This is a composite reflective case. Details have been changed and combined to protect privacy. It is not a record of an identifiable patient. The decision tension reflects real Emergency Medicine practice.
What I saw first
She was small and still under the ambulance blanket. Thin limbs. Shallow breathing. Eyes half-closed, not tracking. The handover was brief: elderly woman, care home resident, found unresponsive. Blood pressure low. ReSPECT form in the notes. DNACPR in place.
The department was full. Ambulances waiting outside. The board deep with patients. I was hours into a night shift that had already been heavy.
I took the form out of the notes before I examined her. I read the DNACPR section first. I glanced at the frailty score. I registered the care home address on the transfer letter. Somewhere in that sequence, before I had listened to her chest or checked her blood gas, the shape of a plan had already started forming.
That is the honest part. The part that still bothers me.
What shaped my first decision
Several things were pulling me toward a simpler decision. I want to name them.
Frailty. The transfer letter described a woman who needed help with everything. High frailty score. Dementia. That profile carries clinical weight. It tells you something about reserve. The word frailty had started to do too much work in my mind.
The care home label. Patients arriving from care homes in the early hours carry a set of associations most ED clinicians will recognise. We do not examine those associations carefully enough.
The ReSPECT form. Neatly completed. Present in the notes. Its presence changed the emotional register of the encounter. A patient with a form can become a patient for whom the decisions have already been made. The paperwork was becoming louder than the physiology.
The DNACPR wording. I saw the box. I registered it. Without meaning to, I let a decision about CPR expand into a decision about everything else.
Staff assumptions. The nursing team had already begun using comfort-care language. Nobody was unkind. They were reading the same signals. The environment was confirming a decision I had not properly made.
Fear of causing harm. I did not want to subject a frail elderly woman to burdensome treatment that would not help. That fear was legitimate. It was also doing the thinking for me.
Time pressure. Other patients. Other decisions waiting. The pull toward a quick label was strong.
Limited collateral. The transfer paperwork was incomplete. The ambulance crew had handed over a medication list and a short summary, but there were no recent bloods and no clear account of how the deterioration had started or how quickly it had moved.
I felt the pull toward a simpler decision. It came from a place that felt compassionate. That is what made it dangerous.
"The paperwork was becoming louder than the physiology."
Why I nearly got it wrong
I nearly limited treatment before I had finished my assessment.
I had anchored on the referral. Care home. DNACPR. Frailty. These anchors were factually correct. They were not diagnostically complete.
I was on the edge of assuming this was irreversible dying when I had not yet excluded the common reversible causes of acute deterioration: sepsis, metabolic derangement, drug toxicity, hypoglycaemia, arrhythmia. The differential was wide. I had collapsed it to one possibility without evidence.
I was reading DNACPR as do-not-treat. I know the distinction. Every Emergency physician knows it. The conflation still happens, quietly and repeatedly, in departments across the country.
I had not spoken to the family. I had not properly reviewed her medications against the clinical picture. I had not asked for senior input. I was treating the form instead of the patient.
If I had followed my initial trajectory, I would have placed her on a comfort pathway within the hour. I would have documented it neatly. I would have sounded compassionate in the notes. And I would have been wrong.
What made me pause
It was not one moment. It was a sequence of small things.
Her heart rate was raised. Not the gentle bradycardia of a patient slipping away. A compensatory tachycardia. Her body was fighting something.
The blood gas came back. Metabolic acidosis. Lactate raised. And the potassium was high. That potassium reading changed everything.
The ECG showed peaked T-waves and a widening QRS complex. Not subtle. Not borderline. Dangerous.
I checked the glucose. Normal. I went back to the ambulance handover and the transfer paperwork. The medication list documented an ACE inhibitor and a potassium-sparing diuretic. I reviewed the medications again after the blood gas and ECG findings. The combination made sense of the biochemistry. The electronic records confirmed she had been on both for months. A carer reached by phone said she had been off her food for days and barely drinking.
I rang the family. Her daughter told me her mother had been sitting up and talking the previous week. The decline was days, not weeks. This was not a slow trajectory. This was an acute event.
She had acute kidney injury with hyperkalaemia. A reversible emergency. Treatable with calcium gluconate, insulin and dextrose, intravenous fluids, and stopping the offending drugs.
I started treatment. Within ninety minutes the ECG began to normalise. Her potassium dropped. Her blood pressure improved. Her consciousness lifted.
She had not been dying. She had been acutely unwell with a correctable problem that I nearly missed because I had already decided what the story was.
What the form meant
The ReSPECT form records recommendations about emergency care and treatment. It is the product of a conversation between the patient, their family, and their clinical team. It exists to support decision-making in an emergency. It does not replace it.
The DNACPR section relates specifically to cardiopulmonary resuscitation. It records a decision about whether CPR should be attempted in the event of cardiac arrest. It does not address other treatments.
The form supports proportionate care. It asks clinicians to consider what interventions are appropriate in the context of the patient's values, their clinical condition, and the likely benefit of treatment.
Patient wishes matter. Reversibility matters. Likely benefit matters. The form holds these together. It does not rank them. It does not excuse the clinician from weighing them at the bedside.
What the form did not mean
It did not mean do not assess.
It did not mean do not treat.
It did not mean ignore reversible pathology.
It did not mean abandon clinical judgment.
It did not permit unwanted or non-beneficial CPR.
It did not override a valid and applicable advance decision to refuse treatment.
The form was a starting point for thinking. I had nearly used it as an ending.
The two ways to harm
There are two ways to get this wrong. Both cause harm.
One harm is over-treatment. Subjecting a frail patient to burdensome, invasive intervention against their wishes or without reasonable prospect of benefit. This happens. It causes suffering. Clinicians are right to guard against it.
The other harm is under-treatment. Withholding assessment and treatment from a frail patient because paperwork, age, or place of residence narrowed thinking too early. Because the label attached before the diagnosis. Because the form was read as a conclusion instead of a guide.
Both are failures of care. The skill is holding the tension between them long enough to make a decision that is genuinely proportionate to the patient in front of you.
"Frailty is not futility."
What happened next
No miracle. No dramatic rescue.
She was admitted. She received fluids, potassium correction, renal monitoring, and medication review. Over the following days her kidney function improved. Her potassium stabilised. Her consciousness returned to something close to her baseline.
She went back to her care home. She was not transformed. She was not cured of her frailty or her dementia. She returned to roughly the life she had been living before the acute illness.
I spoke to the family. I explained what had happened and what we had treated. The conversation was honest. I documented my reasoning, the clinical findings, the decision to treat, and the response.
It was not a triumph. It was proportionate care for a reversible illness in a frail patient. That should be unremarkable. The fact that it nearly did not happen is the point of this reflection.
What I learned
I learnt to slow down at the point where everyone else wants a quick label.
I learnt to read the form, then return to the patient.
I learnt to ask whether the illness in front of me was irreversible dying or reversible physiology.
I learnt that frailty is not futility.
I learnt that a blood gas, an ECG, and a phone call can change a decision that felt settled.
I learnt that the pull toward comfort care can come from compassion and still be wrong.
I learnt to sit with the discomfort of uncertainty rather than resolve it too early.
What I would tell a junior doctor
Read the ReSPECT form carefully. The whole form. Not just the tick box.
Clarify what is refused. A decision about CPR is not a decision about antibiotics, fluids, or potassium correction.
Do not turn DNACPR into do-not-treat.
Look for reversible physiology. Check the gas. Check the ECG. Check the glucose. Check the potassium. Check the drug chart.
Ask for collateral history early. Ring the care home. Ring the family. Find out what the patient was like last week.
Escalate uncertainty. If you are not sure, ask a senior. A decision to limit treatment carries weight. It deserves more than one pair of eyes.
Document your reasoning. Write down what you found, what you considered, and why you chose the course you chose.
Speak to family with honesty and humility. They are frightened. They deserve a straight account.
Avoid both harmful over-treatment and harmful under-treatment. The answer is rarely at either extreme. The question should always consider both.