Blog

    Breaking bad news in the ophthalmology ST1 interview

    The ophthalmology ST1 interview has two stations, and this post is about the communication and ethics one. Specifically the hardest job inside it: telling somebody something they do not want to hear.

    It is also one of the few parts of the interview with a published structure behind it, which makes it more preparable than its reputation suggests.

    This post is that structure, taken from SmartOST's interview webinar, with the worked case left out. What follows is the shape of the conversation and what each part of it is for.

    Where SPIKES comes from

    The structure has a name and an address. SPIKES was published in 2000 by Walter Baile, Robert Buckman and colleagues in The Oncologist, written for oncologists telling cancer patients about a recurrence, a treatment that has stopped working, or what happens when there is nothing further to try. Six steps, in the paper's own shorthand:

    • S, setting. Prepare the room and the conversation.
    • P, the patient's perception. Find out what they already understand.
    • I, invitation. Ask permission before you explain.
    • K, knowledge. Give the information clearly.
    • E, exploring and empathy. Respond to the emotion.
    • S, strategy and summary. Give a plan, and check it landed.

    Two things are worth knowing before you build a whole station on it.

    It is recommended here, not required. The Royal College of Ophthalmologists makes breaking bad news a learning outcome in its curriculum, and its 2021 guidance Low Vision: the essential guide for ophthalmologists says the College "recommends various resources including the use of 'SPIKES', the six-step protocol developed within the oncology profession". That guidance has since come off the College's site, and the archived copy carries the sentence at page 23. Recommended among various resources is exactly as far as that goes. It is not a UK requirement, and nothing says an interviewer is holding a copy.

    Its authors were careful about what it is. They wrote in the original paper that the protocol "is not completely derived from empirical data", and the evidence since has not settled the question. The one systematic review of SPIKES itself, by Mahendiran and colleagues in 2023, found that none of the studies it gathered looked at patients at all. The two randomised trials that have looked at patients found no benefit, and one of them found the group whose news was broken this way reported more distress than the group given routine care. It is a well-made scaffold for a hard conversation. It is not a validated instrument, and you should not describe it as one.

    Neither of those makes it less useful here, because an interview is not asking you to prove a protocol works. It is asking you to have a difficult conversation, in front of somebody watching, without forgetting anything. A six-step spine is a good way to do that, and the rest of this post is what each step is actually for.

    Setting is easy to rush and easy to rehearse

    The station is held remotely, which decides what this step actually is. You are not arranging a room, you are saying what you would arrange, and an interviewer hears the whole of it rather than watching you do it. That makes it the one part of the station you can rehearse almost word for word.

    What it covers: privacy, interruptions dealt with, everybody sitting down. An introduction, and who you are. An offer of support before you need it, so ask whether they would like anyone else in the room and whether you can get them anything. Then how the conversation is going to run.

    It is easy to rush, because it feels like throat-clearing before the real conversation starts. Getting to the diagnosis faster trades something certain for something you were going to reach anyway.

    The offer of support is the easiest piece to lose, and it is the one that sets up what kind of conversation this is about to be.

    Perception is finding out what they already think

    Before you tell somebody anything, find out what they already believe. A focused history does some of this, and the rest comes from asking directly.

    Three questions carry it:

    • What is your idea of what is going on?
    • What is concerning you?
    • What were you expecting from our conversation?

    Those are ideas, concerns and expectations, which candidates will have heard called ICE, and the webinar's presenters are blunt about covering them in a communication and ethics station. Asking all three is not the hard part. The hard part is that there can be more than one agenda in the room, and the second one can be the one that matters, so keep a follow-up ready: "Anything else that is worrying you today?"

    This is also where listening does the work. The presenters put active listening at the centre of eliciting ICE, and what you are listening for is a cue rather than an answer: the aside, the hesitation, the detail volunteered out of order. You cannot plan those. You can only leave enough room in your own talking to notice them.

    Invitation is one sentence that is easy to lose

    Ask permission before you explain. "Would it be helpful if I explained what we've found?"

    This step is easy to drop, because it looks like a formality. It is doing three things at once. It respects the patient's autonomy over how much they are told and when. It checks they are ready to hear it, which is not the same as you being ready to say it. And it signposts the change of gear, so the explanation that follows does not arrive out of nowhere.

    All of that for one sentence, which makes it the cheapest thing on this list.

    Knowledge starts with a warning shot

    Fire a warning shot before the news itself. "I'm afraid I have some challenging news." Then pause. The pause is part of it, not a stage direction.

    Then explain, and the constraint is plain language throughout. Link what you found to what they have been experiencing, so the diagnosis explains their symptoms rather than replacing them. Jargon is the failure mode here, and it is the one to plan against, because under pressure the precise word is the one that arrives first.

    Chunk and check as you go. Say a piece, stop, and ask whether that makes sense so far. Then signpost what happens next and how urgent it is.

    Empathy is knowing when to say nothing

    When the news lands, stop talking.

    Silence is a hard instruction to follow, because a gap in an interview feels like a gap you are supposed to fill. It is the opposite. Leaving space is what lets the patient react, and their reaction is what you are then responding to.

    What to do with it: mirror how they are behaving, acknowledge what they are feeling, and play their own words back to them so they can hear that you registered them. Then offer reassurance that is about support rather than about outcome. "I know it's difficult, however we are here to support you."

    The strategy and summary has to include a senior

    The last step is a plan the patient can repeat back. Tests and management, laid out in order. A safety net, which means saying plainly when to come back or seek urgent help. Where you are referring them and why. Leaflets and anything else they can take away.

    Two parts of it are worth rehearsing, because both are easy to leave out and neither takes long to learn.

    The first is escalation. The webinar's presenters are unambiguous that discussing the plan with a senior must be included to be safe, and it is not a sign that you were unsure. Leaving it out is what makes a plan unsafe.

    The second is knowing what the health service actually has. Naming PALS, an eye clinic liaison officer, the relevant charities and the specialist service you are referring into shows you understand that a patient is looked after by more than one person, which is what the presenters mean by demonstrating an understanding of the wider team. You cannot infer those names on the day.

    The order is a checklist, not a script

    The six steps are what a good conversation ends up having covered. They are not the sequence it will take, and treating them as one is a reliable way for a prepared candidate to come across badly.

    Knowledge and empathy in particular go back and forth. You explain, they react, you respond, they ask something, you explain again. Work down a list and you will keep pulling the conversation back to where the list says you should be. The webinar's presenters tell candidates to expect to deviate from the planned structure and order, and this is the situation they mean.

    Build a rough template for each kind of case, so you do not forget a step under pressure. Then be ready to abandon its order the moment the patient takes the conversation somewhere else. The template is insurance against omission, not a running order.

    You are interviewing as an ST1

    This is the frame that makes several of the sections above make sense.

    You are not expected to have deep knowledge. You are expected to have breadth and to be safe. The subspecialty answer is not what the level asks for, and saying you would discuss the case with a senior is the safer move.

    So when you are unsure in the station, the move is not to guess further into the medicine. It is to say what you would do about being unsure, which is almost always the same thing a safe ST1 does in the department.

    Where to start

    Take the six steps and write out, in your own words, what each one is for. Not what to say in it. What it is for. Memorised sentences are what make a candidate sound scripted; knowing why each part of the conversation exists is what lets you build the sentence on the day.

    Then pick the two you would be most likely to drop under pressure. Invitation and the silence inside empathy are the two to watch, because each is one decision rather than one skill.

    All of this is easy to agree with on a page and hard to do out loud. SmartOST's interview practice is built around cases of this shape.