October 8, 2012

Seeking Clinical Decision-Support Standards
By Susan Chapman
For The Record
Vol. 24 No. 18 P. 14

The Health eDecision Project hopes to establish a standard format for CDS interventions to make it easier for EHR vendors to build the technology into their systems.

The Office of the National Coordinator (ONC) for Health Information Technology defines clinical decision support (CDS) as providing “clinicians, staff, patients or other individuals with knowledge and person-specific information, intelligently filtered or presented at appropriate times, to enhance health and health care.” While CDS can be paper based, as in the case of paper order sets and documentation forms, the process is becoming increasingly electronic.

Regardless of form, CDS offers guidance for the practice of medicine in areas such as diagnosis, treatment, and information exchange. The ONC views CDS as being comprised of “computerized alerts and reminders to care providers and patients; clinical guidelines; condition-specific order sets; focused patient data reports and summaries; documentation templates; diagnostic support, and contextually relevant reference information.”

CDS has been lauded by some as a game changer, a key component of the burgeoning HIT scene. “Stage 1 of meaningful use imposed requirements for computerized physician order entry [CPOE] on healthcare institutions. CPOE can be inefficient, so to increase efficiency and adoption, hospitals created order sets and built them into their CPOE systems,” says Stephen Claypool, MD, vice president of clinical development and informatics at Wolters Kluwer Health/ProVation Medical and a member of the Health eDecisions Project.

The concept of CDS extends beyond patient-physician interaction, Claypool says. “Even when a physician is not with a patient, CDS plays a role,” he says. “Care plans, for example. Everyday things that should be done by both physicians and nurses for the patient are part of the EHR and provide guidelines for care all the way to discharge. In this way, all healthcare systems need CDS and, while the process is not yet dynamic—where the physician can enter data that then prompts a list of questions—the potential for a robust system is there.”

One of the larger difficulties with implementing CDS within EHR systems, Claypool says, is that it requires facilities to employ bioinformaticists, clinical and technical experts who help build, implement, and share CDS across an organization. Most healthcare facilities do not have the expertise to decide which CDS to implement, create and maintain the clinical content behind the technology, understand how to utilize it effectively in clinician workflow, and technically implement the system.

This challenge is further complicated by the fact that each EHR system has different mechanisms for implementing CDS because there are no widely adopted standards. Establishing such standards would make it easier for facilities to purchase, share, and implement CDS, Claypool notes.

Creating Structure
To make CDS interventions more timely as well as readily capable of sharing, implementing, and using, the Health eDecision Project, a collaborative working group composed of clinicians, academics, and vendors and supported by the Standards & Interoperability Framework, is developing standards so that such interventions are available either via the Web or CDS systems.

The project aims to identify, establish, and coordinate standards that facilitate systems and services to allow CDS interventions to be implemented and shared. Its target outcomes include alignment with meaningful use, the creation of catalogues and repositories, direct interaction between CDS and EHR technology, widespread adoption and use, and the development of additional CDS interventions.

Finding practical CDS implementations was quite the challenge, says Jonathan Teich, MD, chief medical informatics officer for the health sciences division of Elsevier, who participated in the CDS National Roadmap, which preceded the Health eDecision Project, and several other projects that advanced the cause of shared, reusable CDS. “There were obstacles to translating guidelines into actionable computer logic,” Teich says when describing the success of the various projects. “We have to be able to build an intervention that is usable within an EHR.”

Though CDS interventions are a meaningful use requirement, small providers do not have the resources to do their own informatics. As a result, they could benefit from examples supplied by those organizations with the wherewithal to do so. Additionally, EHR vendors are reluctant to move forward to support shareable CDS if there are no standards to follow.

“We have to be able to have well-specified knowledge, the right rules, and proper implementation,” Teich says. “We need something practical and acceptable. People don’t want to reinvent the same thing numerous times. It’s expensive and time consuming. Standards are very important.”

Teich provides an illustration of the importance of being able to share information. “Say I’m a physician working in one hospital for which diabetes care improvement is important,” he says. “We may make a number of order sets and care plans. I want to share that information on overall treatment. The mechanism of exchange may be CDS repositories, which may be among the things that CDS producers can provide. In this scenario, a second hospital can go to a repository and download the CDS interventions that we created directly into their EHR.”

Such a process also would benefit clinical colleges, healthcare professionals, and other quality-focused organizations that generate CDS items. For example, the American College of Cardiology, which produces clinical knowledge and guidelines, could store its information in a CDS cloud- or Web-based repository for later use by physicians, nurses, pharmacists, and informaticists.

Doug Fridsma, MD, chief science officer and director of the office of science and technology at the ONC, describes the Health eDecision Project as a continuation of the work that has been part of the ONC for several years, including that of the CDS National Roadmap project. This latest effort builds on those blueprints to determine the next steps, with approximately 200 individuals taking part, more than 70 of whom participate in weekly teleconferences.

“We try to define what success is and present it to the community,” Fridsma says. “The community then helps refine it. We want vendors and users to be able to download a CDS intervention into the EHR. We think that would help with uptake of these interventions. In this way, we’re sharing artifacts. We also may want a more interactive system that allows healthcare professionals to query a service, which would send back recommendations based on best practices.”

Fridsma says the Health eDecision Project team operates under two guiding principles: to go beyond a one-size-fits-all solution so that users and vendors will have numerous tools and to harbor no expectations that it will resolve every issue.

“We don’t expect to solve all of the world’s problems,” Fridsma says. “Instead, we’re undertaking an incremental process that is demonstrable, actionable, and testable. We define what we want to work on and what we don’t. We hope to figure out how to structure medical knowledge that is shareable and executable.”

Once the team understands those factors, it agrees on a plan to move forward. “We operate on the path of least regret,” Fridsma says. “We take steps that build upon one another. We recognize that the pieces are there and solve parts of the puzzle incrementally. It’s self-sustaining; we get on the right path and enthusiasm takes over.”

The ultimate goal of improved CDS interventions is to provide the best patient care possible. “Meaningful use is being driven by better healthcare for the nation, and we can use CDS and EHR for better quality of care,” says Jacob Reider, MD, the ONC’s acting chief medical officer. “Stage 1 meaningful use requires a better measure of quality of care. Stage 2 is also a better measurement of healthcare. CDS represents tools with which we can empower the medical community to get As—to be extraordinary for all patients. This project is about building the train tracks for a robust, scalable delivery system for those to measure and improve quality of care. It’s what gets me up in the morning, and I know it’s what motivates everyone involved in this project.”

Challenges Remain
Claypool believes one of the project’s largest hurdles is gaining consensus among its members on what the standards should be. What’s more, once the project succeeds in establishing standards, there will remain several “real-world” barriers that impact hospital efforts to integrate CDS into their systems and clinical workflows.

 “Healthcare facilities find it difficult to implement CDS in their EHR systems. When they finally get it running, they have to keep it up-to-date with changes in medicine. Maintenance is a big issue,” Claypool says. “Furthermore, without physician input, CDS is often not implemented properly and physicians fight back, which is problematic for the system.”

Enabling healthcare systems to be ready to utilize shared CDS or submit patient information to a cloud-based CDS system poses yet another barrier. Claypool points out that it takes a great deal of time and money to get a facility up to speed. “There needs to be an effort to map the content in the EHR to the content in the clinical decision-support intervention,” he says. “Basically, each healthcare facility must have information mapped to an interoperability standard that is used in the clinical decision-support intervention. That content mapping effort is time consuming. Hopefully, stage 3 meaningful use will force facilities to map their data to standards, which will make the sharing of CDS a little easier.”

However, creating and agreeing on standards won’t necessarily solve everything. “Even if there is a standard in place, decision makers still have to choose which CDS to use based on what works in each facility,” Claypool notes.

Fridsma sees the main challenge as having too large a pool of standards from which to choose. “But that can also be an opportunity,” he notes. “We want to reuse things that we’ve defined. We want to identify holes. As we develop artifacts, we need tools, content management, ways to update. All of that may be out of scope for the first iteration, but we look to them for the future.”

Moving Forward With CDS Interventions
CDS is a sophisticated component of HIT, one that requires computable biomedical knowledge, person-specific data, and a reasoning mechanism that combines knowledge and data to generate and present helpful information to clinicians as care is being delivered. The information must be filtered, organized, and presented in a way that supports a facility’s current workflow, allowing the user to make an informed decision quickly and take action.

Fridsma maintains that HIT designed to improve clinical decision making is particularly attractive for its ability to address the growing information overload clinicians face and provide a platform for integrating evidence-based knowledge into care delivery. Although the majority of CDS applications operate as components of comprehensive EHR systems, stand-alone CDS systems are also in place.

To become more attractive to healthcare organizations, CDS must have established standards to help drive down costs. “Right now, it’s like the early days of cell phone service. Not one provider covers all areas so we all have to carry five different phones,” Fridsma says. “CDS is a critical part of the puzzle to improve healthcare. If we can lower the cost for EHR to use CDS and if we do it in a standardized way, developers will have a much broader marketplace not just a specific area of focus.”

Fridsma likens the Health eDecision Project to the creation of the Internet, which standardized the representation of information. Like the Health eDecision Project members, those who created the Web established the basic standards to support it. “Their work eventually enabled people to do things like buy Christmas gifts online,” Fridsma says. “Like that, there may be benefits that we have yet to envision. If we can agree on standards, then we will see benefits that we haven’t even anticipated.”

— Susan Chapman is a Los Angeles-based writer.