Jackie Njoroge is VP, Digital Transformation and Operational Excellence at US Oncology.
Healthcare is complex. Within healthcare, oncology is even more complex. A patient’s journey starts before the first appointment, with a referral that triggers the movement of records across multiple facilities and systems. By the time a patient sits down with an oncologist, the care team may have already navigated prior authorization and assembled clinical documentation from multiple sources. Then treatment begins, bringing coordination across providers and billing.
I lead digital transformation and operational excellence for a physician-owned network that includes oncology and several specialty care disciplines. As we leverage technology to help enable better patient outcomes, we find that some of the most impactful AI work we are doing is happening where care is being delivered.
Where AI Is Already Making a Difference
One of the first areas where we focused was ambient scribing. AI tools handle real-time documentation so that after the visit, the provider has an intelligently transcribed record ready for review and doesn’t have to spend significant time compiling notes by hand. The clinician still signs off on every note, but the process takes much less time and effort, so their attention can be entirely on the patient during the appointment.
We have also built AI tools for smart preparation before visits. If a provider is seeing 10 patients in a day, they need a concise summary of each patient before walking into the room, one that draws from records across multiple systems. What used to fall on the care team to pull together manually can now happen automatically through AI, giving time back to the care teams.
Oncology scheduling can be extremely complex: A patient may need labs, then pretreatment steps, then infusions timed precisely relative to what came before. Managing that sequencing manually takes significant effort to ensure patients get the best care possible. Layering AI on top of scheduling systems can help recognize patterns and identify next best actions more efficiently.
We are building toward a model where a patient calls our contact center, and an agentic AI handles intake and schedules the visit using that same logic. When you have a cancer diagnosis, every day matters. How many days of unnecessary waiting could be eliminated by a scheduling system that sees the entire calendar and sequences automatically?
AI is also changing how we approach billing and care navigation. The thoughtful use of AI in prior authorization and billing workflows can help streamline administrative processes and reduce the delays and friction patients may face at an already difficult time.
These solutions don’t just benefit patients; they also help prevent provider burnout. Our providers need to be fully present with a patient who may be having one of the worst days of their life, and anything that simplifies the process helps make that happen.
Earning Clinician Buy-In
How do you build trust in AI broadly, given that AI-generated outputs can sometimes be inaccurate or incomplete and clinicians operate in high-stakes care settings? Rather than asking clinicians to trust any AI tool blindly, you build systems and processes that earn their trust.
That starts with involving clinicians in building the system. We have strong clinician tech champions who give feedback on our existing workflows and help define what good outputs look like. When we rolled out ambient scribing, early on, some people may have doubted it would ever work. AI systems perform best when they are carefully configured, tested and refined with input from the people who use them. As the industry has matured significantly, and large language models (LLMs) now arrive with substantial base knowledge built in, we have been able to accelerate adoption in a way that is compliant and improves both efficiency and experience.
This Is Transformation, Not A Purchase
The broader lesson for all of healthcare is this: If you approach AI as a technology purchase, it will underdeliver. This is organizational transformation, meaning you need to start with the business problem. Identify the repeatable tasks where AI can free up human capacity for higher-judgment, more empathetic work. Run small technology/AI options alongside deliberate process design.
Organizations that spend too long building the perfect plan will find the tools have moved on before they can deploy them. The real goal is a more intelligent and responsive healthcare organization, with providers who have the tools and bandwidth to give the best care possible every day. With thoughtful and careful implementation, AI can help you get there.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
