All CDS Collaborative Participants Meeting #1

Discussion Topics: (Slides used during the meeting are available at the bottom of this page; an audio recording of this call is available here.)

1. Welcome: Note that project title is not that catchy; we are open to suggestions. Project has been underway since October. Jerry Osheroff, MD is leading the project.

2. Collaborative membership: In the collaborative, we have nearly 130 individuals from the following organization types: care delivery organizations across the country, EHR suppliers, federal agencies, Beacon communities, society partners, CDS analytic suppliers, and clinical transformation consulting firms.

3. Collaborative goals: Accelerate local CDS and Performance Improvement (PI) efforts by documenting, sharing and enhancing target-focused CDS intervention strategies; develop a consensus-based CDS configuration template, processes for filling it out and sharing, and path to realizing local value from this use; scale this work and benefits to many organizations and PI imperatives; and do this in a manner that delivers strong benefits to all stakeholders - especially care delivery organizations, but also EHR vendors, Federal agencies, and others.

4. Private collaborative website: Collaborative members are able to access detailed information about the collaborative and the project through an invitation-only website. To gain access to the site, individuals need to join the collaborative.

5. Engaging more stakeholder input: Project is at a phase where we will begin looking to collaborative participants for their input on whether and how the project work can be useful on a broader scale.

6. Pilot sites: We have engaged 7 inpatient and 2 outpatient sites. They are (inpatient) Ascension/St John Providence Health System (Cerner), Hennepin County Med Ctr (Epic), Lehigh Valley Health Network (GE), MetroHealth System, Case Western Reserve U. (Epic), New York Hospital, Queens (AllScripts), Texas Health Resources (Epic), and University of Pennsylvania (Allscripts); and (outpatient) University of Pennsylvania (Epic) and Veterans’ Administration (homegrown EHR).

7. CDS configuration template: The pilots are testing and adapting the template to real scenarios within their organizations. By the end of the project, we will have the following deliverables: A working draft of a target-focused CDS configuration template, 2 high priority targets identified as focus for pilot efforts, A refined template based on input from pilot sites and the CDS Collaborative, Data about the implications of this work for key stakeholders, A summary project report, and a proposal for the next project phase.

8. Project success metrics: 1) Create a stakeholder-approved template for documenting/sharing CDS configurations; 2) pilot sites use the template; report significant value to local target-focused CDS/PI efforts (Quantitative survey: Likert scale on value and anecdotes: about positive project impact); and 3) at least 50% of participants ask to continue.

9. Actual and anticipated benefits: Makes explicit current local target-focused CDS thinking and deployments, which helps stakeholders (e.g. CDS recipients, organizational leadership, informatics, IT and quality teams) understand and communicate current approach and brainstorm; CDS and PI teams visualize inter-related components of target-focused CDS/PI activities; Highlights opportunities to improve configurations (Workflow backbone/CDS Five Rights); and explain why current CDS approaches not working well. The common template reveals broader experiences, including sharing enhances local CDS/PI learning and progress.

10. Review of template: Jerry walked through the template to show the different components; feedback from Collaborative participants on the template and its use are welcomed.

11. Reports from the 7 most active pilot sites:

• Hennepin County Medical Center: HCMC joined the collaborative because of strong interest in healthcare innovations and emerging delivery models. Have limited resources to pursue process and content of certain CDS initiatives. Easier to do in a collaborative. Have used the flow diagram as a documentation and discussion tool. Have used the CDS grid internally among core team members to evaluate individual CDS support interventions. Believe it will be helpful in leaving a trail for when and how changes are made to the CDS configuration. Flow diagram is very useful to frame high level discussions and document the necessary components for a gap analysis. The CDS 5-Rights document seemed too academic at first use; however, that opinion may change as the document is further used.
• Lehigh Valley Health Network: LVHN signed up for the pilot to support efforts to develop standard work tools to design and deliver care. CDS configuration template seemed like a good way to put information into a standardized format. Will present template next week to the CDS committee to obtain their endorsement on making the template the shared communication tool. The tool is very intuitive and believe it can be used by other committees in the hospital providing a standard form of communication that keeps everyone on the same page. Also, having a shared information library will help in the communication of what our information systems can and cannot do. Lastly, there is a lot more to applying CDS packages to raise organizational awareness of the system capabilities.
• Univ. of Penn - outpatient: Have been working for sometime on patient-centered medical home with population management of those with diabetes as a focus. To capture non-encounter based care, the template helps deconstruct the care process. The template also helps them to think through how to reach out to diabetics. Working through the backbone, grid, and CDS 5-Rights templates was very helpful. The backbone is very good at identifying points of active alerts. Order sets are useful, but only for the duration of the encounter. The data that enables CDS to take place (lab results, etc.) have to be captured in a discrete function so that the CDS knows what actions have occurred on not.
• NY Hospital Queens: Representing hospital CDS workgroup. We joined the collaborative to help develop and adopt best practices for CDS performance improvement. For our organization, the collaborative serves as a starting point for assessing CDS at NYHQ, which has had CPOE since 2006 and has a plethora of order sets and MLMs. The hospital's Anticoagulation Task Force has completed and reviewed the completed template for VTE prophylaxis. The Workflow Template offers a format to review current CDS interventions within the context of clinical care. The CDS 5 Rights Template will be a useful tool to examine individual interventions in greater detail. The project complements the VTE clinical quality measures for meaningful use.
  • Texas Health Resources: Very excited to join the collaborative due to THR commitment to advancing electronic health records. THR is using the template as a workflow road map for CDS intervention. Enables them to step back from the weeds and get a broad view of the intervention. One of the main benefits is the ability to look at the work flow as a whole vs. by single interventions. Can determine if interventions complement each other or are in conflict. From a technical build perspective, can see if a CDS intervention contradicts another one. With VTE prophylaxis, with each level of care, there is a new assessment. If an intervention already ordered, additional assessments aren't needed.
  • Univ. of Penn - inpatient: CDS requests are forwarded to CDS committee to determine if it should be pursued. Penn decided to be a pilot site based on Jerry's long-standing relationship with Penn, as well as the fact that Penn has always benefited from learning from other organizations about their decision support approaches. In addition, hoping to participate in development/refinement of tool that can allow our CDS group to better recognize opportunities where CDS can improve care processes/outcomes. Regarding the tools, the backbone and the table that accompanies it allow you to flesh out where your interventions under consideration can be targeted so that you can best meet your goals. In using the template, we realized that we had an VTE risk assessment when patients first came in to the hospital, and then information on what medicines were ordered for the patient. However, we were missing continuous assessments to gauge if patients were going from low risk to high risk, changing their requirements for VTE prophylaxis. We used the template by taking the VTE prophylaxis decision support and completeing the CDS 5-Rights table. Regarding benefits, most interested in learning how others are using the templates, how they are managing decision support in general, and how we can use the templates to form accessible libraries of CDS strategies used by institutions across the country to address clinical priorities.
  • Veterans' Administration: There was a lot of interest from the VA nationally to be involved in the pilot. However, using decision support may vary from site to site, so decided to select one site only. Have had decision support in place for 15 years. Wanted to see how template could help create more sophisticated interventions. Felt the template would also help optimize existing decision support tools. The benefits from using the tools include: useful in establishing an overarching plan for best practice for diabetes care in VA, helpful in identifying roles of various health professionals and specific CDS tools to support them, and provides a roadmap for a long-term project allowing for incremental improvements and CDS development around specific tasks.

12. Open discussion:

• Clayton Curtis (VA): Observations regarding presentations: Suggest that everyone take away the recurring comment about different types of CDS. There is a mistaken impression that there is only one thing called CDS. Think the template allows people to work into specific places where CDS interventions occur. Also, there has been an attempt by ONC and AHRQ to set up a "Federal CDS Collaboratory" for federal CDS activities to share information on CDS efforts in place. They did not have the structure of a template, so they constructed a list of topics for describing projects - a template such as that used in this project could be useful as an organizing scheme for such efforts. Lastly, some of the places where CDS has worked in the past does not fit comfortably into the template; different CDS use cases should be considered in finalizing the template contents and structure.
• Jonathan Teich (Elsevier, CDS Guidebook, ONC Consultant): Two questions: 1) Backbone was modified twice over a 5-year period to incorporate things the HIMSS CDS guidebook authors had learned. Eager for continued input. 2) Interested if sites using the templates can populate the template and then share it, and then see if another organization can use the first group's input.
• Mary Goldstein (VA): ONC SHARP project: Not sure how this project relates to Bob Greenes' SHARP project. That project is an ONC-funded CDS project to develop an 'Implementer's Workbench' for configuring CDS interventions to local needs. Bob Greenes responded that the main difference is that this CDS/PI Collaborative is user driven to tailor CDS to specific environments. Bob hopes to share tools they are developing with the Collaborative.
• Dan Pollock (CDC): National Healthcare Safety Network: The CDC maintains this database where providers can submit key healthcare safety-related data. Starting to develop module on VTE for purposes of supporting CDS-enhanced performance measurement, reporting and improvement - working with CMS on this. The core activity for the collaborative is to accelerate the local performance improvement initiatives. However, there are a lot of interconnections with these CDC/CMS/federal initiatives that we are eager to cultivate. Floyd Eisenberg mentioned NQF's work with value sets for reporting, which are related to the value sets needed for CDS.
• Nick Hampton (BJC): Evidence to support intervention value? Is there room in the grid for this type of supporting documentation? There is a place in the template to record performance pre- and post-intervention. As far as tracking evidence to support a particular CDS intervention, that is not part of the current effort. However, documenting CDS interventions that are not in place yet would be helpful. Information on VTE prophylaxis would be helpful to share.
• Maxine Ketcham (THR): THR is learning from other pilots, e.g. things that are note helpful for VTE prophylaxis CDS. It is helping them get a bird's eye view their targets and efforts to address them.