AI Scribes Slash Admin Time, Offering a Lifeline for Mental Health Care

OTTAWA, ON – November 25, 2025 – In a healthcare system grappling with unprecedented demand for mental health services and widespread clinician burnout, a new report from Ottawa offers a compelling case for technological intervention. Data from a four-month evaluation at the Ottawa Institute of Cognitive Behavioural Therapy (OICBT) reveals that the use of an AI-powered medical scribe, Heidi, slashed clinician documentation time by a staggering 66 percent and significantly improved work-life balance. The findings signal a crucial development in the quest for sustainable healthcare models, where artificial intelligence acts not as a replacement for human clinicians, but as a vital partner in their work.

The Crushing Weight of Administrative Burden

For years, the narrative around healthcare has been dominated by the twin crises of rising patient needs and a shrinking, exhausted workforce. In mental health, this pressure is particularly acute. Clinicians often spend hours each week on administrative tasks—transcribing session notes, writing referral letters, and completing forms—long after their last patient has gone home. This “pajama time,” as it's often called, is a leading contributor to burnout, detracting from both patient care and personal well-being.

OICBT, a prominent Canadian psychology practice, sought to address this challenge head-on. Their evaluation of Heidi, an AI tool designed to automate clinical administration, was born from a commitment to finding more sustainable ways of working. The results were immediate and profound.

"Heidi is helping our clinicians restore balance and presence in their work," stated Dr. Pete Kelly, a clinical psychologist and the chief financial officer at OICBT. "By introducing efficiencies that reduce after-hours documentation, our team can focus on patient care instead of paperwork. We're seeing more engagement, less burnout, and a foundation for a more sustainable practice." This sentiment captures the core promise of such technology: to give clinicians back their most valuable resource—time—and allow them to reinvest it in direct patient interaction and their own recovery.

A Case Study in Measured Success

The data from OICBT’s evaluation provides a quantitative look at the AI scribe's impact. Clinicians saw their average weekly documentation time plummet from 213 minutes to just 71 minutes. This 66 percent reduction is not just a statistical victory; it represents over two hours of administrative work eliminated from each clinician's week. The effect on morale was equally significant. Using a five-point scale to measure satisfaction with work-life balance, OICBT reported a jump from 55 percent to 83 percent among participating clinicians.

Beyond the numbers, the qualitative feedback highlighted a deeper benefit. Therapists reported experiencing a lower cognitive load during their workday. By offloading the task of real-time note-taking to the AI, they could maintain greater focus and feel more present and engaged during sensitive patient conversations. While the OICBT data comes from an internal evaluation rather than a formal, peer-reviewed study, it serves as a powerful real-world proof-of-concept that is being echoed in other healthcare settings. Similar trials in the UK have also reported drops of over 50 percent in documentation time, suggesting the trend is both replicable and scalable.

Navigating the Ethical Frontier of AI in Therapy

The introduction of AI into the deeply human and confidential space of therapy rightly raises important ethical and regulatory questions. For any AI tool to succeed in this domain, it must not only be efficient but also trustworthy. The primary concern is the potential for technology to compromise the sacred patient-clinician relationship or violate stringent privacy standards.

Heidi and OICBT appear to have navigated this landscape with deliberate care. The platform emphasizes its compliance with all relevant Canadian privacy laws, including the federal Personal Information Protection and Electronic Documents Act (PIPEDA) and Ontario's Personal Health Information Protection Act (PHIPA). Critically for Canadian providers, all patient data is stored within Canada, addressing data sovereignty concerns. Furthermore, the system is designed to obtain patient consent and gives clinicians control over data retention, including options for auto-deletion of sensitive information.

Dr. Thomas Kelly, CEO and Co-Founder of Heidi, framed the technology as a supportive, not disruptive, force. "Therapy demands a sustained focus and presence with patients. As an AI care partner, Heidi is helping ease administrative strain without compromising human connection," he explained. This positions the technology as a tool that operates in the background, much like a human scribe, allowing the therapist to remain the central, empathetic figure in the room. Ultimately, the clinician remains legally and ethically responsible for the final clinical note, reinforcing that the AI is an assistant, not an autonomous decision-maker.

Reshaping the Future of Clinical Practice

The success at OICBT is a microcosm of a broader shift in healthcare. The market for AI medical scribes is expanding rapidly as health systems recognize their potential to build capacity. Heidi itself already supports over two million patient consultations per month in Canada and differentiates itself with features like transcription and translation in over 115 languages—a significant asset in a multicultural society. By offering core functions for free, it also lowers the barrier to entry for smaller practices and solo practitioners who have been historically priced out of major tech upgrades.

Looking forward, the impact of this technology extends beyond simple efficiency. It has the potential to fundamentally reshape clinical practice and training. By automating rote administrative tasks, AI allows clinicians to operate at the top of their license, focusing on complex diagnosis, therapeutic alliance, and care planning. Recognizing this, OICBT plans to expand its use of Heidi into the clinical supervision of early-career psychologists, report writing workshops, and other training programs. This integration signals a future where proficiency with AI tools will be a core competency for healthcare professionals.

The goal is not to replace the human element of care but to unburden it. As AI partners become more integrated into clinical workflows, they could free up the capacity needed to reduce waitlists, improve care quality, and create a more sustainable and rewarding profession for the dedicated individuals on the front lines of mental health.