The coronavirus pandemic has inflamed the sore spots of the global healthcare industry. Namely, it has revealed issues with data sharing and interoperability. The global health community needs structures enabling health IT systems to gather and share information about how the novel coronavirus is transmitted, its symptoms, and testing results. The spread of COVID-19 has also brought multiple privacy, data protection, security, and compliance challenges, all of which rest on healthcare providers’ shoulders.
But COVID-19 hasn’t canceled other serious medical conditions and the necessity of their treatment. Let’s consider how cancer treatment is going in the UK, ranked sixth among countries with the most well-developed healthcare systems in the 2020 Best Countries report by U.S. News & World Report.
Healthcare data sharing issues related to cancer treatment in the UK
In the UK, cancer patients must go to different facilities for diagnostic tests, imaging, and examinations. But clinicians need to access patient data from all these facilities in one place to review test results and decide on an approach to treatment: surgery, chemotherapy, radiation therapy, or sometimes all three. Local Area Teams (LAT) under the National Health Service (NHS) commonly manage this process and organize treatment. But transferring all patient data from an NHS hospital to a private provider turns out to be a challenge.
Patient records, lab and imaging results, and radiotherapy scans are stored on different incompatible systems. This incompatibility has led to a mass of ad hoc solutions and the systemless transfer of paper documents between the NHS and private clinics — alongside the possibility of confusion and delays. Rutherford Health, a UK network of oncology centers, came up with a solution to their internal and external data sharing problems.
Networked multi-center system to tackle interoperability
Understanding the high risk of equipment failure, the Rutherford Health network implemented an advanced networked multi-center proton beam therapy system that’s capable of seamlessly sharing patient information within and beyond the network’s centers.
When individuals receive radiotherapy in a Rutherford Health center, a team of healthcare specialists can examine scans that show the cancer and the area around it to plan treatment. These scans are stored centrally in a secure private cloud. And in case the equipment fails in one center, a patient can quickly be moved to another center with properly functioning equipment. A networked system helps the treatment team immediately access and use patient information to keep treatments on schedule.
This system also allows Rutherford Health to get treatment plans from the NHS and quickly provide those plans to their centers. Additionally, Rutherford Health has signed service-level agreements with a number of private UK healthcare providers to ensure the seamless transfer of radiotherapy information.
Many clinics as well as life sciences, health insurance, telehealth, and remote patient monitoring organizations can also enjoy the benefits of integrated healthcare systems — especially if they’re building these systems from scratch, without dependencies on legacy systems.
Although legacy systems are expensive to maintain and hard to connect with other systems, they’re difficult to replace due to the amount of essential data they hold and the multiple systems they support. But with modern technologies, it’s also possible to improve legacy systems.
Later, we’ll tell you how you can ensure integration with all the systems you need to integrate with. But let’s first pay attention to key problems within healthcare organizations that an effective healthcare integration system might tackle.
What healthcare provider pains might a healthcare integration system heal?
According to ResearchAndMarkets, interoperability issues are a big restraint to providing better customer service in clinical settings. One of these issues is the inability of systems to share data and correctly interpret received data. But there are also other pains healthcare providers need to heal.
Data integration challenges. Medical organizations are failing to integrate data platforms into their software infrastructures, and data integration problems cost healthcare providers money. Operating costs are high because of low operational efficiency. Each system you need to integrate with might store data differently, resulting in difficulties with translating and consolidating data. And the more healthcare software solutions an organization deploys, the more difficult it is to clean up the data mess.
Poor customer experience. Consumers are paying more and more for their medical bills. Meanwhile, they’re used to the top-notch web and mobile experiences provided by companies in the financial and e-commerce industries. No wonder they expect the same experience from healthcare providers. Offering this kind of experience means allowing healthcare consumers to access all their medical data via a central software solution that’s context-aware and doesn’t repeatedly ask the same questions.
Burden of administrative tasks. Healthcare specialists are burdened by time-consuming administrative work and the necessity to use inefficient applications. Physicians and nurses spend time filling out medical forms instead of taking care of patients.
Issues with data security. Medical institutions deal with Protected Health Information (PHI), and US healthcare providers have to comply with strict requirements such as those imposed by HIPAA. Healthcare providers should also make sure they protect patient data well, as cybercriminals often target healthcare systems to get access to PHI.
Data flooding infrastructure. All of the data a medical facility deals with has to be securely stored, processed, and analyzed to improve patient treatments, medical outcomes, and financial results. If not, there’s a high possibility of errors and missing details. Medical errors are dangerous, as they’re the leading cause of death in the US. Incomplete or missing imaging data and relying on memory are some of the causes.
Insufficient use of analytics. Medical institutions often can’t integrate analytics into their clinical and operational workflows. According to a recent survey by Black Book Research, eight in ten interviewed healthcare managers state that their use of advanced analytics for decision-making and strategic planning is negligible.
Lack of a competitive advantage. Competition between healthcare providers is fierce. To survive, a medical organization must build an effective management system. In the US, Medicare and insurance companies prefer to rely on electronic data exchange, which is possible only with an automated system. A clinic that’s able to send and obtain patient data and medical reports electronically always wins over competitors.
Lack of process automation. Without one unified system that securely integrates data, the ability to automate processes and effectively use healthcare information is extremely limited. Process automation reduces operational costs. Conversely, if you don’t automate your processes, your operational costs increase.
The solution to all these problems might be an automated hospital integration system tailored to business requirements. Such a system should integrate with all services of providers, payers, vendors, and other entities as well as with devices to incorporate data into one user interface.
But there are some pitfalls on the road to building such a system.
Healthcare integration implementation challenges
Let’s see what the obstacles are to creating an efficient healthcare integration system and figure out how to tackle them.
Prevalence of legacy systems. Healthcare organizations keep using legacy systems, as they’re expensive and difficult to replace. However, these systems often prove ineffective when it comes to processing huge volumes of data and can’t interoperate effectively with new systems.
Growing volumes of unstructured data. Medical organizations must manage numerous data formats. In addition, all clinical data systems such as EMRs have their own specific data schemas. As a result, it’s likely that data points won’t map properly across systems. Historical digital healthcare data also needs to be processed and transformed to lay a foundation for further analysis.
Need for top-notch security. Health systems require a safe and secure authorization mechanism to exchange data. EHRs are extremely lucrative for criminals to capture. A social security number may cost 10 cents on the black market. But a medical record may cost up to $1,000, which is 10,000 times the value.
Insufficient software adoption by end users. Medical personnel have to focus on situations that require prompt medical decisions rather than on dealing with confusing and complicated software. If a system is too complicated, users need time-consuming workshops on how to use it. If users are merely provided with training materials, the probability of errors remains high.
How can you solve these challenges and what do you need to build a scalable, flexible, and expandable system integration platform for healthcare?
Approaches that have proven effective for building healthcare integration systems
There are a huge number of ready-made healthcare integration systems, but the two most common types are healthcare enterprise service bus (ESB) tools and integration Platforms as a Service (iPaaS).
Enterprise service bus implements an on-premises software architecture model. The idea of the ESB architecture is that a technical team integrates different apps around a communication bus and then allows those apps to talk to the bus. This approach divides systems from each other, letting them communicate without being dependent on or having knowledge of other systems. ESB products like those from Oracle, WSO2, and IBM enable developers to build bus-like architectures but vary in their implementations and capabilities.
Integration Platform as a Service solutions provide developers with a set of integration tools and require no on-premises hardware or software. iPaaS solutions also ensure real-time access to data stored in multiple apps and locations, which is great for real-time analytics and collecting data from different apps and devices. There are many third-party platforms that provide API management, in-box connectors to promptly connect to virtually any healthcare app, integrations with the top EMR/EHR systems, and data mapping to quickly transform data formats. At Yalantis, we design systems that can easily integrate with lots of third-party platforms.
As you can see, both ESB and iPaaS healthcare integration engines have their strengths. But they also have common disadvantages.
Dependence on a specific technology
Picking a ready-made service, whether it be WSO2 or MuleSoft, makes you dependent on one technology. For example, if you choose WSO2, you’ll be able to write your code only using Java.
Moreover, each integration service takes developers much time to learn, and you might have trouble finding the specialists you need in the long run.
Relatively high and unclear total cost
According to IT Qlick, MuleSoft’s basic license starts from $80,000 per year. But to define the exact cost you’ll pay, you should consider various factors including customizations, data migration, training, support, and upgrades. Most likely, you won’t be able to calculate the total cost due to lots of unpredictable factors that may arise once you’ve started adopting the software.
Moreover, each integration platform charges specific fees and sets specific limitations. MuleSoft, for example, allows you to integrate only ten apps for one vCore (a unit of computing capacity for processing on CloudHub, which is equal to one virtual core). Dell Boomi, in turn, charges for connectors.
The ESB approach was introduced in 2002 as the first step towards splitting up the traditional monolithic architecture into small, functionally independent modules.
But the modern IT world is now quickly shifting towards building software architectures based on microservices. According to a 2020 report by Global Industry Analysts, Inc., the global market for cloud microservices is projected to rise from $810 million in 2020 to $2.8 billion by 2027, growing at a CAGR of 19.2 percent over this seven-year period.
What’s the reason for such popularity and how can a microservices-based architecture help to create an integrated healthcare system?
Read also: Testing and Quality Assurance at Yalantis
A microservice architecture is the most flexible way of creating highly integrated healthcare systems
With a custom microservice architecture, a software system is split into a set of services. You can consider microservices as functions or modules that comprise your software solution. Examples of microservices include an audit trail, user authentication, and an email subscription manager.
Many leading brands unrelated to healthcare including Netflix, eBay, and Twitter have long been benefiting from the use of microservices to create and maintain complex systems. Let’s see what benefits the microservice approach promises for building highly integrated healthcare projects.
Make changes easily and lower development costs. Microservices enable code reusability. This reduces the cost of mobile and web development or changing software, ensures the efficient use of resources, and makes it easy to scale software. With a microservices approach, developers make each service as small as possible, ensuring it performs just one business function so they can easily implement any changes or add new features to the system. Developers make sure each microservice is an autonomous part of the entire system. This makes each part of the system less dependent on others, lowering the maintenance cost.
Avoid dependence on a specific technology. A microservice architecture allows different technologies to communicate with each other in a unified system. This enables developers to use the most appropriate languages, frameworks, libraries, and other tools for each part of the system. Also, thanks to the small size of services, developers can easily rewrite them using new and more effective technologies. With such ongoing improvement, your system won’t quickly become outdated.
Ensure a secure and sustainable system. In a properly built microservice architecture, each software unit is isolated. If one unit fails, it doesn’t affect the entire system. Additionally, the use of encryption mechanisms and authentication and authorization tools provides security for data at rest and in transit.
Communicate with legacy systems. With the right approach, developers can successfully migrate data from or integrate with a legacy system. If you just need to migrate data, you can create a utility responsible for data migration and install it on a legacy system. At Yalantis, we have an established process that helps us effectively migrate data. If we need to integrate with a legacy system to ensure further data exchange, we integrate with the system via an API.
To receive all these benefits, we need to enable our microservices to communicate and exchange data. But how can we ensure they do this? There are two common approaches.
Microservice orchestration and microservice choreography
Let’s take a look at two approaches to implementing microservices in healthcare and check their definitions.
Microservice orchestration. Just as a conductor guides musicians in an orchestra, developers implement one central controller that directs all interactions between microservices. The central controller transmits all events and responses directly to microservices.
Microservice choreography. As a dancer in a dance crew is only worried about their own routine, each microservice functions independently without any need for instructions. This approach ensures a decentralized way of broadcasting events. The services interested in those events use them to act accordingly.
The orchestration architecture is easy to manage, as developers centralize all business processes. On the other hand, this architecture allows for dependencies due to coupled services, meaning that if service A fails, service B will never be called.
The choreography approach prevents dependencies between services, as each service independently knows how to react to events. But in this case, business processes are scattered, making it hard to manage the entire process. Given this, how can you make the right choice? The wisest approach is to use a combination of orchestration and choreography.
Hybrid approach as the golden mean
At Yalantis, we’re adherents to the hybrid approach, as it lets us have the best of both worlds by mixing the orchestration and choreography approaches on a case by case basis.
For example, we can follow the choreography approach for implementing services responsible for user authorization or sending emails. As a result, these services will be able to retrieve events from an event broker with no need to know who produced those events.
On the other hand, the choreography approach is ineffective for real-time data processing when we need to ensure direct interactions between services with no intermediary. That’s when we can benefit from the orchestration approach to enable smooth and swift data processing.
By ensuring a hybrid microservices development, we can create long-lasting, scalable custom solutions that are technology-agnostic.
Implementing highly integrated healthcare systems requires a particular developer skill set and experience to be able to use existing best practices. We suggest you work with a mature outsourcing software integration provider with relevant experience in native app development. Such technology vendors have health system integration engineers experienced in creating customized solutions and can efficiently build the healthcare integration ecosystem you need.
Moreover, a dedicated team will make sure your software system is user-friendly and help your staff and partners adapt to it. This includes providing continuous support, gathering end user feedback, making improvements to the system, and training employees on how to work in the new environment.