Dilemma questions are the ones candidates most often answer with a verdict: the patient can refuse, the relative cannot be told, the colleague must be reported. Panels are not scoring the verdict. They are scoring the reasoning: whether you found out what was actually happening, named the principles in tension, heard the other side fairly, saw the options, chose with the patient at the centre, and kept everyone safe. ETHICS is that reasoning in the order panels mark it, and the four pillars sit inside it.
What the letters stand for
Establish the facts. What has actually happened, what the patient or the person wants, and capacity if it is in question. Most dilemmas look different once the facts are known, and a candidate who reasons before finding out reads as hasty.
Think through the principles. The four pillars: autonomy, the person's right to decide for themselves; beneficence, the duty to act in their interests; non maleficence, the duty not to harm; justice, fairness to others and to the use of resources. Add confidentiality where it applies. Name which principles are in tension in this case; that naming is the first mark.
Hear the other side. The relative's reasons, the patient's reasons, the colleague's, and what someone who disagreed with you would argue. Discuss both sides fairly before you choose, because the panel scores whether you can.
Identify the options. Two or three, with what each would mean for the people involved. Rarely is the choice between doing the thing and not doing it; there is usually a third way that respects more of the principles.
Choose and act. The decision, justified, with the patient at the centre. Who you involve, what you do first, and what you would do if it did not work, because the probe is always "and if they still refused?". Show how you reason through uncertainty rather than insisting there is one correct answer.
Safeguard and document. The patient kept safe whatever the decision, the record written, the senior told, the door left open for the conversation to continue.
Why reasoning, not verdicts
Most dilemmas in interviews have no single right answer, and panels know it. A patient with capacity who refuses life saving treatment may do so, and the doctor's distress does not change that; but the answer that scores is not "they can refuse", it is the one that establishes capacity, explores why, names autonomy against beneficence, hears the family, offers alternatives, respects the decision, involves a senior, documents, and leaves the door open. The structure makes the reasoning visible, and visible reasoning is what the rubric rewards.
At different levels
Medical school applicants reason through the friend asking for answers, the patient refusing treatment, and whether paying for faster care is fair; the structure is the same and no medical knowledge is needed. Juniors apply it to the drug error nobody noticed, the relative who asks for information, the patient who wants to leave. Registrars and consultants apply it to the family who want everything done, the advance directive the family dispute, the full unit with two patients for one bed, and the policy questions about resources, with the law and the local processes named at Choose and Safeguard.
Worked example
A relative asks you not to tell the patient their cancer diagnosis.
Establish: what the patient has already been told, what the patient has said they want to know, whether the patient has capacity. Think: the patient's autonomy and right to information against the relative's beneficence toward them; confidentiality, because the information is the patient's. Hear: the relative's fear, often from a previous experience or a cultural expectation, and the case that bad news harms. Identify: tell the patient regardless; agree with the relative not to tell; or find out from the patient what they want to know and follow that. Choose: the third, because a patient with capacity decides what they are told, and most patients want to know; involve the senior and, if the relative's culture is part of it, a liaison worker. Safeguard: the patient's wishes documented, the relative supported, the conversation left open.
Colleague problems
When the dilemma is about a colleague who is impaired, underperforming or unprofessional, SPIES is the structure instead, because its first step is to seek information about a person and its spine is patient safety, initiative, escalation and support. ETHICS is for decisions; SPIES is for people.
Common mistakes
A verdict with no reasoning. Principles named but no tension identified. The other side dismissed rather than heard. Only two options. No next step for the probe. And forgetting to document or to tell a senior.
Practising
Take the dilemma questions for your level and reason through each aloud with the six steps in two minutes, naming the pillars in tension, then have someone play the relative or the patient who insists. The reasoning is what the panel is waiting to hear.