At CHOC Children's, Dr. Steven Martel perceives generative AI not as a revolutionary solution, but as a valuable resource that can alleviate clinician burnout, enhance patient care, and transform the way hospitals approach data.
For many years, hospitals have invested heavily in electronic health records, crafting extensive digital systems aimed at streamlining healthcare delivery. However, this effort has inadvertently left many clinicians overwhelmed by the very technology designed to assist them.
Dr. Martel, who serves as vice president and chief health information officer at CHOC Children's within the Rady Children's Health system, asserts that the emergence of new artificial intelligence tools may finally mitigate these challenges.
"We’ve dedicated decades to embedding electronic health records into healthcare operations," he explained. "Unfortunately, these systems have become primarily structured databases focused on documentation and billing rather than fostering clinical insights." He further noted, "The issue of clinician burnout is largely tied to the cognitive load resulting from trying to extract relevant, actionable data for specific patients."
Martel regards generative AI as more than just a technological upgrade; he sees it as an unprecedented chance to significantly improve the interaction between clinicians and patient data, allowing systems to derive meaningful insights from vast pools of information.
Providing quicker solutions for complicated cases is part of this transformation. "Generative AI systems present one of the most substantial advancements in IT over the past decade that could positively influence clinicians' experiences," said Martel. "Within our health system, we utilize tools that deliver contextually aware insights tailored to specific diseases, enabling clinicians to access structured and unstructured data cohesively."
He elaborated on the tangible benefits already manifesting in their system, noting that complex patient chart reviews, which once took over an hour, can now be efficiently completed in minutes. "Crucially, pertinent information gleaned from various documents is now presented more clearly, complete with annotations and links to original data in a manner that was considered a dream just five years ago."
For healthcare professionals managing busy schedules and critically ill patients, reducing a lengthy review to mere minutes can profoundly influence their workday's rhythm and stress levels. Martel contends that these improvements not only foster efficiency but also enhance patient safety and clinician satisfaction with electronic health records.
However, Martel emphasizes the importance of skepticism regarding AI's potential. "While I am optimistic about the advantages these tools can provide, organizations should approach the narratives surrounding these technologies with caution," he advised. He underlined the necessity for healthcare institutions to comprehend how AI systems are constructed and ensure these tools are suitable for their patient demographics.
"Grasping how these AI models operate, the foundational knowledge informing their development, and evaluating their relevance to your patient population is vital for determining a tool's usefulness," Martel remarked.
This thoughtful approach is becoming increasingly necessary in healthcare, where leaders balance the enthusiasm for AI with concerns about patient safety, inaccuracies, and reliance on machines.
For Martel, the larger challenge involves not just implementing AI technologies but also rethinking the organizational perspective on data itself. "Health system executives must recognize that the real opportunity with AI lies not merely in adoption but in redefining how data is perceived and utilized to generate actionable insights within their organizations," he stated.
To utilize AI effectively, he indicated that healthcare providers need adaptable data platforms and a readiness to innovate, even if certain attempts do not succeed. "Leaders should view data as a valuable asset and establish a flexible data infrastructure that allows for optimal tool utilization," he explained. "Embracing the fact that some tools will fail is more valuable than only celebrating those that work."
Martel added that successful AI implementations often require extensive testing and ongoing refinement. "Cultivating an organizational culture that accepts measured risks, paired with appropriate governance, will enable progress despite the occasional setbacks of some AI projects," he noted. He also pointed out that successful tool deployment frequently originates from iterative testing and adjustments. "Those who expect success simply by activating a tool will likely be disappointed."
Even as AI technology advances, Martel insists that healthcare leaders must prioritize patient safety. "It is crucial that, as AI capabilities expand, we maintain our focus on patient safety within health systems," he stressed. Accordingly, he believes that fully autonomous clinical decision-making should be avoided. "We must preserve the human element in clinical workflows supported by AI, enhancing but not replacing the clinical judgment of caregivers," he advised.
Ultimately, the ongoing discussions within hospitals revolve around trust—trust in the data, in the tools, and in the clinicians who will utilize them. At CHOC Children's, Dr. Martel envisions AI as a means to alleviate the burdens that define contemporary healthcare, provided that hospitals leverage the technology to complement human expertise rather than replacing it.
