There are many reasons why hospital doctors can be slow to engage patients on the question of DNACPR. This is one of them.
A frail patient agrees that a DNACPR form should be put in her notes. Two weeks ago she was admitted with severe pneumonia, and survived a week in ICU on a ventilator. Although she beat the odds, it was made clear to the family that if her chest deteriorated again, she would not benefit from being taken back to ICU.
The DNACPR decision is made a few days after she arrives back on the general ward, once the receiving team have taken the measure of the situation. In the event of a cardiac arrest her pulse might well be regained several minutes into the ALS algorithm (many are, after the elephantine doses of adrenaline that are given), but she would be very unlikely survive to discharge. Indeed, based on the therapeutic ‘ceiling’ set on her discharge from ICU, the option of returning to those same life support machines that are required to bring someone out of a post-cardiac arrest situation has already been ruled out.
DNACPR is clearly appropriate. The form is written. Everyone agrees, including the patient.
Then she vomits up a bowlful of blood. It looks like it could be a stomach or duodenal ulcer. Her blood pressure drops, but improves with some rapidly infused fluid, then some packed red cells. But the blood keeps coming, and it is clear that she needs an endoscopy to confirm the diagnosis and, hopefully, treat it. She is too frail to be brought down to the endoscopy department; her breathing is shallow, her blood pressure is labile. She needs to be stabilised in ICU. A phone call is made. The ICU registrar attends the ward. She flicks through the notes,
“She’s clearly not an ICU candidate. It’s all over the notes.”
“That was written in case of a respiratory deterioration. This is different.”
“It’s a bleed. It could be reversed. I accept that she shouldn’t receive multiple organ support, but she needs two or three days of close monitoring, perhaps a bit of vasopressor support* if her BP drops to keep the kidneys going. If the endoscopist can stop the bleeding she could be out of there in a day and back on the ward.”
“It rarely happens like that.”
“I’m not comfortable with the idea of her bleeding out here. Yes, she is frail, yes if she arrests, I accept it’s over. But there may well be a straightforward, treatable lesion there.”
The registrar withdraws to liase with her senior. He attends. He reads the notes, takes in the DNACPR form, heads in to look at the patient, and emerges from the curtained bay.
“I think we should honour the DNACPR… keep her comfortable…”
The medical consultant stares, trying hard to disguise his disgust.
“’Honour the DNACPR’! The DNACPR has nothing to do with her treatment in this situation. It is not a symbol designed to ensure instant palliative care. This is precisely why we don’t do DNACPRs when perhaps we should, because of this sort of….”
And he goes on, building up steam, accessing his frustration. In the end a compromise is reached… but the point has been made, and it is worth considering in more detail.
The question to consider here is – Do DNACPR forms materially influence a patient’s chances of being treated aggressively in the event of an emergency? On the face of it, they should not. The DNACPR is there to make it clear to whoever happens to be on the ward at the time, that the crash team should not be summoned. Death should be allowed to occur naturally. It says nothing about endoscopies, monitoring on the ICU, short term vasopressors… Moreover, every version of the DNACPR form I have seen emphasises that it is not irrevocable, it can be rescinded if circumstances change.
Yet, surely, there must be a close correlation between DNACPR status and ‘not for ICU’. The chances of surviving an ICU admission are driven by the same physiological parameters as those that determine survival after cardiac arrest. Looking at the figure below**, you can see how CPR sits at the apex of medical interventions, just above ICU.
Although I cannot find the studies to confirm this, in my experience, most if not all patients who are not for CPR are also not for ICU. The two go hand in hand. Therefore, by filling out a DNACPR for my patients, I am also making a statement, that he or she should not be considered for admission to ICU. But what if that is not the message I wish to convey? I can make an explanatory note; I can use a better ‘form’ (eg. ReSPECT, recently promoted in the BMJ, yet to be used nationwide). But I cannot overturn the impression that a DNACPR form makes on many colleagues, that the death of the patient should not come as a surprise, and that efforts to reverse an unexpected but life-threatening problem should not be strenuous.
‘Honour the DNACPR.’ I have heard that phrase. It is because those six initials carry so much weight, and may exclude the possibility of life-extending measures, that I, and others, are sometimes slow to start the conversations around resuscitation.
The larger point I take away from this scenario is that forms, even subtle ones like ReSPECT or other treatment escalation processes, are ill-suited to the multitude of ways in which patients can change suddenly. Although they come with the caveat that they should not be used as blanket statements, they can be read as headline summaries about how to approach patients. Different doctors will read them in different ways. To keep the options open, many doctors will choose to delay allowing those headlines to enter the notes.
* a powerful drug infused to keep up the blood pressure
** Taken from Dahill et al, Improving documentation of treatment escalation decisions in acute care BMJ Qual Improv Report 2013;2
For more on the reasons why DNACPR decisions can be delayed, explore this booklet…