Service-oriented architecture enables health care enterprises to provide sharable, usable data to their consumers.
Practicing medicine was much simpler for Gunsmoke’s Doc Adams, when all that a community provider needed was a black leather bag for tongue depressors and a stethoscope, along with a locked oak cabinet containing quinine, ether and a bottle of rotgut whiskey. The repository for virtually all of the health data relevant to patient care was Doc’s nearly faultless memory.
These days, the health care environment is significantly more complicated, comprised of a complex network of tools and technologies. Health information executives struggle to manage a growing portfolio of clinical and business systems, including patient administration, electronic medical records (EMRs), picture archiving and communications systems, radiology information systems, labs, pharmacy and medications, billing and coding, and many more.
It is not unusual, in fact, for independent delivery networks to support between 100 and 200 discreet computer systems.
As reliance on health information technology (HIT) has grown, so have the numbers of users—or consumers—of captured data. It is no exaggeration to state that virtually everyone affiliated with a health care enterprise—from C-level leadership to doctors and nurses to pharmacy technicians and security—relies to some degree on computer-driven data. At the same time, nearly every user and application needs something from another system. Health care organizations must invest significant effort, time and resources to provide this access—because, typically, the simple exchange of data elements is not enough.
Within this context, the need to share a wealth of information among consumers is growing exponentially, and has amplified the need for integration and interoperability. Health care organizations are beginning to investigate platforms that facilitate data-sharing, thus allowing information to be transformed into intelligence. The most promising approaches are built within a service-oriented architecture (SOA) framework that enables the enterprise to make data available, understandable and useable for a wide variety of consumers.
Disparate systems
From a technical standpoint, SOA is an approach to building applications with distinct units—also known as services—that can be distributed over a network, reused and combined to create new applications. This means an SOA-based platform facilitates interoperability among existing clinical or administrative systems, workflows and processes. Varying applications can consume data from each other, because functionality is agnostic to originating system architecture, format and structure. This is vital, since each of the many applications a health care organization typically deploys has been developed by a different vendor and has its own structure, developmental tools and underlying logic.
The benefits of operating within an SOA framework are wide-ranging. Services written once can be reused by any and all applications. Maintenance items and changes are automatically updated in all applications using the service. Ultimately, different applications can make use of a common set of services for potentially very different purposes.
Although they may not have recognized it as such, consumers today are familiar with the functionality SOA provides. Individuals who book travel online, for example, can access an independent rental car or hotel reservation system from the airline’s flight reservation system. Likewise, retail enterprises may use SOA to manage inventory-updating activities, consuming data across point-of-sale systems, hand-held applications, customer service systems and e-commerce portals.
Delivering transaction-level benefits
While developing its interoperability platform, dbMotion recognized the value SOA could bring to its solution. The SOA framework underlies individual systems and technologies, removing the challenge of data-sharing from the transaction level. It expands usability of data because it facilitates greater functionality than simply moving data from one location to another. The use of SOA as a foundation, enabled the company to develop applications and services that logically connect complex, heterogeneous and non-standard IT environments, creating an agile platform that can adjust to evolving needs.
An SOA domain provides specific benefits such as platform independence; security standard implementation for authentication, authorization and federation; reusability; agility; business process encapsulation; autonomy; and well-defined service boundaries.
For health care organizations, these benefits are particularly important. As reliance on a multitude of clinical systems has grown, HIT executives have struggled with the dilemma of leveraging the enterprise’s investment in technology while supporting an increased need for seamless data sharing.
In response, many executives have attempted to build a monolithic system, utilizing a single to provide a broad array of applications. This approach has distinct drawbacks, however, because it requires the organization to adopt a premature sunset strategy, decommissioning legacy systems that still add value to clinical and administrative operations. Likewise, it limits the technology options available to users, obstructing their ability to adopt best-of-breed systems that precisely meet the distinct requirements of varying departments, units and service lines. Monolithic systems also restrict an organization’s ability to adopt emerging technologies that may offer advanced functionality in the future.
Perhaps the most discouraging aspect of the monolithic approach, however, is that it does not completely solve the interoperability problem. Despite being developed by a single vendor, systems may not adequately support data sharing. Because each application may apply distinct rules and logic to the data gathered and stored, it may ultimately produce different “flavors” of the same data—flavors that may not be familiar or palatable to the receiving system.
Leveraging technology investment
An SOA framework allows the health care enterprise to implement a full complement of widely varied systems to optimize performance, quality and outcomes. All systems are empowered to share data through semantic interoperability. This allows systems to share information, but also enables them to truly understand and make use of incoming data while maintaining the original meaning of that data, even if the data has been created and stored in varying systems, formats and terminologies. This allows each system to not only view other categories of information, but to easily interpret, analyze, compare, rationalize and act upon the data.
Consider, for example, medication reconciliation. A health care organization wants to ensure that its computer system can identify drug interactions, medication and therapeutic duplications. Because information for this service resides in more than one system, reconciliation cannot be effectively accomplished without automatically and semantically referring to multiple systems.
The platform must aggregate data about medications from many sources, such as ambulatory and inpatient medical records. It must then decipher the meaning of incoming information, regardless of the originating source language or format. In other words, the receiving system must understand if the patient is receiving Lipitor (artovostatin) and Zocor (simvistatin), and that they belong to the same therapeutic group of statins. Then the provider must be notified of issues that need to be resolved, including possible overdose, medications suspended during inpatient admission and contraindications.
This level of functionality allows multiple clinical and administrative consumers across the enterprise to have meaningful access to a wealth of data that was previously unavailable to them. Consider the benefits to the following types of consumers:
- Providers have a unified, longitudinal medical record at the point of care, which provides a comprehensive history of diagnoses and therapeutic interventions to enhance the efficacy of care, minimize errors and improve patient safety. In addition, because data is at the provider’s fingertips, the organization eliminates time wasted tracking down missing information, or the ordering of duplicate tests and studies due to unavailable results.
- Quality officers are empowered to aggregate and analyze data across the enterprise more easily. This allows them to efficiently monitor quality initiatives, identify areas for improvement, generate reports for regulatory agencies and participate in pay-for-performance initiatives.
- Clinical researchers can automate the previously labor-intensive, time- consuming process of collecting, organizing and understanding data. This allows organizations to shift the focus of their resources and therefore move far more quickly, efficiently and effectively to the core research itself.
The SOA framework offers great advantages to HIT executives as well. The combination of SOA with standards and a rich set of semantic services enables interoperability products to not only deliver an integrated patient record for use by a clinical viewer, but also to serve as a platform to develop future solutions for new use cases and scenarios.
The flexibility and agility provided by SOA allows decision-makers to adjust HIT plans (e.g., when to replace existing systems, implement upgrades, etc.) without disrupting existing systems.
In other words, changes made to one aspect of the organization’s IT environment—such as adaptations to workflow or EMR processes—need not affect other aspects of the organization. As such, SOA speeds implementation, reduces costs and minimizes errors.
Too often in the past, health care leaders have been a step or two behind the curve, developing solutions to the problems of two or three years earlier. The flexibility and power of SOA, when used in the right way, empowers health care IT platforms and the organizations using them to address the challenges of today and the years to come.
Mr. Ofek is founder and CTO for dbMotion, a provider of health care information integration software. Learn more about SOA at www.dbmotion.com.
This article can be found online at Advance for Health Information Executives.