Seattle Children’s AI Assistant & Governance
- Seattle Children's is leveraging artificial intelligence to enhance clinical decision-making,deploying an AI assistant that achieves over 98% accuracy after tuning and validation.Dr.
- The AI assistant, currently being rolled out in emergency and urgent care settings, doesn't simply provide answers.
- Notably, the tool operates independently of the organization’s EHR. This design choice allows Seattle Children’s to avoid integration complexity while refining the tool’s performance and also facilitates potential...
Seattle Children’s is revolutionizing pediatric care with an advanced AI assistant, achieving impressive accuracy in emergency and urgent care settings. this innovative tool,centered on a clinician-first approach,engages physicians with targeted questions to enhance patient understanding before offering recommendations—a notable leap in clinical decision-making. The AI operates independently of the EHR, streamlining integration and paving the way for future collaboration. Furthermore, they’ve revamped governance structures for faster, more informed decisions, fostering cross-disciplinary input. discover how their EHR support Optimization Program and specialized committees, involving doctors, nurses, and administrators, are driving improved outcomes and decisive action.News Directory 3 can’t help but notice the proactive governance model at play.Learn how seattle Children’s is reshaping the future of healthcare with its human-centered, outcome-driven approach. Discover what’s next…
Seattle Children’s Builds AI Assistant to Advance Pediatric Care
Table of Contents
Seattle Children’s is leveraging artificial intelligence to enhance clinical decision-making,deploying an AI assistant that achieves over 98% accuracy after tuning and validation.Dr. Lin, who still practices primary care pediatrics, emphasized the importance of a clinician-first approach to technology design.
The AI assistant, currently being rolled out in emergency and urgent care settings, doesn’t simply provide answers. Instead, it engages clinicians with questions to better understand a patient’s condition before offering recommendations. This approach ensures that the tool aligns with real-world physician use cases.
Notably, the tool operates independently of the organization’s EHR. This design choice allows Seattle Children’s to avoid integration complexity while refining the tool’s performance and also facilitates potential future sharing with other organizations.
According to Lin, many clinical judgments occur in real-time, before documentation in the EHR. The AI assistant captures these nuances,enhancing the quality of care. While not a diagnostic tool, it redirects clinicians if the provided details doesn’t match the intended pathway. As a notable example, the bruising pathway aids in distinguishing between anticoagulation issues, trauma, and potential abuse through context-driven questions.
Human-Centered Design and IT Culture
Seattle Children’s CIO Zafar Chaudhry has fostered a culture where IT teams shadow clinicians to understand workflows firsthand. This initiative is supported by a “tech pod” at the hospital entrance, where clinicians can receive immediate tech support.
One of Lin’s key achievements is the EHR Support Optimization Program (ESOP), which replaced a post-implementation group of physician superusers with a smaller, outcome-focused team. Under ESOP, providers are funded annually to complete optimization projects within their specialties, receiving mentorship and analyst support.
one endocrinologist, for example, developed a proactive system to remind families of diabetic children to request insulin management letters over the summer, significantly reducing last-minute requests. Another project involved population health management, leading to a disease-specific registry.
Reimagining Governance for Faster Decisions
Lin also helped design a governance structure that emphasizes action and cross-disciplinary input. This includes specialized EHR committees for ambulatory,inpatient,perioperative,and integrated workflows,each co-chaired by a nurse informaticist,a physician informaticist,and an IT applications manager.
These committees include physicians, nurses, administrators, and representatives from revenue cycle, compliance, privacy, and medical records. Decisions are made by majority vote, promoting openness and healthy debate. co-chairs meet in advance to review agenda items, ensuring urgent decisions are addressed promptly.
This approach has created a decisive IT surroundings, with stakeholders now valuing the effectiveness of governance meetings.
Key Takeaways
- Deploy AI responsibly, focusing on validated clinical content.
- Keep clinicians involved in the design and implementation of new tools.
- Streamline support programs to focus on specific, outcome-driven projects.
- Formalize governance with clear voting authority and stakeholder diversity.
- Emphasize decisiveness through majority voting and transparent debate.
Lin emphasized that the focus should always remain on the human element, whether it’s the physician, clinician, or patient.