🩺 Why traditional training isn’t enough
Motivational interviewing works: it had a clinically relevant effect in around three out of four trials from a meta-analysis of 72 studies (Rubak et al., 2005). The problem is learning it. A workshop improves skill, but without subsequent feedback it erodes within about six months; three or four feedback sessions sustain it (Schwalbe et al., 2014). And that feedback requires observing and scoring every conversation — expert time that no centre can give to every group.
Meanwhile, communication failures were involved in 30 % of malpractice claims analysed by CRICO (STAT, 2016).
🔄 Advantages with a simulated patient
Every limitation of traditional training has a response in practice with a simulated patient.
| | 🏛️ Traditional training (workshop, role-play, actors) | 🤖 With an AI simulated patient | | :------------ | :--------------------------------------------------------- | :----------------------------------------------------- | | 🏋️ Practice | A few hours, depending on the group’s schedule | Unlimited, at any time | | ⚡ Feedback | After the session and depending on who observes | Immediate, turn by turn, with explicit criteria | | 🛡️ Risk | Mistakes in front of peers or real patients | Mistakes without consequences | | ⚖️ Equality | Each student has a different experience | The same case for the whole group | | 📋 Assessment | The teacher’s impression | Evidence from every conversation |
In Entrena-T, that feedback doesn’t come at the end, but during the conversation: a set of signals based on motivational interviewing criteria updates with each intervention and shows, in real time, whether the patient feels heard or is becoming defensive.
More practice and better feedback is precisely what evidence identifies as key to sustaining the skill.
💶 Economic justification
The problem is costly: lack of adherence to treatment is estimated to cost around €11.25 billion per year in Spain (CGCOF) and up to €125 billion in the European Union (Frontiers in Pharmacology, 2022). Not everything can be solved by talking better, but conversation is a lever that is almost never trained rigorously.
Training it today is also expensive: actors, rooms and teaching hours grow with every student and every repetition. In a crossover trial at two British medical schools, a session with an AI simulator cost 46 % less per student than with actors (£33.48 vs £61.75), although students rated the actor session more highly (Tyrrell et al., 2025). With a simulator, every additional repetition costs very little and teaching time shifts from scoring to mentoring.
🔬 Rigour
Entrena-T doesn’t replace the actor or the teacher: it complements them, and the final assessment remains human. Moreover, evidence on virtual patients with AI is still young: a 2025 review identified 28 studies and describes the field as early-stage (JMIR, 2025). That’s why we propose starting with a pilot of Entrena-T, with measured outcomes, at no cost.
💭 Reflection
Every clinical skill is learned somewhere, and those who learn it are always indebted to someone. Sutures are practised on silicone; life support on a mannequin. But speech — the most delicate of our tools — has, for generations, been learned on the only material available: the person in front of us, just when they most needed to be heard.
No one did it carelessly. There was almost no other place to make mistakes. And so, without anyone deciding it, we accepted that the cost of learning to listen would be paid by others.
Today there is another place: a space where error harms no one and today’s clumsiness can become tomorrow’s care. What’s at stake isn’t a technology. It’s a moral question about who should bear the cost of our learning.
And you — who paid the price for your learning to listen?
📩 Request your trial
