With the rapid integration of artificial intelligence into healthcare, health systems are navigating a complex landscape of promise and uncertainty. From alleviating physician burnout to unlocking new revenue streams, ambient clinical AI solutions are at the forefront of this transformation. To unravel the real-world impact of these technologies, we sat down with James Maitland, an expert in robotics and IoT applications in medicine. In our conversation, we explored the tangible financial returns health systems are seeing from improved coding accuracy, the delicate balance between reducing administrative burdens and increasing patient loads, and the critical factors that drive successful adoption. We also delved into innovative pricing models that are lowering barriers to entry, the ethical considerations of using AI for billing, and how this technology is profoundly reshaping the sacred physician-patient relationship.
Health systems report that AI scribes can generate over $1,200 in monthly revenue per provider by improving coding accuracy. Can you detail how more complete notes lead to this specific shift in E/M coding and what safeguards are needed to ensure this reflects genuine clinical complexity?
It’s a fascinating and direct consequence of moving from human-led documentation to an ambient AI model. What we’re seeing is not about artificially inflating codes, but about finally capturing the full, nuanced picture of a patient encounter. For instance, at a system like McLeod Health, they observed a significant shift in their Evaluation and Management, or E/M, coding. For established patients, level 3 codes dropped by over 18%, while the more complex level 4 and level 5 codes rose by 7.3% and 5%, respectively. The fundamental reason is that the AI-generated notes are simply more complete. The system captures every detail of the conversation, ensuring that the documentation accurately reflects the true complexity of the patient’s condition and the physician’s medical decision-making. The key safeguard is rigorous internal auditing and physician oversight. The technology is a tool to create a better draft, but the physician must always review and attest to the final note, ensuring it’s a clinically accurate representation of the visit, not just a pathway to a higher CPT code.
A primary goal for adopting AI scribes is to mitigate physician burnout by reducing after-hours documentation, yet some systems also see an organic 20% increase in patient volume. How do you balance giving time back to physicians with the potential pressure to see more patients?
This is the central tension that every healthcare leader grapples with when implementing these tools. The data is compelling; we see a 35% to 65% reduction in after-hours documentation across major health systems. The primary goal, as leaders like Dr. Bryon Frost at McLeod have emphasized, is to combat burnout and restore the joy of practicing medicine. If you simply use that reclaimed time to push physicians to see more patients, you’re not solving the problem—you’re just making the hamster wheel spin faster, which perpetuates burnout. The key is in how you frame the initiative. The most successful adoptions have focused on physician well-being first. The increases in patient volume, like the 18% at McLeod or 22% at FMOL Health, were reported as organic. This means that with less cognitive load from documentation, physicians naturally have more capacity and may choose to see another patient, but it isn’t a top-down mandate. The balance is achieved by making the technology a tool for physician autonomy, not another instrument of productivity pressure.
McLeod Health’s vendor selection process involved a unique bake-off with mock patient encounters. Beyond just the accuracy of the final note, what specific workflow integrations and usability factors were most critical in that evaluation, and how did they predict successful real-world adoption?
That bake-off was a brilliant piece of practical implementation science. While many organizations might focus solely on the final output—the quality of the generated note—McLeod understood that a perfect note is useless if it’s a nightmare to get it into the electronic health record. They brought in actors for mock encounters with real physicians across different specialties, which was a great start. But the second phase was the real differentiator. They tested the final two vendors on their workflow integration directly within their Epic EHR. This is where the true test lies. How many clicks does it take? Is the interface intuitive? Does it interrupt the physician’s natural thought process? In their case, the chosen solution, Suki, absolutely shined in this phase. The reports say that 90% of the physicians chose it based on the workflow alone. This seamless integration is the most critical predictor of adoption because if a tool adds friction to a physician’s day, no matter how good it is, it will be abandoned.
While many AI tools use a flat subscription fee, a utilization-based pricing model has been described as a “game changer.” How does this encounter-based model alter the ROI calculation and help overcome initial cost barriers, especially when trying to achieve system-wide adoption?
The pricing model is absolutely pivotal, and the shift away from a flat per-provider-per-month fee is a massive development. A traditional subscription, which can be $300 or more per license, forces a Chief Medical Information Officer to become a gatekeeper. They have to constantly worry, “Am I getting my money’s worth out of Dr. Smith, who only uses this ten times a month?” It creates an enormous financial and administrative burden. The encounter-based model, where the health system pays a small fee for each use up to a cap, completely flips the script. It democratizes access. A CMIO can provision a license for every single physician in the organization without worrying about the upfront cost or predicting adoption rates. This removes the initial financial barrier and allows the technology to spread organically. The ROI calculation becomes much more direct and less risky; you’re only paying for what is actively generating value, and it allows you to scale the solution from a small pilot to an enterprise-wide standard with incredible flexibility.
Some researchers express concern that ambient AI could inadvertently drive up healthcare spending without improving outcomes. From your perspective, how can health systems leverage this technology to enhance patient care and ensure financial gains are a byproduct of better documentation, not just aggressive coding?
This is a valid and important concern that the industry needs to address head-on. There is a risk, as some researchers have pointed out, that these tools could be used simply to maximize billing without a corresponding improvement in patient health. However, responsible health systems are framing this differently. The financial upside should be seen as a welcome byproduct of the primary goals: reducing burnout and improving the quality of clinical documentation. When a note is more detailed and accurate, it leads to better care continuity, fewer medical errors, and a clearer picture of the patient’s history for the entire care team. Furthermore, these systems can proactively identify care gaps, like an overdue cancer screening, which directly improves patient outcomes while also, yes, generating revenue. The key is a culture of clinical integrity. The focus must remain on using AI to support better medicine, with the financial ROI being the result of that, not the sole driver of the initiative.
Beyond saving time, patient satisfaction scores for “provider listening” reportedly increased by over 6% with this technology. Could you share an anecdote or explain the mechanism by which an AI scribe actually enhances the physician-patient interaction and builds trust during a visit?
It’s one of the most powerful, though often overlooked, benefits. That 6.3% jump in patients feeling their provider is listening and trusting them is incredibly significant. The mechanism is simple and profound: it removes the computer screen as a barrier between two human beings. Think about a typical visit. The doctor is often forced to divide their attention, typing furiously, clicking through boxes, their eyes glued to the monitor. This creates a physical and emotional distance. With an ambient scribe, the keyboard is gone. The physician can turn their full attention to the patient, make eye contact, lean in, and listen actively. They can be fully present. This small change completely transforms the dynamic of the exam room from a data-entry session into a genuine human connection. That active, uninterrupted listening is the very foundation of trust in the physician-patient relationship, and the technology, ironically, is what facilitates that return to a more traditional, personal style of medicine.
What is your forecast for ambient clinical AI?
I am incredibly optimistic that we are just scratching the surface. Right now, the focus is on documentation and coding, which is a fantastic start, but the true evolution will be when these tools mature from scribes into genuine clinical assistants. The future I envision is one where the days of paternalistic medicine are over. Imagine a scenario where a patient newly diagnosed with cancer is sitting with their oncologist. The oncologist outlines a treatment plan, but then they both—doctor and patient—turn to the AI and ask, “Is this a good plan?” The AI could then respond, “Actually, based on the patient’s specific genomic markers, there are two new clinical trials they would qualify for that offer a promising new treatment.” This is the personalized medicine we’ve been dreaming of. It will transform the relationship into a true partnership between the patient, the doctor, and an AI, all working together to find the absolute best path forward. That’s the future these platforms are poised to deliver.
