10 Lessons on Enterprise Imaging Strategic Planning From Duke University Medical Center

Nov. 21, 2016
During a Nov. 17 webinar, Duke University Medical Center's Christopher Roth, M.D., described 10 lessons from Duke’s experience launching an enterprise imaging strategic planning process.

When Duke University Medical Center physicians began their enterprise imaging strategic planning process in 2011, they created a diagram of applications and vendors currently in use broken down by work flow: image interpretation, image results documentation, image distribution and image capture. “Our organization looked at it and saw we had a lot of redundancy here,” said Christopher Roth, M.D., Duke’s vice chair, information technology and clinical informatics and director of its imaging informatics strategy. “Very quickly we started seeing wins, not from buying more stuff, but from paring down the stuff we already had.”

During a Nov. 17 webinar sponsored by the Society for Imaging Informatics in Medicine, Roth described 10 lessons from Duke’s experience for others to consider as they launch an enterprise imaging strategic planning process.

Lesson No. 1: Get everyone to start thinking strategically. Roth and others created documents in PowerPoint and Microsoft Visio that sought to tie imaging informatics goals to larger strategic goals of the health system. “We tried as best we could to tie the things we wanted to do to existing strategy. We also developed our own goals and strategies and socialized them,” he said. “This turned into an icebreaker and a tool to show that we are thinking strategically and expanded our conversations beyond a small group doing enterprise imaging. As we built out these various plans and refined them, it started forcing some forward thinking among finance and IT, the clinical business and other stakeholders because it became clear that the current state was not good enough and we should be doing a lot better.”

Lesson No. 2: The enterprise imaging strategy has to be a living document and iterate over time. “We kept using it in different settings to educate and change culture as best we could across a very large organization,” Roth said. Later Duke Health developed its own overall strategic vision, which breaks down into a series of goals and strategies that enterprise imaging can align with. “This taught me a good lesson. Having a stakeholder plan like this should inform my strategic plan and it will,” he said. “But not having a stakeholder plan should motivate you to write your own plan anyway. Because that is what we started doing and we started changing the culture without having the necessary degree of alignment within our organization.”

Lesson No. 3: Your strategic plan should bring almost everybody together. At Duke, the strategic planning process got radiology and cardiology conversing, he said. “We realized we had similar problems and started working together rather than pushing back on each other.”  The creation of a governance committee has helped allow physicians to guide the process, Roth said. Twenty-seven faculty physicians from 17 different specialties make up the imaging governance committee. Included are teammates from IT and patient safety, but it is a physician-driven governance group where the physicians are making the decisions about where they need to go next. “That has brought everybody together,” Roth said. “In some cases that bringing together has been easy; in other cases more challenging. Some specialties are moving faster than others. It has been helpful to have not just a plan to bring everybody together, but a governance body to do so as well.”

Lesson 4: The strategic plan should be concise and stand without you. It will be shown all over the organization, and sometimes you won’t be there, Roth stressed. It should be understandable by everyone from entry-level analysts to high-level executives. “Saying it in a few words is harder than saying it in a lot of words,” he added. You have to write it so that both physician specialists and those in informatics would be interested.

Lesson No. 5: Your strategic plan should quantify outcomes. Duke made a three- to five-year plan that included efforts such as storing all dermatology images and decision support. “Then, very importantly, we created specific outcomes as success factors for the initiative over the course of development,” Roth said. “Enterprise imaging is not an end unto itself. A new application being installed is not an outcome. Those are initiatives. The outcome has to be something that is a quantifiable gain. It has to have a target number and a target time. It has to be obvious what the benefit is so that people in the C-suite and finance groups can have an a-ha moment and understand why you want money for a component of enterprise imaging. They want to know why you want it. What is the win and ROI?”

Lesson No. 6: Your strategic plan should answer stakeholder needs. For instance, Duke’s board stressed to the CIO that cybersecurity is a key concern of theirs. The enterprise imaging team started looking at issues such as CDs, thumb drives, DICOM studies, and de-identification of data for research purposes and how those fed into an improved cybersecurity posture.

Lesson No. 7: Your strategic plan needs to communicate: where you are today. The process at Duke revealed that many clinicians didn’t know what was happening in imaging informatics outside their specialty. To help share information across specialties, in December Duke will hold the first annual enterprise imaging town hall meeting to allow stakeholders to hash out the plan, so everybody can get on the same page and ask questions of each other. If there are disagreements later, they will want to refer back to the plan and say this is what was agreed upon in a governing body.

Lesson No. 8: The plan should incite a lot of questions. When they mapped out all the systems they had, even just within Duke’s four nuclear cardiology sites, they found they had four different systems for image capture, three different systems that provided a reading work list, and three different systems for documentation. “That is not helpful in any way, so we are centralizing that right now,” Roth said. “It is taking the complexity out of the hands of our IT folks.” These efforts will get people asking questions about the current state.

Lesson No. 9: Your plan should demand uncomfortable choices. The effort can lead to awkward conversations, Roth said. “It is your responsibility as a leader is to be able to have those and ask tough questions,” he stressed. Are we going to prioritize security or education? Hire folks or spend money on applications? Once you get a roadmap, he added, it becomes easier to tell people we can’t do a particular project right now, because they see it on the roadmap in the future.

Lesson No. 10: Your plan should turn the gears of resource and project planning. Don’t create something that gets put on a shelf. “It has to be something living not just as a document, but living and breathing in project management and in a resource planning application,” Roth said. Duke tied it to the IT ticket management application. “To execute on the plan, you have to translate it into whatever part of the organization is executing, with start dates and end dates — for instance, which ones are for calendar year 2017.  Duke’s plan has 57 initiatives and 90 outcomes it is going to be measuring over the next three to five years.

Roth closed by saying he hoped his presentation would spark some ideas for what others might do at their organizations and give them courage to put ideas on paper and start socializing them.