AI Can Ease Administrative Burden in Community Oncology

Artificial intelligence (AI) tools like ambient scribes and clinical decision support are reducing administrative burden in community oncology, according to Debra Patt, MD, PhD, MBA, MPH, who also outlined adoption barriers and automation strategies for tackling prior authorization challenges in her interview with The American Journal of Managed Care^® (AJMC^®).

Patt is the executive vice president at Texas Oncology, president of the Community Oncology Alliance (COA), and a member of the American Society of Clinical Oncology Board of Directors. On September 14, day 1 of the COA Payer Exchange & Innovation Summit, she served as a panelist in both the “Executing Innovative AI Solutions: Lessons from Community Oncology” and “Oncology Innovation Showcase: Prior Authorization Solutions" sessions.

This transcript has been lightly edited for clarity.

AJMC: My first 2 questions pertain to “Executing Innovative AI Solutions: Lessons from Community Oncology.” What are some of the most promising applications of AI you're seeing in community oncology today?

Patt: There are so many applications of AI in community oncology today that I'm really excited about. Probably what I'm most excited about are the things that allow our very talented people to work top of license.

Last year, when we talked about digital transformation in the practice, I talked about the Home for Dinner campaign. I called it the Home for Dinner campaign because we think about how AI supports clinicians, nurses, and administrative professionals in practices. It's really to decrease the administrative burden that we have in practice to allow us to act top of license. The ambient AI scribe service that most of our doctors, our 600 doctors in practice, are using today allows them to be home for dinner. Because they're not sitting after their clinical day that's very challenging and fraught with emergencies, dictating their notes at the end of the day; they're able to leave the practice and go home for dinner.

By doing that and implementing AI, it's not supplanting any of the work that clinicians are doing. It's allowing them to complete the task faster and allowing them to complete the more meaningful aspects of their work. Now, when I'm in clinic, on my clinic day, I get to sit and look at a patient and talk to the patient. I'm not sitting in front of the computer. That's much more meaningful. That's why I went into the joy of medicine. It really allows us to act top of license.

The same is true for our other staff. We've implemented a solution for our triage nurses that allows them to prepopulate some of their notes with clinical decision support, some of the documentation burden. I feel like it's been a transformational shift, where they were nurses 50% of the time and secretaries 50% of the time. By decreasing their administrative burden with the use of AI, it's not telling them what to do; it has allowed them to shift to decrease their decision fatigue and be nurses 80% of the time and secretaries 20% of the time.

I'm really excited about how AI is allowing people to really act top of license and really take joy in the work that they're doing. Caring for cancer patients is a joy; it's incredibly rewarding work. That's usually not what's contributing to burnout, but it's these other things that really have taken the joy out of medicine and clinical practice that are so cumbersome.

In truth, I'm excited about AI everywhere, and it will be in every system that we use. I like to say, as we think about AI in cancer care, it's like the movie Everything Everywhere All at Once, because everywhere from screening to drug discovery to management of the revenue cycle and everywhere in between, we are going to see that these systems are easing the burden of practice, easing the burden for our administrators, and that's really critical to us completing our jobs in a timely way, allowing us to have better work-life balance and have sustainable practices.

AJMC: On the other hand, what are some of the biggest barriers or challenges practices face when adopting new AI technologies, and how can these be overcome?

Patt: There are many barriers and challenges in adopting AI solutions. I will say, I think the first challenge that we have is really being advocates for adopting some of these changes. To adopt these changes, we have to speak to our executive leadership and usually demonstrate return on investment for investing in these solutions.

In addition, we can't really have point solutions. We don't have the luxury of having point solutions. We would have too many. We really need to be thinking about channel solutions and sphere solutions so that, because the environment is changing so rapidly, we can look at a single vendor that we would've previously thought of as a point solution and think about how they will iterate and evolve with us as technology changes and our needs change.

I'll give an example of that. We went live with electronic patient-reported outcomes (PROs) instruments 6 years ago. We now work with a vendor, Canopy, to use electronic PROs when we decided we wanted to do remote patient monitoring and measure temperature in patients in addition to remote therapy monitoring, which we were already doing. They evolved with us and developed technology to support our patients. This has allowed us to have closer monitoring of patients at home.

Cancer care is like that. We are constantly changing, and we really need solutions that are more comprehensive in a channel or in a sphere to evolve with us and continue to iterate. That's just one of the challenges that we have. Obviously, there are a lot of security and compliance challenges. There are natural challenges with AI around bias, hallucinations, and errors. These are all things that you need to understand, manage, and, when possible, overcome. We at the Community Oncology Alliance have developed an AI task force, and we came out with a playbook to help practices navigate this landscape for exactly these reasons: to help them make these choices in a more informed way.

AJMC: Moving to the “Oncology Innovation Showcase: Prior Authorization Solutions," what are some of the biggest challenges prior authorization creates for community oncology practices and their patients?

Patt: Prior authorization presents many challenges for practices and their patients. There are often time delays to have the timeliness of treatment. There are often information delays. There are often denials or even downcoding.

I think one of the most important things in how we use AI systems to support our prior authorization environment is automation. The revenue cycle has always followed key performance indicators. Prior authorization has always followed a lot of key performance indicators. But automating those reports, having them included in a dashboard for easy visualization, and then [seeing] how the process loop is working [helps] to make sure the right people are in the loop to solve the problems that you face.

There are new problems every day. Are you having denials? Are you seeing downcoding? Are you having to manage letters of medical necessity and appeals? I think automated systems, dashboards, and really thinking about the process loop to make sure that you're providing feedback to your teams to solve the problem are some of the critical aspects to selecting how to use these systems. Even the best system is not going to work if you don't have the right infrastructure to make it work well.

AJMC: What tools, programs, or innovations have you found most effective for streamlining prior authorization workflows?

Patt: There are many different tools that people are using to streamline prior authorization workflows. Everything from patient assistance programs to automated ways in which we can track denials and make sure that claims go out clean the first time. Because [without] test edits or some kind of editing software to make sure that claims go out clean, it's a real challenge whenever you're having to refile or manage a denial or manage a letter of medical necessity.

Those, in many of our practices, tend to be highly manual processes, and so, to what degree again we can automate, template, and create a reasonable process loop, I think, is the best way for us to implement any software solution. I think software solutions that are going to be helpful usually manage the whole channel in various aspects or manage the whole sphere, oncology-specific solutions that continue to evolve with you.

AJMC: Bringing both panels together, what advice would you give community oncology practices that are looking to adopt new technologies or solutions while ensuring they address real-world needs and deliver timely, high-quality patient care?

Patt: AI and innovation are central to the discussion of community oncology thriving today. For practices, every practice, I would argue, that struggle with this, come join us at the COA meetings. This was the focus of our meeting in April 2026. This is the focus of our Payer Exchange & Innovation Summit. We have a committee. We've designed a playbook. We are in it with you, trying to solve this with you. We are all solving this as a practice.

I just spent this whole morning with different practices across the country that are trying to solve for AI solutions and how they assist with revenue cycles in the whole cadre of revenue cycle solutions. It was so eye-opening for me, because I think as a big practice, there are so many things we do well, but we also have tremendous vulnerabilities. Being able to have open discourse and discussions with people about what's working for them, what's not working for them, and how we can solve our problems differently has been really critical, and I think it will be critical to the success of community oncology moving forward. We're in it together, and if you too struggle with this, come join us.