The Blog

Do Not Attempt Resuscitation Orders: Are Doctors Devious and Secretive? Why Context Is Important

Warning: this is a long article. But complexities surrounding Do Not Resuscitate forms don't allow for skim reading, and much recent reporting has been sensationalist hogwash

Doctors have been getting a lot of bad press recently. This week is no different, with many aspects of medical practice scrutinized by the watchful eye of the fourth estate; an audit by the Royal College of Physicians has looked at end of life care in hospitals in England, including aspects of communication of treatment decisions. There has been some improvement in end of life care since the last audit in 2013. The journalistic media have focused their attention on Do Not Attempt Cardiopulmonary Resuscitation forms, an evaluation of which formed part of the report. So what were the areas of concern?

The Telegraph, leading on this, reports that about 40,000 dying patients are affected by this, in the journalist's own words, 'secretive' process each year.

First let us look at some of these figures and percentages. The 40,000 patients per year are an extrapolation: they are a best guess figure, not mentioned in the RCP audit itself, but presumably worked out by journalists for the benefit of a headline. The figure itself has been calculated based on the number of deaths captured for this audit, which was 9302. For all of these deaths, the medical notes were used to capture whether discussions had taken place or not. The clinical part of the audit collected data from 1-31 May 2015 from the records of 9,302 patient records across 142 NHS organisations, mostly acute hospital trusts. Participating units were asked to submit up to 80 patient records, so the final number is not the total number of people dying in hospital during that month. In light of this, the figure of 40,000 is at best speculative.

Where an existing DNACPR form was not explicitly discussed with a patient, reasons for not doing so included unconsciousness or lack of mental capacity. In 16% (961/6072) of cases, no reason for not discussing was given. I have not been able to ascertain much more about these 16%, other than that there was no written evidence in the notes regarding which discussions, if any, had been held. It could well be the case that the patient was confused at the time, and the doctor did not document this fact on the form or in the notes, and an audit is a binary process: either it's documented, or it isn't, in which case it will be captured as 'conversation did not happen'. Some of this may have been recorded due to DNACPR forms that were only partially filled in. The doctor may have omitted writing down the fact that he/she discussed DNACPR with the patient- this is bad practice, and should be improved, but what perhaps mattered most to the doctor in his context was that the patient was aware. Nothing secretive in this, very realistic scenario. I routinely suggest to all my healthcare professional colleagues to fill in DNACPR forms from start to finish, with as much free text as possible, in order to give an accurate picture of the actual DNACPR conversation with the patient and proxy. This percentage, and the ambiguity it causes with regard to what actually happened prior to the form being filled in, needs to be addressed and reduced; the context matters hugely, and it is therefore quite difficult to leap to too many conclusions (sensationalist or otherwise) with regard to these 16%.

A Do Not Attempt Cardiopulmonary Resuscitation (DNACPR) order was in place for 94% (8711/9302) of patients' notes at the time of death. Where sudden deaths are excluded, discussion about CPR by a senior doctor with the patient was recorded in 36% (2748/7707). Discussion about the CPR decision with the nominated person(s) important to the patient was documented in 81% of cases. This is as compared to 72% when the last audit was performed in 2013, so there has been improvement.(for the full figures, go to page 31 of the full report here). It is unclear with the remaining percentage of nominated persons who were not informed of the DNACPR form, whether it had in fact already been discussed with the patient. By this I mean that the patient may have known about the DNACPR form, but the relatives did not. Medical teams have primary responsibility discussing medical care aspects like DNACPR with their patient; and usually endeavor to also discuss with family and friends/proxy. But if the closest person to the patient is in Australia, a phone discussion about this sort of subject can also go desperately wrong. Again, context is king, and the individual variations of each situations and cross-over issues are very hard to capture as part of an audit project. But this requires all numbers to be analysed and reported in a very measured way.

And this has not happened. Reporting and summarizing of the pertinent figures outlined above, has varied in quality. Non-disclosure is not perceived as acceptable practice in 21st century medical settings, and quite rightly so. But equally, it is sensationalist and self-serving to create fear by suggesting a kettle of vulture-like doctors who clandestinely place DNACPR orders in people's notes. The above-mentioned Telegraph article by Laura Donnelly is called: "Exclusive: 'Unforgivable' failings in end-of-life care revealed as 40,000 dying patients subject to secret 'do not resuscitate' orders every year". If I'd written this article, I would have named it differently: "Awkwardness, timing and appropriateness in death and dying communication between doctors and patients: DNACPR forms not always discussed. Some improvements, but more needs to be done, says national report.' "

Terrible title, I realize. No prospect of my joining the likes of The Sun or The Telegraph. But on the other side of the extreme, recent 'click-baiting' and attention grabbing headlines, using words such as 'secretive', 'unforgivable' 'failings' and '40,000 dying patients', have made people fearful and frightened and have even served to make some fear a conspiracy, suggesting that all doctors are heartless, cannot be bothered saving lives and perhaps prefer to chase targets on how many DNACPR forms they can achieve.

Furthermore, many of the news articles written on this subject imply a false premise; that for many patients with advanced illness, there is a strictly binary life-or-death decision process: turn to the left, and you will be privileged enough to receive CPR and if you do receive it, you'll walk out of hospital and live happily ever after (in this scenario people don't die); turn right and a doctor will slap a DNACPR order on your record, you will soon deteriorate and the life-saving treatment is inexplicably withheld, (together with all other treatments), in order for the healthcare service to save money. And why bother tell you about the nasty DNACPR form anyway? This of course is journalism at its most hyperbolic and solipsistic: attracting large readership numbers through the medium of fear.

It is far more important that as a modern society we focus on the human factors that occur in face-to-face discussions during episodes of hospitalization and severe illness, which lead to these important discussions not being addressed. And one of the main challenges is talking openly, plainly and frankly about death, when patients have life-limiting, palliative illnesses. So let's start out by being honest about our awkwardness. If you sat opposite me in a train and I started a conversation with 'Where would you like to die and who would be with you when it happens,' you'd probably skip to the next train compartment. Maybe social media are gradually challenging the taboo of talking about death and dying, and so I am glad that many doctors, nurses and patients are now able to contribute their views via hashtags like #EOLCare and charities like Dying Matters.

The report and the reaction in the media coincide with the start of Dying Matters Awareness week. And perhaps this is quite a good starting point for the Big Conversation; there is much work to be done. Part of this will be a discussion about our attitudes to interventional medical procedures towards the end of life. I have given CPR and it has been successful, but in the context of advanced, palliative illness it hasn't, and it has been harrowing to administer and go through. Somehow, we need to communicate this better, and in Wales, for example, we have produced videos as part of the #TalkCPR campaign to bring this across, as well as a video and resource website called

It is also important to state some aspects relevant to DNACPR forms:

• Doctors do not have to give CPR. It is a clinical decision dictated by the arising medical situation.

• Doctors have to ask express permission of a patient to discuss any aspect of their medical care, including DNACPR forms, with their relatives and/or proxy, and patients can express a preference that this does not happen.

• CPR rarely works in patients with palliative illness. I have outlined the low rates of success in another article on this topic in the Guardian

And while we are clarifying some misconceptions in the popular media:

Patients in the UK do not have to sign their own DNACPR forms

If a medical team feels that CPR is not appropriate and would only cause severe suffering during a natural dying event in a patient with advanced, terminal illness, then they can make a decision to allow a natural death free from some aspects of medical intervention, like CPR, but with an emphasis on others, like intravenous antibiotics to reverse a pneumonia. So the patient will still receive all care that is required, which can include blood products, antibiotics, pain relief, even chemotherapy, radiotherapy, dialysis, etc.

Of course, the decision whether to give CPR to someone, or not, has to be very confident and, ideally, made in advance of the need for CPR arising, i.e. anticipatorily. If it hasn't and a patient arrests, then the CPR versus DNACPR decision has to happen very swiftly at that moment in time. Ward 5 in a hospital may put out a 'crash call' for a patient, seconds later the on-call cardiac arrest team arrives. The members of this emergency team do not know the patient, and in most cases a split second decision is made to administer immediate and forceful CPR, whoever the patient is. This is often witnessed by the family, who stand by while the chest is compressed and high voltage electric currents are administered to the bare chest and large needles and cannulas are inserted into veins and arteries. Had there been a well-documented DNACPR order in this patient's notes, it would have constituted in many ways a letter from one doctor to another:

'Please don't give CPR to this lady who I know well. She would not survive it and it would cause her pain. But please consider giving all other treatments to reverse any problems that may occur whilst I'm not looking after her, like blood transfusions, antibiotics etc. However if her heart and breathing stop, don't give the extreme treatment of CPR'.

So traditionally, these orders have been ways of communicating this very important decision from one doctor/nurse/team to another (on call) team. Unfortunately, these messages are sometimes not stated very clearly in medical notes and on DNACPR forms, and we need to improve. This is one very important message to take away from the Royal College of Physicians audit: whilst documentaion and discussion appears to have improved since 2013, more can and should be done.

If you have read until this point, you may be starting to realise how long it takes to explain these complexities and challenges. I can see why journalists can't face delving deeper and just opt for click-baiting headlines that will earn them a million plus hits. Not necessarily having worked in healthcare settings means there are a lot of 'unknown unknowns' for the average journalistic writer, i.e. they are often not aware of their own gaps in knowledge and context. Healthcare professionals need to engage with journalists more, perhaps via blogs and social media and portray accurately what life on the wards and in the community is actually like.

In Wales, we have a national policy on the topic of DNACPR and Allowing Natural and Anticipated Death, that is available for patients and their proxy to look at and read. The policy is called 'Sharing and Involving' and the communication strategy is called 'TalkCPR'. The idea was to make resources and videos about what CPR is and what it isn't available to patients with palliative illnesses. It should never replace a face-to-face discussion, but may inform further and act as an additional information resource. Doctors and nurses who get involved in these discussions need regular training on how to approach these sensitive serious illness discussions, and two of the videos set out to contribute to this learning.

Societal attitudes towards discussions surrounding dying and death in palliative illness are likely to change. I recently saw a lady who had read about CPR and wanted to discuss the ins and outs of the procedure in light of her many illnesses and her own life philosophy. She'd printed out the All Wales DNACPR policy from the internet. We had a really good discussion about it, involving quite a few morbid jokes. And with the advent of social media and patient empowerment, I hope we will see more patients and families coming to doctors and nurses and turning the tables: 'Doctor, can we talk about something really awkward please? My death and whether it will involve CPR? I hope you don't mind.'

Before You Go