Difficult Conversations

Difficult conversations are exactly that: difficult. They may be difficult for numerous factors that are related to the subject matter, the patient, their relatives and friends and yourself. The first time you go through one of these conversation, you may feel drained, raw and emotional or you may not feel affected at all.

Dr James White is not just a GP but also an anaesthetic registrar in training who regularly works across critical care and has held numerous roles in medicine. The podcast goes through some different situations that may form a continuum of discussion from the delivery of a test result and possible life-changing diagnosis, the discussions that may take place when faced with an acutely unwell patient and end of life and DNACPR decisions. The post below discusses the SPIKES model in more detail with input from a discussion on Facebook.

Listen below:

Or take it with you:





You may simply not want to do this. Difficult conversations are difficult for many reasons. It is worth considering the reasons why should you hesitate and what you can do to address these. If you hesitate before going into these conversations or feel uncomfortable then it is worth considering why and how barriers can be overcome.

  • Emotional. Your emotions are important and showing emotion during these conversations can be important however, your emotions may be driven by personal events outside of work. You may have experienced recent bad news yourself or even a bereavement. Sometimes bringing these experiences into a discussion can actually help the conversation but you may feel it is all too raw. Be honest with yourself and your colleagues if such things make you feel unable to undertake these discussions.
  • Fear and anxiety. It is natural to be afraid or anxious about these discussions. You may be fearful that blame is going to be assigned to you (in some circumstances it may be). You may be worried you could be asked questions that you do not have an answer for or be pushed to give information you do not feel able to. It is entirely natural to be anxious about causing emotional harm to a patient when discussing difficult subjects which is why paying attention to how you do this is vitally important.
  • Experience and training. Throughout a doctors career, they will likely undertake thousands of these discussions. If you are at the beginning of your career you may never have even sat through one before and have no frame of reference to guide you. Few will be naturally gifted in these situations and everyone at some point has conversations that do not go well. If you do not feel experienced enough to do this, then tell your seniors so they can help you. It might be you need someone with you for support or you need to sit in when someone else leads the conversation to gain experience.
  • Time. Time is precious in modern and there rarely is enough. These conversations often need a good chunk of time setting aside for them. the time pressures may be patient related as well. The patient or their relatives may only be available for a short period which can put you under pressure to make that time slot, especially if it prior to the weekend or a break. Your workload may need handing over to someone else in order to give you time to do this. Take caution against handing over these conversations though because this will seem that you are trying to dodge out of it and your colleagues may not be thankful for it. If you have received the test result, you should be the one to act upon it.
  • Causing distress. In reality, there is no good time to have these conversations. However, a patient who is already distressed by events may not be in the right mindset to receive and discuss further distressing information. These conversations are often on a continuum and take place in chunks. You may need to go back and repeat a previous discussion if the patients distress was too great to hear what was said. As much as we do not want to cause distress through our actions, we do have to have these conversations and as such must accept that we are likely to cause distress but through careful handling of these situations we can hopefully minimise it.



Medical schools often teach us models to follow with these conversations. I personally, have never tried to strictly follow a model but the discussion with James revealed I actually do. The SPIKES model (1) is one that is often taught and is easy to follow. Remember though that not all steps need to be used each time. You will likely need to adapt your approach each time to ensure you keep it specific to that patient. The SPIKES model was designed for oncology and may not apply in a strict format to every situation we face but provides a great framework to adapt.


Step 1:

Setting up

First and foremost: decide if you or someone else is going to have the conversation with the patient and/or their relatives and if it is someone else, make sure they know this. If you are going to be leading the discussion then sometimes it is worth mentally rehearsing what you plan on saying. You can’t rehearse what other people will say but you can know your words in advance. If someone is coming with you, you can share your mental model with them. This will allow them to add anything beforehand they feel is important and check that what you plan on saying is appropriate. It also allows them to prompt you during the conversation if you miss something.

Sketch nurses

Some points to consider:

  • Privacy. The options for this can be limited. A patient may not be able to get out of bed to go to a room where the door can be closed or a separate room may simply not be available. Curtains, regrettably are not magically soundproof but they offer at least the pretense of attempting privacy. You may need to be mindful of the volume of your voice. Privacy is important after the conversation as well. If patient and/or relatives/friends are upset, pulling back the curtains to show their anguish to the ward is not wise and it may be worth asking if they want the curtains to remain closed. If you’re in a room on a ward, they may wish to stay there to compose themselves but you cannot leave people sat in an official office unattended and may need to vacate the space so others can continue with their jobs.
  • Involvement. The patient, or even their companions, may wish for others to be present for the discussion. This provides support for them but the extra persons may be able to process the information more effectively and ask questions that may not be asked without them. Too many people can hinder a discussion. Gauge the size of the space, your own comfort with numbers and ask who the patient most wants present if numbers are to be limited to a close couple or few.
  • Level. Get down to the level of those you are speaking with and be close to them so you can speak at an even tone. You may find yourself having to crouch or kneel by a bed. If you can, make sure that everyone has a seat and if you must then give up your own and kneel instead. Once you do sit, be comfortable and make sure your body language is open. This time and care tells the patient you are there for them and are not going to be hurrying off as soon as you are finished.
  • Connect. Eye conact and open body language is essential. You may feel uncomfortable maintaining eye contact in these situations but you must force yourself to do so. Let the patient know with your non-verbal cues that you care and are invested in the conversation.
  • Time. Ensure you have the time set aside for this. In our busy working lives, time is a precious commodity. It is crucial that you minimise interruptions which may mean giving your bleep to someone else or even turning it off. Make sure the ward staff know what is happening in the room or behind the curtains so someone else does not accidentally barrel in at a tender moment.

Sketch time


Step 2:


This is your opportunity to find out what is already known by the patient and their significant others about the situation you have come to discuss. Exactly what you ask will vary depending upon the situation at hand. You may start with simply “What do you know so far?” or if they know you are there to discuss a test then “What do you believe was the reason why x test was organised?”. However you ask, it needs to be open and you must give the other person or people time to fully respond and even explore their ideas in more depth.

You may find an undertone of denial or false hopes. You may discover, as is often the case, that they already know what is going on and may have even guessed what you are about to tell them. It may also become apparent that those you are talking with know nothing about what is going on. Try not to linger and force people to answer a question if they cannot or may not want to. They are often scared and voicing the reasons behind a test or situation may bring their fears into reality.

This is also a time where you can gather more information. If you are talking about an acutely unwell patient, you may need to gather some more information about their recent history or past medical history. This extra information will help to frame the discussion and provides the opportunity to rectify any misconceptions they may hold prior to delivering the news you are there to.

Step 3:



This is the part that, depending on what the patient says can give you total relief or extreme concern. Some patients and their close ones may want full and frank information but some may want brief outlines and some may want none at all.

Exactly how you broach this will vary depending on the situation. Often, the situation I am faced with is that of a patient who is extremely unwell or even dead and I must deliver this information to them or their loved ones. After exploring what may be a brief background, we need to know how to proceed in our delivery of information. When the news is dire, in order to avoid euphamisms, I will often ask if I may be blunt and I have yet to have a moment where I am told no. This serves two purposes. Firstly, it gives permission to speak plainly and to use terms we may otherwise wish to avoid such as those associated with “death”. Secondly, those I speak with always stiffen because they know the conversation is going to reach a peak and they are going to be faced with the information they have been dreading.

However, when delivering information such as test results, bluntness may not be possible because not all the facts will be to hand. Instead, you may wish to ask how the patient wants the information delivered. Do they want to know just an outline of the next steps? Do they want to know just what the worst case is? Do they want the initial diagnosis and then to ask their own questions? Do they want to wait until the next appointment to receive further information and ask questions? Be wary though as you may be tempted to speculate which can cause problems.

Keep what you say to facts and certainties and do not be afraid of admitting when you don’t know. However, if you say you don’t know, try and offer a solution of how you or the people you are talking to can find out the answers.

Step 4:



Patients and their close ones are not stupid. By this point in the conversation they wil tend to know that something unpleasant is coming.

If you feel it has not been explicitely appreciated then you may be more comfortable using phrases such as: “I’m sorry but I have some bad news to tell you” or “There is something difficult we must discuss next”.

Sketch wording

The delivery of knowledge is where we can enter our comfrot zone because we have medical things to deliver. Do not enter the trap of using medical jargon which is easy to fall into because we assume other are on our level. We may think words that are common to us mean the same to others. Some examples to consider would be:

  • Metastasise = spread
  • Kidney/heart/liver failure = kindey/heart/liver not working well or potentially not at all
  • Biopsy = sample
  • Culture = growing a sample of blood/urine etc in the lab
  • Sepsis = severe infection that has caused problems with other parts of the body
  • Cardiac arrest = heart stopped beating
  • Benign = usually NOT cancer


If you read the original SPIKES paper (1), it warns against excessive bluntness. I would see this as different to offering to be blunt in what you say when you wish to avoid platitudes or euphamisms. This would be using a term that would cause unnecessary hurt or even anger such as: “You’re disease is terminal and you are going to die”. Instead, even if you have to say it is terminal, tell the patient what you can do. You may have to tell them they will eventually die but despite this you can manage symptoms and keep them in touch with the healthcare support networks. As the messenger, you do not want to be shot.

We are taught to give information in small, manageable chunks and check the understanding of these chunks. In practice, this can be clunky and awkward. Chunks may include a few pieces of information rather than a single piece per chunk. You are still aiming for a conversation not an impromptu quiz. You may even have to accept that no matter how hard you try, some information will not be absorbed. The checking of information may occur in a further consultation with yourself or another that follows you.


Finally, most importantly, do not tell them nothing can be done because no cure is possible. Something can always be done even if it is not a cure. A patient or their loved ones may have priorities of comfort or dignity when they know a cure is not possible and we can certainly provide assistance with those. The patient may have continuing and distressing symptoms that need addressing and you can probably do something to alleviate these.

Alongside ensuring you deliver clear information, avoid any false promises or speculation you may be tempted to provide. You may be pushed for a prognosis or timescale that you cannot provide. People do understand when they are told that it is not possible to speculate. Even when you explain this, you may still be pressed but remain steadfast and continue to explain that you are unable to accurately say and anything you do say is likely to not be true. You can tell them the things you can do. If you know a specialist is coming in a couple of days, tell them this or that you will prescribe medication to combat symptoms as examples.

Step 5:



The most difficult part of any of these interactions is dealing with emotion. This will no doubt be those of the patient or their relatives/friends but it may be your own.

Empathy is essential but difficult.

Empathy may not feel natural. Avoid forced empathy at all cost and show what you can. This may be sitting, silently and patiently in the room whilst people cry. It may be offering a cuppa or box of tissues. It may even be giving the other person a gentle hold of your hand.

One important aspect to consider is the reason behind the emotion. This, most often, is obvious. If the reaction you witness is outside what would be expected, you can delicately explore the reasons. If the reaction is anger, what has caused this? If a person is silent, give them time and avoid pressuring a response but if they remain silent for a prolonged period, ask open questions to try explore their silence.

Empathy can be shown through a sincere, understanding response to their emotion. This may be through simple statements such as “I understand this news has come as a shock and may be difficult to process” or “I know this was what you feared may happen and I too had hoped it would not be the case”. Even the way you greet those you are speaking with can establish empathy and tell them bad news is coming: “I’m sorry we have to meet like this”. Make what you say sincere and it will be appreciated.

Dealing with emotions requires time. Time may not be a luxury in the middle of a busy working day. Having a nurse or another professional with you may allow you to ask forgiveness to take your leave but allow your companion to stay and help deal with the emotional response.

nurse post

Step 6:


Strategy & Summary

The strategy of how you proceed following this conversation ideally should be arrived at through collaboration. In some situations though, collaboration may not be possible and the final part may be an agreement about the next steps. In a situation where the patient may feel completely out of control, giving the chance to be involved in decision making grants some power to the patient.

Some patients and their relatives may not be ready to discuss strategy immediately after receiving bad news. In this case, follow up is vital. This follow up may be the next outpatient appointment or it may be the specialist referral you have made or your consultant ward round the following day. Where you can, be specific about when this will be so the patient and their relatives can be prepared should they have particular things to discuss.

Some people may have very unrealistic expectations about the capabilities of medical treatments. They may not fully appreciate the gravity of what has been said. After all, part of the cycle of grief is denial.

This final part is where, if needed, you can explain what is and is not possible. You can also provide a guarded opinion about the likely outcome if you are able to. Again though, you may be pressed to prognosticate. I find by this point I’ve likley been asked multiple times about a likely timeframe or if everything is going to be ok. I find I tend to frustrate people by refusing to do this. I use words such as “probably” very often. I explain each time that I cannot give an accurate prognosis or timeline. The last time I tried, I was very wrong and it caused hurt for those I provided this speculation to. I explain that to speculate in the face of uncertainty could cause hurt if I guessed and got it wrong. We all know patients who we expect to die soon but live for much longer.

The very last thing you want to do is give someone false hope and confusion.

Not only that, but if you tell someone they are going to be okay when you know it is unlikely to be true, you may find yourself faced with significant anger afterwards. The desire to give this assurance is a human one and is born from compassion. We may wish to alleviate suffering, we may feel inadequate in our role as a healer or it may present a convenient way to escape the bad news already delivered.

When attempting to summarise you can either do it yourself or ask the patient/relative to. This can feel unnatural. Ways to do it include:

  • “We’ve spoken about a lot already, would you mind if I go over the main parts again?”
  • “I appreciate what I’ve said has been difficult for you, could you tell me what you undertand from what has been said?”

This needs to be a brief summary, not a complete redo of the whole conversation. Asking the patient to summarise may reveal further worries of theirs through where they place emphasis, especially if it differs from what you would emphasise. Some situations prohibit excessive repetition. For example, telling someone a patient is dead does not need repeating otherwise you may appear cruel. Though if you have explained that a patient is dying and it is not appreciated then you may need to reiterate this is terms that are clear.

A final point is to remember that just because a patient has been told they will die of one thing, it does not stop a different condition from causing their death.

Sketch multip



“Do not attempt cardiopulmonary resuscitation” needs a special mention. It is a conversation that is frequently shied away from because it is unpleasant, difficult and carries many misconceptions about our duties.

The most important aspect to remember is that CPR is a medical intervention and the decision to provide or not is ultimately a clinical decision but have every attempt made to discuss it.

If you are a Foundation Doctor or even CT1/2 level you may not be able to sign DNACPR forms. This may vary between hospitals but increasingly it is only a registrar or above who can do this. These forms also need verifying by a consultant at the earliest opportunity which does get forgotten.

Commonly, a decision is made but the form is not completed whilst awaiting a discussion with the patient or their next of kin and relatives. These decisions may be made at any time during the working day or night and the appropriate discussion can happen at any time too. There have been high profile court cases involving DNACPR decisions not being discussed (2, 3) and when inappropriately applied, deemed as breaching Article 8 of the Human Rights Act (2).

The GMC has issued guidance on the subject of making and discussing DNACPR decisions and is worth reading to familiarise yourself with (4). In this written piece, we are not looking to give advice regarding the decision making process but to help you with the discussion. Making the decision should be considered alongside the patient and their relative/next of kin’s wishes, the multi-disciplinary team and if you are junior, you should not be making it in isolation. The GMC guidance does state though that a DNACPR decision should be made as early as possible if a patient is admitted acutely unwell to hospital and are at risk of cardiac arrest (4).

We do have a duty to discuss DNACPR decisions.

We must not make assumptions that a patient would not want to engage in discussion and should sensitively explore what they do and do not wish to know. You may be surprised by how many people do want to discuss this and what their actual views are. In the event the patient does not want to know, you need to obtain their permission to discuss it with those close to them but you should not pressure them into the discussion.


We cannot avoid this conversation because we or our team dislike it.

In the event a patient lacks capacity, you should make every reasonable effort to discuss with those close to the patient. If this is not possible then a legal advocate which may be someone who holds power of attorney for health or even an IMCA should be sought.

You must be clear that we are not asking the patient’s proxy to make the decision themselves. Often they think we are asking them to decide and in truth, we are relaying information about a decision that we have made as the medical team.

The discussion itself needs to be clear and devoid of euphamisms. Too often the patient is asked to make a choice from a poorly worded question:

“What do you want us to do if your heart stops?”

“Start it again!”

This question is vague regarding our message and it shifts the decision making burden to the patient from us. This is a clinical decision regarding an aggressive, distressing and traumatic intervention that is often followed by an existence that may be intolerable to some. Unless this is explained, how can the patient possibly undertand? Especially when what is portrayed by popular media is a very rosy image of a brief intervention with good success and quick recovery. The reality is a brutal intervention followed by what may be a prolonged stay on ITU that may ultimately result in the patient dying days, weeks or even months later in a position that they would never have wanted.

Unfortunately, the message we are conveying can be misinterpreted to mean that we are giving up. After explaining what a DNACPR decision means, it is wise to explain what treatments you can provide and how they can help the patient.

dnacpr fb post


Take Home Points

Difficult conversations are exactly that: difficult. The SPIKES model was designed for oncologists however applies well to any situation involving a difficult conversation. Models are only useful if applied properly and with compassion. What people want is a human and not a robotic conversation model. Pay attention to those you encounter that do these conversations well and try take on board what they do well to enhance your own ability. If you do have another person in the room, ask them afterwards what they though and what they suggest you can improve.

DNACPR decision discussions can be tough. Be explicit and make every effort to ensure the discussion happens.


Speak soon



GET IN TOUCH! Either through the comments section or email if there are any comments or corrections or if you want to be involved with future topics.

RATE US on iTunes to help others find the resource more easily!

SHARE US with others you think would benefit or even your training leads to help us distribute the resource further! Twitter: @FOAMdation



  1. Baile W et al, SPIKES—A Six-Step Protocol for Delivering Bad News: Application to the Patient with Cancer. The Oncologist 2000; 5:302311 
  2. GMC (2010), Treatment and care towards the end of life: good practice in decision making: Cardiopulmonary resuscitation (CPR). Para 128-146. Available at: Last accessed: 19/08/2018

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.