Key Takeaways

• Oncology APPs can help translate research into practice by leading measurable quality improvement efforts, expanding care delivery roles, and generating data that demonstrate effects on clinical outcomes, patient experience, and operations.
• Successful implementation of innovations such as ePROs, AI tools, and APP-led care models requires more than evidence alone, with reimbursement, workflow integration, institutional support, physician collaboration, and scalable infrastructure all influencing adoption.
• As AI and decentralized care models continue to evolve, oncology APPs will need stronger skills in data interpretation, implementation science, interdisciplinary collaboration, and research dissemination while maintaining appropriate human oversight of new technologies.

Madeline Merrill, AGNP-BC, MSN, discusses how oncology research can move from promising evidence to sustainable changes in clinical practice. Drawing on research presented at the 2026 ASCO Annual Meeting, Merrill highlights the importance of reimbursement, workflow integration, institutional leadership, physician collaboration, and replication across care settings in determining whether innovations such as electronic patient-reported outcomes (ePROs), AI, APP-led clinical trial models, and decentralized care are successfully adopted. She encourages oncology APPs to identify problems within their own practices; conduct focused pilots; measure clinical, patient experience, and operational outcomes; and pursue mentorship and opportunities to publish and present their findings. These skills can position APPs to play a larger role in care delivery innovation while helping ensure that emerging technologies and workforce models translate into meaningful improvements in access, quality of life, and patient outcomes.

Transcript

Hi, I’m Madeline Merrill. I’m a nurse practitioner at Memorial Sloan Kettering Cancer Center in New York City. I work primarily in research. I’m a clinical trials nurse practitioner in the phase 1 and 2 group working in solid tumors. The service is Early Drug Development, and I am also an education specialist in terms of research for all APPs at the institution.

I think that there’s a couple of things that determine whether something actually changes practice. The first thing that I would have to say is reimbursement. I think that’s where the ePRO is very exciting because the technology itself has 15-plus years of trial data, but the adoption really only scaled, and it has now been presented on a very large scale, once the CPT codes made it billable. Evidence alone doesn’t necessarily change practice, but the evidence plus the reimbursement and the sustainable payment model does. That’s, I think, one important area, making sure that reimbursement reflects the science.

I think beyond that, there’s also public health policy. There was one study as well from a 2026 ASCO abstract that found that local and paid state sick leave was associated with earlier treatment initiation among working-age, early-stage breast cancer patients, particularly in areas with low female employment. I think that that’s another thing to keep in mind, that translating research into practice sometimes means translating research to legislation as well, not just clinical protocols.

I think that you have to be able to see the replication across multiple cancer care settings. Not just a single strong one-institution result, but specifically the ePRO data from the Texas Two-Step and the Canopy data from Arkansas. Those show that the ePRO benefit is consistent across different communities. It showed benefit years apart, as well as in different states and on different platforms. That kind of really gives clinicians the confidence to adopt versus just one promising pilot.

I think something that was a big discussion and conversation specifically about AI this year was that the tools that are available have to fit with existing workflows rather than kind of running parallel. There was one study by Jahanzeb* [see Author's note] who made a point that AI is really universal, but finding the tool that aligns with the current workflows and embeds a workflow into a clinical practice is what will actually be the one that is used. That study specifically looked at, if you flag a result or flag something in the patient’s chart, that’s great. But then if there is not a clear clinical workflow of how to make that work, then that makes it really difficult to actually implement.

I think beyond that, there has to be institutional support, and there has to be a leadership investment in anything that comes out. Specifically with the Braun-Inglis study from the University of Hawaii, looking at APPs and clinical trials, they had strong institutional support. They had structured education. They had standardized resources. They had physician collaboration, and that leadership investment really built the ability for their APPs to function at a very high level and then ultimately increase trial accrual and clinical trial access for patients in the community.

I think that all of those things together help innovation be implemented. And you really need kind of all of them. You need it to be both at the institution level and then at the community level, at the national level, and beyond to really have something that is truly adopted and utilized within a clinical care pathway.

Barriers that we have certainly seen are administrative and particularly reimbursement burden. There was specifically an abstract that looked at this. Odia found that prior authorization, when we’re talking about reimbursement, prior authorization denial rates cost institutions pretty substantially. I think they found that administrative costs from prior authorizations exceeded $90 billion annually when this was mostly borne by providers and patients and not payers. That, I think, is one that has to be recognized, that administrative and reimbursement burdens can really limit implementation.

I think that there is also, specifically looking at Vella and Braun-Inglis’ studies, physician hesitancy to hand off responsibility for consults and for clinical trials and research to APPs. That was explicitly named as a barrier in Eileen Vella’s study from The US Oncology Network. But I think that having physicians on board and being supportive of APPs can really help access to care.

I think that there’s also resource requirements that don’t necessarily scale to smaller practices. The Hospital-at-Home model and some of the CRS management are also a safety infrastructure. It’s realistic for larger systems to implement something like this, but a small community practice just may not have the resources to partner with their local hospital and with local paramedics to improve access to care for patients who are being monitored. I think that’s also something that we have to take into consideration.

We are seeing patients in the clinic. We are oftentimes the first people to notice a pattern. We are able to start with kind of defined questions and do our own contained pilots with measurable outcomes.

Every successful example that I discussed today of utilization of APPs built its case with clear before-and-after metrics: time to appointments, accrual percentages to trials, hospitalization rates. There’s not just a qualitative buy-in, but a quantitative measure. I think starting small with the question that you see about a pattern and then looking at the literature and looking at your own metrics is a great place to start.

I think beyond that, seeking out mentorship and structured education pathways rather than assuming that the scope expansion will happen by itself. There’s a SWOG MAPP program, there’s SWOG’s in-core approach that are built for formal mentorship within kind of their models of care. We should be creating structural change, looking for how best we can overall within institutions change this rather than having kind of ad hoc buy-in.

I think tracking and reporting the data yourself, looking at your patient volume, looking at your own hospitalization rates, and tracking that is really important. And then engaging physician champions. I don’t think any of the care pathway changes happen without institutional and physician buy-in. I think that physicians can help mentor. They can help kind of nominate you to take on projects. It’s not just APP enthusiasm that moves institutional change, but also our colleagues and our physician supporters.

I think that there’s no single metric that really looks to see if the care value is successful. But a really successful care delivery innovation shows movement across at least 3 categories. One is hard clinical and utilization metrics, one patient experience metric, and one operational or provider-facing metric. I think that if you are just saying that there is a utilization or a clinical metric, you haven’t made the full case yet. I think that that is something to look at. Again, 3 major metrics to think about: the clinical or utilization metric, the patient experience metric, and the operational metrics as well. I think that those are good things to monitor the success of the program.

I think probably the biggest place to lead innovation is through workforce and access redesign. This is where we historically have had the clearest and most replicated evidence, specifically with the stuff out of The US Oncology Network, asking to take on consults, asking to be a part of a higher level as principal investigators on trials. I think that that is a place that really can lead innovation moving forward.

I think data fluency, specifically as AI becomes more utilized, as different AI tools are used, understanding and being able to read and question both a study, but also the data across a dashboard or validation data, so that we know when to trust these tools and when to push back.

But I also think that we need to make sure that we’re doing quality improvement methodology. Studies that succeeded, succeeded because someone built structured education, mentorship, and measurement into a rollout, not because just an idea alone was good. That’s kind of an implementation science skill set, not just a clinical skill set.

I think additionally, we have to factor in cross-disciplinary communication. Championing change requires bringing physicians, administrators, and policymakers along, which we saw in every successful example this year.

I think we need to be publishing and presenting data. It’s not just anecdotes. We need to track the data and present it at programs like ASCO and as an abstract. I think that we need to be building mentorship structures. And then I think we need to be getting involved in professional societies such as ASCO, such as AACR, as well as the APP-specific ones such as APSHO.

I think that these are all places that we’re uniquely positioned to really impact patient outcomes and really impact the patient experience and drive changes within the health care delivery system. And I think that we can be doing more to publish and present and to be involved in a lot of the science that is happening around us.

I think short term, in the next 3 to 5 years, I’m really looking at kind of the maturation of the AI clinical decision support. I think that that’s really interesting. But there’s just so much out there. There’s so many AI companies that are competing to see how they can best impact patient outcomes. From clinical trial matching to treatment decisions to even prognostication and increasing the number of patients that choose best supportive care over treatments based on predicting a response, I think that that is a really interesting and rapidly advancing area that I will keep an eye on.

Beyond that, I also think that I really want to keep my eye on APP-led care models that are scaling beyond just their pilot sites. I think that that has a really big potential to impact patient quality of life and care, keeping patients close to families to let them receive lifesaving treatments and to manage their toxicities near their homes and keeping them in the workforce and things like that.

All of these things together will impact patient outcomes and impact quality of life for patients. I think specifically, The US Oncology Network is doing a lot to try to bring care that is APP-driven into community oncology practices, with the hopes of really impacting patient quality of life and the patient experience.

I think it will not be a single flashy tool, but it will actually be the unglamorous infrastructure work. The reimbursement pathways, the workflow integrations, and the workforce policy are going to have to catch up to what’s been proven effective. I think that that’s going to include APP role expansion and all of the kind of community-based, at-home models that will exist. They’re sitting on strong evidence, but what they’re missing right now is really structural work that makes them the default instead of the exception.

I would encourage every oncology APP to track their outcomes on their quality initiatives as though they’re preparing to publish. I think that there’s a huge gap that we are able to fill in terms of identifying the trends for patients, identifying the symptoms, symptom management. I think that in terms of actually publishing and disseminating our work, that is an area where we are not necessarily well trained and could really benefit from further mentorship and structured education around this.

I would want to encourage APPs to start looking at their own practice and figuring out what is worth publishing to them, looking at their own metrics from that with the idea that they will eventually publish.

Beyond that, I would advise every oncology APP to also stay appropriately skeptical of any AI tool right now, or even if it’s not AI, just being appropriately skeptical of tools that are reducing work without clear ownership of what happens next. I think that a big underlying theme this year at ASCO was that AI exists, we’re refining it, we’re making it work, but that there always needs to be a physician or provider in the loop. There always needs to be a human that is double-checking it, and there needs to be clear responsibility for that. We need to make sure that we close the loop between a signal and an action. I think that approaching the technology with a skeptical lens will also be really beneficial to APPs in the near future.

I think that the 3 most important lessons are that APPs really can drive changes in care. I think that that was really exemplified by the 2 abstracts from Eileen Vella and Christa Braun-Inglis. I think that there is the ability to work truly independently and that that should be something that we’re aspiring to.

I think as well that another important lesson is learning how to use AI tools and other digital tools. And then I think also that there is a shift to prophylaxis instead of responding to symptoms of treatment. That is another area in which APPs are really well prepared to have a large impact in terms of improving patient outcomes.

*Author's Note: The study cited at 2:37 was erroneously attributed to Saria, while the correct study is by Jahanzeb.