Does Physician Education on New NCCN Guidelines Change Behavior? Data From Ontada Say, “Yes”

This week, thousands of oncologists will travel to Chicago for annual meeting of the American Society of Clinical Oncology (ASCO), to hear the latest evidence from clinical trials. At least some of the data presented will be touted as “practice changing,” meaning that clinicians should start offering new tests or treatments in response to the results.

How quickly that happens, of course, depends on many factors: how long does it take for new evidence to be absorbed into guidelines of the National Comprehensive Cancer Network (NCCN)? In turn, how quickly do physicians learn about these new standards? And finally, how rapidly do payers make their own shifts, offering coverage to address the new standard of care?

With this in mind, researchers with Ontada, part of McKesson, recently presented data at ISPOR 2026 in Philadelphia that shows the connection between physician education on an important new standard and subsequent behavioral change.

Their abstract, “Improving Real-World Adoption of HER2-Low Therapies Through Targeted Educational Initiatives,” evaluated data from the electronic health record (EHR) used in practices of The US Oncology Network, called iKnowMed, to measure activity among oncologists who received an EHR-based, on-demand peer consultation and podcasts about new NCCN guidelines on use of HER2-low therapies in metastatic cancer.¹

“With the increasing availability of HER2-low targeted therapies for patients with metastatic disease, it is crucial to understand real-world testing and treatment patterns in community oncology settings,” the researchers wrote.¹

According to the abstract presented at ISPOR 2026, researchers evaluated activity among 4536 providers who voluntarily used the education resources within 7 months of their launch and compared the patterns with those of physicians who did not engage with the educational materials. The Ontada team noted that HER2 testing was already “nearly ubiquitous” before and after the educational activity.

Nonetheless, results showed that guideline-concordant therapy increased from 62% to 71% for HER2-low patients who were treated by providers taking part in the education activities but did not change in the comparator group (63% before the initiative vs 64% post-initiative).¹

For insights into these findings, The American Journal of Managed Care^® (AJMC^®) turned to Jessica Paulus, ScD, vice president for Real World Research at Ontada, who was senior author and presenter of the study. Paulus’ responses to AJMC’s questions appear below.

AJMC: Your data reveals that education results in more testing: it increased the absolute difference in the percentage of patients treated by physicians who used testing. Did this surprise you, given how common testing already is?

Paulus: We found that HER2 testing was already very high at baseline and this was not surprising. Testing at baseline was in the mid-90% range and it stayed high over the course of this initiative. This wasn't necessarily surprising because HER2 is a biomarker that's been around for a while and there's already been so much work to standardize the use of these tests across patients with breast cancer.

At first glance, it might look like there's not much opportunity there, but where the nuance comes in is that the testing is no longer just a binary yes or no, but that HER2-low represents a more granular category, engenders differences in how results are interpreted, documented and handled by the clinician. While we did see modest changes in testing patterns, the bigger takeaway was that, even in a space where testing is largely standardized, education made a big difference in how those results are actually applied in clinical decision-making.

We saw an approximate 10 percentage-point difference in the improvement of guideline-concordant treatment, which indicates that educational interventions really helped physicians know what to do with the testing results. While we know that testing is really being done ubiquitously, it indicated that there was a need for better support for clinicians to interpret the testing, especially in the landscape of a quickly evolving treatment space where HER2-low represents a kind of new paradigm in personalized medicine for patients with breast cancer, rather than having just a binary “yes or no” value for HER2.

AJMC: The data revealed that education does move the needle on guideline-concordant care, and by a significant percentage in a short period of time. Did the degree of difference between the physicians receiving education and those who did not surprise you?

Paulus: To be transparent, I was very pleasantly surprised and hopeful that there would be a meaningful change in terms of those physicians that were receiving educational support. The degree of change was very clinically meaningful, measurable and exceeded my expectations because we know medical oncologists are extremely busy professionals and that behavior change is incredibly difficult. Being able to support and promote a clinically significant increase and meaningful changes or influences on clinical behavior in such a short timeframe was honestly surprising.

The fact that we saw the change in guideline-concordant therapy limited to providers who engaged with the educational resources—while there was essentially no change in the comparator group—tells us 2 things: First, the adoption of new treatment paradigms is not automatic, even when guidelines are updated. Second, we learned that relatively light touch and highly targeted educational initiatives that were delivered in the workflow can meaningfully influence clinical behavior in a short timeframe. The magnitude of the difference reinforces that there's still a gap between awareness and consistent application, and perhaps that's where these types of initiatives can have the biggest impact.

AJMC: In the study’s limitations, you note that physicians who engage with education resources may be different from those who do not. Are there any defining differences (demographic or geographic, for example) between physicians who engage with education vs those who do not?

Paulus: That's an interesting question, and in the study, we did not formally characterize differences between the providers who did engage with education versus those who do not, so we can't definitively say whether there are demographic or geographic drivers. However, this was done in a sense on purpose, because this was not a research and educational-coupled program that was intended to draw causal inferences. This was a real-world quality improvement initiative that accomplished its goal, because we saw a very clinically meaningful improvement in NCCN guideline-concordant care for patients with breast cancer. We did not analyze the causal mechanism of why it happened, though it is certainly interesting and from a methodological standpoint, we have to assume there are underlying differences.

It's easy to speculate that providers who choose to engage with educational resources may be earlier adopters, more interested in treatment guideline updates, or more inclined towards continuous learning and that's why we were careful to avoid over-interpretation of causal mechanism which again, was not the intent of this research program. This was a quality improvement research initiative looking to meaningfully improve guideline concordance which, most importantly, was accomplished.

AJMC: Is this type of research important to measure effectiveness of education programs? When leadership sees this degree of difference in prescribing following an education program, does it build the case to require education when there is a major guideline update?

Paulus: This type of research that's integrating rapid insights from real-world data with various physician education and engagement programs, is increasingly important. I would frame it differently than requiring education, instead, what this work shows is that we can measure first, where we should dedicate educational efforts, and then also measure whether those educational efforts have translated into changes in real-world treatment decisions.

This has historically been very difficult to do for a number of reasons: One is that it's very difficult to meaningfully influence behavior but also the data must be high quality and rapidly available to assess behavior change or changes in real-world treatment decisions quickly and reliably. From a leadership perspective, this is very powerful because organizations can move beyond simply delivering education to evaluate which approaches are most meaningful, which are most effective, and where the gaps may remain.

The goal is not to mandate education universally, and certainly there are so many demands on the time of medical oncologists, that it's not necessarily practical here. Our goal was to design highly targeted and high-quality, evidence-informed educational material and approaches that fit naturally into the clinical workflow and address some of the gaps in care that are most important to clinicians.

We find that often, clinicians are seeking easily accessible and high-quality support in areas where there are very rapidly evolving treatment guidelines, especially in the community oncology setting, where there may be certain indications or cancer types that are uncommonly encountered. These approaches don't feel like a “box to check” or requirements but instead are much needed support for every medical oncologist's goal—to provide the right treatment for the right patient at the right time.

Reference

Berini CR, Stevens LAS, Guo J, et al. Improving real-world adoption of HER2-low therapies through targeted educational initiatives. Presented at: ISPOR 2026, May 17-20, 2026; Philadelphia, PA. Poster No. HSD11. https://bit.ly/4vd1fy0