What is a Patient Digital Twin?

A Patient Digital Twin is a digital counterpart of one person's journey through care.

It holds what is known about that person's situation, follows them through an episode of illness or treatment, and can talk with them and with the people looking after them.

The word "twin" comes from engineering. A jet engine has a digital twin: a model fed with the real engine's data, so engineers can see how it is doing and what might happen next. Medicine has borrowed the idea in two directions. One builds twins of the human body, detailed models of a heart or a lung used in research and to test treatments in silico. The other builds twins of the patient's journey. That is what this page is about and what llaama is building.

Not a model of the body, a companion for the journey

A Patient Digital Twin does not simulate organs. It knows that a person is having a knee replaced on the twelfth, that they take a blood thinner, that they live alone and were anxious at the pre-operative visit. It knows what the surgical team wants them to do before and after the operation, and what signs should prompt a call.

With that knowledge it can do three simple things well. It can explain, in the person's own language and at their own pace, what is happening and what comes next. It can notice, because it hears from the person between visits, when something is drifting from the expected path. And it can escalate, by bringing a nurse or a doctor into the conversation at the moment they are needed.

It does not diagnose, and it does not prescribe. Those remain decisions for clinicians. The twin's job is presence: someone who has read the file, who is there at two in the morning, and who knows when to fetch a person.

Where it adds value

Most of a care episode happens where no clinician is present. A patient spends an hour a month with the care team and the rest of the month at home, with questions, worries and symptoms that do not wait for the next appointment. That gap is where preventable problems grow, and where a Patient Digital Twin earns its place.

Before surgery, when instructions are long and the patient is frightened. Explaining the same things patiently, as often as needed, and confirming that they have been understood.

After discharge, when complications begin quietly. Asking how the wound looks, whether the pain is different today, and passing an early signal to the team instead of waiting for the emergency department.

In chronic illness, where the work is daily and the visits are quarterly. Keeping the thread between appointments, so that the next consultation starts from what actually happened rather than from what the patient remembers.

In clinical studies, where participants must follow a protocol for months. Keeping them informed and engaged, and capturing how they feel when they feel it, not at the next site visit.

The value is not that a machine replaces a nurse. It is that the nurse hears about the right patient at the right time, and that no patient is alone with a question.

What a twin must never be

A twin that pretends to be human, that answers beyond what its clinicians have approved, or that acts without a person deciding, is not a twin. It is a liability.

A trustworthy Patient Digital Twin draws only on knowledge its care team has approved, says so when a question is outside its role, hands over to a human at the edge of that role, and keeps a complete record of every exchange. Patients, clinicians and regulators can each check what it said and why.

How it relates to the European Virtual Human Twin

The European Commission's Virtual Human Twin initiative works towards digital twins of human physiology for research, treatment planning and regulatory science. Patient Digital Twins sit at the care-delivery end of the same idea: while science builds the twin of the human body, the Patient Digital Twin accompanies the patient's journey today. The two will meet. LLAAMA is a signatory of the European Virtual Human Twins Manifesto.

Frequently asked questions

Is a Patient Digital Twin a chatbot? It can talk, but talking is not what defines it. A chatbot answers whoever types. A Patient Digital Twin knows one patient's situation, follows a care plan set by clinicians, keeps a record, and brings people in when needed.

Does it make medical decisions? No. It explains, listens, notices and escalates. Diagnosis, prescription and every change to a patient's care remain decisions made by clinicians.

Who sees what the twin knows? The patient and their care team. What the twin may say, and what it must never say, is set by the clinicians responsible for the patient.

Is it the same as the digital twins used to model organs? No. Those model the body for research and treatment planning. A Patient Digital Twin follows the person through care. The two approaches complement each other.

Does it replace nurses or doctors? No. It extends their presence to the hours and places where they cannot be, and makes sure they hear about the right patient at the right time.

Where is it used today? In surgical pathways, from preparation and consent through recovery, in the follow-up of chronic conditions, and in clinical studies where participants are accompanied through the protocol.