user Admin_Adham
20th Jul, 2026 12:00 AM
Test

When the Algorithm Knows the Risk but Misses the Fear

How can we preserve the doctor-patient relationship at a time when AI is making its way into general practice offices? During a session at the 30th WONCA Europe conference 2026, held in Paris, France, Antonio Parata, a general practitioner in Liège and a member of the Scientific Society of General Medicine, advocated for a humanistic approach to AI. In his view, these technologies should enhance clinical practice without ever overshadowing empathy, which he considers a true medical skill.

The doctor begins his discussion with a familiar clinical scenario. Faced with an anxious patient convinced she is destined to suffer the same fate as her father, who died of a heart attack, an algorithm concludes that her cardiovascular risk is low. Two truths then coexist: the statistical truth of the algorithm and the relational truth of the patient’s lived experience.

“Which truth do we trust first?” he asked. A question that, in his view, sums up one of the main challenges of integrating AI into general practice.

The Risk for ‘Soft Dehumanization’

Parata warned against what he calls a “silent peril”: a gradual dehumanization of the consultation. Digital scribes, automated triage systems, and chatbots can certainly improve the organization of care, but they also risk subtly shifting the doctor’s attention from the person to the screen.

“The patient’s story becomes a set of data to be scanned, and the person ends up being replaced by a clinical case,” he summarized. Compounding this trend is significant cognitive overload for practitioners, who are already grappling with the growing complexity of chronic diseases and rising levels of burnout.

SUGGESTED FOR YOU

Empathy as a Clinical Skill

For Parata, empathy is not solely a matter of the doctor’s personality. It is a clinical skill in its own right, drawing on cognitive, emotional, and compassionate dimensions.

He illustrates his point with a patient for whom an AI system suggests depression. Empathic analysis, on the other hand, allows for the integration of factors absent from algorithms — such as financial difficulties or the burden of family caregiving — in order to develop a truly tailored care plan.

AI can, in fact, help strengthen this empathy when it relieves the doctor of certain administrative tasks, for example through the use of medical scribes. Conversely, it can also weaken it if it creates an illusion of certainty or generates so many alerts that it dulls clinical intuition.

A Vulnerability Now Shared by All

Parata proposes thinking of care as a “new ecology” based on three interdependent actors: the patient, with their uncertainty; the physician, faced with complex decision-making and work overload; and the sociotechnical system, with its biases and opacity.

“AI is not neutral,” he pointed out. Incomplete or erroneous data can disrupt the continuity of care and steer decisions in the wrong direction.

An Ethical Compass in 5 Questions

To guide physicians in using these tools, the general practitioner proposes an “ethical compass” structured around five questions:

  1. Does the tool preserve the quality of the listening process?
  2. Who remains responsible for the decision and for explaining it to the patient?
  3. Is the AI capable of accounting for cases of multimorbidity?
  4. Can the clinician easily flag a false positive or a system error?
  5. Does the tool risk exacerbating inequalities in access to care?

In his view, these questions must be considered when evaluating any AI solution deployed in clinical practice.

Learning to Master AI

When asked during the Q&A session about the future role of AI in the training of general practitioners, Parata believes that learning to use it will become essential.

AI is already becoming a part of our lives. We must learn to master it so that it doesn’t master us. He emphasized, however, that it must remain a tool designed to enhance the physician’s capabilities, not to replace them.

When asked about the possibility of “coding” empathy, he was more reserved. While certain technologies, such as emotion recognition or stress level detection, can support the patient-physician relationship, they will never be able to replicate what makes human interaction unique. “Let’s start by coding our own humanity. Our vulnerability is not a weakness; it is the foundation of our moral responsibility,” concluded Parata.

This story was translated from MediQuality, part of the Medscape Professional Network.


Share This Article

Comments

Leave a comment