Health

Weaving the AI Threads Together: Essential Conversations for Integrating AI in Healthcare Transformation

The following is a guest article by Demetri Giannikopoulos, Vice President of Innovation at Aidoc

In the ever-evolving landscape of AI, 2024 stands out as the year where artificial intelligence is poised to claim the coveted title of Merriam-Webster’s word of the year. Nowhere is the buzz around AI more palpable than in healthcare circles, where the promise of AI to revolutionize a complex and, in many ways, fragmented system looms large.

Despite notable successes, many AI solutions have fallen short, showcasing potential only in specific areas or departments. Why? Often, AI is treated as a technological bandage. Teams chasing algorithms promising immediate impact, akin to buying a tool without a clear plan for its utilization. This approach overlooks the strategic groundwork essential for successful AI integration. Like healthcare, pathways to clinical AI integration are diverse and nuanced, not one-size-fits-all. The true power of AI lies not in the algorithms alone but in building an overarching AI strategy.

To effectively implement clinical AI, health systems must engage in four pivotal conversations with their external AI partners: human factors, infrastructure considerations, emerging AI regulations, and AI effectiveness. These conversations form the cornerstone for unlocking AI’s transformative potential in healthcare. And yes, we say partners versus vendors, given the vital role these companies play in enabling a health system’s success with AI, and the value they need to deliver for healthcare today – not five years in the future.

Considering Human Factors

First and foremost, AI must be useful, usable, and used. While the initial implementation of AI solutions may spark enthusiasm, long-term success depends on robust adoption. The irony with AI is that for a tool designed to help people, it requires people to be successful. One of the key determinants is the day-to-day user experience: Does it simplify tasks or add complexity? Is it accurate or does it require additional effort? Is it user-friendly or cumbersome?

Inadequate integration with existing workflows, a lack of transparency in AI decision-making, and inconsistent user interfaces can stifle adoption rates. Merely installing a solution and expecting users to embrace it is not a recipe for success. It’s essential to partner with an AI company that offers an effective success team with a clear process. A process that enables change management and provides ongoing support with a focus on demonstrating the true value of AI within your environment.

A customer success team must work hand-in-hand with healthcare professionals to ensure that the AI solution is not only integrated seamlessly into existing workflows but also continuously optimized for maximum efficiency and user satisfaction. This proactive approach can significantly enhance the chances of successful AI adoption and long-term benefits for healthcare organizations.

Evaluating Infrastructure Realties

Shifting our focus to the second critical conversation – the infrastructure imperative. The delivery of AI outcomes marks just the beginning. To truly impact clinical workflows, AI must dovetail with current systems—from PACS to EHRs to scheduling—and offer enhancements where integration isn’t adequate. It should also bridge traditional information silos, creating an all-encompassing patient overview and profile. 

With finite resources and growing AI complexity, health systems need a lean method for technology management and deployment. This entails understanding how the AI partner will support the implementation of AI at scale and ensure adoption by end-users without requiring any ‘tricks’ to make it work. A reliable partner should demonstrate how they minimize downtime, reduce IT lift, and mitigate risk. This includes showcasing a solid record in real-world clinical settings and implementing AI grounded in proven applications.

Preparing for Regulation

Regulation is the third conversation, and if your partner isn’t bringing it up, it might be time to consider other partners. The pace of AI regulation is accelerating, with guidelines from professional societies and existing or imminent government regulations adding complexity. The President’s focus on AI, particularly in healthcare, through initiatives like the Executive Order and the payer/provider task force, underscores the importance of compliance.

Having a knowledgeable partner who not only follows evolving regulations but also adheres to suggested guidelines, even those that don’t directly apply, is crucial to safeguarding a hospital’s AI investment. Recent updates like the 9 questions in ONC’s HTI-1, the well-thought value assessment questions from the recent “Multi-Society Statement from the ACR, CAR, ESR, RANZCR and RSNA and prior publications like the ECLAIR guidelines provide a great starting point for questions health systems should be asking.

Measuring AI’s Impact

Finally, organizations must see the impact of AI implementation. This requires a partner who can collaborate to assess not only immediate effects, such as faster notification and more efficient management, but also the second and third-order effects that may not be immediately apparent. For example, detecting an aneurysm incidentally could prompt more active management, potentially avoiding adverse outcomes in a certain percentage of patients. Understanding these pathways and outcomes is crucial for fully quantifying the financial impact of AI interventions.

Consider how early intervention due to AI could lead to preventive treatments and strategies, compared to late-stage diagnoses requiring intensive care. Understanding these connections allows for the recognition of patient and financial betterment rooted in AI-initiated early detection and proactive treatment. Taking a holistic approach to these value touchpoints enables organizations to attribute patient and financial outcomes to the initial detection and active management facilitated by AI.

Understanding how your AI partner collaborates with healthcare facilities to analyze pathways and patient outcomes or how they assess the downstream impact of their AI solutions cannot only shift AI from a cost to a profit center but also keep the patient at the center of care.

As healthcare evolves, AI will play a critical role in driving innovation and improving patient outcomes. Historically, AI has fallen short due to a lack of clear criteria to vet external partners. By concentrating on these crucial discussions and selecting an AI partner who can deliver on these key areas, health systems can ensure they are prepared to harness the power of AI to transform healthcare delivery.

About Demetri Giannikopoulos

Demetri Giannikopoulos, VP of Innovation, brings over two decades of experience in healthcare technology, specializing in implementation, workflow optimization, and interdisciplinary team development. He is actively involved in the healthcare community and has served on various committees, including the Coalition for Imaging & Bioengineering Research (CIBR) executive steering Committee (2018-2020) and the American College of Radiology (ACR) Patient and Family Advisory Council. Demetri graduated cum laude with a B.Sc. in Computer Science from Florida State University.

Get Fresh Healthcare & IT Stories Delivered Daily

Join thousands of your healthcare & HealthIT peers who subscribe to our daily newsletter.