Discussing Racism in Health Care Without Platitudes

Every applicant agrees that racism is wrong. Which is precisely why saying it earns you nothing. It is the shared starting line, not a contribution, and an answer built on it runs out of material in about forty seconds.

What separates strong answers here is analytical, not moral. It is being able to say what kind of racism is at work in the scenario in front of you, because the two main kinds call for completely different responses.

Two kinds, and why the difference does the work

Interpersonal racism is what one person says or does to another. A comment, a refusal, a dismissal, a joke in a tea room. It has an author, and you can usually point at it.

Structural racism is the pattern of outcomes produced by how a system is designed, funded, staffed and located. Its defining feature is that it does not require anybody in the building to hold a hostile belief. A service can be full of decent people and still deliver worse care to a particular population, because the decisions that shaped it were made somewhere else, decades ago, by people who were not thinking about them.

That last sentence is the one to be able to say cleanly. It is the difference between an applicant who thinks racism is a personality flaw and one who understands why the profession keeps talking about it long after individual attitudes have improved.

There is a third category worth a mention: unconscious or implicit bias, the associations that shape a rapid judgement under time pressure. Useful to name, dangerous to over use. Bias explains why good clinicians make patterned errors. It does not excuse them, and an answer that treats it as absolution has taken the concept exactly backwards.

Why candidates only ever discuss the first kind

Because it is easier and more flattering. An interpersonal scenario has a villain, and you get to be the person who intervenes. Structural racism has no villain and no satisfying five minute resolution, so it gets skipped.

Give it thirty seconds anyway. Even in a scenario about a single comment, one line noting that the individual behaviour sits inside a system that produces its own patterns shows range. It is the same instinct that lifts most ethical stations in Australian MMIs: handle the immediate situation, then place it in a larger frame without abandoning the person in front of you.

What structural racism looks like in a health service

Abstract descriptions are why this half of the topic feels unusable. Concrete examples fix that.

• Where services are built and how they are funded, which determines who can physically reach care.
• Who works there. A workforce that looks nothing like the community it serves changes what patients expect before a word is said.
• Interpreters that exist on paper and are not booked, because booking one makes a clinic run late.
• Whose reported symptoms get believed. There is a substantial research literature on differences in how pain and distress are assessed across patient groups, and it is worth referring to that literature rather than quoting a figure at it.
• Who ends up labelled difficult, non compliant or aggressive in a handover, and what that label does to the next clinician's judgement.
• Guidelines and evidence built on populations that did not include everyone the guideline is now applied to.
• Practical design: opening hours, forms, appointment systems, and whether leaving before treatment is finished is recorded as a patient's choice or examined as a service failure.

The patient who refuses a clinician

This is the hardest common scenario, and candidates often collapse into one of two easy positions: patients can choose whoever they like, or the patient should be refused care. Neither survives contact with the situation.

Some useful anchors. Patient autonomy covers accepting or declining treatment. It does not extend to directing who is employed or abusing staff. Clinicians have a right to a workplace free of racial abuse, and health services generally have policies covering exactly this, which is where you should say you would look rather than inventing one.

Then the qualifiers that show judgement. Emergencies are different: nobody is left untreated while this is negotiated. Capacity matters, since delirium, dementia and acute distress can produce speech a person would never otherwise use. And a request that looks like preference rather than prejudice, such as wanting a female clinician for an intimate examination, is a different conversation entirely.

The move that most candidates miss is whose job this is. The targeted clinician should not have to negotiate their own dignity. The senior clinician or manager owns the conversation with the patient, and the first thing anyone should do is ask the colleague what they want to happen, rather than deciding for them.

The comment in the tea room

A colleague says something about a group of patients. Everyone laughs slightly or nobody says anything, and the moment passes. This is the scenario where most people, honestly, say nothing in real life, and interviewers know that.

Do not promise a heroic confrontation you would not deliver. Describe something proportionate and actually usable. A short, low heat interruption in the moment: I do not think that is fair on them. Or a quiet word afterwards. Or, if the person is senior and the risk is real, raising it with someone who can. What matters is that your answer contains an action rather than a feeling of discomfort.

Calling in tends to work better than calling out for a one off comment from someone who is otherwise reasonable. Repeated behaviour, or anything a patient could hear, moves up the scale quickly. Saying that you would calibrate, and naming what would make you escalate, is worth more than either extreme.

Traps

• Performed outrage. Volume is not analysis, and a station that turns into a denunciation has stopped reasoning.
• Colour blindness. I do not see race sounds generous and means you will not see the pattern when it appears in your own outcomes.
• Making the targeted person do the work. Asking a colleague to educate everyone after they have just been abused adds to the injury.
• Treating it as a problem of a previous generation. Structures outlive attitudes, which is the entire reason the structural category exists.
• Quoting statistics you cannot source. Describe the direction and name where the current data is published instead.

If it has happened to you

Plenty of applicants have their own experience of this, and you are under no obligation to use it. If you choose to, keep it factual and keep the focus on what you did and what you would want a system to do, rather than on the hurt. You are not there to prove anything, and an assessor who hears a controlled, specific account of something difficult will draw the right conclusion without being asked to.

Equally, if you have never experienced it, do not manufacture proximity. You can reason about something you have not lived, and pretending otherwise is transparent.

Practising it

Take three scenarios, one interpersonal, one structural, one where a patient is the source, and answer each in ninety seconds. The discipline of the clock stops you retreating into general statements, which is where this topic goes to die. Our overview of what an MMI is in Australia and New Zealand covers the station format, and since designs differ and change, check the current admissions page of the university you are sitting.

Then rehearse it out loud with something that answers back, because tone carries an unusual amount of the weight here. MasterMed's live AI interviewer runs timed MMI stations and marks you against a rubric. The first speaking station is free on the trial, no card, and the trial never converts by itself.

Rubric domains vary by school, and our guide to how MMI scoring works in Australia lays out the ones that recur. None of them are checking whether you disapprove of racism. They are checking whether you can name what is happening precisely, act proportionately, and stay useful in a situation where the easy options are silence or a speech.