Does Artificial Intelligence (#AI) Chatbot Outperform Physicians in Patient Experience?

JAMA Article Draws Fire for Its Research Biases - But Should We Ignore Its Conclusions Altogether?

A recent Journal of the American Medical Association (JAMA) study (summary hyper-linked) found that #ChatGPT outperforms physicians in counseling patients. The complete research compares written responses from physicians and ChatGPT to real-world health patient-directed questions. It's rocked quite a few boats in the medical community. Some within that community are threatened, and others are reflective.

A panel of licensed healthcare professionals preferred ChatGPT responses 79% of the time and rated ChatGPT responses as higher quality and more empathetic. Gulp. Understandably, some doctors are not happy with this study. And many were not pleased with me for not diving deeper into the complexities inherent in the research in my initial LinkedIn post. Message heard. Understood!

The news headlines and the initial study callouts overplay the immediate importance of ChatGPT in the physician-patient relationship. Physicians do not fair poorly.  However, the authors provide an inflection point that should not be ignored and must be acknowledged – Communication is Part of the Care and Cure! Physicians must be trained and have time to deal with patient curiosity and urgencies. #Patientexperience is different. They do not want to sit idle or silent. They are curious and concerned.

As generations have become more familiar with technology in their day-to-day lives, perhaps they place more trust in machines' “unbiased nature” over humans. That assumption has led to the rise of "#misinformation." We believe our Twitter feeds if we don't explore the facts further. But, our screens reduce the press of needing to engage with people at the moment - they give us time to think and check in with this "on-call" information aggregator. To let the information sink in without being confronted about the next step. Doctors are too often pressured into an eight-minute per-patient provider reimbursement model. It's not their fault - it's the system that they must co-exist within. But that tilted system leads to the consumer seeking - needing - alternatives. If so, even imperfect ChatGPT4 and beyond will be a go-to.

There are changes afoot that we need to make happen sooner rather than later by moving minds, systems, and behaviors so that life-sustaining and life-saving approaches to patient care may eventually tip the scale of human survival toward health and wellness. However, we see data from a human perspective – sometimes self-interests or emotional needs for control. ChatGPT is the aggregate of data and human input. It is not divorced from us but a faint mirror of the human experience.

Yes, this study is worth reading. Yes, many have criticized its design and the intent of the authors. Yes, many are fearful that machines may replace physicians. But, the latter assumption is doubtful. Reading between the lines reinforces that, as industry colleague Ritesh Patel often says, “If it moves, digitize it!” People get their information in ways that are quick and convenient. That is a reality everyone in the health community must face!

The medical community and health communicators must rise to the moment if they want to harness this technology. Learn about ChatGPT and how it operates - its prompts. Also, read words from experts on the digital health news platform Medika Life including the insightful words by innovation theorist John Nosta . John will rock your boat; however, often, he points to where this is going.  Read the words of Tom Lawry , former head of Microsoft's AI team, author of the best-seller Hacking Healthcare, and a global counselor on the practical application of AI.

Almost one year ago, I penned a piece titled: Health Possibilities We Cannot Afford to Block. There were 10 ideas/technologies included in that piece - #1 was #AI. That's the heart of ChatGPT. Fixing one part of the healthcare puzzle is encouraging - but is it transformational? What can we do to make things work better for patients? Medicine can harness the power of ChatGPT to make it work even better for patients seeking healing solutions. Perhaps we can give physicians more time to help patients feel their doctors have and always are among their greatest advocates. We can also bring technology companies and leading medical associations together to talk about ChatGPT influence on trusted people-to-people connections, particularly with physician-patients.

Why do consumers turn to machines instead of people for medical counsel?  Well, we haven't been able to clone or at least develop teaching models drawing upon the many outstanding physicians who demonstrate incredible patience and empathy for patient woes and questions - doctors like WebMD's John Whyte and NHS's and Microsoft's Junaid Bajwa —many answers to consider. Among the most important are skill, collaboration and empathy.

Consumers may feel that devices are better listeners and work with them in partnership. We should expect this outcome due to the fragmented health ecosystem that consumers must navigate with difficulty. We must recognize that ChatGPT's interest and popularity among health information seekers didn't just happen. It is possible to realize that these same information seekers feel they are not getting what they seek.

Keep learning! This is not the end of humanity and the beginning of the Matrix - where people, software and machine battle for survival. The world will be changing in amazing ways in the short years ahead. Collaboration and communications go hand-in-hand as essential tools for healing.