to infiltrate that culture within your organization. Although certain tools are required to execute a successful DevOps initiative, the larger barrier to success is more people and culture-oriented – a lesson we learned. Gartner has predicted that through 2022, 75% of DevOps initiatives will fail to meet expectations due to issues around organizational learning and change. George Spafford, Senior Director Analyst at Garner, says, “People-related factors tend to be the greatest challenges – not technology”.
The lack of resources, including time, energy, budget, talent and organizational knowledge and understanding of DevOps culture discourages teams from making the jump – even if they can recognize the potential positive business impact down the road.
Joining forces with a DevOps solutions partner with experience in healthcare can allow your organization to reap the benefits of DevOps, but without the same degree of implementation risk since the team in partnership would be familiar working in an agile environment and managing risk associated with developing software for the healthcare industry. Two examples of what a DevOps partnership could look like include:
1. A DevOps partner is consulted to help you build your own internal DevOps team and practice
A 2018 survey of those who had implemented or plan to implement DevOps into their organization showed that 52% were likely to invest in redesigning their processes to drive a DevOps approach over the next year as part of their implementation initiative and 55% would invest in more training for DevOps personnel. An experienced partner can evaluate your current technology systems and processes, and identify the changes that need to be made and the required investment in resources and tools (including talent and technology). They can then develop a tailored action plan with you outlining the steps to successful implementation including training and strategies for adopting DevOps culture across teams.
2. A DevOps partner is onboarded to act as your DevOps team and implement practices
Your team could onboard a DevOps partner as a team extension within your organization. This partnership would allow you to offload repetitive tasks that could be managed and supported by the oboarded DevOps team, leveraging their expertise and integrating with your organizational processes in a way that’s least disruptive so that your team can focus on delivering more value to patients. These tasks might include configuration checks, installing software updates, performing database backups, monitoring workloads, monitoring infrastructure health, etc.
From our experience building our own DevOps function at Macadamian, there are bumps and bruises along the way to successful implementation. It’s not a speedy journey; getting everyone on board and aligned from a process perspective and managing change can be quite an undertaking. However, it can be done – perhaps with the help of an experienced DevOps partner with a background in healthcare. The result will keep your healthcare organization competitive in a world where patients’ expectations of their healthcare experience continue to rise as other industries ride a fast-moving digital wave. With DevOps, healthcare IT teams have the capacity to improve both patient experience and outcomes through technology.
Download the Guide to Developing Engaging Digital Health Software
This 90+ page guide focuses on the intersection of your user, technology choices, and the regulatory environment you face that will play a critical role in the success of your application.
Health Information System Integration
In this webinar, we discuss interoperability in healthcare and answer attendee questions on Health Information System integration. Download the webinar Now.Read More
Guide to Creating Engaging Digital Health Software
This guide shares our knowledge and insights from years of designing and developing software for the healthcare space. Focusing on your user, choosing the right technology, and the regulatory environment you face will play a critical role in the success of your application.Read More
Accelerate Time To Market Using Rapid Prototyping
In this webinar, you will learn how to leverage rapid prototyping to accelerate your products time to market in one week, agile sprints.Read More