Medical Journal Webinar Shows Ambivalence On AI, Patient Partnership

A webinar on artificial intelligence and the role of patients that was sponsored by one of the world’s most prestigious medical journals showed marked ambivalence about genuine partnership.

On the one hand, the webinar by NEJM AI, a spin-off of the renowned New England Journal of Medicine, had a title redolent of old-school beneficence: “AI in Health Care – Putting Patients First.” A brief introduction extolled the opportunity for attendees to learn how to enhance the “overall patient experience.” The message about respective roles was unmistakable.

On the other hand, throughout the meeting patient advocates were given the chance to present eye-opening “ground truths” about patients’ AI use. There was, for instance, a blunt declaration of independence by Hugo Campos, a prominent patient advocate who has pioneered the use of digital health tools to promote patient autonomy.

AI, he told the group, enables patients to access reliable information outside the traditional care system. “The shift from a hierarchical to a collaborative model empowers patients with greater autonomy and alters the doctor-patient relationship,” said Campos. “The introduction of generative AI is the most significant step in this evolving relationship.”

Added Andrea Downing, co-founder of the Light Collective, “We need to be at the center of design of AI solutions, to make sure AI is actually improving our care.” In June, the collective published an important monograph entitled, AI Rights for Patients. In addition, she and other advocates urged attendees to follow the hashtag #PatientsUseAI.

The words of Campos, Downing and two other patient advocates (Grace Cordovano and Nina Tousch) particularly resonated with me because of a commentary I wrote for The BMJ in 2017 about the impact of digital health advances on the doctor-patient relationship. Noting that individuals could increasingly “find, create, control and act on an unprecedented breadth and depth of information,” I urged physicians to respond to the erosion of the traditional information asymmetry by embracing a new kind of “collaborative health” partnership. That partnership would be anchored in shared information, shared engagement and shared accountability.

The commentary was rejected by the New England Journal, but was an “Editor’s Choice” at The BMJ (formerly the British Medical Journal). An updated article examining the impact of data democratization and the burgeoning capabilities of AI was recently published by the Petrie-Flom Center at Harvard Law School.

Tellingly, none of the non-patient panelists or audience questioners directly confronted the full implications of the patients’ remarks about AI-enabled empowerment. Dr. Isaac “Zak” Kohane, editor of NEJM AI and chair of Harvard Medical School’s Department of Biomedical Informatics, attempted to get a group of “executive sponsors” of the journal to do so in a panel on “AI and the Patient Experience.” Perhaps in part due to the implicit slant of the panel’s title, he was unsuccessful.

Kohane asked Nancy Colton, vice president of digital platforms and AI at Elevance Health, how the company would use its deep store of data plus AI to help patients navigate their options. She responded that customer service representatives would have an AI assistant.

Kohane asked Jared Pelo, a physician and co-founder of Bionic Health, which is affiliated with Microsoft, about offering equivalent access to its intelligent assistants for patients and clinicians. Pelo responded that it is now easier for patients to access their electronic health record, which will help the relationship between patient and care team.

And when Rajeev Ronanki, chief executive officer of Lyric, predicted that “every one of us down the road will have an AI assistant,” by “every one of us,” he seemed to mean clinicians.

After Downing told the group, “Trust is what’s missing, as is partnership,” I found myself thinking of the words of medical ethicist Dr. Jay Katz, who authored The Silent World of Doctor and Patient some 40 years ago.

“I believe patients can be trusted,” Katz wrote. “If anyone were to contest that belief, I would ask: Can physicians be trusted to make decisions for patients? This book has argued that both must be trusted, but that they can only be trusted if they first learn to trust each other.”