When thinking about artificial intelligence (AI) in healthcare, it may seem counterintuitive to see its potential to humanize the field—but that’s precisely the promise of ambient listening and AI-powered scribe technologies. These powerful tools can alleviate hours of documentation burden for clinicians, allowing them to focus less on paperwork and more on making meaningful connections with their patients.
I often find myself advocating for the importance of staying updated with new technologies. Last year, I was taken aback by the overwhelming interest from my colleagues in a virtual demonstration of an AI scribing tool. This technology has been expanding at a pace rivaling, if not surpassing, that of virtual care platforms during the COVID era. For years, doctors have voiced their frustrations about the documentation burden that encroaches on their personal time, even haunting their dreams. Now, at last, there's a solution that promises to reclaim our time.
Improving documentation efficiency: ambient listening and AI scribes
Ambient listening and AI scribe tools work in tandem to capture and document conversations between patients and their healthcare providers. Ambient listening employs microphones and sensors to “listen” to the conversation while AI models collect and distill the pertinent information in real time into structured notes, summaries, or actionable insights. This data can be added to the patient’s electronic health record (EHR) via an app on the provider’s phone, tablet, or computer.
While dictation tools have been used for some time, they’ve historically required the provider to dictate directly to the program. This either hinders the natural flow of conversation or requires the provider to wait until the patient has left the room to make their dictation. By contrast, newer solutions primarily work passively in the background. Studies show family physicians spend an average of 45% of their clinical hours interacting with a computer rather than a patient, so outsourcing documentation to AI could yield significant improvements in clinician well-being and more meaningful clinician-patient interactions.
The rise of AI scribes: benefits, challenges, and the future of healthcare
According to Dr. Scott MacDonald in a recent episode of Nordic’s "Designing for Health" podcast, “A lot of patients feel [AI has] improved the engagement and interaction with the physician. A lot of times, I don’t even log into the computer, I’m just sitting straight in front of a patient, looking them in the eye. I know something else is taking care of my documentation, so I don’t need to worry about that.”
Providers who use AI-powered scribes also experience a notable increase in wellbeing, with one study showing that 65% of physicians surveyed felt that using a scribe had enhanced their well-being.
Automating documentation helps clinicians “put patients before paperwork,” reduces the administrative burden, and improves care continuity by ensuring comprehensive documentation.
AI models, however, aren’t immune to misinterpretation. To ensure the accuracy of information in a patient’s EHR, it's still essential for providers to review and adjust the AI analysis. On occasion, it may be necessary for the provider to speak directly to the program (like older dictation technologies). As noted by Dr. MacDonald, important information is often communicated non-verbally, such as a patient pointing to a body part to indicate where they’re experiencing pain, and AI models can’t interpret these non-verbal cues. The need to speak directly to the program can be awkward and change the dynamic of the appointment.
However, unlike dictation technologies, AI has the potential to provide real-time analysis and recommendations, which could lead to faster diagnoses and treatment planning. For example, future iterations of these systems might detect early signs of conditions like sepsis or mental health crises during routine consultations, allowing for proactive and early treatment.
Overcoming the challenges of AI implementation in Canadian healthcare
While these tools help ease the burden on strained healthcare workers and have excellent potential for improving patient care, successfully implementing them across the Canadian health system comes with significant challenges, including cost, the need for regulation, and training and education for providers and patients.
Organizations implementing AI technologies will be responsible for training their employees. Leaders will also need to manage the cultural shift of integrating AI into clinical workflows, providers will need to understand the risks and benefits to ensure patients understand and are comfortable with their use and are able to confidently provide their consent. Transparency around how the technology works, its benefits, and how the organization will store and protect patient information is essential. Demystifying AI and emphasizing its role in enhancing, rather than replacing, human care will be essential for providers and patients alike.
Another significant challenge in implementing AI technologies is the lack of government regulation. While Canada has federal guidelines for generative AI, according to the Canadian Medical Association Journal, AI scribes are currently unregulated. This means “clinicians must understand how clinical information captured by the software is stored, retained, accessed, and subsequently used...The onus is on the clinician or their institution to understand privacy implications and potential for patient harm, and to ensure the software performs as intended.” Until government regulations are in place, this could be a prohibitive burden for smaller organizations and clinics.
Lastly, the cost of implementing these systems is a major hurdle—costing a minimum of $20,000 USD for a minimum viable product to up to $1,000,000 for a custom solution—and most deployments are limited to physicians, sidelining nurses and other healthcare professionals. Democratizing access will be essential to maximizing the technology’s impact.
Conclusion: recovering the human touch in healthcare through AI
Clinician burnout is a major challenge in Canada. While there are several factors contributing to this, including understaffing and post-pandemic exhaustion, AI scribe and ambient listening technologies are showing great promise for helping to ease the administrative burden many feel and enhance the patient-provider connection. To realize the full potential of these solutions, we must clear a path for proper training, change management, and governance, and figure out ways to make the technology more cost-effective and accessible for all.
Interested in integrating AI tools into your clinical workflows, but don’t know where to start? Contact a Nordic consultant today to discuss your goals and priorities.