How AI Integration at The Ottawa Hospital Enhances Healthcare Efficiency and Patient Care

Clinician and patient satisfaction remain critical challenges in healthcare, with patients often facing difficulties accessing care and clinicians overwhelmed by administrative burdens. The Ottawa Hospital (TOH) addressed these issues by integrating Microsoft’s DAX Copilot, an AI-powered ambient assistant, into their clinical workflow. This innovative solution has already demonstrated substantial benefits, including saving seven minutes per patient encounter, reducing clinician-reported burnout by 70%, and achieving a 93% patient satisfaction rate with care experiences.

DAX Copilot is seamlessly integrated with the Epic electronic health record (EHR) system, widely used in healthcare settings. It captures physician-patient conversations via a mobile app and generates draft clinical notes in real time. Physicians can focus on patient interaction without the distraction of manual documentation, as the AI extracts relevant details such as symptoms, diagnoses, treatment plans, and follow-ups. Clinicians then review and finalize the notes, significantly reducing after-hours charting and documentation.

Microsoft continuously refines DAX Copilot’s large language models (LLMs) using a vast repository of curated clinical data. These models are optimized by medical specialty to ensure accuracy and relevance. Feedback loops compare AI-generated notes with clinician edits to improve performance over time, reducing the need for manual corrections and enhancing trust in the system.

TOH’s robust evaluation framework leverages Microsoft Power BI dashboards to monitor clinician feedback, patient surveys, and EHR data monthly. This continuous assessment guides improvements and ensures the tool effectively reduces administrative workload across physician categories. By alleviating documentation burdens, clinicians experience less burnout and cognitive load, enabling more meaningful patient engagement and increased patient throughput per shift.

Patient consent and privacy are paramount; all patients agree to recordings and have access to their notes via the MyChart portal. Patient receptivity has been overwhelmingly positive, with 97% reporting experiences as good as or better than traditional visits. Enhanced physician-patient interaction facilitated by AI contributes to this satisfaction, despite the compressed nature of clinical appointments.

Looking forward, TOH plans to expand AI capabilities to include biomarker detection and assessment of social determinants of health, addressing factors like nutrition and transportation that impact patient outcomes. Additionally, automating post-visit actions such as orders, pre-authorizations, and referral letters is a key goal to streamline clinical workflows further.

Beyond documentation, TOH is pioneering “digital teammates” like Sophie, an AI avatar capable of multilingual communication and interpreting patient sentiment and behavior. Sophie can detect discrepancies in patient-reported symptoms and engage patients with clarifying questions, enhancing clinical insight. Another avatar is set to assist patients with healthcare navigation and pre-screening, addressing the global healthcare staffing crisis by supporting patient engagement and follow-up.

These digital teammates aim to maximize human touch by optimizing resource allocation and ensuring patients receive comprehensive support throughout their healthcare journey. Future integration with Epic will enable these avatars to interact directly with patient records, enhancing proactive care models rather than reactive responses.

The Ottawa Hospital’s pioneering use of AI demonstrates the transformative potential of ambient AI assistants and digital avatars in healthcare. By reducing administrative burdens, enhancing patient engagement, and addressing staffing shortages, AI is reshaping healthcare delivery toward more efficient, patient-centered, and proactive care.