A recent discussion between ACCC and Highlands Oncology centered on how artificial intelligence can be implemented thoughtfully in cancer care to support clinicians and operations without compromising clinical judgement or patient care.
Artificial intelligence (AI) has moved beyond small pilots and isolated use cases in oncology. From documentation tools embedded within electronic medical records to scheduling and operational platforms, AI is already reshaping how work gets done. As adoption accelerates, the real question becomes how cancer programs integrate AI responsibly so that it supports clinicians, strengthens operations, and ultimately improves patient care.
These questions were at the center of a recent conversation between the Association of Cancer Care Centers (ACCC) and Terence Pierce, PMP, director of IT and Project Management, and Jessica Reddington, LPN, clinical informaticist, at Highlands Oncology Group.
Rather than focusing on future-state possibilities, the discussion centered on how AI can be implemented thoughtfully today as a tool that supports clinicians and operations without compromising clinical judgment or patient care.
At Highlands Oncology, AI is deliberately positioned as an assistant—not a decisionmaker. This guiding philosophy informs every step of how new tools are evaluated, selected and introduced into practice.
While many tools are marketed as intelligent or autonomous, Highlands leaders view most current applications as decision support tools—useful for surfacing information, reducing administrative burden, and flagging potential issues, but not for replacing human expertise. Framing AI this way has helped build trust among clinicians and set clear expectations about its role in care delivery.
This philosophy is reinforced by a structured governance approach that remains uncommon across much of health care. Highlands established a multidisciplinary committee to evaluate AI tools, with IT and clinical informatics involved from the outset. This proactive approach has positioned the organization well ahead of many of its peers.
In many programs, IT is brought in late—after key decisions have already been made. At Highlands, bringing technical expertise into the process early has been a game changer, ensuring workflow realities, data considerations, and downstream impacts are addressed before tools ever reach end users.
Just as important is the organization’s commitment to starting small. New tools are typically piloted with a limited group, often a single physician or small team, before broader rollout. This measured approach has helped protect clinicians from burnout and allowed the organization to learn and adjust before scaling.
Rather than applying AI broadly, Highlands has focused on high-impact areas where volume, repetition, and complexity create clear opportunities for improvement. Current use cases include clinical documentation support, chart preparation, tumor registry workflows, referral intake, and scheduling.
One area where the impact has been especially visible is chemotherapy infusion scheduling. Coordinating infusion schedules requires balancing chair availability, nursing coverage, treatment length, and provider schedules. By leveraging AI-enabled scheduling tools to help manage these variables, Highlands has been able to balance daily peaks and reduce bottlenecks in the infusion center.
This approach has resulted in more evenly distributed patients across available resources. In practice, this has contributed to infusion wait times averaging around 10 minutes, while also creating more balanced nurse workloads and more predictable schedules.
Across these examples, the goal is not automation for its own sake. It’s removing friction so clinicians and staff can spend more time on work that requires judgment, coordination, and patient interaction.
The conversation also acknowledged that not every AI introduction went according to plan. Earlier experiences with underdeveloped tools, particularly in documentation, served as important reminders about timing and readiness. When clinicians are asked to adopt tools that are not yet mature, trust can erode quickly.
“The end users [physicians] aren’t there to develop the tool,” Reddington noted. “They need something that is developed enough that they can start using it and then…tweak it from there to customize it to us.”
Those lessons have reinforced the importance of realistic rollouts, controlled pilots, and choosing tools that are sufficiently developed before bringing them into clinical workflows.
Identifying and vetting AI vendors, especially those specific to oncology, remains a challenge for many programs. Tools are often developed by smaller startups and marketed broadly across health care, making it difficult to be aware of what’s on the market and distinguish meaningful differences between products.
At Highlands, vetting has focused less on what tools claim to do and more on how vendors partner with care teams. That includes evaluating data use practices, customization capabilities, reporting functionality, and what happens when recommendations don’t align with clinical or operational reality.
Another lesson that emerged was the importance of working with vendors who will provide access to data about their tool’s performance. Leaders need visibility into where friction occurs and when tools fall short of real‑world needs.
“The big thing for us is making sure before we choose a product, we also have the reporting and data we're going to need to address [potential] problems,” shared Pierce. “And then from there, it's going back to the vendor and saying, hey, this keeps happening, can you figure out why?”
Access to performance metrics and reporting has allowed the team to catch issues early, refine workflows, and have more productive conversations with vendors. Just as important, it establishes shared accountability.
As AI capabilities continue to evolve, so too will the questions that cover autonomy, oversight, and accountability. For now, Highlands Oncology’s experience offers a grounded perspective to other cancer programs: Meaningful AI adoption is less about speed and more about integration.
When asked what would be helpful as Highlands continues to move forward in its AI adoption journey, Pierce said, “Better resources around vendor selection and clearer guidance on which AI‑specific checkboxes organizations may not have considered before.”
This feedback is part of why ACCC has developed new AI resources, including a tip sheet to help cancer programs think more deliberately about how and when to introduce AI tools. Rather than promoting specific technologies, the resources are designed to help teams ask better questions about readiness, safeguards, governance, and long‑term value.
For programs navigating AI for the first time, having a practical framework can make the difference between thoughtful integration and rushed implementation—missteps that can quickly carry 6‑figure costs or higher.
As oncology continues to grapple with when, where, and how to use AI, Highlands Oncology’s experience offers a clear takeaway: Progress comes not from moving fastest, but from building the right structures—and asking the right questions—from the start.
Learn more about ACCC’s AI in Cancer Care initiative at accc-cancer.org/ai. Explore Highlands Oncology’s approach to AI in greater depth during their upcoming webinar on May 21 at noon EDT. Don’t miss the opportunity to hear directly from their team: register here.
Highlands Oncology is an independent, physician-led cancer program based in Northwest Arkansas. Founded in 1996, the practice delivers multidisciplinary oncology care across the cancer continuum, with a longstanding focus on coordinated, community-based care.
Highlands was recognized with a 2025 ACCC Innovator Award for its work in remote patient monitoring, highlighting the organization’s ongoing emphasis on practical, scalable innovation in cancer care delivery.
