Who pays when AI steers your doctor wrong?

HISTORY IS REPEATING ITSELF

Imagine how epic EHRs would have been if the current companies were founded by healthcare providers, designed by healthcare providers, and owned by healthcare providers with controlling interests in the companies. As a nation we might have actually seen meaningful use achieved. The promise of less administrative burdens, improved efficiency, and improved outcomes in healthcare may have actually been met.

HEALTHCARE PROVIDERS NEED TO RAPIDLY ENGAGE WITH AI

https://www.politico.com/newsletters/future-pulse/2023/11/30/if-ai-steers-a-doctor-wrong-who-pays-00129242

This article was something I first saw Sanat Dixit MD, MBA, FACS post about last week. Politico focused primarily on the liability aspects of using AI in healthcare and who would compensate patients. That’s unfortunate because in a world using AI appropriately, this question becomes a moot point.?

Appropriate use of AI in healthcare involves humans in the loop and those humans need to already be skilled at what they do - new patient care coordination, social work, nursing, advanced practice care, surgery, etc. I see far too many companies out there who hire these humans as consultants, frequently after primary design, and far too few who employ these providers in the day to day optimization of their products.

HEALTHCARE PROVIDERS NEED TO DEMAND THEIR VOICES BE HEARD IN THE DESIGN AND IMPLEMENTATION OF THESE SYSTEMS

Incorporating humans in the loop in AI systems, especially in healthcare, can significantly mitigate the risks associated with AI errors and enhance the reliability of the care delivered.?

Here's how humans in the loop help address the problem of AI steering doctors wrong:

VERIFICATION AND OVERSIGHT:

By having healthcare professionals verify AI recommendations before making clinical decisions, the risk of errors can be significantly reduced. This human oversight ensures that AI-generated advice is scrutinized and validated against medical knowledge and patient-specific contexts, thereby preventing potential mistakes from translating into clinical actions.

COMBINING HUMAN EXPERTISE WITH AI:

Humans in the loop bring a level of understanding, empathy, and contextual judgment that AI currently cannot replicate. By combining AI's data-processing capabilities with human expertise, the overall decision-making process becomes more robust, nuanced, and patient-centered. This collaborative approach leverages the strengths of both AI and human intelligence, enhancing the quality of care.

CONTINUOUS LEARNING AND IMPROVEMENT:

AI systems with humans in the loop can continuously learn from the oversight provided by healthcare professionals. Each intervention by a human can be used as a feedback mechanism to improve the AI's algorithms, making the system more accurate and reliable over time. This iterative process ensures that the AI system evolves in alignment with real-world clinical scenarios and standards of care.

REDUCING LIABILITY AND ETHICAL CONCERNS:

The presence of humans in the loop can address legal and ethical concerns related to AI in healthcare. By maintaining human responsibility and accountability in critical decision-making processes, this approach can clarify liability issues when errors occur, ensuring that there is always a human accountable for clinical decisions.

ENHANCING PATIENT TRUST:

Knowing that a qualified healthcare professional is overseeing and verifying AI-generated recommendations can increase patients' trust in the healthcare system. This human element reassures patients that their care is not solely in the hands of an algorithm, fostering a more trusting patient-provider relationship.

ADAPTABILITY TO COMPLEX CASES:

AI algorithms might not perform well in atypical or complex cases where training data is limited. Humans in the loop can identify such cases and take over decision-making, ensuring that patient care remains consistent and high-quality even when AI falls short.

ADDRESSING ETHICAL AND SOCIETAL VALUES:

Healthcare decisions are often influenced by ethical considerations and societal values, areas where AI might not be fully equipped to navigate. Human involvement ensures that decisions are made in a manner that respects ethical norms and aligns with societal expectations.

EMPOWERING HEALTHCARE WITH PROVIDER-CENTRIC AI INTEGRATION

Reflecting again on the trajectory of Electronic Health Records (EHRs), it's evident that with foresight, we could have navigated around the pitfalls that have contributed to the current challenges in healthcare—physician burnout, escalating administrative burden, and diminished access to care. These issues stem from systems designed without direct input from those in the trenches of healthcare, pushed by those prioritizing profit over patient and provider welfare.

Incorporating healthcare providers as integral components in the AI framework, "humans in the loop," presents a transformative solution. This approach bridges the gap between AI's potential and the intricate realities of healthcare, ensuring AI complements rather than replaces professional expertise, improving safety, reliability, and integration in medical practices.

The question now is, are you ready to pioneer this vital shift? Engaging directly with AI use and development can catalyze a healthcare environment where technology truly serves the patient and the provider, aligning with the core mission of medical care. Your expertise and firsthand experience are invaluable in shaping AI tools that enhance, rather than hinder, healthcare delivery. This is an invitation to lead the charge in redefining healthcare through a harmonious blend of human insight and artificial intelligence.

Fire up!