Geisinger’s “Soft Demo” Proves Key for a Successful Clinical Pathways Rollout

An administrator in the Geisinger Cancer Institute discusses the rollout of ClinicalPath.

Clinical pathways in cancer care are beyond the “new” phase, but oncology practices and health systems are still learning all the ways they can bring greater standardization to care, eliminate unnecessary care, improve data collection, and drive value.

That’s what Mark Wojtowicz, MS, MBA, the administrative director for Oncology Research and Innovation at Geisinger Cancer Institute, shared during a session of the virtual meeting of the Association of Community Cancer Centers, held March 1-5, 2021.

Geisinger implemented Elsevier’s ClinicalPath in medical and radiation oncology in April 2019, starting with a “soft demo” and a weeklong training process. The pathways product was rolled out at all sites a month later. The health system decided to integrate ClinicalPath into its Epic electronic health record (EHR) management system, because incorporation into the workflow was seen as critical to getting physicians to adopt pathways.

Pharmacists oversaw a process in which the pathway regimen library was “harmonized” with Epic, to systematically update new workflows in hematology and oncology. This was a key process, because Geisinger actively takes part in clinical trials.

“Accruing to clinical trials is an important part of what we do, and we know that it means a great deal that our clinicians have clinical trial information at that decision-making point,” Wojtowicz said. To have all options available when clinicians were talking to patients, “we wanted to implement across all these specialties, and we wanted to do it as rapidly as possible.”

“A lot of the credit goes to the team that really made that happen,” he said.

The soft demo was key, Wojtowicz said. “Everybody was at the starting line, the tool was actually launched; it just wasn't promoted to the broad practice that it was live and accessible.”

During this period, the physician champions worked with the pathways to achieve buy in among the others—and allow time for tweaks and integrating with other workflows. And then a month later, it was time to flip the switch. Over a week, the pathways team met with clinicians and let them try the pathways system to give them opportunities for training and practice.

Wojtowicz shared a decision tree slide that showed each step, starting with whether a diagnosis was or was not on a pathway. The mapping system was critical, he said.

We know that clinical pathways oftentimes will not cover every single scenario that's out there, or every decision that an oncologist needs to make. So, you have to be able to map that out—what’s that pathway look like? Or, what do those workflows look like when it's a non-pathway related decision versus pathway related.”

This “visualization” process allowed the team, and by extension, physicians, the chance to think through what the workflow would look like and helped with implementation.

Integrating workflows in medical oncology with the EHR definitely presented some challenges, but physicians were determined to have one way to interact with the system. In the beginning, extra information had to be tailored and removed. “And that did require that [for] some regimens they had to differentiate, and we had to build new ones. But I think that was a worthwhile exercise, because everyone required consensus, discussion, and agreement around that direction.”

The final step was to ensure that the pathways were multidisciplinary—that physicians could work with a system that would allow transparency between medical and radiation oncology so that treatment options could be presented together to patients. “What worked for us was that the system allowed for this transfer of data,” so if a decision was made in medical oncology, it was shared with radiation oncology.

“That was a huge thing as far as workflow goes, because we really had to be prepared to have not only the individual clinician be able to handle the workflow, but then you kind of get to that next level where its, “How about the multidisciplinary workflow?’” he said. “You’ve got to make it simpler for those things to happen.”

Going forward, the next step is to make use of new data on the pathway system that allows artificial intelligence to use predictive analytics to inform decision makers in the practice.

“It doesn’t really stop there,” he said. More value will come from investments health systems are making in predictive analytics.

“We started bringing our clinical pathways decision data to the discrete elements that come out of that decision, we started bringing that into data tables, and linking it back with our clinical information to gain better insights about when decisions were made, and what's happening after that.” This allows clinicians to decide what care plans should look like and how much monitoring should occur.

Geisinger also opted to include prognostic questions. Tough as it is, physicians are asked if they would be surprised if the patient were to die in the next year. The answer helps drive more palliative care referrals.

The health system is cognizant that it is asking doctors to take time to gather data, to add clicks to their day, Wojtowicz said. Adoption rates have been good, and the process continues to improve.

From the health system’s perspective, “they’re extracting data in discrete fashion, and merging it back with some of our clinical data to make the most out of it.”