AHRQ Projects Developing Clinical Decision Support Tools for Shared Decision-Making About Chronic Pain Care
Two ongoing projects supported by the federal Agency for Healthcare Quality and Research are developing patient-facing and clinician-facing shareable clinical decision support (CDS) to improve chronic pain care and reduce opioid dependency. During a Jan. 15 webinar, project leaders discussed the challenges of implementing ambitious technical goals, including using FHIR and the SMART project’s CDS Hooks.
Since 2016 AHRQ has been supporting efforts to make CDS more shareable, standards-based, and publicly available. One component of the initiative, CDS Connect, focuses on prototype infrastructure for sharing CDS resources. In 2018 CDS Connect focused on chronic pain management, including developing an interoperable pain management summary or “dashboard,” that consolidated patient-specific, pain-related information normally scattered throughout an EHR into a single view for clinicians. The two projects supported by AHRQ build on the CDS Connect resources.
During the webinar, Barry Blumenfeld, M.D., of RTI International, described progress on the multi-organizational team effort he and the Patient-Centered CDS Learning Network are leading to develop shareable CDS to promote shared decision-making in primary care clinics—including data from state-level prescription drug monitoring programs (PDMP), patient-reported outcomes, and the capability to document in local electronic health records. RTI International is working with Vanderbilt University Medical Center and the University of Chicago to implement and evaluate CDS for both patients and clinicians in the area of chronic pain management.
The specific aims of the research are to create CDS that:
• Is interoperable and publicly shareable
• Meets the needs of both patients and clinicians through both patient-facing and clinician-facing channels and formats
• Has demonstrable impact by evaluating with appropriate measures and outcomes.
Bloomfield described the system under development: The patient-facing side is called MyPAIN for Chronic Pain, which includes a patient-reported outcome measure (PROM)-based pain assessment. Through the patient portal, a patient would receive an invitation to open an application and engage in a pain assessment, educational materials for treatment options, information on opioids, and a pre-visit assessment questionnaire that may include questions about social determinants. This information is melded with the clinician-facing dashboard called Pain Manager to be used in shared decision-making sessions. The clinician can see pain assessments and social determinants information as well data available via interface to the state PDMP program.
“One part of the system is facing the patient and the other the physician, but they are tightly coupled,” Bloomfield said. The two-year grant program faces many challenges, he added. Working with the Vanderbilt University Medical Center and the University of Chicago, he said, it has been difficult to come up with a decision tree rubric to decide what content to incorporate across two sites. “Both are running Epic EHRs, but each will require localization,” he said. “We are developing two applications and user interfaces that have to interact in different clinical settings.”
There are ethical and legal challenges around licensing content, he added. “Building one PDMP solution across two states with different regulatory requirements has been challenging,” he added. There also are technology-related issues around deploying CDS Hooks and Smart on FHIR and FHIR resources because they aren’t fully supported in EHRs yet.
The second two-year effort is being led by the MedStar Health Research Institute, along with partners including Georgetown University and George Washington University. Kristen Miller, Dr.PH,, the scientific director of the MedStar Health National Center for Human Factors in Healthcare, said the effort would help primary care clinicians and patients work on opioid tapering. Their human factors engineering approach involves developing use cases, personas, journey mapping and story boards to understand who the user of the system is, what actions they should take and how the technology should respond to their actions. They have worked to capture patient-reported outcome data and creating data visualizations for users.
Like Bloomfield, Miller mentioned working on ethical and legal policy issues. For instance, patients entering data may think providers are seeing it in real time and responding. Providers have to decide at what threshold they need to engage with the patient. This is related to legal liability issues.
This effort involves a three-phase roll-out to 15 primary-care sites representing three different EHR vendors. One challenge is that local EHR customization may be required for vendor sites that have not adopted FHIR standards, and CDS Hooks is not implemented in most EHRs yet.
